Online evaluation method for lithium battery capacity fading considering capacity fading disturbance and error compensation

Abstract

In response to the accurate and rapid prediction of capacity fading in lithium-ion batteries, this paper proposed an online evaluation method for lithium battery capacity fading considering capacity fading disturbance and error compensation. Firstly, considering the significant disturbance caused by environmental changes and instrument measurement errors in the measurement of influencing factors, mathematical modeling of the measurement errors of influencing factors is carried out using uncertainty methods to establish a capacity degradation disturbance model. Next, features such as the time difference of reaching the cut-off voltage, temperature peak value, and the time to reach the peak temperature are extracted as health features. A data-driven error compensation model based on convolutional neural networks is constructed to dynamically compensate for the battery capacity fading online evaluation values based on model-based methods. Finally, the proposed method was validated in the NASA PCoE and Oxford battery datasets, achieving a MAPE as low as 3.44 % under ambient temperature conditions on the NASA PCoE dataset and 1.23 % on the Oxford dataset. It also exhibited excellent performance under both high- and low-temperature conditions. These results highlight the method's robustness and wide applicability across different datasets and operating environments.