Development Of Host Computer Software Of Li-ion Power Battery Testing System And Study On Classifying Algorithm Of Li-ion Batteries

As the Li-ion battery has many advantages, such as high working-voltage, large capacity, long circle-life and non memory effect, it becomes a research hotspot for novel power battery. Li-ion batteries must be packed so as to satisfy the demands of automotive field for battery performances, and the uniformity of single batteries is a key influencing factor of the performance of a battery pack, in order to improve it, batteries must be classified on producing process. The classifying platform is battery formation test system. Because lower computer mainly finishes functions of data sampling and charge-discharge execution, that of classifying mainly realized by host computer software of the equipment.At present, the host computer procedures of domestic Li-ion battery formation equipment have some disadvantages, such as single function, complicated operation, and lack of intelligent classifying function. Aiming at the above-mentioned problems, the host computer procedure of the formation and testing equipment designed by ourselves for power battery was developed by VB and the classifying method of Li-ion batteries was also studied. The main contents in the paper are as follows:(1) According to the requirements of practical application, the planning and design of host computer procedure was carried out. The bar code retrospection solution was proposed against the difficulty in retrospecting battery formation data emerged in the present detecting equipment software. The software is composed of main controlling module, data recording and processing module, battery classifying module and industrial-step editing module. The test results of the software show that it has advantages of easy and stable operation.(2) As the formation curve is the best reflection of battery character, it is employed as the parameter for classifying. After classifying eligible batteries according to their formation curves by the method of Fuzzy C-means clustering, in comparison with the battery pack classified only by capability, the uniformity of single batteries was enhanced on the storage performance, and the differences between each single battery were less; capacity fading rate of battery pack became slower apparently in 200 times cyclic process. Moreover, the charge and discharge performance of battery pack was stable; the first 100 times cyclic superposition curves of the battery pack were compact.(3) By adopting mixed programming mode with VB and MATLAB, the fuzzy clustering method independent on MATLAB program environment was realized in upper host classifying system. This solves the problems of code-compiling complexity and low operating efficiency in the program compiled solely by VB.