Research and Application of BP Neural Network PID Control Algorithm in Three Tank System
|School||Liaoning University of Science and Technology|
|Keywords||Three-tank water level BP-PID SCM Modbus OPC|
Level is one of the main object of modern industrial process control parameters, liquid level control in the iron and steel metallurgy, petrochemical and food processing industry are very common. Traditional PID control method is difficult to achieve good control effect Level in complex level systems, it is severely restricting further of the product quality and economic efficiency. Three-tank system is a good simulation of a complex industrial level control objects. So three-tank system as research object, this paper researches BP-PID intelligent control method of complex level system. This paper both researches the BP-PID theory and practical application of the algorithm.This paper firstly introduced the theory of traditional PID algorithm and neural network theory and in-depth study of the BP neural network PID control algorithm. This paper established mathematical models of the three tank level system for simulation and compared with the traditional PID algorithm to verify the BP-PID algorithm effectiveness; Secondly,this paper designed the three-tank system microcontroller to complete the data acquisition and control functions. And the standard MODBUS communication protocol was written; Again this paper researched communication methods between single chip controller and KEPWARE OPC server, completed the VB OPC client application design, implementation of the VB OPC server software and communications capabilities to code VB-based BP neural network algorithm, combined with single chip controller implementation of the BP-PID control algorithm; Control system is given the results of experimental runs, the results show that, BP-PID control algorithm is adaptive than traditional PID control algorithm and it can achieve good results for the control of complex fluid systems.