Life Cycle Online Management And State Estimation Of Vehicle Power Battery Pack

Lithium-ion batteries are gradually becoming dominant in electric vehicles energy storage systems due to their advantage,such as high energy densities,long lifespan.The inconsistency of the power battery continues to deterioration due to environmental factors,dynamic conditions and internal characteristics during the entire life cycle of the battery.To guarantee safe and reliable battery operation,a battery management system?BMS?is required to monitor and control lithium-ion battery which provides a longer lifetime of the battery.One key technology of the BMS is to estimate the state-of-charge?SOC?of the battery.Accurate SOC estimation can be used for preventing over-charge and over-discharge operations of the battery,thus avoiding the harm cause to the battery.This paper aims at a type of vehicle LiFePO₄ battery,the battery management system of life cycle is researched and designed.The main works are as follows:The basic working principle and performance indicators of LiFePO₄ battery are deeply analyzed,and the dynamic charge and discharge characteristics of LiFePO₄battery are tested based on the experimental equipment.According to the defects of the traditional SOC definition method,the corrected SOC definition is given by analyzing relevant factors that affect battery state estimation.The accurate SOC estimation depends on the established battery model,a second-order RC equivalent battery model is established based on the traditional equivalent circuit model that accurately reflects internal characteristics.Then offline and online method are adopted to identify resistor and capacitor model parameters in an equivalent circuit model,and the result shows that the online identification method has higher identification accuracy.The main idea of state estimation needs to combine the open-circuit voltage method to feedback the voltage parameters.This paper uses polynomial fitting to fit the relationship between the open-circuit voltage and the SOC curve of the battery.The basic characteristics of unscented kalman filter?UKF?and particle filter?PF?are analyzed in SOC estimation,and use complementary advantages to propose an unscented particle filter?UPF?algorithm with high estimation accuracy and fast convergence speed.The superiority of the joint estimation algorithm is verified by MATLAB simulation and contrast analysis,and simulation results show that the estimation accuracy is less than 2.5%,the convergence time is less than 200s under dynamic conditions.To improve the portability of the battery management system and meet the requirements of different power level,a distributed architecture of master-slave is proposed.The control strategy consists of some slave module for gathering information and achieving cell balance,a master module for processing and analyzing data.This paper has completed the design of the main control board,voltage/current acquisition from the control board,inspection and balancing slave control board,and combined with the study of the estimation algorithm in this paper,the management system software has designed.Through experimental comparison data of different working conditions,the feasibility of the battery management system and accuracy of SOC estimation are verified.