基于优化BP神经网络的PID控制器研究

提出了一种新型的基于优化BP神经网络结构的PID控制器（PID－NNC），该控制器将神经网络和PID控制规律融为一体，既具有神经网络自学习，自适应及逼近任意函数的能力，又具有常规PID控制器结构简单，可靠性高的特点，且控制器的算法采用的是优化的BP算法，可以避免网络陷入局部极小点，也可以加快网络的训练速度，所以该控制器可以对具有非线性，时变性和不确定性等复杂系统实行控制。利用MATLAB软件对非线性系统进行了仿真研究，其仿真结果表明该控制器具有很好的控制效果。

Study of PID Controller Based on Improved BP Neural Network

A new type PID-controller based on improved BP neural network is presented, which is a combination of the neural network and the PID rule. It has merits of both neural network and PID-controller. Moreover, improved BP algorithm in controller can make study speed of network faster and can eliminate the disadvantages of BP network. So it is suitable for the process with nonlinear, uncertain, and time-varying parameters. The computer simulation on an object with nonlinear shows that the controller can produce the perfect control effect.