Research On Uniformity Of Lithium Ion Battery

In lithium-ion battery pack applications, early failure even safety problems maybe occurred because of ununiformity of single battery characters and performances. Therefore how to classify single batteries in order to improve their uniformity and reduce the character changes of batteries in cycling have become a important issue in lithium battery application. But the existing classifying method is not inefficient and criterion of classifying is not in reason. The present prediction method is unpractical. From this point of view, the source of ununiformity, classifying method and predict model of battery capacity fade were presented in this paper.Various influence factors of uniformity were investigated, especially the temperature was studied in detail. Rate of capacity fade, charge and discharge performance, component and modality of electrode surface related to temperature were analyzed. The result shows that temperature is a critical factor to influence the uniformity of batteries.The uniformity of the single battery usually causes the imbalance in battery pack, then initial capacity loss and early battery pack failure was occurred. Therefore it is necessary to check out the uniformity of batteries and classify batteries according to the measurement result before batteries are assembled into battery pack. In common sense charge and discharge curve is the best reflection of battery character, the batteries with different character and performance have different charge and discharge curve. It's a reasonable classifying parameter. With the help of software Statistical Analysis System(SAS), classifying is accurate and efficient. After classifying the performance of battery pack was improved. The initial capacity loss and early battery pack failure were restrained. The charge and discharge performance of battery pack is stable.According to the data collected by Battery Management System (BMS), using the time series method, the predict model of battery capacity fade was established. This model can predict capacity fade of battery cycled at different current and temperature. The predict error of short-term prediction of this method was less than 3%.