7

ROOT LOCUS TECHNIQUE

7.1 Introduction

The closed-loop poles or the roots of the characteristic equation of the system ­determines the absolute and relative stability of linear control systems. Investigating the trajectories of the roots of the characteristic equation is an important study in linear control systems. Root loci introduced by W. R. Evans in 1946 are the trajectories of the roots of the characteristic ­equation or the graphical representation of the closed-loop poles when certain system parameter varies. In addition, it is a powerful method of analysis and design for stability and transient response.

The location of closed-loop poles in the control system depends on the ­variable loop gain in the system. Hence, it is necessary to have a clear knowledge about how the closed-loop poles move in the Eqn1-plane when the loop gain is varied. The closed-loop poles can be moved to the desired locations by making simple adjustment in the gain. Hence, it is ­important to select an appropriate gain value such that the transient response characteristics are ­satisfactory.

In this chapter, the technique for the construction of root locus for the characteristic ­equation of the control system with the help of simple rules is discussed.

7.2 Advantages of Root Locus Technique

The advantages of root locus technique are:

  1. It gives a clear idea about the effect of variable loop gain adjustment with relatively small effort.
  2. The stability of the system can be understood clearly.
  3. It gives an idea about the transient response of the system, i.e., whether the ­system is over-damped, under-damped or critically damped.
  4. Root locus technique can be applied to the discrete control system using z-transform.
  5. Root contours (Root loci for multivariable parameters) are obtained by varying one parameter at a time.
  6. It helps in designing the system accurately by selecting gain Eqn4 for a particular ­damping ratio.
  7. Gain and phase margins can be determined using the root locus technique.

7.3 Categories of Root Locus

The general root locus problem is formulated by referring to

Eqn5

where Eqn6 is the nth order polynomial of Eqn7 given by Eqn8

Eqn9 is the mth order polynomial of Eqn10 given by Eqn11

n and m are positive integers and Eqn12 is a variable loop gain that can vary from Eqn13

Depending on the variable loop gain in the closed-loop control system, the root locus is categorized as given in Table 7.1.

Table 7.1 ∣ Classification of root loci based on K

tbl1

7.3.1 Variation of Loop Gain with the Root Locus

The block diagram of a closed-loop control system is shown in Fig. 7.1(a) and its resultant block diagram is shown in Fig. 7.1(b). The transfer function of the closed-loop control system is given by

Eqn19

where Eqn20

C07F001a

Fig. 7.1(a) ∣ A simple closed-loop system

C07F001b

Fig. 7.1(b) ∣ Reduced form of a closed-loop system

It is clear that the closed-loop poles of the control system depend on the gain K. But the root locus depends on the values of closed-loop poles. Let us see how the closed-loop poles and hence the root locus varies with closed-loop gain.

The values of two poles (Eqn21 and Eqn22) for different values of gain Eqn23 are given in Table 7.2. The corresponding plot for the poles is shown in Fig. 7.2.

Table 7.2 ∣ Values of poles for different gain values K

tbl2

From Table 7.2, it is clear that one of the poles moves from left to right and the other one moves from right to left as closed-loop gain increases. But when Eqn24 = 100, both the values of poles are equal. If Eqn26 is increased further, poles moves into the complex plane where one pole moves in the direction of positive imaginary axis and the other one moves in the direction of negative imaginary axis.

The root loci for the above closed-loop control system with the closed-loop poles are shown in Fig. 7.2. The gain Eqn26 = 100, is the breakaway point.

The changes happening in the transient response of the system as the gain value varies can be inferred from the root locus. From Table 7.2, the following information is inferred and tabulated in Table 7.3.

Table 7.3 ∣ Nature of the system based on gain K

tbl3
C07F002

Fig. 7.2 ∣ Root loci of the closed-loop system

7.4 Basic Properties of Root Loci

The closed-loop transfer function of a simple control system is

Eqn29(7.1)

and its characteristic equation is given by

Eqn30(7.2)

The formulation of the root loci is based on the algebraic equation expressed as

Eqn31(7.3)

If Eqn36 contains a variable parameter Eqn37, then the function Eqn38 can be rewritten as

Eqn39(7.4)

Substituting Eqn. (7.4) in Eqn. (7.2), we obtain

Eqn40

Hence, the characteristic equation of the system becomes

Eqn41

Here, the numerator is similar to the general algebraic equation used for formulating the root loci. Thus, the root loci for a control system can be identified by writing the loop transfer function Eqn42 in the form of Eqn43.

7.4.1 Conditions Required for Constructing the Root Loci

The characteristic equation of the closed-loop control system is expressed as the general algebraic equation used for constructing the root locus.

Here Eqn44 whereEqn45 does not contain any variable parameter Eqn46 and s is a complex variable.

Hence, the characteristic equation of the system becomes

Eqn47

Therefore, Eqn48

From the above equation, two Evans conditions for constructing the root loci for the given system are obtained as follows:

Condition on Magnitude:

Eqn49

Conditions on Angles:

  1. Eqn50

    = Odd multiple of Eqn51 radians or 180°

  2. Eqn52

    = Even multiple of Eqn53 radians or 180°

    where Eqn54.

7.4.2 Usage of the Conditions

The following points help in the construction of root locus diagram in s - plane based on the above conditions.

  1. The trajectories of the root loci are determined by using the conditions on angles.
  2. The variable parameter Eqn56 on the root loci is determined by using the ­condition on magnitude. The value of Eqn57 is determined once the root loci are drawn.

7.4.3 Analytical Expression of the Conditions

Based on the knowledge of the poles and zeros of the loop transfer function, Eqn58, the root locus for the system is graphically constructed. Hence, the conditions required for ­constructing the root loci are to be determined in the form of poles and zeros.

Let us express the transfer functions Eqn59 as

Eqn60

where Eqn61 and Eqn62

The poles and zeros of Eqn63 can be real or purely imaginary or complex conjugate pairs.

Using the previous conditions for constructing the root loci, the conditions in terms of poles and zeros of Eqn64 are determined.

  1. Eqn65

    The conditions on angles for different root loci are

  2. For Eqn66 (RL):

    Eqn67

  3. For Eqn68 (CRL)

    Eqn69

    where Eqn70

Consider a point Eqn71 on the root loci drawn for a system whose characteristic ­equation is Eqn72.

For a positive Eqn73, any point on the RL should satisfy the following condition:

The difference in radians or degrees between the sum of angles drawn from the zeros and sum of angles drawn from poles of Eqn74 to any point s1 in the RL should be an odd multiple of Eqn76.

For a negative Eqn77, any point on the CRL should satisfy the following condition:

The difference in radians or degrees between the sum of angles drawn from the zeros and sum of angles drawn from poles of Eqn78 to any point s1 in the CRL should be an even multiple of Eqn80.

7.4.4 Determination of Variable Parameter K

The variable parameter Eqn82 in the characteristic equation Eqn83 can be determined for each point Eqn84 in the root loci by

Eqn85

The value of Eqn86 at any point s1 in the root loci is obtained by substituting the value s1 in the above equation. The numerator and denominator of the above equation is graphically interpreted as the product of the length of the vectors drawn from the poles and zeros of Eqn89 respectively.

If the point s1 lies on the RL, then the variable parameter Eqn91 has a positive value. In addition, if the point s1 lies on the CRL, then the variable parameter Eqn93 has a negative value.

Example of Constructing a Root Loci

Consider the function Eqn94.

Here the ­number of zeros Eqn96 and number of poles Eqn97 are 2 and 3 respectively.

Consider an arbitrary point s1 in the Eqn100-plane to verify the conditions. The ­vector lines are drawn from that point to each pole and zero present in the function as shown in Fig. 7.3.

C07F003

Fig. 7.3 ∣ Vector lines from an arbitrary point

The angles from each pole or zero to that arbitrary point s1 are determined with respect to positive real axis and are denoted as Eqn102 for poles and Eqn103 for zeros as ­represented in Fig. 7.3.

If the arbitrary point s1 is located in the RL, the angle condition for constructing the root loci is given by

Eqn105

Eqn106

where Eqn107.

Similarly, if the arbitrary point Eqn108 is located in the CRL, the angle condition for constructing the root loci is given by

Eqn109

Eqn110

where Eqn111

Using the arbitrary point s1, the magnitude of the variable parameter K is obtained by

Eqn114

From Fig. 7.3, the lengths of vectors drawn from each pole and zero are noted down. Now, the variable parameter Eqn115 is given by

Eqn116

The sign of the variable parameter K depends on whether the root loci are RL or CRL.

In general, for the function Eqn118 with multiplying factor as the variable parameter K with the known poles and zeros of the function, the construction of root loci follows the two steps as given below:

  1. The arbitrary point Eqn119 in the Eqn120-plane is identified, which satisfies the respective angle condition for different root loci (RL or CRL).
  2. The variable parameter Eqn121 on the root loci is obtained using the formula.

Example 7.1: The loop transfer function of the system with a negative feedback system is given by Eqn122. Consider an arbitrary point Eqn123 in s-plane. Determine whether the point s1 lies in the RL or CRL.

Solution: The phase angle of the loop transfer function at Eqn125 is

Eqn126

= Eqn127

= Eqn128

= Eqn129

Since this angle is neither an odd multiple nor an even multiple of Eqn131, the arbitrary point Eqn132 does not lie either on RL or CRL.

7.4.5 Minimum and Non-Minimum Phase Systems

If all the closed-loop zeros of the system lie in the left half of the Eqn133-plane, then the system is called non-minimum phase system; and if any one of the closed-loop zeros of the system lies in the right half of the Eqn134-plane, then the system is called minimum phase system.

7.5 Manual Construction of Root Loci

The trial-and-error method of determining the variable parameter K and an arbitrary point Eqn135 in the s-plane that satisfies the angle condition for different types of root loci (RL or CRL) is a tedious process. To overcome these difficulties, Evans invented a special tool called Spirule (which assists in addition and subtraction of angles of vectors). But this method also has its difficulties to determine the unknown parameter K.

7.5.1 Properties / Guidelines for Constructing the Root Loci

It is necessary to understand the properties, guidelines or rules for constructing the root loci for a system manually. The properties are based on the relation between the poles and zeros of Eqn136 and the roots of the characteristic equation Eqn137.

Property 1: The points on the root loci at Eqn138 and K = ±

  1. The points on the root loci corresponding to the variable parameter Eqn140 are the poles of Eqn141.
  2. The points on the root loci corresponding to the variable parameter Eqn142 are the zeros of Eqn143.

Consider a system whose characteristic equation is given by

Eqn144

When K = 0, the roots of the given equation are Eqn146,Eqn147 and Eqn148. Similarly, when Eqn149, the roots of the equation are Eqn150. Since the order of the equation is 3, the third root lies at Eqn151.

The above equation can be written as

Eqn152

Therefore, Eqn153

Thus, the poles and zeros of the given system are Eqn154 respectively.

Hence, it is clear that the roots of the characteristic equation at Eqn155 are the poles of the transfer function Eqn156 and the roots of the characteristic equation at Eqn157 including one at Eqn158 are the zeros of the of the transfer functionEqn159.

Property 2: Branches on the root loci

The locus of any one of the roots of the characteristic equation when the variable parameter K varies from Eqn161 is called the branch of the root loci.

The number of branches present in the root loci is equal to the number of poles or zeros present in the loop-transfer function depending on the values. Therefore, the condition by which the number of branches can be determined is given below:

Let Eqn162 be the number of poles present in the loop transfer function and

Eqn163 be the number of zeros present in the loop transfer function.

The number of branches of root loci is determined as:

If Eqn164, number of branches of root loci = n

If Eqn166, number of branches of root loci = m

If m = n, number of branches of root loci = m or n.

For the same characteristic equation of the system discussed in the previous property, Eqn169, the number of branches of the root loci is equal to 3, since the order of the characteristic equation is 3.

Property 3: Root locus symmetry

The root locus of any system is symmetrical on the real axis of the s-plane. The root locus of the ­system may be symmetrical about the poles and zeros of the system or to any point on the real axis of the s-plane.

The root locus of any system may be symmetrical to one or more points on the real axis of the plane. It is also possible that the root locus may be symmetrical to a complex conjugate point.

Property 4: Intersection of asymptotes with the real axis of s-plane

All the asymptotes present in the root loci of any system will intersect at a particular point on the real axis of the s-plane. The particular point on the real axis where the asymptotes intersect is called centroid.

The formula for determining the centroid is given by

Eqn170

The centroid obtained using the above formula is a real number even when the poles or zeros are either real or complex conjugate pairs.

Property 5: Asymptotes of root loci and its angles and behaviour of root loci at Eqn171

We know that the characteristic equation of any system Eqn172 can be replaced by Eqn173, i.e., Eqn174. If the order of the polynomials Eqn175 are not equal (i.e., Eqn176), then some of the branches of the root loci approach infinity in the s-plane.

Asymptotes are used to describe the properties of root loci that approach infinity in the s-plane when Eqn177. The number of asymptotes Eqn178 for a system when Eqn179 will be Eqn180, which describes the root loci behaviour at Eqn181. The angle of asymptotes is determined by

Eqn182(7.5)

where Eqn183, n is the number of finite poles of G(s)H(s) and m is the number of finite poles of G(s)H(s)

The asymptotes are drawn from the centroid Eqn185 by using the angle of asymptotes.

The asymptotes, the angle of asymptotes and the centroid (i.e., intersection of ­asymptotes on the real axis) for a transfer function Eqn186 are shown in Fig. 7.4.

For the given example, the number of asymptotes Eqn187, is 3(Eqn188).

In Fig. 7.4, A1, A2 and A3 are the asymptotes whose angles are 60°, 180° and 300° which are obtained using Eqn. (7.5).

C07F005

Fig. 7.4 ∣ Asymptotes, angle of asymptotes and centroid

Figs 7.5(a) through (c) show the asymptotes, the angle of asymptotes and the centroid for different types of system.

tbl4

Property 6: Root loci on the real axis

In the root loci, each branch on the real axis starts from a pole and ends at a zero or infinity. Each branch on the real axis is determined based on a condition given below by assuming that we are searching for a branch of root loci on the real axis from right to left.

For a branch to exist on the real axis after a particular point, the summation of number of poles and zeros of Eqn207 to the right of that particular point should be an odd number.

The point at which the above condition is checked may be pole or zero which lies on the real axis. This condition for determining the branches of root loci on the real axis does not get affected by the complex poles and zeros of Eqn208.

For the system whose transfer function is given by Eqn209, the branches of root loci are indicated in Fig. 7.6.

C07F007

Fig. 7.6 ∣ Indication of branches of root loci on the real axis

Here, if we look to the right of the point −1 (zero of the system), only one pole exists (i.e., odd number). Hence, a branch of root loci exists between −1 and 0. Now, if we look from infinity, the total number of poles and zeros existing to the right of the point is 5 (odd number). Hence, a branch of root loci exists between Eqn210.

Figures 7.7(a) and (b) indicate the branch of root loci existing on the real axis for two general systems.

tbl5

From Figs. 7.7(a) and (b), it can be inferred that the addition of one pole at the origin changes the branch of root loci existing on the real axis.

Property 7: Angle of departure and angle of arrival in root loci

In a root locus, the angle of departure or arrival is determined for a complex pole or complex zero respectively. This angle indicates the angle of the tangent to the root locus.

The angle of departure Eqn215 and angle of arrival Eqn216 is determined using

Eqn217

Eqn218

For the system whose poles are denoted as shown in Fig. 7.8(a), the angle of departure of the complex pole is determined using the formula.

The angle of departure for the pole Eqn219 is Eqn220 and the angle of ­departure for the pole Eqn221 is Eqn222.

For the system whose zeros are denoted as shown in Fig. 7.8(b), the angle of arrival of the complex zero is determined using the formula.

C07F009a

Fig. 7.8(a) ∣ Angle of departure

C07F009b

Fig. 7.8(b) ∣ Angle of arrival

The angle of arrival for the zero Eqn223 is Eqn224 and the angle of arrival for the zero Eqn225 is Eqn226.

Property 8: Intersection of root loci with the imaginary axis

The intersection point of the root loci on the imaginary axis and the corresponding variable parameter K can be determined using two ways.

  1. Using Routh−Hurwitz criterion.
  2. Substituting Eqn227 in the characteristic equation and equating the real and complex parts to zero, the intersection point on the imaginary axis Eqn228 can be obtained and its corresponding gain value K.

Property 9: Breakaway or break-in points on the root loci

The point on the s-plane where two or more branches of the root loci arrive and then depart in the opposite direction is called breakaway point or saddle point. The point on the s-plane where two or more branches of the root loci arrive and then depart in opposite direction is called break-in point. The illustrations of different types of breakaway points are given in Table 7.4.

Table 7.4 ∣ Types of breakaway points

tbl6

The root locus diagram for a particular system can have more than one breakaway point/break-in point and it is not necessary to have the breakaway point/break-in point only on the real axis. Due to the complex symmetry of root loci, the breakaway point/break-in point may be a complex conjugate pair.

The breakaway/break-in points on the root loci for a particular system whose ­characteristic equation is represented by Eqn229 must satisfy the following condition:

Eqn230(7.6)

It is noted that all the breakaway/break-in points on the branch of the root loci must ­satisfy the above condition, but not all the solutions obtained using the equation are ­breakaway points. In other words, we can say that the above condition is necessary but not sufficient. Hence, in addition to the above condition, the breakaway point must also satisfy the ­condition, Eqn231, for some real variable parameter K.

We know that Eqn232

Taking the derivative on both sides of the equation with respect to the variable s, we obtain

Eqn233

Thus, the condition for breakaway point can also be written as

Eqn234(7.7)

All the solutions obtained by solving Eqn. (7.6) or Eqn. (7.7) cannot be valid breakaway/break-in points. Let Eqn235 be the solutions obtained by solving the equation. Each ­solution Eqn236 is said to be a valid or invalid breakaway point/break-in points based on the condition given in Table 7.5.

Table 7.5 ∣ Validation of breakaway point

tbl7

Inference about the breakaway point

  1. If the solutions obtained using Eqn. (7.6) or Eqn. (7.7) are all real values and the value of Eqn248at those real values are positive, then those values are called breakaway points on the root loci.
  2. If the solutions obtained using Eqn. (7.6) or Eqn. (7.7) are complex conjugate values and if they also satisfy the condition Eqn249 (i.e., angle of Eqn250 at the breakaway point should be an odd multiple of 180°), then it is said to be a breakaway point of the root loci.

Gain at the breakaway point

Eqn251

Property 10: Root sensitivity

The root sensitivity is defined as the sensitivity of the roots of the characteristic equation when the variable parameter K varies. The root sensitivity is given by

Eqn252

Hence, at the breakaway point, the root sensitivity of the characteristic equation becomes infinite (since Eqn253 at breakaway point). The value of K should not be selected to operate the system at breakaway point.

Property 11: Determination of variable parameter K on the root loci

If the root locus for the system is constructed manually, then the variable parameter K at any point Eqn254 on the root loci can be determined by

Eqn255

But graphically, the variable parameter Eqn256 is determined by

Eqn257

The properties to plot the root loci are tabulated in Table 7.6.

Table 7.6 ∣ Properties to construct root loci

tbl8

7.5.2 Flow Chart for Constructing the Root Locus for a System

The flow chart for constructing the root locus for a system whose loop transfer function is known is shown in Fig. 7.9.

C07F010a
C07F010b

Fig. 7.9 ∣ Flow chart for construction of root locus

7.6 Root Loci for different Pole-Zero Configurations

The root loci for different pole-zero configuration are given in Table 7.7. The shape of the root loci for the system depends on the relative position of poles and zeros of the system.

Table 7.7 ∣ Root loci for different pole-zero configuration of the system

tbl9

Example 7.2: Sketch the root locus of the system whose loop transfer function is given by Eqn553.

Solution:

  1. For the given system, poles are at Eqn554, i.e., n = 4 and zeros do not exist, i.e., Eqn555.
  2. Since Eqn556, the number of branches of the root loci for the given system is Eqn557
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn558
    2. Angles of asymptotes,

      Eqn559

      Therefore, Eqn560

    3. Centroid, Eqn561

      Eqn562

  4. Since complex poles exist, angle of departure exists for the system.

    To determine the angle of departure Eqn564:

    1. For the pole Eqn565,

      Eqn566

      Eqn567

    2. For the pole Eqn568, Eqn569
  5. To determine the number of branches existing on the real axis:

    If we look from the pole p = 0, the total number of poles and zeros existing on the right of 0 is zero (nor an even number or an odd number). Therefore, a branch of root loci does not exist between 0 and Eqn570.

    Similarly, if we look from the pole p = −4, the total number of poles and zeros existing on the right of −4 is three (odd number). Therefore, a branch of root loci exists between 0 and −4.

    Hence, only one branch of root loci exists on the real axis for the given system.

  6. The breakaway and break-in points for the given system are determined by

    Eqn571(1)

    For the given system, the characteristic equation is

    Eqn572(2)

    i.e., new3

    Therefore, Eqn573(3)

    Differentiating the above equation with respect to Eqn574and using Eqn. (1), we obtain

    Eqn575

    Eqn576

    Upon solving, we obtain

    Eqn577

    Substituting s = −2 in Eqn. (3), we get K = 64. Since Eqn580 is positive, the point Eqn581 is a breakaway point.

    Since two of the three obtained solution are complex numbers, it is necessary to check the angle condition also. Therefore,

    Eqn582

    Also, Eqn583

    Eqn584

    Since the angle for the given system at the complex numbers are odd multiple of 180°, the obtained complex numbers are also breakaway points.

    Hence, Eqn585 and Eqn586 are the breaking away points for the given system.

  7. The point at which the branch of root locus intersects the imaginary axis is determined as follows:

    From Eqn. (2), the characteristic equation of the system is

    Eqn587(4)

    Routh array for the above equation,

    Eqn588

To determine the point at which root locus crosses the imaginary axis, the first element in all the rows of Routh array must be either zero or greater than zero. i.e.,

Eqn589. Therefore, Eqn590

Substituting K in the characteristic equation, we obtain

Eqn591

Substituting Eqn592 in the above equation, we get

Eqn593

Eqn594

Equating the imaginary part of the above equation to zero, we get

Eqn595

Eqn596

or Eqn597

Therefore, the point at which the root locus crosses the imaginary axis is Eqn1852.

The entire root locus for the system is shown in Fig. E7.2.

C07E005e

Fig. E7.2

Example 7.3: A pole-zero plot of a loop transfer function with gain K in the forward path of a control system is shown in Fig. E7.3(a). Sketch the root locus for the system.

C07E006a

Fig. E7.3(a)

Solution:

  1. In the pole-zero plot shown in Fig. E7.3(a), the poles are at Eqn1853 i.e.,Eqn600 and zeros are at Eqn601, Eqn602 i.e., Eqn603. Therefore, the loop transfer function, Eqn604.
  2. Since Eqn605 the number of branches of the root loci for the given system is Eqn606.
  3. For the given system, as the number of poles and number of zeros are equal, the number of asymptotes for the given system is zero. Hence, it is not necessary of ­calculating the angle of asymptotes and centroid for the given system.
  4. As all the poles and zeros are real values, there is no necessity in calculating the angle of departure and angle of arrival.
  5. Though the number of branches of the root loci for the given system is 2, it is ­necessary to determine the number of branches of the root loci existing on the real axis.

    If we look from the point p = −1, the total number of poles and zeros existing on the right of −1 is zero (neither odd nor even). Therefore, a branch of root locus does not exists between −1 and Eqn607.

    Similarly, if we look from the point p = −3, the total number of poles and zeros ­existing on the right of −3 is one (odd number). Therefore, a branch of root loci exists between −3 and −1.

    Similarly, if we look from the point z = −4, the total number of poles and zeros existing on the right of −4 is two (even number). Therefore, a branch of root loci does not exist between −4 and −3.

    Similarly, if we look from the point z = −5, the total number of poles and zeros existing on the right of −5 is three (odd number). Therefore, a branch of root loci exists between −5 and −4.

    Similarly, if we look from the point at −∞, the total number of poles and zeros existing on the right of the point is four (even number). Therefore, a branch of root loci does not exist between −∞ and −5.

    Hence, two branches of root loci exist on the real axis for the given system.

  6. The breakaway/break-in points for the given system are determined by,

    Eqn610(1)

    For the given system, the characteristic equation is

    Eqn611(2)

    i.e., new4

    Therefore, Eqn612(3)

    Differentiating the above equation with respect to Eqn613 and using Eqn. (1), we get

    Eqn614

    Eqn1854

    i.e., Eqn615

    Upon solving, we get

    Eqn616

    Substituting the values of s in Eqn. (3), we get the corresponding values of K as

    K = 0.202 for s = −2.42 and K = 143.07 for s = −4.38

    The breakaway point exists only for the positive values of K. Hence, s = −2.42 and s = −4.38 are the two breakaway points that exist for the given system.

  7. The point at which the branch of root loci intersects the imaginary axis is determined as follows:

    From Eqn. (2), the characteristic equation for the given system is

    Eqn620

    i.e., Eqn621

    Routh array for the above equation is

    Eqn622

    For Eqn623Eqn624Eqn625

    For Eqn626Eqn627Eqn628

    For Eqn629, Eqn630Eqn631

    As there exists no positive marginal value of K, the root locus does not get intersected with the imaginary axis. Also, the complete root locus lies in the left half of s-plane.

    The final root locus for the system is shown in Fig. E7.3(b).

    C07E006b

    Fig. E7.3(b)

Example 7.4: A unity feedback control system shown in Fig. E7.4(a) has a controller with a transfer function Eqn632, which controls a process with the transfer function Eqn633. Find the loop transfer function of the system and construct the root locus of the system.

C07E007a

Fig. E7.4(a)

Solution: For the given system Eqn634 and Eqn635. Therefore, the loop transfer function Eqn636.

  1. For the loop transfer function of the system, the poles are at 0, Eqn637 i.e., Eqn638 and zero does not exist i.e., Eqn639.
  2. Since Eqn640 the number of branches of the root loci for the given system is Eqn641.
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system is Eqn642.
    2. Angles of asymptotes

      Eqn643 for Eqn644

      Therefore, Eqn645

    3. Centroid

      Eqn646

  4. Since complex poles exist, the angle of departure exists for the system.

    To determine the angle of departure Eqn647:

    1. For the pole Eqn648,

      Eqn649

      Eqn650

    2. For the pole Eqn651, Eqn652
  5. To determine the number of branches existing on the real axis:

    If we look from the point p = 0, the total number of poles and zeros existing on the right of 0 is zero (neither odd nor an even number). Therefore, a branch of root loci does not exist between 0 and Eqn653.

    Similarly, if we look from the point at −Eqn654, the total number of poles and zeros ­existing on the right of the point is three. Therefore, a branch of root loci exists between −Eqn655 and 0.

    Hence, only one branch of root loci exists on the real axis for the given system.

  6. The breakaway or break-in points are determined by

    Eqn656(1)

    For the given system, the characteristic equation is

    Eqn657(2)

    i.e., new5

    Therefore, Eqn658

    Differentiating the above equation with respect to Eqn659 and using Eqn. (1), we obtain

    Eqn660

    Eqn661

    Upon solving, we obtain

    Eqn662

    Since the above points are complex conjugate values, the angle condition should be checked.

    i.e., Eqn663

    Eqn664

    Also, Eqn665

    Since the angle of Eqn666 is not an odd multiple of Eqn667, the points obtained using Eqn. (1) is not a breakaway point.

    Hence, no breakaway point exists for the given system.

  7. The point at which the branch of root loci intersects the imaginary axis is determined as follows:

    From Eqn. (2), the characteristic equation for the given system is

    Eqn668

    Routh array for the above equation is

    Eqn669

    Depending on K, the system can be stable or marginally stable or unstable.

    From Routh array, if the value of

    1. K is between 0 and 3, the system is stable.
    2. K is greater than 4, the system is unstable.

    Now, if K = 4, the third row of Routh array becomes 0. Hence, the ­auxiliary equation is Eqn670i.e., Eqn671.

    Solving the above equation, we obtain Eqn672, which is the intersection point on the imaginary axis.

    The final root locus for the system is shown in Fig. E7.4(b).

    C07E007b

    Fig. E7.4(b)

Example 7.5: Develop the root locus for the system whose loop transfer function is given by Eqn673

Solution:

  1. For the given system, the poles are at 0, −1 and −4 i.e., Eqn674 and zero is at −2 i.e., Eqn675.
  2. Since Eqn676, the number of branches of the root loci for the given system is Eqn677
  3. The details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn678.
    2. Angles of asymptotes

      Eqn679 for Eqn680

      Therefore, Eqn681

    3. Centroid

      Eqn682

  4. As all the poles and zeros are real values, there is no necessity to calculate the angle of departure and angle of arrival.
  5. To determine the number of branches existing on the real axis:

    If we look from the pole p = −1, the total number of poles and zeros existing on the right of −1 is one (odd number). Therefore, a branch of root loci exists between −1 and 0.

    Similarly, if we look from the pole p = −4, the total number of poles and zeros ­existing on the right of −4 is three (odd number). Therefore, a branch of root loci exists between −4 and −2.

    Hence, two branches of root loci exist on the real axis for the given system.

  6. The breakaway point for the given system is determined by

    Eqn686(1)

    For the given system, the characteristic equation is

    Eqn687(2)

    i.e., new6

    Therefore, Eqn688

    Differentiating the above equation with respect to Eqn689 and using Eqn. (1), we obtain

    Eqn690

    Eqn691

    The above equation can be solved using Lin's method (trial-and-error method). The third-order polynomial of the above equation will have at least one real root. The trial and error method for determining the roots is shown below:

    tbl10

    Since the remainder obtained in the fifth trial is Eqn697, the point Eqn698 is one of the roots of the equation.

    Hence, rewriting the above cubic equation, we obtain

    Eqn699

    Solving the above equation, we obtain

    Eqn701

    Substituting s = −0.549 in Eqn. (3), we get K = 0.5888. Since Eqn704 is positive, the point Eqn705 is a breakaway point.

    As the other two points are complex conjugate points, it is necessary to determine the angles of the system at those points.

    i.e., Eqn706

    Also, Eqn707

    Since the angles obtained are not odd multiple of Eqn708, the complex points obtained using Eqn. (1) are not a breakaway or break-in points.

    Hence, only one breakaway point exists for the given system.

  7. The point at which the branch of root loci intersects the imaginary axis is obtained as follows:

    Using Eqn. (2), the characteristic equation for the given system is

    Eqn709

    i.e., Eqn710

    Routh array for the above equation is

    Eqn711

    To determine the point at which the root locus crosses the imaginary axis, the first element in the third row must be zero, i.e., Eqn712. Therefore, Eqn713. As K is a negative real value, the root locus do not cross the imaginary axis.

    The final root locus for the system is shown in Fig. E7.5.

    C07E008

    Fig. E7.5

Example 7.6: Sketch the root locus for the system whose loop transfer function Eqn715. Also, (i) determine gain Eqn716 when Eqn717, (ii) determine the settling time Eqn718 and peak time Eqn719 using the gain value obtained in (i).

Solution:

  1. For the given system Eqn720, the poles are at 0, −2 and −4 i.e., Eqn721 and zeros do not exist i.e., Eqn722.
  2. Since Eqn723, the number of branches of the root loci for the given system is Eqn724.
  3. The details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn725.
    2. Angles of asymptotes

      Eqn726 forEqn727

      Therefore, Eqn728

    3. Centroid

      Eqn729

  4. As all the poles and zeros are real values, there is no necessity in calculating the angle of departure and angle of arrival.
  5. To determine the number of branches existing on the real axis:

    If we look from the pole p = −2, the total number of poles and zeros existing on the right of −2 is one (odd number). Therefore, a branch of root loci exists between −2 and 0.

    Similarly, if we look from the point at Eqn731, the total number of poles and zero existing on the right of the point is three (odd number). Therefore, a branch of root loci exists between Eqn732 and −4.

    Hence, two branches of root loci exist on the real axis for the given system.

  6. The breakaway/break-in points for the given system are determined by

    Eqn733(1)

    For the given system, the characteristic equation is

    Eqn734(2)

    i.e., new7

    Therefore, Eqn735(3)

    Differentiating the above equation with respect to Eqn736 and using Eqn. (1), we obtain

    Eqn737

    Eqn738

    Solving for s, we get

    Eqn739

    Substituting the values of s in Eqn. (3), we get the corresponding values of K as

    K = 3.0792 for s = −0.8453 and K = −3.0792 for s = −3.1547.

    The breakaway point exists only for the positive values of K. Hence, s = −0.8453 is the only one breakaway point that exists for the given system.

  7. The point at which the branch of root loci intersects the imaginary axis is determined as follows:

    From Eqn. (2), the characteristic equation for the given system is

    Eqn746

    i.e., Eqn747(4)

    Routh array for the above equation is

    Eqn748

    To determine the point at which the root locus crosses the imaginary axis, the first ­element in the third row must be zero, i.e., Eqn749. Therefore, Eqn750

    Substituting K in Eqn. (3), we get

    s3 + 6s2 + 8s + 48 = 0

    Upon solving, we get

    s = −6, j2.828 , −j2.828.

    Hence, the root locus crosses the imaginary axis at Eqn753.

    The complete root locus for the system is shown in Fig. E7.6(a).

    C07E009a

    Fig. E7.6(a)

    1. It is given that the maximum peak overshoot for the system is Eqn755. We know that the maximum peak overshoot for a second-order system is

      Eqn756

      Substituting the known values, we obtain

      Eqn757

      Eqn758

      Solving the above equation, we obtain

      Eqn759

      For this damping ratio Eqn760 a straight line can be drawn from the origin with the angle, Eqn761, obtained by Eqn762. The point at which the straight line crosses on the root locus is Eqn763 as shown in Fig. E7.6(b).

      C07E009b

      Fig. E7.6(b)

      We know that the magnitude condition at any point of the system

      Eqn765

      i.e., Eqn766

      Therefore, Eqn767

      Therefore, gain K corresponding to the given peak overshoot, Eqn768 is K = 9.174.

    2. The point at which the line for Eqn769 cuts the root locus is Eqn770. The settling time and peak time of the system is obtained as

      Eqn771

      Eqn772

Example 7.7: Prove that the breakaway points for a system in the root locus are obtained by solving for the solutions of Eqn1857 where K is the ­open-loop gain (variable parameter) of the system whose open-loop transfer function is Eqn1858.

Solution: The characteristic equation of the system whose open-loop transfer function of the system given by Eqn805 is obtained as Eqn1859 which is the combination of s terms and open-loop gain K terms. It can be arranged as

Eqn806(1)

where P(s) is the polynomial containing K and s terms and

KQ(s) is the polynomial containing K and s terms

Now at breakaway points, multiple roots occur. Mathematically, this is possible if Eqn809.

Differentiating Eqn. (1) with respect to s, we get

Eqn810

Therefore, Eqn811(2)

Substituting Eqn. (2) in Eqn. (1) at breakaway point, we can write

Eqn812(3)

Solving Eqn. (3) for s, breakaway points can be obtained.

Now, equating Eqn. (1) to zero and solving for K, we get

Eqn813

Therefore, Eqn814(4)

i.e., Eqn815(5)

This is same as Eqn. (3) which yields breakaway points. This proves that the roots of Eqn816 are the actual breakaway points.

Example 7.8: (i) Prove that the root locus of the system Eqn817 is a circle. (ii) Assuming a = 4 and b = 6 for the above loop transfer function, develop the root locus. (iii) Determine the range of variable gain, K for which the system is ­under-damped. (iv) Determine the range of variable gain, K for which the system is critically damped and (v) Determine the minimum value of damping ratio.

Solution:

  1. Let Eqn818 be a complex point on the root locus for the given system.

    Therefore, Eqn819

    To prove that the root locus of the system is a circle, the angle of the system at any point must be 180°.

    i.e., Eqn820(1)

    Therefore, Eqn821

    Eqn822

    Eqn823

    Eqn824

    Eqn825

    Eqn826

    From Eqn. (1),

    Eqn827

    Therefore, Eqn828

    Since Eqn829, we obtain

    Eqn830

    Therefore, Eqn831

    Adding and subtracting Eqn832 on both sides, we obtain

    Eqn833

    Therefore, Eqn834

    The above equation is the equation of a circle with centre Eqn835 which is the location of open loop zero and radius Eqn836. This proves that root locus of the given system is a circle.

  2. Assuming a = 4 and b = 6, the loop transfer function of the system becomes Eqn837.
    1. For the given system, the poles are at 0 and −4, i.e., Eqn838 and zero is at −6, i.e., Eqn839.
    2. Since Eqn840, the number of branches of the root loci for the given system Eqn841
    3. For the given system, the details of the asymptotes are given below.
      1. Number of asymptotes for the given system Eqn842.
      2. Angles of asymptotes

        Eqn843for Eqn844

        Therefore, Eqn845

      3. Centroid

        Eqn846

    4. Since no complex pole or complex zero exists for the given system, there is no ­necessity of calculating the angle of departure or angle of arrival.
    5. To determine the number of branches existing on the real axis:

      If we look from the pole p = −4, the total number of poles and zeros ­existing on the right of the point is one (odd number). Therefore, a branch of root loci exists between −4 and 0.

      Similarly, if we look from the point at Eqn848, the total number of poles and zeros existing on the right of the point is three (odd number). Therefore, a branch of root loci exists between −6 and Eqn849

      Hence, two branches of root loci exist on the real axis for the given system.

    6. The breakaway and break-in points for the given system are determined by

      Eqn850(1)

      For the given system, the characteristic equation is

      Eqn851(2)

      i.e., new8

      Therefore, Eqn852(3)

      Differentiating the above equation with respect to Eqn853 and using Eqn. (1), we obtain

      Eqn854

      Eqn855

      Upon solving, we obtain

      Eqn856

      Substituting the values of s in Eqn. (3), we get the corresponding values of K as

      K = 1.071 for s = −2.5358 and K = 14.928 for s = −9.4641.

      The breakaway point exists only for the positive values of K. Hence, s = −2.5358 and s = −9.4641 are the two breakaway points that exist for the given system.

    7. The point at which the branch of root loci intersects the imaginary axis is determined as follows:

      From Eqn. (2), the characteristic equation for the given system is

      Eqn860(4)

      Routh array for the above equation is

      Eqn861

      To determine the point at which the root locus crosses the imaginary axis, the first element in the second row must be zero, i.e., Eqn862.

      As K is a negative real value, the root loci does not cross the imaginary axis.

      Hence, the final root loci for the system are shown in Fig. E7.8.

      C07E013b

      Fig. E7.8

  3. For the system to be under-damped, the range of gain K is

    Eqn874

  4. The system is critically damped when the gain K = 1.0717 or 14.928.
  5. Determination of minimum value of damping ratio Eqn875

    For the root locus of the given system a tangent is drawn from the origin to the root locus and the angle from the real axis to the tangent is measured as Eqn876 as shown in Fig. E7.8.

    Therefore, the damping ratio, Eqn877.

Example 7.9: Plot the root locus for the system whose loop transfer function is Eqn879. Also, determine the variable parameter K(i) for marginal stability of the system and (ii) at the point s = −4.

Solution:

  1. For the given system, the poles are at 0, −1 and −3 i.e., Eqn880 and zero does not exist i.e., Eqn881.
  2. Since Eqn882, the number of branches of the root loci for the given system is Eqn883.
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn884.
    2. Angles of asymptotes

      Eqn885 for Eqn886

      Therefore, Eqn887

    3. Centroid

      Eqn888

  4. Since no complex pole or complex zero exists for the given system, there is no ­necessity of calculating the angle of departure or angle of arrival.
  5. To determine the number of branches existing on the real axis:

    If we look from the pole p = −1, the total number of poles and zeros existing on the right of −1 is one (odd number). Therefore, a branch of root loci exists between −1 and 0.

    Similarly, if we look from the point atEqn889, the total number of poles and zeros existing on the right of the point is three. Therefore, a branch of root loci exists between Eqn890 and −3.

    Hence, two branches of root loci exist on the real axis for the given system.

  6. The breakaway/break-in points for the given system are determined by

    Eqn891(1)

    For the given system, the characteristic equation is

    Eqn892(2)

    i.e., new9

    Therefore, Eqn893(3)

    Differentiating the above equation with respect to Eqn894 and using Eqn. (1), we obtain

    Eqn895

    Eqn1870

    Upon solving, we obtain

    Eqn896

    Substituting the values of s in Eqn. (3), we get the corresponding values of K as

    K = 0.63311 for s = −0.451 and K = −2.11 for s = −2.215

    The breakaway point exists only for the positive values of K. Hence, s = −0.451 is the only one breakaway point that exists for the given system.

  7. The point at which the branch of root locus crosses the imaginary axis is determined as follows:

    From Eqn. (2), the characteristic equation for the given system is

    Eqn905a

    Therefore, Eqn905(4)

    Routh array for the above equation is

    Eqn1871

    To determine the point at which the root locus crosses the imaginary axis, the first ­element in the third row must be zero, i.e., Eqn906 Therefore, Eqn907

    Substituting K in Eqn. (3), we get

    s3 + 4s2 +3s +12 = 0

    Upon solving, we get

    s = −4, j1.732 , −j1.732.

    Hence, the root locus crosses the imaginary axis at Eqn910.

    The root locus plot for the system is shown in Fig. E7.9.

    C07E014

    Fig. E7.9

  1. The marginal value of variable gain K = 12.
  2. The variable gain K at any point s is obtained by

    Eqn912

    Hence, at s = −4, we obtain

    Eqn913

    K = 12.

Example 7.10: Sketch the root locus of the system whose loop transfer function is given by Eqn914. Also, determine ω at Eqn916.

Solution:

  1. For the given system, poles are at Eqn917 and new i.e., Eqn918 and zero is at −2 i.e., Eqn919.
  2. Since Eqn920, the number of branches of the root loci for the given system is Eqn921.
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn922
    2. Angles of asymptotes

      Eqn923 for Eqn924

      Therefore, Eqn925.

    3. Centroid

      Eqn926

  4. Since complex poles exist, the angle of departure exists for the system.

    To determine the angle of departure θd:

    1. For the pole Eqn927

      Eqn928

    2. For the pole Eqn929Eqn930.
  5. To determine the number of branches existing on the real axis:

    If we look from point at Eqn931, the total number of poles and zeros ­existing on the right of the point is three (odd number). Therefore, a branch of root loci exists between Eqn932 and −2.

    Hence, one branch of root loci exists on the real axis for the given system.

  6. The breakaway/break-in points for the given system are determined by

    Eqn933(1)

    For the given system, the characteristic equation is

    Eqn934(2)

    i.e., new10

    Therefore, Eqn935(3)

    Differentiating the above equation with respect to Eqn936 and using Eqn. (1), we obtain

    Eqn937

    Eqn938

    Upon solving, we obtain

    Eqn939 and −3.73

    Substituting the values of s in Eqn. (3), we get the corresponding values of K as

    K = 10.62 for s = −3.73 and K = −1.464 for s = −0.26

    The breakaway point exists only for the positive values of K. Hence, s = −3.73 is the only one breakaway point that exists for the given system.

  7. The point at which the branch of root locus crosses the imaginary axis is determined as follows:

    From Eqn. (2), the characteristic equation of the system is

    Eqn948(4)

    Routh array for the above equation is

    Eqn949

    To determine the point at which root locus crosses the imaginary axis, the first element in the second row and third row must be zero, i.e., Eqn950 and Eqn951. Solving for Eqn952, we obtain Eqn953 and Eqn954. Since Eqn955 lies within Eqn956, we take Eqn956.

    As K is negative real value, the root loci do not cross the imaginary axis.

    The final root locus for the system is shown in Fig. E7.10.

    C07E015

    Fig. E7.10

To determinine damping ratio ω

The characteristic equation of a second-order system is

Eqn958(5)

where Eqn959 is the undamped natural frequency of the system and Eqn960 is the damping ratio.

The characteristic equation of the given system after substituting Eqn961, is

Eqn962(6)

Comparing Eqs. (5) and (6), we obtain

Eqn963

Given K = 1.33, Hence new11 i.e., ωn= 2.379

Then, Eqn964

Therefore, the damping ratio, Eqn965

Example 7.11: Plot the root locus of the system whose loop transfer function is given by Eqn994.

Solution:

  1. For the given system, poles are at 0, −1, −3 and −5 i.e., Eqn995 and zeros do not exist i.e., Eqn996.

    Since Eqn997, the number of branches of the root loci for the given system is Eqn998.

  2. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn999.
    2. Angles of asymptotes,

      Eqn1000 for Eqn1001

      Therefore, Eqn1002

    3. Centroid

      Eqn1003

  3. Since no complex pole or complex zero exists for the system, there is no necessity of determining angle of departure/angle of arrival.
  4. To determine the number of branches existing on the real axis:

    Since the total number of poles and zeros existing on the right of −1 and −5 are 1 and 3 (odd numbers), a branch of root loci exists between −1 and 0, and −5 and −3.

    Hence, two branches of root loci exist for a given system on the real axis.

  5. The breakaway/break-in points for the given system are determined by

    Eqn1007(1)

    For the given system, the characteristic equation is

    Eqn1008(2)

    i.e., new12

    Therefore, Eqn1009(3)

    Differentiating the above equation with respect to Eqn1010 and using Eqn. (1), we obtain

    new1

    Upon solving, we obtain

    Eqn1011

    Substituting the values of s in Eqn. (3), we get the corresponding values of K as

    K = 2.878 for s = −0.425, K = 12.949 for s = −4.253 and K = −6.035 for s = −2.07

    The breakaway points exist only for the positive values of K. Hence, s = −0.425 and s = −4.253 are the two breakaway points that exist for the given system.

  6. The point at which the branch of root locus crosses the imaginary axis is determined as follows:

    To find the crossing point on the imaginary axis, substitute Eqn1020 in

    Eqn1021

    Eqn1022

    Eqn1023

    Equating imaginary part to zero, we obtain

    Eqn1024

    Eqn1025

    Eqn1026

    or Eqn1027

    Hence, the root locus intersects at Eqn1028.

    Equating real parts to zero, we obtain

    Eqn1029

    K = 35.55.

    Hence, the gain of the system at which the root locus intersects the imaginary axis is 35.55.

    The complete root locus for the system is shown in Fig. E7.11.

    C07E017

    Fig. E7.11

Example 7.12: Sketch the root locus of the system whose loop transfer function is given by Eqn1031.

Solution:

  1. For the given system, poles are at 0, −1 and −2, i.e., Eqn1032 and zeros do not exist i.e., Eqn1033.

    Since Eqn1034, the number of branches of the root loci for the given system is Eqn1035.

  2. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn1036
    2. Angles of asymptotes

      Eqn1037 for Eqn1038

      Therefore, Eqn1039

    3. Centroid

      Eqn1040

  3. Since no complex pole or complex zero exists for the system, there is no necessity in determining the angle of departure/angle of arrival.
  4. To determine the number of branches existing on the real axis:

    If we look from the pole p = −1, the total number of poles and zeros existing on the right of −1 is one (odd number). Therefore, a branch of root loci exists between −1 and 0.

    Similarly, if we look from the pole p = −Eqn1042, the total number of poles and zeros existing on the right of the point is three (odd number). Therefore, a branch of root loci exists between −Eqn1043 and −2.

    Hence, two branches of root loci exist on the real axis for the given system.

  5. The breakaway/break-in points for the given system can be determined by

    Eqn1044(1)

    For the given system, the characteristic equation is

    Eqn1045(2)

    i.e., new14

    Therefore, Eqn1046(3)

    Differentiating the above equation with respect to Eqn1047 and using Eqn. (1), we obtain

    Eqn1048

    Upon solving, we obtain

    Eqn1049

    Substituting the values of s in Eqn. (3), we get the corresponding values of K as

    K = 0.384 for s = −0.42 and K = −0.375 for s = −1.5

    The breakaway point exists only for the positive values of K. Hence, s = −0.42 is the only one breakaway point that exists for the given system.

  6. The point at which the branch of root locus crosses the imaginary axis can be ­determined as follows:

    From Eqn. (2), the characteristic equation of the system is

    Eqn1055(4)

    Routh array for the above equation is

    Eqn1056

    To determine the point at which root locus crosses the imaginary axis, the first ­element in the third row must be zero, i.e., Eqn1057. Therefore, Eqn1058

    Substituting K in Eqn. (4) and solving forEqn1059, we obtain

    Eqn1060

    Hence, the root locus crosses the imaginary axis at Eqn1061.

    The final root locus for the system is shown in Fig. E7.12.

    C07E018

    Fig. E7.12

Example 7.13: The characteristic polynomial of a feedback control system is given by Eqn1063. Plot the root locus for this system.

Solution: The characteristic polynomial for the given system is

Eqn1064

i.e., Eqn1065

Eqn1066

Eqn1067

i. e., Eqn1068

We know that the characteristic equation is Eqn1069.

Therefore, Eqn1070

  1. For Eqn1071, poles are at 0 and −1, i.e., Eqn1072 and zeros are at −2 and −3, i.e., Eqn1073.
  2. Since Eqn1074, the number of branches of the root loci for the given system is Eqn1075
  3. Since the number of poles and number of zeros present in the system are equal, no asymptote exists for the system and hence the angle of asymptotes and centroid need not be determined for the system.
  4. Since no complex pole or complex zero exists for the system, there is no necessity of determining angle of departure/angle of arrival.
  5. To determine the number of branches existing on the real axis:

    Since the total number of poles and zeros existing on the right of −1 and −3 are 1 and 3 (odd numbers), a branch of root loci exists between −1 and 0 and −3 and −2.

    Hence, two branches of root loci exist on the real axis for the given system.

  6. The breakaway and break-in points for the given system can be determined by

    Eqn1079(1)

    For the given system, the characteristic equation is

    Eqn1080(2)

    i.e., new15

    Therefore, Eqn1081(3)

    Differentiating the above equation with respect to Eqn1082 and using Eqn. (1), we obtain

    Eqn1083Eqn1084

    Eqn1085

    1.

    Upon solving, we get

    Eqn1881

    For Eqn1086 and Eqn1087, the values of K are 0.0717 and 13.93 respectively. Since the values of Eqn1088 are positive, the points Eqn1089 and Eqn1090 are breakaway point and break-in point.

  7. The point at which the branch of root loci crosses the imaginary axis is determined as follows:

    From Eqn. (2), the characteristic equation of the system is

    Eqn1091(4)

    Routh array for the above equation,

    Eqn1092

    To determine the point at which root locus crosses the imaginary axis, the first ­element in all the rows must be zero, i.e., Eqn1093. Therefore, Eqn1094.

    Since K is negative, the root locus does not cross the imaginary axis.

    The complete root locus for the system is shown in Fig. E7.13.

    C07E019

    Fig. E7.13

Example 7.14: Sketch the root locus of the system whose transfer function is given by Eqn1158.

Solution: We know that, the closed-loop transfer function of the system with the feedback transfer function Eqn1159 and open loop transfer function Eqn1160 is given by

Eqn1161

Converting the given transfer function to the above form, we obtain

Eqn1162

Hence, the function Eqn1163 for which the root locus has been drawn in the subsequent steps:

  1. For the given system Eqn1164, poles are at Eqn1165 i.e., Eqn1166 and zeros do not exist i.e., Eqn1167.
  2. Since Eqn1168, the number of branches of the root loci for the given system is Eqn1169.
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn1170
    2. Angles of asymptotes

      Eqn1171 for Eqn1172

      Therefore, Eqn1173

    3. Centroid

      Eqn1174

      Eqn1175

  4. Since complex poles exist, the angle of departure exists for the system.

    To determine the angle of departure Eqn1176:

    1. For the pole Eqn1177

      Eqn1178

      Eqn1179

    2. For the pole Eqn1180, Eqn1181
  5. To determine the number of branches existing on the real axis:

    Since the total number of poles and zeros existing on the right of −4 is 3 (odd number), a branch of root loci exists between −4 and 0.

    Hence, one branch of root loci exists on the real axis for the given system.

  6. The breakaway point/break-in points for the given system can be determined by

    Eqn1185(1)

    For the given system, the characteristic equation is

    Eqn1186(2)

    i.e., new18

    Therefore, Eqn1187(3)

    Differentiating the above equation with respect to Eqn1188 and using Eqn. (1), we obtain

    Eqn1189

    Upon solving, we obtain

    Eqn1882

    For Eqn1190, Eqn1191 is 21.01. Since Eqn1192 is positive, the point at Eqn1193 is a breakaway point.

    Since the other two points obtained by solving Eqn. (1) are complex numbers, it is necessary to check the angle condition also.

    Hence,

    Eqn1194

    Eqn1195

    Similarly, Eqn1196

    Since the angles obtained for the given system at the complex numbers obtained on solving Eqn. (1) are not odd multiple of 180°, the complex numbers are not ­breakaway points.

    Hence, only one breakaway point exists for the given system.

  7. The point at which the branch of root loci intersects the imaginary axis can be ­determined as follows:

    From Eqn. (2), the characteristic equation of the system is

    Eqn1197(4)

    Routh array for the above equation is

    Eqn1198

    To determine the point at which root locus crosses the imaginary axis, the first ­element in all the rows must be zero, i.e., Eqn1199. Therefore, Eqn1200.

    Substituting K in the characteristic equation, we obtain

    Eqn1201

    Substituting s = jω in the above equation, we obtain

    Eqn1202

    Eqn1203

    Equating the imaginary part of the above equation to zero, we obtain

    Eqn1204

    Eqn1205

    or Eqn1206

    Therefore, the point at which the root locus crosses the imaginary axis is Eqn1207.

    The complete root locus for the system is shown in Fig. E7.14.

    C07E021

    Fig. E7.14

Example 7.15: Sketch the root locus of the system whose loop transfer function is given by Eqn1209. Also, determine gain K

(i) when the damping ratio is ξ = 0.707 and (ii) for which repetitive roots occur.

Solution The given function can be rewritten as

Eqn1211

Replacing Eqn1212 by K' in the above equation, we obtain

Eqn1213

  1. For the given system, poles are at 0 and −2 i.e., Eqn1214 and zeros are at Eqn1215 i.e., Eqn1216.
  2. Since Eqn1217, the number of branches of the root loci for the given system is Eqn1218.
  3. Since the number of poles and zeros present in the system are equal, the ­given system does not have any asymptote. Also, for the given system, it is not ­possible to determine the angle of asymptotes and centroid
  4. Since complex zeros exist, the angle of arrival exists for the system.

    To determine the angle of arrival Eqn1219:

    1. For the zero Eqn1220,

      Eqn1221

      Eqn1222

    2. For the zero Eqn1223, Eqn1224
  5. To determine the number of branches existing on the real axis:

    Since the total number of poles and zeros existing on the right of −2 is 3 (odd number), a branch of root loci exists between −2 and 0.

    Hence, one branch of root loci exists on the real axis for the given system.

  6. The breakaway/break-in points for the given system can be determined by

    Eqn1227(1)

    For the given system, the characteristic equation is

    Eqn1228(2)

    i.e., new19

    Therefore, Eqn1229(3)

    Differentiating the above equation with respect to Eqn1230 and using Eqn. (1), we get

    Eqn1885

    Eqn1886

    Eqn1887

    Upon solving, we obtain

    Eqn1888

    Substituting the values of s in Eqn. (3), we get the corresponding values of K as

    K = −1.6 for s = 2 and K = 0.5 for s = −1

    The breakaway point exists only for the positive values of K. Hence, s = −1 is the only one breakaway point that exists for the given system.

  7. The point at which the root loci intersects the imaginary axis can be obtained as follows:

    From Eqn. (2), the characteristic equation of the system is

    Eqn1236

    Routh array for the above equation is

    Eqn1237

    To determine the point at which root locus crosses the imaginary axis, the first element in all the rows must be zero, i.e., Eqn1238. Therefore, Eqn1239

    Thus, Eqn1240

    Since K is negative, the root locus does not cross the imaginary axis at any point.

    The complete root locus for the system is shown in Fig. E7.15.

    C07E022

    Fig. E7.15

  1. Determinine K:

    Given the damping ratio Eqn1242angle Eqn1243Eqn1244.

    Draw a line from origin at an angle of 45° so that the line bisects the root locus.

    Determine the point at which the root locus has been bisected. For the given problem, the root locus gets bisected at Eqn1245.

    Hence, the magnitude of the given transfer function at that particular point is 1.

    Eqn1889 atEqn1246

    Eqn1247

    Eqn1248

    Eqn1249

    Eqn1890

    Therefore, the value of K for the given damping ratio is Eqn1250

  2. Finding gain at which the repetitive roots occur:

    For a given system, the repetitive roots occur only at breakaway or break-in points.

    Therefore, Eqn1251

    i.e., Eqn1252

    Eqn1253

    Eqn1254

    Therefore, the value ofEqn1255at which repetitive roots occur isEqn1256.

Example 7.16: Construct the root locus of the system whose loop transfer function is given by Eqn1302.

Solution:

  1. For the given system, poles are at 0, 0 and −10 i.e., Eqn1303 and zero is at −2 i.e., Eqn1304.
  2. Since Eqn1305, the number of branches of the root loci for the given system is Eqn1306
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn1307
    2. Angles of asymptotes

      Eqn1308 for Eqn1309

      Therefore, Eqn1310 and Eqn1311

    3. Centroid

      Eqn1312

      Eqn1313

  4. Since a complex pole or complex zero does not exist for the given system, there is no angle of departure/angle of arrival.
  5. To determine the number of branches existing on the real axis:

    If we look from the pole p = −10, the total number of poles and zeros existing on the right of −10 is three (odd number). Therefore, a branch of root loci exists between −10 and −2.

    Hence, only one branch of root loci exists on the real axis for the given system.

  6. The breakaway/break-in points for the given system can be determined by

    Eqn1317(1)

    For the given system, the characteristic equation is

    Eqn1318(2)

    i.e., new20

    Therefore Eqn1319(3)

    Differentiating the above equation with respect to Eqn1320 and using Eqn. (1), we obtain

    Eqn1321

    Eqn1322

    Eqn1323

    Eqn1324

    Upon solving, we obtain

    Eqn1325

    Substituting Eqn1326 in Eqn. (3), we obtain Eqn1327. Hence, the point is not a valid breakaway point.

    For the other two points, i.e., Eqn1328, it is necessary to check the angle condition.

    Eqn1329

    Similarly, Eqn1330

    Since the angle of the system at the complex numbers is not an odd multiple of 180°, these complex numbers are not valid breakaway points.

    Hence, no breakaway point exists for the given system.

  7. The point at which the branch of root loci intersects the imaginary axis can be ­determined as follows:

    From Eqn. (2), the characteristic equation of the system is

    Eqn1331(4)

    Routh array for the above equation is

    Eqn1332

    To determine the point at which root locus crosses the imaginary axis, the first element in all the rows must be zero i.e., Eqn1333. Therefore, Eqn1334.

    Substituting the value K = 0 in Eqn. (4), we obtain

    Eqn1335

    Upon solving, Eqn1336.

    Since all the values of Eqn1337 are real, the root locus does not cross the imaginary axis.

    The complete root locus for the given system is shown in Fig. E7.16.

    C07E024

    Fig. E7.16

Example 7.17: The loop transfer function of a system is given by Eqn1380. Construct the root locus of the system.

Solution:

  1. For the given system, poles are at 0 and 1, i.e., Eqn1381 and zero is at −1 i.e., Eqn1382.
  2. Since Eqn1383, the number of branches of the root loci for the given system is Eqn1384.
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn1385.
    2. Angles of asymptotes

      Eqn1386 for Eqn1387

      Therefore, Eqn1388

    3. Centroid

      Eqn1389

      Eqn1390

  4. Since a complex pole or complex zero does not exist for the given system, there is no angle of departure/angle of arrival.
  5. To determine the number of branches existing on the real axis:

    If we look from the pole p = 0, the total number of poles and zeros existing on the right of 0 is one (odd number). Therefore, a branch of root loci exists between 0 and 1.

    Similarly, if we look from the point at Eqn1392, the total number of poles and zeros existing on the right of that point is three (odd number). Therefore, a branch of root loci exists between −1 and Eqn1393.

    Hence, two branches of root loci exist on the real axis for the given system.

  6. The breakaway and break-in points for the given system can be determined by

    Eqn1394(1)

    For the given system, the characteristic equation is

    Eqn1395(2)

    i.e., new21

    Therefore, Eqn1396(3)

    Differentiating the above equation with respect to Eqn1397 and using Eqn. (1), we obtain

    Eqn1398

    Eqn1399

    Therefore, Eqn1400.

    For Eqn1401 and Eqn1402, the values of Eqn1403 are 0.1715 and 5.828 respectively. Since the values of Eqn1404 are positive, Eqn1405 and Eqn1406 are the breakaway and break-in points respectively.

  7. The point at which the branch of root loci intersects the imaginary axis can be ­determined as follows:

    From Eqn. (2), the characteristic equation of the system is

    Eqn1407(4)

    Routh array for the above equation is

    Eqn1408

    To determine the point at which root locus crosses the imaginary axis, the first element in all the rows must be zero, i.e., Eqn1409 Therefore, Eqn1410.

    Substituting K in Eqn. (4), we obtain

    Eqn1411

    Upon solving, we get

    Eqn1412

    Therefore, the root locus crosses the imaginary axis at Eqn1413.

    The complete root locus for the system is shown in Fig. E7.17.

    C07E026

    Fig. E7.17

Example 7.18: Develop the root locus of the system whose loop transfer function is given by Eqn1415 using the properties of root loci.

Solution:

  1. For the given system, poles are at Eqn1416 i.e., Eqn1417 and zero does not exist i.e., Eqn1418.
  2. Since Eqn1419, the number of branches of root loci for the given system is Eqn1420.
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn1421.
    2. Angles of asymptotes

      Eqn1422 for Eqn1423.

      Therefore, Eqn1424.

    3. Centroid

      Eqn1425

      Eqn1426

  4. Since complex poles exist, the angle of departure exists for the system.

    To determine the angle of departure Eqn1427:

    1. For the pole Eqn1428

      Eqn1429

    2. For the pole Eqn1430, Eqn1431
  5. To determine the number of branches existing on the real axis:

    If we look from the pole p = −25, the total number of poles and zeros ­existing on the right of −25 is one (odd number). Therefore, a branch of root loci exists between −25 and 0.

    Hence, only one branch of root loci exists on the real axis for the given system.

  6. The breakaway/break-in points for the given system can be determined by

    Eqn1435(1)

    The characteristic equation of the given system is

    Eqn1436

    i.e., new22

    Therefore, Eqn1438

    Differentiating the above equation with respect to Eqn1439 and using Eqn. (1), we obtain

    Eqn1440

    Upon solving, we obtain

    Eqn1441.

    Substituting Eqn1442, Eqn1443 and Eqn1444, the values of Eqn1445 are Eqn1446, Eqn1447 and Eqn1448 respectively. Since Eqn1452 is positive only at Eqn1450, the point Eqn1451 is a valid breakaway point.

  7. The point at which the branch of root locus crosses the imaginary axis is obtained by substituting Eqn1457 in the characteristic equation and then comparing the real and imaginary parts of the equation.

    Eqn1458

    Comparing the imaginary parts on both sides, we obtain

    Eqn1459

    or, Eqn1460.

    Hence, the root locus crosses the imaginary axis at Eqn1461.

    The complete root locus plot for the given system is shown in Fig. E7.18.

    C07E027

    Fig. E7.18

Example 7.19: Sketch the root locus of the system whose loop transfer function is given by Eqn1463 by varying the ­parameter, a = −4, −2, −1 and 0. Also, comment on the root locus of the different systems.

Solution: The values obtained for constructing the root loci for different values of the parameter, Eqn1464, are given in Table E7.19.

Table E7.19 ∣ Root loci for different systems

tbl11

The root loci of the systems with Eqn1497 are shown in Figs. E7.19(a), (b), (c) and (d) respectively.

C07E028a

Fig. E7.19(a)

C07E028b

Fig. E7.19(b)

C07E028c

Fig. E7.19(c)

C07E028d

Fig. E7.19(d)

7.7 Effect of Adding Poles and Zeros in the System

Consider a system whose loop transfer function is given by, Eqn1506. The root loci for the given system are shown in Fig. 7.10. For designing a controller to the system given above, it is necessary to have a clear study over the effects of root loci when extra poles or zeros or both get added to the loop transfer function. The above system is stable for all the values of variable gain K > 0.

C07F011a

Fig. 7.10 ∣ Root locus for new2

7.7.1 Addition of Poles to the Loop Transfer Function, G(s)H(s)

The effects of addition of extra poles in the root loci of the existing system are studied for three cases and they are given in Table 7.8.

Table 7.8 ∣ Effect of addition of poles to the system

tbl12

7.7.2 Effect of Addition of Poles

If a pole is added to the system, the following effects take place in the root loci of the original system.

  1. Root locus gets shifted towards the right-half s-plane.
  2. Stability of the system gets gradually decreased.
  3. Nature of the system response becomes more oscillatory.
  4. Marginal value of Eqn1522 decreases.

7.7.3 Addition of Zero to the Loop Transfer Function

The effects of addition of an extra zero in the root loci of the existing system are studied and given in Table 7.9.

Table 7.9 ∣ Effect of addition of zero to the system

tbl13

7.7.4 Effect of Addition of Zeros

If a zero is added to the left half of system, the following effects take place in the root loci of the original system.

  1. Root locus gets shifted towards the left-half of s-plane.
  2. Stability of the system gradually increases.
  3. Nature of the system response becomes less oscillatory.
  4. Marginal value K increases.

Example 7.20: Sketch the root locus of the system whose loop transfer function is Eqn1096. Also, plot the root locus when a zero at −1 is added to the existing system.

Solution:

  1. For the given system, poles are at 0, 0 and −2, i.e., Eqn1097 and zeros do not exist, i.e., Eqn1098.
  2. Since Eqn1099, the number of branches of the root loci for the given system is Eqn1100
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn1101
    2. Angles of asymptotes

      Eqn1102 for Eqn1103

      Therefore, Eqn1104.

    3. Centroid

      Eqn1105

  4. Since no complex pole or complex zero exists for the system, there is no necessity of determining angle of departure/angle of arrival.
  5. To determine the number of branches existing on the real axis:

    If we look from the point Eqn1107, the total number of poles and zeros ­existing on the right of the point is three (odd number). Therefore, a branch of root loci exists between −Eqn1108 and −2.

    Hence, only one branch of root loci exists on the real axis for the given system.

  6. The breakaway/break-in points for the given system can be determined by

    Eqn1109(1)

    The characteristic equation of the system is

    Eqn1110(2)

    i.e., new16

    Therefore, Eqn1111(3)

    Differentiating the above equation with respect to Eqn1112 and using Eqn. (1), we obtain

    Eqn1113

    Upon solving, we obtain

    Eqn1114

    For Eqn1115 and Eqn1116, the values of K are 0 and −1.18 respectively. Hence, the point Eqn1118 is a breakaway point and Eqn1119 is not a breakaway point.

  7. The point at which the root locus crosses the imaginary axis is determined as follows:

    From Eqn. (2), the characteristic equation of the system is

    Eqn1120(4)

    Routh array for the above equation,

    Eqn1121

    To determine the point at which root locus crosses the imaginary axis, the first element in all the rows must be zero, i.e., −Eqn1122. Therefore,Eqn1123.

    Substituting K in the characteristic equation and solving for s, we obtain

    Eqn1124.

    Hence, the root locus of the system does not cross the imaginary axis.

    The final root locus for the system is shown in Fig. E7.20(a).

    C07E020a

    Fig. E7.20(a)

  8. When a zero at s = −1 is added to the existing system, the transfer function of the system becomes

    Eqn1126

    1. For the given system, poles are at 0, 0 and −2 i.e., Eqn1127 and zero is at −1 i.e., Eqn1128.
    2. Since Eqn1129, the number of branches of the root loci for the given system is Eqn1130.
    3. For the given system, the details of the asymptotes are given below:
      1. Number of asymptotes for the given system Eqn1131.
      2. Angles of asymptotes

        Eqn1132 for Eqn1133

        Therefore, Eqn1134 and Eqn1135

      3. Centroid

        Eqn1136

    4. Since no complex pole or complex zero exists for the system, there is no ­necessity of determining angle of departure/angle of arrival.
    5. To determine the number of branches of root loci on the real axis:

      If we look from the pole p = −2, the total number of poles and zeros existing on the right of the point is three (odd number). Therefore, a branch of root loci exists between −2 and −1.

      Hence, one branch of root loci exists on the real axis for the given system.

    6. The breakaway/break-in points for the given system can be obtained by

      Eqn1140(1)

      The characteristic equation of the given system is

      Eqn1141(2)

      i.e., new17

      Therefore, Eqn1142

      Differentiating the above equation with respect to Eqn1143 and using Eqn. (1), we obtain

      Eqn1144

      Upon solving, we obtain

      Eqn1145

      For Eqn1146, K is 0. Since Eqn1147 is neither positive nor negative, the point Eqn1148 is not a valid breakaway point.

      Since the other two points are complex numbers, it is necessary to check the angle ­condition also.

      Eqn1149

      Similarly, Eqn1150

      Since the angle of the transfer function Eqn1151 at the complex numbers is not an odd multiple of 180°, the complex numbers are not valid breakaway points.

      Hence, no breakaway points exist for the given system.

    7. The point at which the branch of root loci intersects the imaginary axis can be determined as follows:

      From Eqn. (2), the characteristic equation of the system is

      Eqn1152(4)

      Routh array for the above equation is

      Eqn1153

To determine the point at which root locus crosses the imaginary axis, the first element in the third row must be zero, i.e., Eqn1154. Therefore, K = 0.

Substituting K = 0 in the characteristic equation, we obtain

Eqn1155

Upon solving, we obtain Eqn1156. As there is no imaginary part in the solution, the branch of root loci does not cross the imaginary axis.

The final root locus for the system is shown in Fig. E7.20(b).

It is inferred that, when a zero is added, root locus shifts left and stability increases.

C07E020b

Fig. E7.20(b)

7.8 Time Response from Root Locus

The pole-zero configuration of a loop transfer function, Eqn1523, plays a major role in ­determining the time response of the system. The closed-loop transfer function of the system with forward loop transfer function Eqn1524 and negative feedback transfer function Eqn1525 is

Eqn1526

where Eqn1527 is the Laplace transform of the output signal and Eqn1528 is the Laplace transform of the input signal.

For a unity feedback system, Eqn1529 and if there existsEqn1530 closed-loop poles and m closed-loop zeros, then the transfer function Eqn1532 can be expressed as

Eqn1533

Eqn1534(7.8)

where Eqn1535 is the closed-loop gain which can be expressed in terms of Eqn1536 and Eqn1537.

The time response of the closed-loop transfer function, Eqn1538 approaches 1 as Eqn1539 if an unit step input is applied to the system. Applying final value theorem, we get

Eqn1540

Here, Eqn1541 and Eqn1542

Therefore, Eqn1543

Eqn1544

Eqn1545

or Eqn1546(7.9)

Substituting Eqn. (7.9) in Eqn. (7.8), we obtain

Eqn1547

If the closed-loop poles existing in the system are distinct, then

Eqn1548

Eqn1549

Taking inverse Laplace transform, we obtain

Eqn1550(7.10)

where Eqn1551.

Using the above two equations, the response of the system can be obtained if the closed-loop poles-zeros are known.

Example 7.21: The pole-zero plot of a loop transfer function of a system is shown in Fig. E7.21. Determine the response of the system when it is subjected to unit step input.

C07E029

Fig. E7.21

Solution:

From Fig. E7.21, the poles are at Eqn1_1 and zero is at Eqn2_1

Using Eqn. (7.10), we obtain the response of the system for the unit step input as

Eqn3_1

where Eqn4_1

To determine, Eqn5_1:

Eqn6_1

Eqn7_1

Eqn8_1

Eqn9_1

Therefore,

Eqn10_1

Eqn11_1

Eqn12_1

Eqn13_1(since Eqn14and Eqn15_1)

7.9 Gain Margin and Phase Margin of the System

The gain margin and phase margin for a system can be determined using root locus method.

7.9.1 Gain Margin of the System

The maximum allowable loop gain value in decibels (dB) before the closed-loop system becomes unstable is called gain margin. The formula for determining the gain margin in dB is

Eqn16

where Eqn17_1

Eqn18_1 gainEqn19_1 at which the root locus crosses the imaginary axis and

Eqn20_1required gain

It is to be noted that if the root locus of the system does not cross the imaginary axis, then Eqn21_1 is Eqn22_1 and hence the gain margin of the system is also Eqn23_1.

7.9.2 Phase Margin of the System

The maximum allowable phase angle in degrees of Eqn24_1 before the closed-loop ­system becomes unstable is called phase margin. The phase margin of the system can be calculated using the following steps:

Step 1: Substitute Eqn25_1 in the given loop transfer function and determine Eqn26_1

Step 2: Substitute Eqn27 in the given loop transfer function and determine the frequency Eqn28_1, for which Eqn29_1.

Step 3: Determine the angle of Eqn30_1 at the frequency obtained in the previous step.

Step 4: Determine the phase margin of the system as

Phase margin, pm = Eqn31_1.

Example 7.22: Determine gain margin and phase margin of the system whose loop transfer function is given by Eqn32_1 and the required/desired gain value is 3.

Solution: The final root locus of the system, whose loop transfer function is Eqn33_1 is shown in Fig. E7.12. Also, from Example 7.12, it is known that the root locus crosses the imaginary axis at Eqn34_1. Substituting Eqn35_1 in the characteristic equation of the system, we obtain the marginal gain Eqn36_1 as 6. Hence, the gain margin of the system in dB is

Eqn37_1

where Eqn38_1

In this case, Eqn39_1 and Eqn40_1. Therefore,

Eqn41_1

Eqn42_1 dB

To determine phase margin of the system:

Step 1: Substituting Eqn43_1 in the loop transfer function, we obtain

Eqn44_1

Step 2: Substituting Eqn45_1 in the above equation and using the condition Eqn46_1, we get the value of frequency Eqn47_1.

Eqn48_1

Squaring on both sides of the equation, we obtain

Eqn49_1

Therefore, Eqn50_1

Substituting Eqn51_1 in the above equation, we obtain

Eqn52_1

Upon solving, we obtain

Eqn53_1

Neglecting the complex numbers and negative numbers, we obtain the frequency as

Eqn54_1 rad/sec.

Step 3: Eqn55_1

Eqn56_1

Step 4: Phase margin of the system, Eqn57_1

7.10 Root Locus for K < 0 Inverse Root Locus or Complementary Root Loci

From Table 7.1, it is clear that the complementary root loci (CRL) or inverse root loci for a system is the simple root locus with gain value K < 0. In the previous sections, we have discussed the properties of root loci and steps for constructing the root loci manually for a negative feedback system with the gain value K > 0. Hence, the characteristic equation taken for constructing the root loci is Eqn59_1

But the characteristic equation for constructing the inverse or complementary root loci is Eqn60_1 (i.e., negative feedback system with the gain value K > 0). Also, if the root loci constructed for the positive feedback system whose characteristic equation is Eqn61_1, which is called complementary or inverse root loci.

The rules for constructing the CRL or inverse root loci are same as the rules for ­constructing the ordinary root locus with small modifications.

The following are the properties that differ from the ordinary properties of the root loci.

Property 1: Number of branches of root loci on the real axis

If the total number of poles and zeros lying to the right side of any point on the real axis is an even number, then a branch exists on the real axis.

Property 2: Angle of each asymptote present in the system is determined by

Eqn62_1

where Eqn63_1.

Property 3: Angle of departure/angle of arrival for the complex pole and complex zero is determined using the following formula:

Eqn65_1

7.10.1 Steps in Constructing the Inverse Root Loci Manually

  1. Plot the poles and zeros of the system Eqn66_1 or determine and plot the points at Eqn67_1. The points at Eqn68_1 are marked using Eqn69_1 respectively (i.e., poles and zeros of the system). Let n be the number of finite poles of Eqn70_1 and m be the number of finite zeros of Eqn71_1
  2. The total number of branches of the root loci is determined by

    Number of branches = Number of poles or number of zeros of the ­system, whichever is greater

  3. The total number of asymptotes in the root loci of the system is determined by

    Number of asymptotes Eqn72_1 = Eqn73_1

  4. The intersection of the asymptotes with the real axis (centroid) is determined by using

    Eqn74_1

  5. Angles of asymptotes for different root loci are to be determined.

    Eqn75_1

  6. Determine of branches of root loci existing on the real axis
  7. Determine of angle of departure or angle of arrival if complex pole or complex zero exists for the system

    Eqn76_1

  8. Determine the breakaway points in the root locus of the system and check for validity of breakaway points. Also, determine gain K at the breakaway point by using

    Eqn77_1

  9. Determine the intersection point of the root loci with imaginary axis and its ­corresponding K is obtained using Routh−Hurwitz criterion.

Example 7.23: Sketch the root locus of the positive unity feedback system Eqn78_1.

Solution:

  1. For the given system, poles are at 0 and −4 i.e., Eqn79_1 and zero does not exist i.e., Eqn80_1.
  2. Since Eqn81_1, the number of branches of root loci for the given system is Eqn82_1.
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes Eqn83_1
    2. Angles of asymptotes

      Eqn84_1

      Therefore, Eqn85_1 and Eqn86_1.

    3. Centroid

      Eqn87_1

      Eqn88_1

  4. Since complex pole/complex zero does not exist for the given system, there is no angle of departure/angle of arrival.
  5. To determine the number of branches existing on the real axis:

    If we look from the pole p = 0, the total number of poles and zeros existing on the right of 0 is zero (nor an even number or a odd number). Therefore, a branch of root loci exists between 0 and Eqn89_1.

    Similarly, if we look from the point at −Eqn90_1, the total number of poles and zeros existing on the right of −Eqn91_1 is 2 (even number). Therefore, a branch of root loci exists between −Eqn92_1 and −4.

    Hence, two branch of root loci exists on the real axis for the given system.

    C07E031

    Fig. E7.23

  6. The breakaway or break-in points for the given system are determined by

    Eqn77_1(1)

    For the given system, the characteristic equation is

    Eqn94_1(2)

    i.e., new23

    Therefore, Eqn96_1(3)

    Differentiating the above equation with respect to Eqn97_1 and using Eqn. (1), we obtain

    Eqn98_1

    Therefore, Eqn99_1.

    We know that the breakaway/break-in points for a system should lie in the branches of root loci on the real axis. But the point Eqn100_1 does not lie in the branch of root loci on the real axis.

    Since Eqn101_1 is not a breakaway point for the given system, the breakaway or break-in point does not exist.

  7. As the branches of root locus lie only on the real axis, there is no point on the imaginary axis at which the root locus crosses.

    The entire root locus for the system Eqn102_1 is shown in Fig. E7.23.

Example 7.24: Develop the root locus for a positive unity feedback system whose loop transfer function is given by Eqn104_1.

Solution:

  1. For the given system, poles are at 0, −2 and −5, i.e., Eqn105_1 and zero does not exist i.e., Eqn106_1.
  2. Since Eqn107_1, the number of branches of root loci for the given system is Eqn108_1.
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes Eqn109_1
    2. Angles of asymptotes

      Eqn110_1 for Eqn111_1

      Therefore, Eqn112_1, Eqn113_1 and Eqn114_1

    3. Centroid

      Eqn115_1

      Eqn116_1.

  4. Since complex pole/complex zero does not exist for the system, there is no angle of departure/angle of arrival.
  5. To determine the number of branches existing on the real axis:

    If we look from the pole Eqn117_1, the total number of poles and zeros existing on the right of 0 is zero (nor an even number or an odd number). Therefore, a branch of root loci exists between 0 and Eqn118_1.

    Similarly, if we look from the pole Eqn120_1, the total number of poles and zeros existing on the right of −5 is 2 (even number). Therefore, a branch of root loci exists between −5 and −2.

    Hence, two branches of root loci exist on the real axis for the given system.

  6. The breakaway or break-in points for the given system are determined by

    Eqn124(1)

    For the given system, the characteristic equation is

    Eqn125_1(2)

    i.e., new24

    Therefore, Eqn127_1(3)

    Differentiating the above equation with respect to Eqn128_1 and using Eqn. (1), we obtain

    Eqn129_1

    Therefore, Eqn130_1 and Eqn131_1.

    We know that, the breakaway/break-in points for a system should lie in the branches of root loci on the real axis. But the point Eqn132_1 does not lie in the branch of root loci on the real axis.

    Hence, Eqn133_1 is not a breakaway point and for the given system, the valid breakaway point is Eqn134_1.

  7. As the branches of root loci follows the asymptotes for the given system, it is clear that the no branch of root loci crosses the imaginary axis.

    The complete plot of the system is shown in Fig. E7.24.

    C07E032

    Fig. E7.24

7.11 Pole-Zero Cancellation Rules

The root locus of a system is constructed/developed based on the loop transfer function of the system Eqn137_1. If the open-loop transfer function of the system Eqn138_1 and ­feedback transfer function of the system Eqn139_1 are given separately, there is always a possibility of cancellation of pole-zero between Eqn140_1 and Eqn141_1. But simply cancelling the common poles and zeros between Eqn142_1 and Eqn143_1 is not a advisable one. Therefore, two rules in cancelling the common poles and zeros existing between Eqn144_1 and Eqn145_1 are given below with proof.

Rule 1: The poles present in Eqn146_1 should not be cancelled by zeros present in Eqn147_1.

Rule 2: The zeros present in Eqn148_1 can be cancelled by poles present in Eqn149_1.

Proof for rule 1:

Consider a system with Eqn150_1 and Eqn151_1. If Eqn152_1 and pole-zero cancellation is allowed, then Eqn153_1.

Now, the characteristic equation of the system with the loop transfer function G(s)H(s) is

Eqn155_1

Eqn156_1(7.11)

The closed-loop transfer function of the system is

Eqn157_1

Eqn158_1.

The characteristic equation for the above function is

Eqn159_1(7.12)

Comparing Eqn. (7.11) and Eqn. (7.12), it is clear that in Eqn. (7.11), a pole at Eqn160_1 is ­missing which is due to the pole-zero cancellation. Therefore, pole-zero cancellation of the system is not allowed.

Hence, the loop transfer function of the system is Eqn161_1 instead of Eqn162_1.

Proof for rule 2:

Consider a system with Eqn163_1 and Eqn164_1. If Eqn165_1 and zero-pole cancellation is allowed, then Eqn166_1.

Now, the characteristic equation of the system with the loop transfer function Eqn167_1 is

Eqn168_1

Eqn169_1(7.13)

The closed-loop transfer function of the system is

Eqn170_1

Eqn171_1.

The characteristic equation for the above function is

Eqn172_1(7.14)

Since Eqn. (7.13) and Eqn. (7.14) are same, zero-pole cancellation of the Eqn173_1 and Eqn174_1 is allowed.

Example 7.25: Consider a system with open-loop transfer function Eqn175_1 and a feedback transfer function Eqn176_1. Plot the root locus for the system.

Solution: Although there exists a common point where pole and zero exist, due to rule 1 discussed in Section 7.11, the loop transfer function of the system is considered as Eqn177_1

  1. For the given system, poles are at −1 and −2 i.e., Eqn178_1 and zero is at −3 i.e., Eqn179_1.
  2. Since Eqn180_1, the number of branches of the root loci for the given system is Eqn181_1.
  3. For the given system, the details of the asymptotes are given below:
    1. Number of asymptotes for the given system Eqn182_1.
    2. Angles of asymptotes

      Eqn183_1 for Eqn184_1

      Therefore, Eqn185_1.

    3. Centroid

      Eqn186_1

      Eqn187_1.

  4. Since a complex pole or complex zero does not exist for the system, there is no angle of departure/angle of arrival.
  5. To determine the number of branches existing on the real axis:

    If we look from the pole p = −2, the total number of poles and zeros existing on the right of −2 is one (odd number). Therefore, a branch of root loci exists between −2 and −1.

    Similarly, if we look from the point at Eqn189_1, the total number of poles and zeros existing on the right of that point is three (odd number). Therefore, a branch of root loci exists between −3 and Eqn190_1.

    Hence, two branches of root loci exist on the real axis for the given system.

  6. The breakaway and break-in points for the given system can be determined by

    Eqn191_1(1)

    For the given system, the characteristic equation is

    Eqn192_1(2)

    i.e., new25

    Therefore, Eqn194_1(3)

    Differentiating the above equation with respect to Eqn195_1 and using Eqn. (1), we obtain

    Eqn196_1

    Eqn197_1

    Therefore, Eqn198_1

    For Eqn199_1 and Eqn200_1, the values of Eqn201_1 are Eqn202_1 and Eqn203_1 respectively. Since the values of K are positive, Eqn205_1 and Eqn206_1 are the breakaway and break-in points respectively.

  7. The point at which the branch of root loci intersects the imaginary axis can be determined as follows:

    From Eqn. (2), the characteristic equation of the system is

    Eqn207_1(4)

    Routh array for the above equation is

    Eqn208_1

    To determine the point at which root locus crosses the imaginary axis, the first ­element in all the rows must be zero, i.e., Eqn209_1 or Eqn210_1.

    Since K is negative, the branch of root loci does not cross the imaginary axis at any point.

    The complete root locus for the given system is shown in Fig. E7.25.

    C07E033

    Fig. E7.25

7.12 Root Contours (Multi-Variable System)

The root locus problems discussed so far has a single variable parameter Eqn212_1, which can be varied for designing the system accurately. But in a general control system problem, there exist number of varying parameters whose effects are to be investigated.

In this section, let us discuss how a root locus problem is carried out for a system with two variable parameters present in it. Root locus is called root contour when one or more parameter is varied. Root contour will posses all the properties of root locus and hence the method of construction of root loci is also applicable in constructing root contour.

The general root-contour problem is formulated by referring to the equation:

Eqn213_1(7.15)

where

Eqn214_1 is the nth order polynomial of s given by Eqn215_1

Eqn216_1 is the m1th order polynomial of s given by Eqn217_1

Eqn218_1 is the m2th order polynomial of s given by Eqn219_1

where n, m1 and m2 are positive integers and K1 and K2 are variable loop gains which vary from Eqn220_1

The following steps are used to draw the root contour for the multi-variable system:

Step 1: Set variable parameter K2 to zero which makes Eqn. (7.15) as

Eqn221_1

Step 2: The equation now has a single variable parameter K1 and the root locus of the single parameter system is determined using the equation Eqn222_1 as the equation is similar to Eqn223_1.

Step 3: Now, dividing the original Eqn. (7.15) by Eqn224_1, we obtain

Eqn225_1

with which the root loci for the system with the single parameter Eqn226_1 can be constructed as the equation is similar to Eqn227_1.

Review Questions

  1. Define root loci.
  2. What are the advantages of root locus technique?
  3. How the root loci of a system are categorized?
  4. Discuss the necessary conditions for formulating the root loci for a system.
  5. What are the properties to be known in constructing the root locus manually?
  6. Define the following terms:

    (i) asymptotes, (ii) centroid and (iii) breakaway point.

  7. Explain the steps involved in calculating the breakaway points.
  8. Explain the procedure in constructing the root locus for a particular system.
  9. Explain the procedure for obtaining K for a specified damping ratio Eqn228_1.
  10. The loop transfer function of a unity feedback system is Eqn229_1. Determine the breakaway and break-in points for the system.
  11. Sketch the root locus of the system Eqn230_1.
  12. Construct the root locus of the system Eqn231_1.
  13. Plot the root locus of the system whose characteristic equation is given by Eqn232_1.
  14. The characteristic equation of a feedback control system is given by Eqn233_1. Construct the root locus for Eqn234_1 and discuss about the ­stability of the system.
  15. Sketch the root locus of the system Eqn235_1.
  16. The block diagram of a control system is shown in Fig. Q7.16. Construct the root locus of the system with Eqn236_1, Eqn237 and Eqn238.
    C07Q0016

    Fig. Q7.16

  17. The block diagram of a control system is shown in Fig. Q7.17. The transfer function of a system Eqn239_1. (i) When the switch, Eqn240_1 is open, construct the root locus for the system and discuss about the stability of the system when Eqn241_1 varies and (ii) when the switch Eqn242_1 is closed, plot the root locus for the system with Eqn243_1 and discuss about the stability of the system when Eqn244_1 varies.
    C07Q0017

    Fig. Q7.17

  18. The characteristic equation of a feedback control system is given by Eqn245_1 where Eqn246_1 are the variable parameters that vary from Eqn247_1. Construct the root contours when (i) Eqn248_1 and (ii) Eqn249_1 is constant.
  19. The loop transfer function of a unity feedback system is given by Eqn250_1.
    1. Plot the root locus diagram as a function of Eqn251_1.
    2. Determine gain K when the damping ratio of the system Eqn252_1.
  20. Construct the root locus of the system Eqn253_1. Also, determine gain K when the damping ratio of the system Eqn254_1.
  21. Sketch the root locus of the system Eqn255_1. Also, determine gain K when the damping ratio of the system, Eqn256_1.
  22. The loop transfer function of a unity feedback system is Eqn257_1. Construct the root locus for the system and determine gain Eqn258_1 (i) for which repetitive roots occur, (ii) for which the closed-loop system becomes under-damped and (iii) for which the system will have Eqn259_1.
  23. Explain the steps how the gain margin and phase margin of a system can be determined.
  24. Plot the root locus of the system whose loop transfer function is Eqn260_1. Determine Eqn261_1 for (i) critical damping, Eqn262_1 and for (ii) zero damping, Eqn263_1. Also, ­determine the gain margin of the system for Eqn264_1.
  25. Determine the gain margin and phase margin with Eqn265_1 for the system whose loop transfer function is Eqn266_1 .
  26. Plot the root locus of the system whose loop transfer function is Eqn267_1. Determine the gain margin and phase margin of the system for (i) Eqn268_1 and (ii) Eqn269_1.
  27. The loop transfer function of the system is given by Eqn270_1. Determine the stability of the system.
  28. The characteristic equation of a system is Eqn271_1. Determine the loop transfer function of the system and construct the root locus for the same.
  29. A system with a unity negative feedback system has the loop transfer function Eqn272_1. Construct the root locus for the system and (i) determine gain K for Mp = 20 per cent (ii) determine settling time, Eqn273_1 and peak time, tp for gain K obtained in (i).
  30. The loop transfer function of a system is Eqn274_1. Construct the root locus for the system and determine gain Eqn275_1 when the phase margin of the system is Eqn276_1.
  31. The loop transfer function of a system is Eqn277_1. Plot the root locus for the system and determine gain Eqn278_1 for an oscillatory response of the system.
  32. Show that the root locus of the system Eqn279_1 is a circle with radius Eqn280_1 and centre (−3, 0).
  33. The loop transfer function of the system is Eqn281_1. Construct the root locus for the system and determine gain Eqn282_1 for a stable closed-loop system.
  34. The loop transfer function of a system is given by Eqn283_1. Sketch the root locus for the system with Eqn284_1 and Eqn285_1.
  35. The unity feedback system has a loop transfer function Eqn286_1. (i) construct the root locus for the system, (ii) show how the addition of zero at Eqn287_1 stabilizes the system and (iii) plot the root locus with Eqn288_1 and determine gain Eqn289_1, ­damping ratio Eqn290_1, un-damped natural frequency Eqn291_1 which gives greatest damping ratio in ­oscillatory mode.
  36. The loop transfer function of a system is Eqn292_1. Construct the root locus of the system.
  37. A unity feedback system has a loop transfer Eqn293_1. Plot the root locus of the system.
  38. Discuss how the root locus of a system will be affected by adding poles to the existing system.
  39. Discuss how the root locus of a system will be affected by adding zeros to the system.
  40. What are the effects of adding poles and zeros to the existing system?
  41. How the time response of a system can be determined using root locus of a system?
  42. The block diagram of a system is shown in Fig. Q7.42. Construct the root locus for the system with Eqn294_1.
    C07Q0042

    Fig. Q7.42

  43. What are the rules to be considered in determining the loop transfer function of a system when Eqn295_1 and Eqn296_1 are given?
  44. The transfer function of a system Eqn297_1 and feedback transfer function, Eqn298_1. Determine the loop transfer function of the system and construct the root locus for the same.
  45. Explain the concept of root contour.
  46. For the following systems, determine the number of asymptotes, angle of asymptotes and centroid.
    1. Eqn278
    2. Eqn279
    3. Eqn280
    4. Eqn281
  47. Determine the angle of departure or angle of arrival or both the angles for the following systems.
    1. Eqn331
    2. Eqn332
    3. Eqn333
    4. Eqn334
    5. Eqn335
  48. Determine the branches of root loci existing on the real axis and hence determine the breakaway or break-in points for the systems given below:
    1. Eqn386
    2. Eqn387
    3. Eqn388
    4. Eqn389
    5. Eqn390
  49. Plot the root locus of the system whose loop transfer function is given by Eqn773.
  50. Sketch the root locus of the system whose loop transfer function is given by Eqn968.
  51. Sketch the root locus of the system whose loop transfer function is given by Eqn1257
  52. 52. Construct the root locus of the system whose loop transfer function is given by Eqn1339
..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
18.191.144.194