Design and application of decoupled controller based on the BP neural network

Abstract

Composed by the neural network and traditional PID controller, a new BP neural network PID parameter self-adjustment decoupled controller is designed to realize the control of multi-input and multi-output systems in this paper, as well as to solve the problem such that the traditional PID controller can not achieve satisfied control effectiveness in multi-input and multi-output strong coupled control systems. The BP neural network PID controller is not only owning non-linear mapping capability, but also owning the PID capability of dealing with dynamical information. Because of the simple structure and algorithm, the BP neural network PID control method is more suitable for industry application.