The Study Of Central Air-Condition System's Fault Diagnosis Based On Fuzzy Neural Network

When a certain equipment of the central air-condition system is out of work, it is not easy to make sure the failure fleetly to the manger. On the other side, the equipment is out of work in case, the technician or expert must be invited from the foreign land. The economic and personnel burden increase consumedly. So, we must use the intellectualized failure diagnose technology in the central air-condition system to find failure estimate failure and forecast failure.This study is based on the air-condition system of a curative factory. Firstly analysis the failure phenomena of the central air-condition that it have happened, make sure five failure phenomena and eleven failure cause and the relationship among them. In this chapter, firstly introduce the basic conception and model briefly. Then up build the failure diagnoses model of BP NET and emulate in the MATLAB, from the result we know that the effect of BP NET is not approving. So, in the sixth chapter, Build a model of four layers FNN and introduce the node of every layer. Have gove the relationship function of the five kinds failure phenomena. Finally use the MATLAB to emulate the FNN.By the analysis on the fuzzy theory and the ultimatum of the ANN, establish the failure diagnose model based on the ANN and FNN. I used the same training sample and examine sample in the emulation based of the ANN and FNN, the result prove that FNN have faster convergent rate and more exact diagnose power than ANN.