Research On Soc Estimation And Cycle Life Prediction Methods For Lithium Ion Battery

Lithium-ion batteries are widely used in various fields due to their advantages such as high output voltage,high energy density,long cycle life,low self-discharge rate and good environmental protection.With the development of lithium-ion batteries,there have been higher standards for battery reliability and service life.As the important parameters in the Battery Management System(BMS),the State of Charge(SOC)and cycle life of the battery have a certain relationship with each other,which has a direct impact on the working performance of BMS,and objectively reflects,for example,the driver's grasp of the battery status and driving experience.Based on existing research,this paper studies these two key technical issues and uses the combination of improvements and algorithms to further explore.This paper starts with the application of lithium-ion batteries in the electric vehicle industry,the development status of the electric vehicle industry and the battery management system,and briefly reviews the current research status of battery SOC and cycle life at home and abroad.In this paper,lithium iron phosphate battery is selected as the main research object,and a series of related experiments are designed for the battery through an experimental platform independently established in the laboratory.Based on the experimental data,the key factors affecting the charge state and cycle life of the battery are analyzed in depth.Then,for the estimation of battery SOC,an algorithm combining improved circuit model and extended Kalman filtering is selected for real-time estimation.Open-circuit voltage and Hybrid Pulse Power Characteristic(HPPC)experiments were used to identify the battery circuit model parameters,the experimental platform was used to simulate the Dynamic Stress Test(DST)and Federal Urban Driving Schedule(FUDS)test data of the actual operating conditions of electric vehicles,and the state space model was recognized?verified.The simulation results show that the error under both simulation conditions is within 1.1%,indicating that the algorithm has high accuracy and feasibility.Finally,the cycle life of the lithium-ion battery is predicted.This paper specifies the end of life when the battery capacity decays to 80% of the standard capacity,and the number of cycles from the initial capacity to the end of life is used as the research data object.A method based on improved particle swarm algorithm and generalized regression neural network was proposed to predict the cycle life,Adding improved learning factors and inertial search weights to particle swarm optimization.The public data of lithium ion batteries of National Aeronautics and Space Administration(NASA)?Massachu-setts Institute of Technology(MIT)and the cycle data measured by an independently built laboratory platform were selected to verify the algorithm Precision.