A Nonlinear PID Control for Time-varying Batch Processes by Using Particle Swarm Optimization

Abstract

Nonlinear batch processes in physical and chemical industries often experience time-dependent parameter changes and dynamic perturbations, which make the common PID method insufficient for achieving high-quality control effects. A particle swarm optimization (PSO)–based nonlinear PID (NLPID) control strategy is introduced to handle the nonlinear behavior and time-varying properties commonly observed in batch processes. At the beginning of this study, a nonlinear function is used to replace the tracking error in the PID method, thereby achieving NLPID control. Secondly, considering a performance index of integral square error (ISE), the optimal NLPID control parameters can be obtained by using PSO algorithm (PSO-NLPID). The proposed PSO-NLPID method is ultimately validated in batch process applications through comparative analysis with current methods. To demonstrate the validity of the proposed control strategy, a typical fermentation process is simulated in the batch cycle. Regarding the parameter time-varying problem in nonlinear batch processes, the PSO-NLPID strategy outperforms existing methods, as evidenced by its lowest ISE value of 1.1293, and a performance enhancement exceeding 65% over standard PID control, which indicates that PSO-NLPID has better control performance.