Research And Implementation Of Failure Diagnosis Expert System For Battery Pack

Much more emphasis has been placed on electric vehicle (EV), due to consideration of energy development, energy safety and environment protection. Hybrid electric vehicle (HEV) has already realized its incipient industrialization and become one of the main directions of vehicle industry.Battery application is the key technology that restricts the development of EV, of which battery management and diagnostics plays a more and more important role. With battery diagnosis system it is possible to predict battery failure earlier, thus the faultless working time of battery pack can be increased.This paper describes the characteristics of batteries, analyzes the main failures of them, introduces the modules of BMS(battery management system). Then diagnosis algorithm of battery pack failures and the structure of failure diagnosis expert system are presented. At last , some diagnostic results are gived.The solution to membership degree of symptoms is the key algorithm of batteries failures diagnosis. It is considered that the cells which performance is differ from other cells of the battery exist failures in the same condition. So the difference of performance of the cells is taked into account for estimating membership degree of symptoms. Then membership degree of the failures is evaluated by membership degrees of the symptoms. At last, DOH(degree of healthy) of the cells was estimate by membership degrees of the failures.The algorithm of on-line fuzzy diagnosis estimate DOH of the cells by using the rules of battery diagnosis and data of the cells. Some information of maintenance and DOH of the cells are outputed by expert system of the faults fuzzy diagnosis. The experiments result shows that the diagnosis results of expert system is consistent with the fact, the percentage of accuracy of explosive faults diagnosis is close to 100% and the system satisfies all requirements of task.