刨花板热压控制系统BP神经网络整定PID控制

Particleboard Hot press Control Based on BP Neural Networks PID Method

针对刨花板热压控制系统中热压控制存在的非线性、纯滞后和时变性等现象,根据BP神经网络PID和常规PID控制的控制思想,提出了BP神经网络PID的控制策略,实现了对PID参数的在线自整定,完善了PID控制器的性能,提高了系统的控制精度。仿真结果表明,与常规PID控制器相比,该控制器明显提高了热压控制系统的动态性能和稳定性。

According to some phenomena such as non-linear characteristic, the special process requirements of steady control without overshoot, and base on BP neural networks control, self-tuning control and PID control theory, we adopt the BP self-tuning PID strategy to control the slab thickness. The method on the PID parameters by BP optimization implement online self-tuning, which can consummate the performances of PID controller and improve control precision and stabi- lization of the electro-hydraulic position servo system. It achieves the control characteristics of hot press technological requirements. Furthermore, the simulation results also verify that the BP self-tuning control all can get the non-overshoot, fast and stable response in the two different sets of PID initial parameters circumstances.