Research On State Of Charge Estimation Of Li-Ion Battery Used In Electric Vehicles Based On Self-Discharge Technology

Owning to the characteristics such as large specific energy, long cycle life, good safety, small self-discharge rate and fast charge, Lithium iron phosphate battery has been regarded as one main source of power for new energy vehicles and is widely used. State of Charge of Lithium battery (SOC) is one key indicator to measure the remaining battery power, but complex application environment of battery pack and inconsistent degree of deterioration of battery pack cause that the accuracy of SOC estimation of battery pack can’t met the needs of practical applications. Therefore, the SOC estimation methods to be able to reflect the degree of deterioration of the lithium battery have become one key technology of online monitoring of performance of battery. In this paper, a targeted research is conducted on the impact of self-discharge current, AC resistance, the degree of deterioration and the operating temperature and other factors of lithium battery on the estimation of SOC, EKF (extended Kalman filter) and SCKF-STF (cubature Kalman filter based on Strong tracking filter) SOC estimation strategy and uncertainty of battery equivalent circuit mode, including the following aspects:Based on the introduction of SOC influencing factors of lithium battery pack, make a comprehensive and detailed analysis on the relationship between the electrochemical characteristics of lithium battery such as self-discharge current, AC internal resistance, discharge depth and life and SOC; on the basis of a large number of experiments, adopt the equivalent circuit model being able to reflect the electrochemical characteristics of lithium battery as the SOC estimation model to propose the method of correction model of self-discharge current, temperature and degree of deterioration..For the shortage of SOC estimation accuracy caused by SOC attenuation which is the result of self-discharge and other factors, analyze the existing SOC estimation methods and propose one EKF SOC estimation strategy. The strategy corrects the influence of shelving, aging, temperature and other conditions of battery on SOC estimation, gives the realization of EKF SOC estimation, and carries out experiments at room temperature and low temperature and shelving experiment. The experimental results show that the EKF estimation method of model correction is able to meet the accuracy requirements of SOC estimation and greatly improve the adaptability of degree of deterioration, self-discharge rate and operating temperature of battery.For the actual discharge requirements of battery pack and uncertainty of equivalent circuit model of the battery at low temperature, conduct an experiment, analyze the influence brought by the uncertainty of model in accordance with the experimental data, propose the SCKF-STF SOC estimation strategy, carry out an experimental verification of SOC estimation at low temperature when the battery voltage is below discharge cut-off voltage at normal circumstances. The comparison with EKF algorithm shows that SCKF-STF algorithm is able to effectively resolve uncertainty of equivalent circuit model brought about by SOC estimation error and eliminate the impact of model uncertainty.In order to solve life decrease of battery pack, study the grouping technology of lithium battery, explores the impact of the electrochemical properties of the battery on the battery life of battery pack, and considering the physical characteristics and electrochemical characteristics of the battery, propose a fuzzy C-means clustering (FCM) support vector machine (SVM) classification method of battery. The experimental results show that the same class battery got by this classification method has better capacity fading consistency and electrochemical consistency.For low SOC estimation accuracy and poor real-time of lithium battery under actual working conditions, in combination of the characteristics of Battery Management System(referred to as BMS), set the minimum sampling point of BMS as the basic cell unit, study the charge and discharge characteristics of the basic cells of battery pack, establish a mathematical model based on battery cell, and adopt the SCKF-STF algorithm to conduct experiments under stimulated working conditions on basic battery cells. The experimental results show that for the frequent transform of charge and discharge operations of battery pack, the SOC estimation strategy is able to meet the SOC estimation accuracy requirements of lithium battery pack for electric vehicles. And on this basis, a CAN bus based battery management system is developed to propose the overall function of management battery of HEV lithium battery and programs involving hardware and software.