A Distributed Framework for Smart Grid Modeling, Monitoring, and Control |
|
CONTENTS
5.3 Web Services—Based Framework for SG Management
5.4.1 Data Acquisition Web Service
5.4.2 Job Submission Web Service
5.4.3 Computational Web Service
5.4.4 Data Storage Web Service
Electrical power systems were traditionally characterized by the presence of numerous utilities, heterogeneous standards, overlapping territories, and a general lack of integration. Recently, under pressure from deregulated electricity markets and new environmental policies, the main limitations intrinsic to this environment have become clear and the structures of modern power systems have evolved to accommodate large changes induced by these new energy policy trends.
In this emerging scenario, the large-scale deployment of the smart grid (SG) paradigm could play a strategic role in supporting the evolution of conventional electrical grids toward active, flexible, and self-healing web energy networks composed of distributed and cooperative energy resources.
From a conceptual point of view, the SG is the convergence of information and operational technologies applied to the electric grid, providing sustainable options to customers and improved security [1].
SG technologies include advanced sensing systems, two-way high-speed communications, monitoring and enterprise analysis software, and related services to get location-specific and real-time actionable data in order to provide enhanced services for both system operators (i.e., distribution automation, asset management, advanced metering infrastructure) and end users (i.e., demand-side management, demand response) [2,3].
Advances in research on SG could increase the efficiency of modern electrical power systems by (i) supporting large-scale penetration of small-scale distributed energy resources (DERs); (ii) facilitating the integration of renewable energy sources; (iii) reducing system losses and greenhouse gas emissions; and (iv) increasing the reliability of electricity supply to customers.
From this perspective, a crucial issue is how to increase the penetration of DERs into the SG by properly coordinating their operation and by restricting their negative impact on grid operations and control.
It is well known that the integration of DERs into running distribution systems affects the power flows and voltage conditions of customer and utility equipment by inducing a number of side effects (e.g., bidirectional power flows, increased fault current levels, and a need for new concepts of network protection [4]). This increases the complexity of the control, protection, and maintenance of distribution systems, which are typically designed to operate radially without any generation at the distribution line or customer side [5]. This requires extensive analysis aimed at (i) identification of the optimal generation schedule that minimizes production costs and balances the demand and supply that come from both micro sources and the distribution feeder; (ii) online assessment of the SG security and reliability levels; and (iii) identification of proper preventive and corrective measures that mitigate the effects of critical contingencies. The main limitations that should be addressed in order to overcome these issues originate mostly from the following intrinsic features characterizing the SG operation:
• The rising level of DERs, which makes the distribution systems more vulnerable with respect to dynamic perturbations
• The increasing number of smaller geographically dispersed generators, which could significantly increase the number of power transactions
• The difficulties arising in predicting and modeling market operators’ behavior, governed mainly by unpredictable economic dynamics, which introduce considerable uncertainty into short-term SG operation
• The need for more detailed security studies assessing the impact of multiple contingencies on the distribution system (i.e., N-2 security criteria)
• The high penetration of generation units powered by renewable energy sources, which induces considerable uncertainty in SG operation
• The low level of upgrade and interoperability of existing distribution management systems that are typically based on low-scalable and proprietary software platforms characterized by different information technologies and standards for the data exchange
• The need for standard interfaces between the distributed tools and components characterizing the power system’s operation environment; these interfaces must be standardized so as to work in a “plug-and-play” concept similar to the hardware used in substation automation
In this context the adoption of web services, a new paradigm for enterprise application integration built on the foundation of open standard and common infrastructure could contribute to overcoming some of these limiting aspects.
A web service is an interface that describes a collection of modular, self-contained applications that are network-accessible through standardized extensible markup language (XML) messaging. The interface hides the implementation details of the service, allowing it to be used independently of the hardware or software platform on which it is implemented and also independently of the programming language in which it is written [6,7]. The use of standard XML protocols makes web services technology platform-, language-, and vendor-independent, and they are thus an ideal candidate for supporting the deployment of the SG paradigm [8].
In agreement with these arguments, this chapter outlines the design of a prototype web services—based framework for integrated SG modeling, monitoring, and control.
The architecture of the proposed framework is based on the following web services:
1. GISinterfaceWS acquires georeferential information on the current network topology and the control devices asset by interacting with a geographical information system (GIS).
2. DataAcquisitionWS acquires the network field data interacting with the entire set of supervisory control and data acquisition (SCADA) systems and field electrical measurements (FEM) by standardized XML messaging.
3. eAssessmentWS develops the distributed generation/storage unit capability assessment by interacting with a network of intelligent units.
4. ComputationalWS employs a parallel solution engine, based on a computational grid, to develop the calculations needed to support the entire set of SG services in terms of state estimation, network analysis, online security assessment, and so on.
5. DataStorageWS is used to permanently store historical data and alarm conditions that can be analyzed off-line using a web browser. The service can be remotely accessed by using data access middleware.
In order to prove the effectiveness of the proposed framework, the results of field activities developed on an experimental test bed are presented and discussed.
The experimental results obtained show that the web service technologies in the SG are an effective tool to integrate monolithic and hard-to-customize power system control and monitoring software tools, allowing system operators to easily manage proprietary hardware instrumentation and multiple software modules.
The control and operational issues of SGs are presented and discussed in [9,10]. The analysis in this chapter shows that the required control and operational strategies of an SG can be observably and conceptually different from those of conventional power systems. These differences will depend on the type and depth of penetration of DER units, load characteristics, power quality constraints, and market participation strategies.
Consequently, novel control strategies and ad hoc energy management systems should be deployed for this new application. These facilities would support such applications as DER and demand response dispatch, distribution automation and substation automation, adaptive relaying, energy management, market pricing, grid modeling, operator displays, and advanced visualization systems [9].
SGs could play an important role in deregulated energy markets, as outlined in [11]. In this context, SGs could be operated in order to serve the total load demand, using local production as much as possible, without exporting power to the upstream transmission grid. This strategy represents a benefit for the transmission system operator, since the SG relieves possible network congestion during peak demand by partly or fully supplying the system’s energy needs.
As outlined in [12], operational control and energy management systems for the SG should be implemented through the cooperation of various controllers, located at all these levels, on the basis of the communication and collection of information about distributed energy systems and control commands. This could be deployed according to a centralized or decentralized control paradigm.
In a centralized control paradigm, the SG controllers optimize the power exchanged between the distributed grid system (DGS) and the main grid, thus maximizing local production depending on market prices and security constraints. This is achieved by issuing control set points to DERs and controllable loads in order to optimize local energy production and power exchanges with the main distribution grid [12].
The decentralized control paradigm is intended to provide maximum autonomy for the DER units and loads. This implies that local controllers are intelligent and can communicate with each other to form a larger intelligent entity. In a decentralized control framework, the main task of each controller is not to maximize the revenue of the corresponding unit but to improve the overall performance of the SG. Thus, the architecture must be able to include economic functions, environmental factors, and technical requirements, for example, black start. These features indicate that a multiagent system (MAS) can be a prime candidate for decentralized SG control [13,14,15]. Armed with such a vision, [14,15] present a general framework for the control of DERs organized in SGs. The proposed architecture is based on the agent technology and aims at integrating several functionalities, as well as being adaptable to the complexity and size of the SG.
Another important issue to address for effective SG operational control is dealing with the problem of security assessment in both steady-state and dynamic scenarios [10]. The main steady-state security concerns for the operation of SGs are satisfying voltage constraints and maintaining power flows within thermal limits. Dynamic security concerns the SG operation under a number of contingencies both within and above it (i.e., during a transition between the grid-connected and islanded modes of operation). Effective SG security assessment requires detailed analysis of phenomena that can compromise SG operation. In this connection, it is expected that online time-domain-based analysis such as voltage stability and transient angular stability can be carried out in real time [16].
All of these functions require the design of reliable, resilient, secure, and manageable standards-based open communication systems [12]. These infrastructures represent a key issue in SG deployment, as they would provide the fundamental backbone for connecting the SG elements, the data providers, and the decision-making entities in an open and interoperable framework.
They should support ubiquitous connectivity between decision-making points and dispersed and heterogeneous data sources characterized by varying degrees of transport, security, and reliability requirements [12,17].
They must ensure the capability of SG control systems to transmit data to and from distributed intelligent controllers; moreover, the prospect of the development of innovative energy market policies imposes the need for the realization of efficient and reliable widespread bidirectional communication links between SG operators and customer premises.
In the light of these needs, IEC 61850 provides a standardized framework for substation automation integration that specifies the following: the communication requirements, the functional characteristics, the data structure within devices, how applications interact with and control other devices, and how conformity to the standard should be tested [18,19]. In the IEC 61850 standard, part 7-420, the information exchange between DER units and monitoring and control devices [20,21] is considered. The standard also defines some measures for the success of an integrated mobile gateway (MG) communication network design [22]:
• Flexibility to adjust and grow the system topology as requirements change
• Performance, especially quality of service (QoS), to enable effective prioritization among competing applications and to meet critical requirements of the most important protection and control functions
• Reliability, for critical protection systems, but also because so many different systems are relying on the same infrastructure
The deployment of a low-cost and high-performance communication infrastructure allowing the cooperation of distributed controllers on the basis of communication and collection of information about distributed energy systems and control commands is a challenge that must be faced by the emerging technologies [12].
The literature analysis discussed above reveals that the present power system operation environment is primarily composed of many distributed tools and components, each addressing only a specific functionality (i.e., energy management, voltage control, network modeling, security assessment, data acquisition, market interface). They do not deal with the definition of an integrated platform capable of easily executing very complex applications, built by composing required functionalities in a standardized, easy-to-use, and well-defined way. To address this problem, service-oriented architectures (SOAs), in particular those based on the standard web service technologies, seem particularly suitable for SG control and monitoring due to their ease of application programming, portability, maintenance, and integration with legacy services [6].
5.3 WEB SERVICES—BASED FRAMEWORK FOR SG MANAGEMENT
The proposed framework for SG control, modelling, and monitoring is based on well-defined interfaces and contracts between services, according to the SOA principles. The interface is defined in a neutral manner that should be independent of the hardware platform, the operating system, and the programming language in which the service is implemented. This allows services built on a variety of such systems to interact with each other in a uniform and universal manner. The main benefits of an SOA system are the improvement of interoperability, the integration of new and legacy applications, the ability to survive evolutionary changes in structure, and the implementation of the internals of each service.
Even though many technologies can be adopted to implement an SOA, for example, the common object request broker architecture (CORBA) and message-oriented middleware systems, web services technologies are emerging as particularly suited for ensuring a high degree of integration among existing or newly created services available on the web. Web services, in fact, extend the advantages of software components, making it possible to employ an existing low-level middleware infrastructure based on web servers and hypertext transfer protocol (http). From a programming point of view, a web service is a set of operations that can be easily accessed independently of the service deployment details.
The World Wide Web Consortium (W3C) is standardizing many protocols and languages for supporting web services technologies. At low level, simple object access protocol (SOAP) represents the exchange format based on XML for data and specific data types in a specific web service call. This protocol is used over the http to issue a call to a remote service whose interface is described through web service description language (WSDL), an XML-based language that is more dynamic and flexible than interface description language (IDL), which characterizes remote method invocation (RMI)-based middleware such as CORBA.
Figure 5.1 shows the overall distributed meta-architecture for SG monitoring and control. The core component of the meta-architecture is the SG engine, which is responsible for the execution of SG control, modeling, and monitoring functions in a geographically distributed scenario. It includes high-level components, mainly the submission service, the operation service, and the notification service.
The submission service is responsible for the handling of user submission requests and is designed to simplify the submission phase performed by a nonexpert network operator. The SG operation service is responsible for the execution of SG control and monitoring. Finally, the notification service is responsible for the asynchronous management of output data of an application able to notify specific events. A basic and fundamental component of the SG engine is the workflow enactor, which manages the execution of the SG service and adds some functionalities related to the specific monitoring application. Here the workflow enactor is traditionally described using its business logic description, and it is written in a certain workflow language.
FIGURE 5.1 Overall web services-based architecture.
The WSDL interfaces of the three kinds of web services defined above are the following:
1. GISinterfaceWS: for georeferential information acquisition
2. DataAcquisitionWS: for real-time data acquisition
3. eAssessmentWS: for distributed generation/storage unit capability assessment
4. ComputationalWS: to process the mathematical computations required by the SG control and monitoring functions
5. DataStorageWS: related to the data storage functionalities
In Figure 5.1, GISInterfaceWS and DataAcquisitionWS are abstracted by the generic DataExchangeWS interface, whereas eAssessmentWS can be bound to the eAssessment module shown in the figure.
GISinterfaceWS was defined for the acquisition of georeferentiated information on the SG. It delivers the following services:
• GetNetworkdata: acquires SG structural data concerning the geographical map, the network topology, and the sources of power supply
• GetControlDevicesAsset: type, location, and characteristic parameters of the available control and monitoring devices
• GetElectricalNetworkDetails: line parameters, power transformer rating and numbers, impedance values, bus bar scheme, and circuit breaker type and installation
• GetOperationalParameters: substation equipment status, feeder breakdowns, failure of distribution transformers, tripping on feeders/lines, consumer outages
DataAcquisitionWS was defined for the real-time acquisition of the entire set of measurable SG field data.
This comprises, in particular, the SG topology, the available measurements of the electrical nodes parameters, and the actual state of the generator/storage units computed locally by intelligent electronic devices (IEDs). All possible kinds of measurable parameters are well defined; for each kind of parameter, a code and a basic measurement unit are defined according to IEC 61850-7-420 [20,21]. This standard defines the communication and control interfaces for all SG devices, and develops DER object models. The main benefits of using this standard are:
• Consistent data models
• Easy maintenance of data models
• Sharing interoperability of communication based on IEC 61850
• Seamless integration into the station automation and the power control system
Each measurement value returned by the DataAcquisitionWS is correlated to a time stamp, which indicates the time at which the value was acquired.
The returned type by the described portType is a string that corresponds to the uniform resource locator (URL) of the file containing the required information, which can be downloaded at a convenient time.
A reliable assessment of the actual and future generator/storage unit capabilities appears to be crucial for reliable SG operation.
This demands the design of mathematical models able to predict—given the actual unit state and the forecasted environmental conditions—the evolution of the supply capabilities [23]. These predictive models should also exhibit adaptive features, to deal with the time-varying phenomena affecting the generation/storage unit performances (i.e., aging, parameter drifts) and low computational requirements in order to ensure effective hardware implementation.
To address this problem, the proposed framework integrates a network of distributed IEDs to dynamically assess the supply capability of the main SG components. These IEDs are fully managed by a dedicated web service.
This web service delivers the getCapabilitydata portType, which acquires the georeferentiated input data for the capability calculations by interfacing with dedicated web-based public services (forecasted environmental variables), and the submitCapabilitydata portType, which submits the right input data to the distributed IEDs for the capability calculations.
To support the SG’s monitoring and control functions, intensive online network computations are required. These comprise system state estimation, optimal power flow studies (i.e., voltage control), security assessment (i.e., contingency analysis), and economic analysis (i.e., the generators must be dispatched based on an economic assessment of fuel cost, electric power cost, weather conditions, and anticipated process operation).
The online solution of these tasks is computer-intensive activity, especially in the presence of a large number of DGSs, the extension of electrical networks, and the depth of analysis. So, to address these computationally demanding analyses in useful execution times, parallel algorithms to be executed in a distributed environment have to be adopted [8,24]. In this context, the computational engine has to ensure high reliability, flexibility, and scalability. A practical solution is to exploit the interesting features offered by cloud computing providers. A cloud system is able to provide several services at different abstraction levels: infrastructure as a service (IaaS), platform as a service (PaaS), and software as a service (SaaS). By exploiting these services, customers can satisfy every kind of information and communications technology (ICT) need by simply exploiting a high-bandwidth network connection. In particular, the availability of a large number of resources in cloud systems and sophisticated software infrastructures for handling the available hardware makes it possible to ensure high levels of scalability, reliability, and performance. Therefore, unlike the approaches used in the past for solving similar problems, such as middleware for clusters of computers or geographically dispersed computers networked through the Internet, by using cloud systems, computational resources can be handled by exploiting the features of the PaaS level of the cloud stack, where specific middleware can benefit from the security, scalability, reliability, and performance of the overall infrastructure. This kind of middleware is abstracted in our approach by the specific implementation of the JobSubmissionWS interface, can be adopted by the SG engine to execute the computations supporting the monitoring services, and so can be easily integrated into the overall architecture.
The JobSubmissionWS interface exports a set of brokering functionalities that are related to the interaction with a computational engine and are necessary for the submission of an application and the management of execution results in both asynchronous and synchronous ways. In particular, following the grid broker pattern proposed in [25], the following portTypes are defined:
1. setJob. This is used to define all the necessary information for the application execution (which includes information on the execution code and on the input data file), information on the application structure and its execution requirements in relation to the resource configuration (such as the operating system, a particular run-time environment, library, or transport communication layer, etc.), to QoS parameters, such as the maximum execution time, and, in the case of an economics-driven grid system, the maximum budget that can be spent. A call to the setJob portType returns an application ID, which can be used successively to perform actions on it as required. For generalization purposes, we considered the application setup made through a compressed file, structured in a well-defined way, and understandable by the specific underlying grid-computing middleware. This file will typically contain source and executable files or both, input data files application, and QoS requirement description files.
2. submitJob. This is used to verify that the computational engine will be effectively able to execute the application, respecting all the necessary hardware, software, and QoS requirements.
3. discardJob. When the submitter does not judge the execution offer of the computational engine to be sufficient, he or she can remove it from the list of submitted applications.
4. executeJob. This is used to request the execution start of the application.
5. statusMonitor. This is used to continuously verify online the status of the application execution, such as information on the state of the overall execution (ready, running, waiting, ended, suspended, etc.).
6. collectResults. This is used for synchronous waiting for the normal execution termination, in order to receive contextual information on the output data.
7. terminateJob. This is used to abruptly terminate the application execution.
Storage systems are required to permanently store alarm conditions that can be analyzed off-line, for example, by using a web browser or a dedicated application. Since they can be adopted as a reference in network planning or as a knowledge base for expert systems based on decision support systems, these historical data are extremely useful in power system analysis. In order to be compatible with the rest of the framework, these storage systems have to export a DataStorageWS WSDL interface. So, a standardized and convenient method of accessing and managing distributed and heterogeneous data storage resources, such as a remotely accessed realtime database management system (RDBMS) or XML repository, must be defined.
The workflow enactor composes and integrates web services that deliver different functionalities, related mainly to (1) real-time electrical data acquisition from different sources; (2) supply of the high computational power to develop SG monitoring and control functions; and (3) storage functionalities to handle huge amounts of data.
The workflow enactor handles the execution request of the monitoring/control service described in a certain workflow language, which describes the temporal composition of the web services delivering the basic functionalities, submitted by means of the submission service. Figure 5.2 shows the action sequence performed by the workflow, referring to a single execution of the monitoring/control application [8].
Typically, the monitoring/control application has to be executed continuously, requiring for each execution the acquisition of current data on the SG’s state, particularly the following:
• Phase 1. getNetworkTopology and 2. getControlAsset are performed to obtain the necessary SG description and the characteristic data of the available control and monitoring devices.
• Phase 3. getCapabilitydata and 4. submitCapabilitydata are performed to acquire the input data for the generator/storage supply capability calculations, to arrange these data for each IED as a function of its geographical position, and to submit the corresponding information to the distributed IEDs.
• Phase 5. getNodesParameters is performed to obtain the real-time data of each node and is made to interact with the service implementing the DataAcquisitionWS interface, typically delivered by the electrical network for analysis.
FIGURE 5.2 Main actions of the workflow enactor.
• Phase 6. setJob, 7. submit, and 8. collectResults are performed to request the execution of the control and monitoring computations by a computational engine, and, in particular, are made to interact with a service implementing the JobSubmissionWS interface exported by a grid computing system. These computations comprise: (1) the SG’s state estimation, computed by processing the network data acquired and solving the system state equations; (2) the checking of the distribution network operational constraints; (3) the assessment of the system’s security and reliability levels; and (4) the control and regulation functions of the local SG (i.e., power factor, frequency and voltage control, and load management).
Finally, Phase 9, update, is executed to save the output data of the analysis in a data storage system interfaced through a service implementing the DataStorageWS interface.
In the following section, the implementation details of the basic web services and the workflow enactor for SG modeling and control are presented.
5.4.1 DATA ACQUISITION WEB SERVICE
Field data can be acquired by means of geographically dispersed resources (i.e., SCADA systems, IED, and field energy meters [FEM]). A prototype web service implementing the DataAcquisitionWS interface considering the FEM units located on each SG unit (load/generation/storage), and the direct interaction with them for real-time acquisition of the electrical parameters, was developed. In particular, a network of distributed FEMs based on ION 7330-7600 units was considered. They can measure demand on any instantaneous value and record peak (maximum) and minimum demand with a date and time stamp per second. The units are equipped with an onboard web server, which supports the XML format and the protocols ION, Modbus transmission control protocol (TCP), and Telnet, for their full remote control. The FEMs were physically connected to a fast ethernet local area network (LAN), configured with an Internet protocol (IP) address. The authors prototyped the interaction through the http protocol to access, in particular, an XML file, called realtime.xml, which contains the most recent measured field data.
To assess the supply capability curves of the SG components, a prototype version of an IED based on the ZWORLD-BL2100 unit (an advanced single-board computer that incorporates the Rabbit 2000 microprocessor) was adopted. These IEDs are installed directly on the SG components. They allow us to continuously monitor their state and to dynamically predict the corresponding supply capability curves as a function of the forecasted climatic variable profiles. These functionalities are fully available via an IP-based connection [23].
The authors are currently developing another implementation of the DataAcquisitionWS interface, considering the interaction with proprietary SCADAs. This new functionality will allow the integration of all the information coming from existing distributed measurement systems, and so decrease the response time of the service.
5.4.2 JOB SUBMISSION WEB SERVICE
The first implementation of the JobSubmissionWS was based on the hierarchical metacomputer middleware (HiMM), version 1.1, developed at the University of Sannio [26]. The main features of HiMM that are useful in this context are (1) the virtual hierarchical topology, which, allowing applications to exploit dedicated networks or clusters that are not directly accessible by the Internet, makes the system highly scalable; and (2) the support for the object-oriented programming paradigm and, in particular, for the Java language, which simplifies application development and easy integration with web technologies.
In addition, HiMM supports a grid broker service, which allows an application to exploit computing resources in a simplified way, hiding the complexity of the underlying system. The grid broker delivers a specific service to easily submit a master/slave application by using its nearly sequential version and some information necessary for deployment. It simplifies distributed programming, since it supports the separation of concerns related to functional (algorithmic issues, such as the creation of high-level domain-dependent abstractions in the form of objects and classes) and system (lower-level tasks such as object distribution, mapping, and load balancing) aspects of programming. The grid broker is responsible in particular for resource discovery and resource selection on the basis of QoS parameters specified by a user (such as the desired deadline and the maximum budget), task mapping on the selected resources, and task scheduling according to the hierarchical master/slave parallel model. In order to submit an application, the following information is required: (1) application code; (2) input data; and (3) application description and QoS requirements. For (3), descriptors based on XML called job description format (JDF) and user requirements description format (URDF) are to be defined.
5.4.3 COMPUTATIONAL WEB SERVICE
The computational grid used for performing the online computations was deployed on a test bed organized according to a logical hierarchical and heterogeneous topology of two levels, as shown in Figure 5.3 [27]. The first level consists of a set of directly accessible heterogeneous resources composed of eight computers, and two front-end computers that were used as masters to coordinate the computations of two pools of computers located in two different buildings and representing the second level of the network. The first pool consists of a cluster of workstations (COW) composed of eight homogeneous computers interconnected by a fast ethernet LAN through a hub 3Com TP16C. The second pool consists of a network of eight heterogeneous workstations (NOW), interconnected by a fast ethernet LAN through a Switch HP Procurve 2524. The first-level machines are interconnected by a fast ethernet LAN.
The mathematical solution engine adopted to support the online computations is based on a JavaRuntime Environment equipped with specific power systems analysis features. It allows system operators to investigate steady-state and dynamic issues in electrical networks by performing load flow analysis, time domain simulations, and optimal economic dispatch.
FIGURE 5.3 Computational grid adopted for online MG monitoring and control.
Various control and monitoring functions could be executed in parallel by the computational web service. It adopts a brokering service, based on an economy-driven model, to satisfy the QoS constraints specified by the user (i.e., a time deadline to solve a specific computation). By using an economy-driven model, the broker is able to (1) identify the optimal asset of the computational resources needed to satisfy the user requirements (i.e., resource discovery and resource selection process); (2) transparently split a sequential object-oriented task (i.e., a set of contingencies) into subtasks (i.e., a subset of contingencies) according to a hierarchical master/slave computing model; and (3) automatically distribute subtasks to the set of selected computational resources.
In order to evaluate the parallel processing performances of the computational web service, many experimental tests were executed by varying the number of machines used for the computational engine in order to evaluate its speedup factor. The computation time and the speedup obtained by increasing the slaves in powers of two (two, four, and eight slaves) were evaluated. Each computation was performed several times in order to use the average. The speedup obtained by using a two-level architecture is nearly linearly related to the number of slave machines and is very close to the ideal case.
5.4.4 DATA STORAGE WEB SERVICE
In this first phase, the authors used a relation database, installed on a Linux workstation and based on the Postgres RDBMS.
The main implementation issue related to the workflow enactor is the choice of a workflow language and its related technologies, which can be usefully adopted to describe the wireless asset monitoring (WAM) application in the context of a web services-based approach. Many proposals have been made in the literature, especially by the major web services software providers, and many of them are continuously evolving. In these proposals the application workflow is usually described with a textual scripting language (typically the XML format) that is used to provide the “glue” that links involved services together.
Among the possible solutions, we chose business process execution language (BPEL). BPEL is a programming abstraction that allows developers to compose multiple discrete web services into an end-to-end process flow. At the present time it could be considered the most complete, and has risen in popularity since many vendors are developing their toolkits, many of which are also becoming commercial. Among the available BPEL engines, we chose to integrate in our framework the workflow enactor delivered by BPWS4J, since it is free and sufficiently stable.
The BPWS4J toolkit includes a workflow execution platform, a tool that validates BPEL4J documents, and an Eclipse plug-in that provides a simple editor for creating and modifying BPEL4J files. In particular, the workflow enactor of our platform is based on the workflow execution of BPEL4J integrated in Apache Tomcat, version 5.0.24, and is specialized to process the online power systems security analysis (OPSSA) workflow.
For each process, the BPWS4J engine takes in a BPEL document that describes the process workflow to be executed, a WSDL document (without binding information) that describes the interface that the process will present to clients, and WSDL documents that describe the services that the process may or will invoke during its execution. From this information, the process is made available as a web service with a SOAP interface. The SG workflow describes the composition of the web services involved, where each service is defined as a <partner>. Service composition in the SG workflow is defined using a sequence tag, which defines how the partners will be sequentially executed. The specification of a sequence includes definitions of the input message (the receive tag), of the service invocation (the invoke tag), and of the output message (the reply tag). The assign tag establishes the relationship between the output message of a service and the input message of the service in the workflow.
Similarly to the individual web services, which are described through the corresponding WSDL file, the SG web service is also described in a WSDL file (Figure 5.4). Messages, portTypes, and operations are also defined, and there is an additional section (serviceLinkType tag) to define the role of each service during their interaction.
FIGURE 5.4 WAM WSDL file.
The web services implemented, in particular the data acquisition and the job submission web services, were written using the Java language and the Apache Axis toolkit. The Axis toolkit essentially included an SOAP engine, that is, a framework for constructing SOAP processors such as clients and servers, a simple stand-alone server, a server that plugs into servlet engines such as Apache Tomcat, and extensive support for WSDL.
The chapter proposed an advanced web services–based framework for integrated SG control, modelling, and monitoring. The core component of the proposed framework is the SG engine, which is responsible for the execution of the SG management functions in a geographically distributed scenario. It includes a network of remotely controlled units distributed in the most critical SG sections for field data acquisition and advanced protective functionalities; a GIS interface for the acquisition of georeferentiated information on the SG; a grid computing–based solution engine for the online computations; and a web-based interface for graphical synoptic and reporting development.
The web services–based framework has been deployed on a computational grid. The first experimental results have emphasized the important role of web services technology in SG control, modelling, and monitoring.
The application of this technology is very promising, since it makes the proposed framework platform-, language-, and vendor-independent, and thus an ideal candidate for an effective integration in existing energy and distribution management systems (EMS/DMS).
The future directions of our research are oriented toward the conceptualization of a fully decentralized nonhierarchical architecture for SG monitoring and control that will move away from the older centralized paradigm to a system distributed in the field with an increasing pervasion of intelligence devices (smart sensors). The adoption of smart sensors is expected to lead to a more efficient task distribution among the monitoring system resources and, consequently, to significantly less demand on centralized computing resources.
1. S. M. Amin and B. F. Wollenberg, Toward a smart grid, IEEE Power and Energy Magazine, 3(5), 34–38, 2005.
2. A. Schroder, I. Laresgoiti, K. Werlen, B. Schowe-von der Brelie and A. Schnettler, Intelligent self-describing power grids, in Proceedings of the 20th International Conference and Exhibition on Electricity Distribution (CIRED), 8–11 June 2009, Prague, Czech Republic.
3. R. L. King, Information services for smart grids, in Proceedings of 2008 IEEE Power and Energy Society General Meeting—Conversion and Delivery of Electrical Energy in the 21st Century, 20–24 July 2008, Pittsburgh, PA.
4. N. Hadjsaid, J.-F. Canard and F. Dumas, Dispersed generation impact on distribution networks, IEEE Computer Applications in Power, 12(2), 22–28, 1999.
5. R. Comfort, A. Mansoor and A. Sundaram, Power quality impact of distributed generation: Effect on steady state voltage regulation. Presented at the PQA 2001 North America Conference, 2001, Pittsburgh, PA.
6. J. Zhu, Web services provide the power to integrate, IEEE Power and Energy Magazine, 1(6), 40–49, 2003.
7. Q. Liu, J. Wen and H. Sun, Implementing power system management via semantic web services composition, in Lecture Notes in Computer Science, Springer, Berlin, 2009.
8. Q. Morante, N. Ranaldo and E. Zimeo, Web services workflow for power system security assessment, in Proceedings of IEEE International Conference on e-Technology, e-Commerce, e-Service, IEEE Computer Society Press, 29 March–1 April 2005, Hong Kong.
9. J. R. Roncero, Integration is key to smart grid management, in IET-CIRED Seminar on SmartGrids for Distribution, 23–24 June 2008, Frankfurt.
10. H. Johal, W. Ren, Y. Pan and M. Krok, An integrated approach for controlling and optimizing the operation of a power distribution system, in 2010 IEEE PES Innovative Smart Grid Technologies Conference Europe (ISGT Europe), Institute of Electrical and Electronics Engineers, 11–13 October 2010, Gothenburg.
11. B. Ramachandran, S. K. Srivastava, D. A. Cartes and C. S. Edrington, Distributed energy resource management in a smart grid by risk based auction strategy for profit maximization, in Power and Energy Society General Meeting, IEEE, 25–29 July 2010, Minneapolis, MN.
12. A. Mahsoom, O. Olusola and L. Jonathan, Exploring control systems to implement SMART GRID solutions, in Proceedings of 10th IET International Conference on Developments in Power System Protection (DPSP 2010), 29 March–1 April 2010, Manchester, UK.
13. T. M. Overman and R. W. Sackman, High assurance smart grid: Smart grid control systems communications architecture, in 2010 First IEEE International Conference on Smart Grid Communications (SmartGridComm), 4–6 October 2010, Gaithersburg, MD.
14. M. Pipattanasomporn, H. Feroze and S. Rahman, Multi-agent systems in a distributed smart grid: Design and implementation, in IEEE/PES Power Systems Conference and Exposition, PSCE ‘09, 15–18 March 2009, Seattle, WA.
15. V. M. Catterson, E. M. Davidson and S. D. J. McArthur, Agents for active network management and condition monitoring in the smart grid, in First International Workshop on Agent Technologies for Energy Systems, 11 May 2010, Toronto, ON.
16. F. Li, P. Zhang and N. Bhatt, Next generation monitoring and control functions for future control centers, in IEEE PES General Meeting, 26–30 July 2009, Calgary, AB.
17. J. G. Cupp and M. E. Beehler, Implementing smart grid communications, TECHBriefs, No. 4, 5–8, 2008.
18. D. Baigent, M. Adamiak and R. Mackiewicz, IEC 61850 Communication Networks and Systems in Substations: An Overview for Users, SIPSEP, Monterrey, Mexico, 2004.
19. A. Apostolov, C. Brunner and K. Clinard, Use of IEC 61850 object models for power systems quality/security data exchange, in Quality and Security of Electric Power Delivery Systems. CIGRE/IEEE PES International Symposium, pp. 155–164, 8–10 October, Montreal, Canada, 2003.
20. M. C. Janssen, C. A. Andersen, C. Brunner and J. Dall, Modelling of DER schedules using IEC 61850, in Proceedings of 20th International Conference on Electricity Distribution, 8–11 June 2009, Prague, Czech Republic.
21. C. Brunner, IEC 61850 for power system communication, in Proceedings of IEEE/PES Transmission and Distribution Conference and Exposition, 21–24 April 2008, Chicago, IL.
22. P. Lloret, J. L. Velasquez, L. Molas-Balada, R. Villafafila, A. Sumper and S. Galceran-Arellano, IEC 61850 as a flexible tool for electrical systems monitoring, in Proceedings of the 9th International Conference on Electrical Power Quality and Utilisation, EPQU, 9–11 October 2007, Barcelona, Spain.
23. A. Vaccaro and D. Villacci, Prototyping of a distributed JAVA based client/server architecture for e-Assessment of power equipments loading capability, in Proceedings of IEEE-Powertech, 22–26 June 2003, Bologna, Italy.
24. M. Di Santo, N. Ranaldo, A. Vaccaro and E. Zimeo, Java-based distributed architectures for intensive computations in electrical grids, in Proceedings of International Workshop on Java for Parallel and Distributed Computing, April 2004, Santa Fe, NM.
25. M. Irving, G. Taylor and P. Hobson, Plug in to grid computing, IEEE Power and Energy Magazine, 2(2), 40–44, 2004.
26. M. Di Santo, N. Ranaldo and E. Zimeo, A broker architecture for object-oriented master/slave computing in a hierarchical grid system, in Parallel Computing, September 2003, Dresden, Germany.
27. Q. Morante, N. Ranaldo, A. Vaccaro and E. Zimeo, Pervasive grid for large-scale power systems contingency analysis, IEEE Transactions on Industrial Informatics, 2(3), 165–175, 2006.