Study On On-line Contamination Monitoring Method Of Electric Vehicle Batteries

For the time being, more and more countries join the development of alternative energy vehicle in order to settle global environment change and energy short situation, one of those developments is electrical vehicle which have significant impact. Based upon certain research, EV car will be long-term solution which will replace oil energy by electrical power. The hot topics include-powertrain system for EV car, battery charge technology, power station surveillance system and power grid harmonic analysis during battery charge. All above mentioned topics are related to EV car battery. While main technology for EV energy is electrochemistry, which are represented by plumbic acid, nickel-cadmium, sodium sulfur, nickel-metal hydride and lithium ion batteries. These kinds of technology have disadvantage in terms of specific energy, specific power, charge, life-time, safety and cost, hence globally these development are paid more attention to and already received break-through recently.Lithium-ion battery is widely used in EV industry, but from specific cases worldly, Lion battery is the main cause for certain traffic accident and vehicle failure. From that aspect, battery safety research is critical while those study are related to system analysis not only include individual battery cell safety but also need to consider overall assembly process for battery pack. These can be only controlled during production process stage but is hard to monitor during battery working status, although we already have BMS (battery management system), it can only collect single sample data from testing stage without continuity of detection. Hence, main direction of EV car battery on line surveillance and failure diagnose research is for battery application field, for example, BMS development to monitor real-time data.Nowadays on line battery surveillance technology is not well developed, this study is based on various of previous research and decide to choose peak of voltage, virtual value of the leakage current, frequency of leakage voltage, peak temperature and battery insulator for main factor and use fuzzy logic to build up subordinate function for determination. Based on such calculation and analysis,1st level result can present EV car battery real time status and 2nd level data will be the reference whether it’s needed for battery maintenance. We pick three models and calculation results for validation, but the fuzzy logistic matrix is created upon previous experience data with limited extension. This study improve this fuzzy logistic matrix applying fuzzy neural network to enhance accuracy of on line battery surveillance system.