基于模糊神经网络的伺服系统PID控制

Study of PID Controller to Servo System Base on Fuzzy-Neural Network

针对某装备伺服系统具有复杂的非线性和时变性,传统的PID控制存在自适应能力差的缺点,设计了一种模糊神经网络PID控制器。利用神经网络对模糊控制的控制规则进行优化,根据偏差E和偏差变化EC在线调整PID控制器的3个参数。通过与传统的PID控制器仿真实验对比,可以看出模糊神经网络PID控制器能够有效地提高该伺服系统的性能。

Because the servo system of equipment has non -linearity and time -variation characteristics and the traditional PID control has the shortage of self-adapting,a fuzzy neural network PID controller is designed. The rules of fuzzy control are adjusted base on the neural network. According to error and error variance,the three parameters of PID controller are adjusted on-line. The consequence of simulation indicates that compared with the traditional PID,the performance of the servo system is highly improved when the fuzzy neural network adaptive PID controller is used.