基于改进PSO-FNN-模糊PID算法的多品种造纸机定量控制OA

Aiming at the problems of traditional fuzzy Proportional Integral Derivative(PID)control in the basis weight control of multi-variety paper machines—such as parameters tuning relying on experience,weak adaptive ability of fuzzy rules,as well as the standard Particle Swarm Optimization(PSO)algorithm being prone to premature convergence and insufficient local optimization ability-a basis weight control method for multi-variety paper machines based on an improved Particle Swarm Optimization-Fuzzy Neural Network-Fuzzy PID(PSO-FNN-Fuzzy PID)was proposed.An adaptive adjustment strategy for the inertia weight was designed to improve the PSO algorithm,thereby balancing the global optimization and local convergence capabilities of the algorithm,which was further used to optimize the connection weights of the Fuzzy Neural Network(FNN).The FNN optimized by the improved PSO was combined with fuzzy PID to construct an improved PSO-FNN-Fuzzy PID controller,realizing the collaborative optimization of PID parameters and fuzzy rules through the synergy of FNN and Fuzzy PID.Taking the multi-variety paper machine as the control object,the paper basis weight deviation and deviation change rate were selected as inputs,and the PID parameters correction amount as the output.The control performances of four control methods-traditional fuzzy PID,Proportional Integral Derivative-Extreme Learning Machine(PID-ELM),adaptive fuzzy sliding mode control,and the improved PSO-FNN-Fuzzy PID-were compared through simulation experiments.The results show that the improved control method achieves 36 convergence steps,an average paper uniformity of 89.7％,and an overshoot of 5.3％,exhibiting significantly superior performance indicators compared to the other algorithms.This proposed method can effectively adapt to the nonlinear and time-varying characteristics of multi-variety production and significantly improve the basis weight control accuracy.