Life Cycle SOC Estimation Method Of Li-Ion Battery Based On Interactive Multiple Model

In order to reduce the environmental pollution caused by traditional fossil energy,electric vehicles powered by lithium-ion batteries have gradually replaced the internal combustion engine vehicles powered by fossil energy.The basis for the normal operation of lithium-ion batteries is accurate state estimation,but with the increase of the number of battery cycles,the degree of aging increases,and the internal parameters will change nonlinearly,resulting in high difficulty in accurately estimating the state of charge(SOC)of the battery during the whole life cycle.Therefore,accurate monitoring of SOC throughout the life cycle of lithium-ion batteries is important for the safe operation of electric vehicles.Taking the SOC estimation of the whole life cycle of lithium-ion batteries as the starting point,this thesis conducts capacity testing,pulse discharge testing and open circuit voltage(OCV)test analysis on commercial lithium iron phosphate batteries,draws the OCV-SOC curve,and performs charge-discharge tests and aging tests on lithium-ion batteries.To explore the changes of SOC and State of Health(SOH)of lithium-ion batteries under different cyclic test conditions.On the basis of analysis of the characteristics of multiple equivalent circuit models of lithium-ion batteries,the second-order RC equivalent circuit model is selected as the research model,which can well reflect the working characteristics of the battery,has high accuracy and does not require much calculation.The offline identification method based on hybrid pulse capability characteristic(HPPC)and the recursive least squares method(FFRLS)based on forgetting factor were studied respectively.Since FFRLS can solve the problem that the model parameters change with the change of battery operating conditions,the estimation accuracy of SOC can be improved.The effectiveness of FFRLS online parameter identification was verified by dynamic stress test(DST)condition.In order to improve the robustness and accuracy of SOC estimation,an improved SOC estimation algorithm-Strong Tracking Kalman Filter(STKF)algorithm is proposed.Compared with the Extended Kalman Filter,STKF is more robust when dealing with uncertain models.It has a good tracking effect on slowly changing and suddenly changing state quantities,and is more suitable as a filtering algorithm for battery SOC estimation.In order to solve the problem of SOC estimation accuracy degradation in lithium-ion batteries under the whole life cycle conditions,a joint SOC and SOH estimation algorithm is constructed in combination with the interactive multiple model(IMM).In IMM,the Markov transfer probability matrix controls the transformation and information interaction between each aging model,calculates the model probability of matching the current battery and each model according to the residuals,and then fuses the state estimate output of each model and the corresponding model probability to realize the real-time estimation of SOC and SOH of lithium-ion batteries.The IMM-STKF algorithm is experimentally verified by using multiple sets of randomly aged battery data obtained from the test.Under the static cycle test,the maximum error of SOC is estimated to not exceed2.07%,and the maximum usable capacity error does not exceed 3.4%.Under the DST test,the maximum error of SOC is estimated to not exceed 1.98%,and the maximum usable capacity error is not more than 2.2%.The results show that the IMM-STKF algorithm can well achieve accurate tracking of lithium-ion battery SOC and real-time estimation of battery capacity,and realize SOC estimation in the whole life cycle of lithium-ion batteries.