A Novel Optimization of Fractional Order PID Controller Using Chaotic Maps Based Atomic Search Optimization for pH Control in Continuous Stirred Tank Reactor

Purpose The initialization problem of the physics-inspired metaheuristic Atomic Search Optimization (ASO) is overcome by the usage of chaotic maps as the random number generators (RNGs) in this study. This method of initialization improved the convergence speed and avoided the local optimum stagnation of ASO. Methods The proposed Ch ASO with the improve dinitial population using three chaotic maps namely tent, circle and logistic maps were tested on four different benchmark functions. The statistical representatives of the proposed Ch ASO were studied and compared with the basic ASO. Tuning of thefractional-order controllers is a necessity in controller design and the methods used for tuning are mostly theoretical. So the proposed Ch ASO algorithm is used as a tuning methodology to obtain the optimal tunable parameters of the Fractional Order Proportional Integral Derivative (FOPID) controller namely Kp, K-i, K-d, lambda, mu. The objective function (OF) used for the optimization are the integral time absolute error (ITAE), and Zwe-Lee Gaing's (ZLG) to identify the best OF. Results The FOPID controller tuned by Ch ASO was applied to the experimental pH control of a Continuous Stirred Tank Reactor (CSTR). The transient response analysis, frequency response analysis and robustness analysis were carried out for the proposed algorithm-controller combination and compared with ASO-PID, ASO-FOPID. The algorithm-controller combinations were validated using both theoretical and experimental methods. Conclusion The Tent map based ASO-FOPID controller perfromed better the oretical and experimental methods. [GRAPHICS] .