Improved harmonic loss – History gated unit recycling for online state of charge and state of energy co-estimation of lithium-ion batteries for large-scale energy storage stations

Accurate estimation of the state of charge (SOC) and state of energy (SOE) of a battery is critical for battery system management to improve the reliability and safety of battery operation. Since both SOC and SOE are cumulative quantities of time, historical factors greatly influence them. Therefore, this paper aims to propose a research method based on harmonic loss-history gated cell cycling (HL-HGUR) to obtain accurate estimation results. The algorithm is based on the gated recurrent unit (GRU) and takes into account the influence of historical factors, while the loss function is improved by combining the basic forgetting curve to improve the estimation accuracy of SOC and SOE. To further validate the estimation performance of the algorithm, this study uses the Beijing Bus Stress Test Environment (BBDST) and the Dynamic Stress Test Environment (DST) to confirm the efficacy of the method. Under these two conditions, the HL-HGUR algorithm achieves a maximum MAE of 1.29 % and a maximum RMSE of 1.49 % for SOC, and a maximum MAE of 1.09 % and a maximum RMSE of 1.25 % for SOE, which are all better than the traditional LSTM, GRU, and HGRU algorithms. Therefore, the HL-HGUR estimation method can realize the accurate estimation of SOC and SOE, and effectively improve the estimation efficiency of large-scale energy storage lithium-ion batteries.