Optimization of fast charging method based on a reduced order electrochemical model for lithium ion batteries
Type of DegreePhD Dissertation
MetadataShow full item record
Long charging time is one of the technical barriers that should be overcvome for wide acceptance of electric vehicles (EVs) in the market. The charging time can be simply reduced using increased charging current that adversely reduces lifespan and deteriorates safety of batteries. Therefore, design of an appropriate charging protocol is a challenging issue. I propose a fast charging method designed based on a reduced order electrochemical model (ROM) considering degradation effects predominantly caused by side reaction. Different charging protocols were generated by considering different limiting factors such as surface ion concentrations, state of charge, cutoff voltage, and side reaction rate, which were tested in real time using a Battery-In-the-Loop (BIL) system. Experimental results have shown that the proposed charging method considering the limitations of side reaction rate and ion concentration yields the best performances among others, where the charging time is reduced more than 40% compared with normal charging method while the degradation is comparable. On the other hand, lithium plating is another cause for degradation, specifically in the working conditions of high charging currents and low temperatures, which is considered for design of an optimal charging method. Since the model is strongly nonlinear, a nonlinear model predictive control (NMPC) was employed to optimize the charging protocols. The objectives are to reduce charging time and at the same time minimize degradation speed. The charging time and degradation speed are traded off and optimized by varying the weighting factors. In addition, the charging protocol is constructed with not only constant currents, but also pulse discharging currents that promotes lithium stripping, so lithium ions can be recovered out of the plated lithium. Firstly, the proposed protocol was determined at a constant temperature and implemented into a real time capable BIL test station and compared with constant current constant voltage (CC/CV) charging method. Experimental results have shown that the proposed charging method significantly reduces the charging time while the cycle life is extended. Then, the effects of varying temperature on side reaction and lithium plating were considered in the optimal design of charging method. In fact, the side reaction and lithium plating rates are strongly affected by C-rates and temperatures, which is numerically analyzed using the validated degradation model. This analysis allows for determination of optimal temperatures with the longest cycling life at different C-rates. The last method was verified in the BIL system, where the temperature was controlled by a designed thermostat system to track the optimal temperatures. The experimental results have shown that the designed protocol can further reduce the capacity fade while the charging time can be kept.