SOC Estimation And Charging Optimization Of Power Li-ion Battery Based On Improved PNGV Model

Recent years, the LiFePO4power li-ion battery becomes a hot topic in the fieldof electric vehicles. However, the development and application of power li-ionbattery is not perfect enough, and the phenomenon restricts the popularization ofelectric vehicles. How to improve the safety and the efficiency of power li-ionbatteries has become an urgent problem in electric vehicles.In my paper, the equivalent circuit model is used to model the power li-ionbattery, and the PNGV equivalent circuit model is selected to simulate thecharacteristics of the battery. According to the battery characteristics and experimentresults, the PNGV model has been improved. The improved PNGV model reflectsthe dynamic characteristics of li-ion power batteries better, which is used in thesetwo filed in this article: SOC estimation and quick charging.It is difficult to estimate SOC accurately by using traditional algorithm. As aresult, the battery management system (BMS) can’t works as better as we need. Inorder to solve this problem, the improved PNGV model is chosen. After identifyingthe parameters of the chosen model, we use extended Kalman filter (EKF) andparticle filter (PF) algorithm to estimate SOC. The accuracy of SOC estimation hasbeen improved obviously by taking the algorithm and the model into consideration.Power li-ion battery quick charging takes an important role in the field ofelectric vehicle, and it is more convenient when using quick and deep charging inelectric vehicles. The research of quick charging lithium battery is still in a fledgingperiod, and specific principles and application rules have not been researched indetail, though various types of fast charging scheme have been proposed. Aim at thisproblem, the proved PNGV model and its parameters have been used to build thecharging math model. With the proposed charging math model and algorithm ofNSGA-II, the multi-objective problem of quick charge is solved and the front part ofPareto is obtained. Based on the specific requirement, the appropriate choice isselected from the quick charging schemes library. For example, when the chargingtime is determined, the least charging lose solution is selected. The simulationresults of quick charging is tested and verified by using the instrument of theNEWARE company and the experimental platform.