针对压延机辊筒多段温度控制以及常规PID控制在非线性的、时变系统中控制效果的局限性,提出了一种基于BP神经网络整定的PID控制方法,给出了计算机控制系统设计,建立了比例、积分和微分三种参数自学习的PID控制器。对压延机辊筒的温度控制实验结果表明,用该方法整定的PID控制系统,逼近精度高、适应性好。
A PID control method based on BP neural network was proposed, concerning the multi-layered temperature control of calenderer rollers and the limitations of the conventional PID control which resulted from the control of the nonlinear and time-varying system. The system design and software development were also presented. Possessing such merits as the ability of nonlinear mapping, self-learning, self-adaptation and etc., this neural network could help to realize the PID control with the best combinations by means of the understanding of the system performance. The network had high approximate precision and good robustness. The experiment results showed the method was feasible and effective.