Study On Aging Characterization And Internal Temperature Estimation Of Electric Vehicle Power Battery

Electric vehicles have alleviated the huge pressure caused by energy shortage and exhaust pollution.However,in recent years,the spontaneous combustion of electric vehicles has occurred frequently.The reason is that individual single cells work abnormally,causing the internal temperature to rise instantaneously and start to burn.Therefore,grasping the internal temperature of each single cell accurately is of great significance for the safe driving of electric vehicles.In addition,the state of health of the power battery packs is inevitably different.In order to improve the accuracy of temperature estimation among the entire battery pack,it is necessary to study the battery aging characteristics and explore the online health state estimation method of the battery.In this paper,from the perspective of Electrochemical Impedance Spectroscopy and Incremental Capacity curve,the characterization method of the state of health of LiFePO₄ battery was qualitatively analyzed,and the method of estimating the internal temperature of the battery was also studied.Finally,the test was designed to verify the accuracy of the internal temperature estimation in the battery.Firstly,the effects of internal temperature and the state of health on the curve were investigated based on the Nyquist curve under different working conditions.The results showed that the Nyquist curve can't be used to characterize the state of health and internal temperature of the battery.Then,based on the Bode curve under different battery states,the frequency interval in which the phase shift value is least affected by the state of charge and the state of health is determined to be 10 Hz to 100 Hz.In this frequency range,the specific mapping relationship between the absolute value of the phase shift angle and the internal temperature of the battery was found.We tried to use two functions to describe the relationship between phase shift angle and the internal temperature.The test results showed that the estimation accuracy is about 2? at the lower temperature of 5?¹5?,but the accuracy error of the estimation is kept within 1? when the working temperature of the power battery is in the range of 15?⁵0?.Secondly,Cyclic Voltammetry and Incremental Capacity method were used to study the online characterization method of LiFePO₄ battery.Good correlation between the current peak on the CV curve and the state of health of battery were found.There are three types of feature points on the IC curve.The second type of feature points can be used to characterize the state of health of the battery.Then,the common relations between the feature points on the two types of curves and the state of health were analyzed in the same coordinate.Based on the theory to study the similarity characteristics of the mathematical states,the second characteristic point on the IC curve was determined as the best online characterization method of the state of health.Finally,based on the IC curve of the battery at different internal temperatures,the relationship between different feature points and temperature was explored.It was found that the corresponding voltage value of the first type of characteristic points on the curve has a good correlation with the internal temperature of the battery.The exponential function was used to describe the mapping relationship between the first type of characteristic point voltage value and the internal temperature,and the original voltage value was substituted into the fitting function.The results showed that average estimation error of four batteries can be controlled within 2? in the range of 0?⁵5?.Secondly,the mean fitting method was used to determine the average fitting coefficient,and the internal temperature of the four batteries was estimated using the unified function expression.The results showed that most of the estimated error values of the battery can be kept within 3?.