Fault-tolerant control systems. Design and practical applications (Q612970)

A fault-tolerant control system is a control system specifically designed with potential system component failures. A fault-tolerant control may not offer optimal performance in a strict sense for normal system operation, but generally it can mitigate effects of system component failures without completely jeopardizing the mission or putting the user at risk. The philosophy of fault-tolerant control systems design is different from other design methodologies. Consequently, their behavior under system component failures will also be different. The present monograph is one of very few books in this area that considers practical aspects of fault-tolerant control (see for example [\textit{C. Hajiyev} and \textit{F. Caliskan}, Fault diagnosis and reconfiguration in flight control systems. Cooperative Systems 2. Boston, MA: Kluwer Academic Publishers (2003; Zbl 1066.93039)], [\textit{H. Benítez-Pérez} and \textit{F. García-Nocetti}, Reconfigurable distributed control. London: Springer (2005; Zbl 1101.93001)], [\textit{M. Blanke, M. Kinnaert, J. Lunze} and \textit{M. Staroswiecki}, Diagnosis and fault-tolerant control. With contributions by Jochen Schröder. 2nd ed. Berlin: Springer (2006; Zbl 1126.93004)], [\textit{F. Caccavale} and \textit{L. Villani}, Fault diagnosis and fault tolerance for mechatronic systems: Recent advances. Springer Tracts in Advanced Robotics. 1. Berlin: Springer (2003; Zbl 1017.00013)]). The basic concepts of fault-tolerant systems are introduced in the first chapter. The design of a Fault-Tolerant Control (FTC) system requires quick Fault Detection and Isolation (FDI) for adequate decision making and these follow various knowledge-based system strategies: quantitative models [\textit{M. Darouach, M. Zasadzinski} and \textit{M. Hayar}, IEEE Trans. Autom. Control 41, No. 7, 1068--1072 (1996; Zbl 0857.93016)], qualitative models and historical data. Among quantitative models, FDI for linear/nonlinear systems remains a challenge due to the problem of discriminating between disturbance and faults through a wide range of operating conditions. In the literature, FTC methods are developed considering that the FDI is already achieved. In order to eliminate the actuator faults effect which occur on the system, various method have been proposed. Fault-tolerant control system design and analysis against actuator and sensor failures have been treated in detail in the second chapter. Because many industrial systems are complex and nonlinear, a linearized system around an operation point is proposed. Two interesting cases of MIMO (Multiple Inputs and Multiple Outputs) systems are presented: unknown input observer system with linear and affine cases, respectively. This chapter ends with a very representative FTC scheme. Chapter 3 is devoted to the application of fault-tolerant control strategies on a physical lab-scale winding machine. First, is treated the linear case. Then, a nonlinear case is taken. A discrete linear time-varying model is used to start analysis for the nonlinear case. A well known three-tank system is used in Chapter 4 to illustrate design and analysis techniques for fault-tolerant control systems. The authors have also included some \texttt{MATLAB} scripts to guide the readers through the process. Also, first a linear case is analyzed. In order to extend the proposed active FTC approach based on the FDI results on the whole range of operating conditions, the nonlinear model is considered rather than a multiple (linear) model approach. As methods used in this case we mention: an input-output linearizing and input-output decoupling law via a static state-feedback (see [\textit{A. J. Fossard} and \textit{D. Normand-Cyrot}, Nonlinear systems. Vol. 1: Modeling and estimation. London: Chapman \& Hall (1995; Zbl 0937.93500)], [\textit{L. Guzzella} and \textit{A. Isidori}, Int. J. Robust Nonlinear Control 3, No. 3, 261--276 (1993; Zbl 0792.93047)], [\textit{A. van der Schaft}, \(L_2\)-gain and passivity techniques in nonlinear control. Lecture Notes in Control and Information Sciences. 218. Berlin: Springer (1996; Zbl 0925.93004)]). Finally, fault-tolerant control system design and analysis against sensor failures in an active suspension of a full vehicle system have been considered. Detailed mathematical description of the suspension is derived first. The originality of the work in this chapter is the breakdown of the entire suspension system into several interconnected subsystems, each subsystem has its own local controller and its own fault diagnostic module. In conclusion, the authors have successfully presented some of the most important concepts and procedures in fault-tolerant control system design and analysis. They have done this with elegance of mathematics, as well as in-depth physical understanding of the limitations of handicapped actuators and sensors.