Abstract
With the continuous development of technology and industrial automation, the using of industrial and comfortable central air-conditioning increased quickly. The refrigeration system get more and more complicated than before. As a result, the latency trouble points increased, it demands more ability in the system’s trouble shooting. At present, the trouble shooting by manual work is no longer met the actual needs. so the computer aided trouble shooting technology gets more and more pop and gets much economic value by its widely used.
In this paper, the method through neural network and expert system approach to fault diagnosis of air-conditioning technology is studied. In the neural network approach for fuzzy information processing in the neural network method, the basic fuzzy neurons and fuzzy neural network model for a brief introduction. And then construct a fuzzy neural network based on the central air-conditioning air processor fault diagnosis model. The fuzzy neural network a total of 4 layers of neurons in each layer to determine a detailed introduction, and given the phenomenon of category 5 fault membership function. Finally, the MATLAB simulation of fuzzy neural networks. Expert system approach to analysis of the compression refrigeration cycle principle, an analysis of operating characteristics of the refrigerator, cold water drawn from the theoretical unit of the performance characteristics of common faults. On the structure and characteristics of expert system is introduced. And then the depth of theoretical knowledge that has been done on ways to obtain and build the knowledge base and inference control and management have done that way.
Keywords: Centralair-eondition,ANN,Expert system,Failurediagnose