基于锂离子电池容量增量曲线半峰面积的容量在线估计方法

The development of electric vehicles is my country's national strategy and a powerful starting point to achieve the goal of "carbon neutrality and carbon peaking". The rapid growth of electric vehicles has led to the rapid development of the lithium-ion battery industry. Lithium-ion batteries (LiBs) are widely used in electric vehicles and energy storage fields because of their advantages such as high voltage, long cycle life, environmental friendliness, and high energy and power density. However, LiBs decay immediately after use, and capacity fading leads to performance degradation. For example, safety and reliability gradually decline with aging. Accurate estimation of battery capacity is the basis for efficient and safe management of the entire life cycle of LiBs. However, it is not feasible to conduct real-time capacity calibration testing of full charge and discharge of the battery system in practical applications. Therefore, the research on capacity estimation algorithms has become the key to battery management systems. To this end, this manuscriptproposes a battery capacity estimation method that uses a half-peak area as a health factor. The aging characteristic interval is determined by analyzing the capacity decline mode, and the half-peak area of the charging capacity increment curve is calculated using the recursive update method, and limited calculation is implemented on the vehicle side. Online extraction of health factors under storage resource conditions enables universal capacity estimation under a wide temperature range and multi-rate conditions. First, a lithium-ion power battery test platform was built to conduct capacity calibration tests, open circuit voltage tests, and accelerated aging tests on the five commercial LiFePO₄ batteries. Ⅰn this way, we can obtain battery multi-temperature and multi-rate characteristic data, which are helpful for the development of lithium-ion power battery maximum available capacity estimation algorithm and algorithm effect verification.Secondly, by analyzing the change of the characteristic peak of the incremental capacity curve with the number of aging cycles, it was found that there is almost no loss of active material in the positive and negative electrodes during the aging process of these batteries. Ⅰn contrast, the loss of lithium ions continues as the number of cycles increases. Ⅰt is concluded that the main factor affecting the capacity fading of these LiFePO₄ batteries is the loss of lithium ions rather than the loss of active materials. The loss of lithium ions corresponds to the characteristic peak2. Finally, the half-peak area of the characteristic peak of the incremental capacity curve is extracted as the health factor, and an online estimation algorithm for lithium-ion power battery capacity is developed. Considering the influence of ambient temperature and charging current rate on the calculation of half-peak area, temperature, and rate correction functions are proposed to correct the health factor extracted online to achieve versatility under a wide temperature range and multiple rate conditions. Ⅰt was verified based on battery data at different temperatures and rates, and the results showed that the maximum absolute error in capacity estimation was less than 3%. Considering engineering applications, the maximum available capacity estimation algorithm of lithium-ion batteries proposed in this article can online extract the characteristic peak half-peak area of the incremental capacity curve to implement capacity estimation. Ⅰt has low requirements for the computing and storage capabilities of hardware equipment. Ⅰt is expected to be used in new energy vehicles and the edge side of grid energy storage power stations.