Simulation Study On PID Control System Based On BP Neural Network For Greenhouse Environment

The study on greenhouse environment automatic control system is started relatively late in China, and the level is also very low. There is lack of intelligence method on control algorithm to improve its control performance. Previously greenhouse temperature and humidity control were adopted the conventional PID control, which is simple and easy to realize. However, greenhouse is a nonlinear, strong time-varying, strong coupling and complicated object with the parameter changing in a big scope, and also the control parameters setting is more difficult, so it must be combined with the corresponding intelligent algorithm to solve the problem of complex object of greenhouse environment. In the conventional PID control even setting a group of control parameters in small control range will take good control effect, but when the characteristics of controlled object changing, it is hard to make its stability and control quality guaranteed, so control effect of conventional PID is not very satisfactory.This paper mainly studies temperature and humidity monitoring in greenhouse environment, which puts forward the overall design method on temperature and humidity control in greenhouse environment, especially on the problems of parameters hard to control, long settling time and big over-shoot in the conventional PID control greenhouse control system. This paper also analyzes advantages and disadvantages of BP(back propagation) neural network algorithm and the conventional PID control, and based on the conventional PID controller with BP neural network to make the system performance to achieve the best PID control parameters through neural network learning on the system performance, which will be directly transferred to the conventional PID controller so as to obtain PID controller based on BP neural network adjusting. The controller with PID parameters set by neural network can solve what’s the conventional PID parameter difficult to set. Modern greenhouse environmental parameters, mainly intelligence of temperature and humidity control are researched in this paper. On the basis of previous studies, this paper analyzes greenhouse control process is a nonlinear, time delay complex process and establishes temperature and humidity model in the greenhouse environment. Simulations are made to compare with conventional PID control and BP neural network PID control through MATLAB, which shows that PID controller based on BP neural network has better control effect. Meanwhile, STC89C52microcontroller as control core combined with LabVIEW graphical programming technology, it designs a practical greenhouse environment automatic monitoring system that can achieve real-time monitoring and control of the two environmental parameters of temperature and humidity in the greenhouse environment and improve the performance of the greenhouse environment control and control effect,making a beneficial attempt for automation and intelligence of greenhouse environment control technology in the application of control system.