SOC Estimation Based Active Balance Strategy Research For Lithium Battery

For electric vehicle and energy storage,lithium batteries are often in series and parallel connection to meet driving voltage and power demand.However,due to a variety of reasons,the problem of inconsistencies in the battery pack will be more and more severe,the inconsistencies of the battery pack will not only impair battery’s capacity,but also cause safety issues.So it is important to study the cell balancing strategey in battery pack.The widely used passive balancing technology has many disadvantages,such as low energy transfer efficiency,small balance current,as result in long balancing time.On the contrary,active balancing technology has many advantages:high efficiency and large current,and it has become very popular for battery pack during recent years.The cell voltage value in battery pack is commonly used as a reference for the cell balancing.However,the charge/discharge curve of the battery is relatively flat during 10%-90%voltage,which leads to large balancing error.Therefore,State of charege(SOC)value is used as the new reference for battery balancing in this thesis.The research work are shown as follows:1)A new SOC estimation method based on T-S fuzzy model for lithium battery is proposed.It is difficult to design an observer to estimate SOC with the battery nonlinear second-order circuit equivalent model.Thus,by using T-S fuzzy model to describe the battery nonlinear model,a robust observer is proposed to estimate the battery SOC,which has good performance and effectiveness.2)Bidirectional active equalizer between the cells is designed.Compared with passive balancing and the active balancing topology of transferring charge between adjacent cells,the bidirectional equalizer implemented by symmetric flyback circuit allows bidirectional charge flow between the single cell and the battery module,making the balancing control more flexible and more efficient.3)In order to reducing the computing,model predictive control with integer programming is used to optimize the the equalizer control and tackle the large calculation problem to get optimal control sequences.The proposed active equalizer allows multiple transformers to charge or discharge a single cell at a constant current simultaneously,which makes the control of the equalizer become an integer programming problem.The least-square method and the branch-and-bound method are used to get the optimal equalizer control sequence,which greatly reduces the calculation burden,4)Finally,the active equalizer and balancing strategy are verified in experiments,and the result of SOC estimation and the balancing method are discussed.The experiment resutls show that the equalizer and balancing strategy proposed in this thesis can achieve efficient celling balancing in lithium battery pack.