Research on Fuzzy-Neural Networks Controller in Thermostatic and Humidistatic Aircondition System

摘要

The thermostatic and humidistatic aircondition control system is a typical nonlinear, time-variant system with great lag and strong interference. It is difficult to control complicated process to get nice control precision by classical control technology or modern control theory, due to rigorous mathematics model.Non optimized control method reduces the efficiency of aircondition, causes enormous loss of energy. A new method of the perfect combining the fuzzy control and the neural Networks control is proposed in this paper. The fuzzy control could generate control laws for structural fuzzy controller based on the expert experience, without the need of the accurate mathematical model of the object. The neural Networks control can adjust the fuzzy rule base online and fine tune parameters in real time to achieve satisfactory precision and good energy saving. The result of the simulation shows, in the condition of the system parameter's change or the external disturbances, these methods of improvement show high precision and obviously lower energy expenses.