基于时间卷积网络的锂离子电池健康状态估计

With the rapid development of electric vehicles and energy storage systems, the estimation of state of health SOH of lithium-ion batteries becomes one of the key technologies to ensure system safety and reliability. However, the nonlinearity and complexity of baUery degradation process make it diificult for traditional aasessment methods to meet the requirements of high accuracy and real-time performance. Therefore, a battery SOH estimation method based on TCN (temporal convolutional network) and multi-health feature extraction was proposed. This method first extracted eight health features related to time, energy, capacity and incremental capacity from the battery's charge-discharge data. Then, grey relational analysis was used to evaluate the correlation between each feature and SOH, and those with con-elation coefficients greater than 0. 7 were selected as inputs for the model. Finany, the TCN model was employed to capture the causal relationship between the selected features and battery Boh, enabling accurate SOH estimation of battery. The experimental results show that the mean absolute error and root mean square error of this method on four batteries are all within 2,, and the maximum absolute error is less than 0. 05, demonstrating high S0H accuracy and robustness.