Realization Of Air Conditioner Fault Diagnosis Expert System

With the improvement of living level, air conditioning has already been into common families step by step. But when people enjoy the comfortable inhabitation, as time goes, aging of the equipment and other causations, air conditioning may inevitably have some trouble. Nowadays, servicing of air conditioning has mostly depended on the sensorium and experience of maintenance man. So maintenance man's intelligence and experience may have influence on the quality of servicing. If the trouble can not been obviated in time, maintains cost will increase and work efficiency will decrease. To cultivate an eligible maintenance man, much theory and practice is necessary. What's more, cost is high. Thus it is necessary to develop an expert system, which provides on-line monitor and fault diagnosis, meanwhile maintain decision-making.The thesis takes air conditioning as research subject, aims at its working condition and discusses about the monitor technology. It analyses common trouble of air conditioning and the obviations. Additionally, architecture, features, knowledge base design of diagnosis expert system are discussed in this thesis, and on base of these works, we develop a diagnosis expert system about air conditioning.Firstly, the thesis represents expert system and introduces its structure and character. Secondly, it probes into on-line monitor technology and analyses our system architecture. The signals have been classified into two types, that is, slow variation signal and fast variation signal. Thirdly, the thesis treats representation and obtaining of deep and shallow knowledge, discusses the design and management of knowledge base in detail. Finally, after analyzing lots of air conditioning faults we develop a fault diagnosis expert system including knowledge base design of production rule, uncertainty inference engine, explanation mechanism, which realizes fault diagnosis and maintain consultation.In the end, the thesis summarizes all the research work and points out the aspects of the fault diagnosis expert system need to be improved. In addition, the future tendencies for such an expert system are forecast.