基于模糊神经网络的PID张力控制系统

Tension control system based on fuzzy-neural network PID

由于卷绕张力控制系统是一个复杂、联动、时变、非线性系统,采用传统PID控制不能解决系统的非线性时变和PID参数的在线整定难等问题,为此提出一种控制算法——模糊神经网络PID复合控制方式,可根据系统的偏差及其变化率实时对PID的3个参数进行优化,达到具有最佳组合的PID控制,从而实现PID控制的自适应和智能化性能。通过MatLab软件,进行传统PID控制与模糊神经网络PID控制动态性能的仿真比较,结果表明系统采用模糊神经网络PID控制具有更好的动、静态特性和自适应性。

Since the winding tension control system is a complicated, tandem driving, time change and non-linear one, using traditional PID control can not settle the problems of dynamic change of the non-linear system and PID parameter adjustment online. The paper gives one control algorithm-the combination method of fuzzy-neural network PID control, which can optimize three parameters of PID in operation according to error and error rate at real time, achieve the PID control with the best combinations, and realize the adaptive and intelligent performance of PID control. The fuzzy-neural network PID control is compared with traditional PID by simulation through MatLab. The results show that the system has better dynamic, static and adaptive performances by using the fuzzy-neural network PID control strategy.