MDF施胶系统模糊神经网络PID控制

Fuzzy Neural Network PID Control of Medium Density Fiberboard glue system

针对中密度纤维板(MDF)施胶系统中存在的非线性、纯滞后等现象,提出一种模糊神经网络PID控制方法。系统采用模糊神经网络控制器和PID控制器相结合的结构,改进了模糊神经网络自学习的能力,能够实时整定PID控制器的参数,提高控制系统的精度。仿真结果表明,模糊神经网络PID控制系统能够提高MDF施胶系统的响应速度和抗干扰能力,并且使MDF施胶系统具有较好的控制效果与控制精度。

Aiming at the problems of the Medium Density Fiberboard （MDF） glue system such as nonlinear characteristic, pure delay, a fuzzy neural network PID control scheme was presented. In this scheme it has the structure that combines a fuzzy neural network controller with PID controller enhance the self-learning abilities of fuzzy neural network to set the parameters of PID controller in real-time and also improve the accuracy of control system. Simulation results indicate that： the fuzzy neural network PID control system can improve response rapidity of the MDF glue system and anti-interference ability, and also making the MDF glue system have better control of performance and precision.