Case Studies

Abstract

AbstractThis chapter presents step-by-step the optimization of controllers for two canonical benchmark problems of interest to the control community, i.e., the inverted cart-pendulum (ICP) and the automatic voltage regulator (AVR) systems. Literature surveys show that these systems’ control tuning has been intensively studied, but there is still control needs that metaheuristic optimization can fulfill. The principal features of metaheuristics for controller tuning are demonstrated through the optimization by GA and SA of different controllers for the ICP and the AVR system. In particular, simulation results show that metaheuristics provide high-quality solutions regarding dynamic performance and robustness and stability properties. A new robust optimization framework for controller tuning by metaheuristics is proposed. Moreover, analyses on components that influence metaheuristics performance on control tuning are presented. These analyses highlight the major elements that require more attention to applying metaheuristics properly to controller tuning. Also, SA and GA performances are compared. Elements for proper algorithm benchmarking are detailed. Three sections divide this chapter. The first two sections present the application of GA and SA to controller tuning for the inverted cart-pendulum (ICP) and the automatic voltage regulator (AVR) system, respectively. The last section summarizes the major advantages of metaheuristics for controller tuning along with the components to consider applying metaheuristics properly to controller tuning.