Recent Metaheuristic-Based Optimization for System Modeling and PID Controllers Tuning

: Recently, several methods and techniques, including the metaheuristic algorithms, have been developed, to identify and control systems. In this paper, four recent algorithms, such as Ant Lion optimizer (ALO), Differential Evolution (DE), Bat Algorithm (BA), and Harmony Search (HS), are chosen and considered for a one-paper comparison for the first time and exclusively applied to four different types of behaviors. The present contribution concerns the systematic analysis and comparison between the mentioned algorithms for the two tasks of system modeling and for the tuning of proportional, integral, and derivative (PID) controllers. Comparisons with conventional methods, such as Least Squares (LS) for identification and Reference Model (RM) for control, are made with different instructions to highlight the efficiency of this methods Further, the details on their performance metrics in terms of premature convergence and dynamic searches are provided. Simulations results demonstrate how accurately they help to obtain optimal solutions and show the most reliable method for the two main tasks of control and identification. Moreover, the present results confirm that the Differential Evolution strategy has the best performance, stable convergence feature, robustness, and insensitivity to disturbance and signal excitation .