1
C H A P T E R 1
Introductions
Intelligent vehicles have been gaining increasing attention from both academia and industrial
sectors [1]. e field of intelligent vehicles exhibits a multidisciplinary nature, involving trans-
portation, automotive engineering, information, energy, and security [2–5]. Intelligent vehicles
have increased their capabilities in highly and even fully automated driving. However, unre-
solved problems do exist due to strong uncertainties and complex human-vehicle interactions.
Highly automated vehicles are likely to be on public roads within a few years. Before
transitioning to fully autonomous driving, driver behavior should be better understood and in-
tegrated to enhance vehicle performance and traffic efficiency [6–9]. To address these challenges,
researchers have explored advanced driver assistance systems (ADAS), and human-machine in-
terface (HMI) from a variety of points of view [10, 11]. However, since the dynamic rela-
tionships between driver and vehicle are highly complex, satisfactory driver-vehicle interactions
should go beyond the present ADAS and HMI systems. Human-vehicle interactions have al-
ready been considered in a high-level closed loop, where driving style, driving feel, and vehicle
performance are considered [12]. Driving style plays a very important role in vehicle energy effi-
ciency and ride comfort, thus significantly impacting controller synthesis [12–14]. For instance,
control objectives and control protocols should be adaptively adjusted according to different driv-
ing styles. Based on the findings reported in [13], a better understanding of driving styles could
help improve ADAS performance and further reduce vehicle’s fuel consumption through driver
feedback. In [14], an enhanced intelligent driver model was developed, and then it was used
to investigate the impact of different driving strategies on traffic capacity. In [15], an adaptive
cruise control strategy considering the characteristics of different driving styles was developed,
and the proposed strategy could automatically adapt to different traffic situations. Nevertheless,
advanced control and optimization of vehicle systems with characterized driving styles are still
open challenges and worthwhile exploring.
In the meantime, the ever-growing attention to the environment and energy conservation
requires automobiles to be cleaner and more efficient [16–18]. In this study, an electric vehicle
(EV) is chosen as the platform to conduct our research in cyber-physical vehicle systems. Based
on existing studies, small changes in driving style can cause unnecessary energy waste and sub-
optimal performance of an EV [19, 20]. Moreover, regenerative braking capability of EVs can be
enhanced by prior knowledge of driving style. Hence, an optimal energy management strategy
can be obtained with knowledge about the entire driving cycle, environment, and driver be-
haviors. erefore, the information of operating scenarios, driver behaviors, and driver-vehicle
2 1. INTRODUCTIONS
Environment
Controller
Structure
Human Vehicle
Driver Operations
Controller Para.
Adaptation
Controller-
Plant
Interactions
Structure Para.
Adaptation
Human-Vehicle Interactions
Vehicle Plant
Vehicle
Performance
Figure 1.1: Schematic diagram of the CPVS.
interactions are crucial and should be integrated to enhance the energy efficiency of automated
electric vehicles.
A Cyber-Physical System (CPS) is a distributed, networked system that fuses computa-
tional processes (cyber world) with the physical world. An intelligent electric vehicle is a typical
example of Cyber-Physical Vehicle System (CPVS). In details, an automated electric vehicle
involves the following subsystems: the controller, representing the “Cyber” world, the physical
vehicle plant, the driver, the “Human,” and the environment. ese different parts, which are
highly coupled, decide the vehicle’s behavior and final performance, as Fig. 1.1 shows. e main
drawback of the conventional implementations in vehicle design and control is the lack of global
optimality in the selection of architecture, parameters, and variables [25]. For instance, by using
the conventional design method, which deals with different subsystems independently, even if
the controller is very well designed, the improvement of vehicle performance could be limited,
since the physical architecture and parameters are not optimized in sync with the controller,
and the system potential is not fully explored. In this context, the emerging co-design method
provides the capability to extend system design space and further enhance the performance of
CPS [24–28]. In [24], a platform-based design method utilizing contracts to do the high-level
abstraction of the components in a CPS was proposed, and it is able to offer support to the
overall design process. In [26], co-design optimization of a cyber physical vehicle system, which
considers task time, actuator characteristics, energy consumption, and processor workload, was
investigated. In [27], a CPS-based control framework was developed for vehicle systems to min-
1. INTRODUCTIONS 3
imize the car-following fuel consumption and ensure inter-vehicle safety. Besides the cyber and
the physical worlds, we also need to take “Human” of an automated vehicle into consideration.
us, the interactive impacts between the vehicle plant, control variables, multi-performance,
and driver styles should be well understood [29–31].
To further advance the existing CPS methods as well as their applications in vehicle engi-
neering, the following topics will be explored for CPVS in this book: (1) a novel co-design opti-
mization methodology for CPVS; (2) dynamic estimation of hybrid states for online monitoring
of CPVS; and (3) advanced control synthesis for CPVS for improving multiple performance.
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.