Research Of Neural Network Predictive Control In Water Source Heat Pump Air-condition Control System

The main research work of this dissertation is to resolve the time delay problem which occurs when the Water Source Heat Pump Air-condition Control System (WSHPACS) developed by North China University of Technology (NCUT) is working and to study the optimal control of central air-condition based on neural network.The WSACS developed by NCUT works practically in one flat building of Shandong province now. The running of WSACS shows that the dew will occur on ground in refrigerating state, which is absolutely forbidden and influences the working of WSHPACS in summer.With regards to this practical problem, a predictive control model is designed base on the WSACS under the constrained condition of critical dew curve which is found under the condition of temperature and humidity of the environment.The WSHPACS is a large inertia, large time-delay and nonlinear system which is easily jammed by all kinds of outside factor. Sometimes these jamming are very serious and will influence the stability of the system if these jamming are not deal with properly.Because of the complex, nonlinear and large time-delay characteristics of the WSHPACS, an experiment platform of Temperature Control Box (TCB) is designed to stimulate the WSHPACS.Fist the traditional PID control strategy is implemented in the TCB, which shows the structure of the system is reasonable and each function unit works properly. And the expectant control effect occurs.With regards to the characteristics of nonlinear, large inertia and large time-delay system, this paper puts forward a predictive control strategy based on neural network.