PID Controller Design for MIMO Processes Using Improved Particle Swarm Optimization
Wei-Der Chang,Chih-Yung Chen
DOI: https://doi.org/10.1007/s00034-013-9710-4
2013-12-06
Abstract:This paper aims at the PID control system design for multivariable input and multivariable output (MIMO) processes. An improved version of a particle swarm optimization (PSO) algorithm is utilized to design PID control gains in MIMO control systems. In addition to the individual best and the global best particles, the velocity updating formula of the developed algorithm includes a new factor, the best particle of each sub-population, to enhance the search capacity. Based on the improved particle swarm optimization (IPSO), a complete design strategy is proposed for MIMO PID control systems. All control gains will be evolved to the optimal values by minimizing the system performance criterion. To show the efficiency of the proposed design method, a multivariable chemical process system with two inputs and two outputs is illustrated. Some experiment results, including different algorithm parameter settings and comparisons with other methods, are given. Numerical simulations indicate that the proposed method is superior to other optimal methods.
engineering, electrical & electronic