A novel state of charge estimation method of lithium‐ion batteries based on the <scp>IWOA‐AdaBoost‐Elman</scp> algorithm

Abstract

Lithium-ion (Li-ion) battery is a very complex nonlinear system. The data-driven state of charge (SOC) estimation method of Li-ion battery avoids complex equivalent circuit modeling and parameter identification, which can describe the nonlinearity of the battery more directly and accurately. To address the problems of low generalization ability, local miniaturization, low prediction accuracy, and insufficient dynamics in the prediction process of a single feedforward neural network, an IWOA-AdaBoost-Elman algorithm-based SOC estimation method for Li-ion batteries is proposed. The method introduces an improved whale optimization algorithm (IWOA) to continuously optimize the nonlinear weights of the Elman neural network during the iterative process. Using the AdaBoost algorithm, multiple weak IWOA-Elman predictors are recombined into one strong SOC estimator by successive iterations. The combined strong predictor has strong generalization ability, estimation accuracy, and dynamic characteristics. To verify the rationality of the model, the SOC estimation is performed under dynamic operating conditions. The experimental results show that the proposed method is more accurate and stable compared with other optimization models. In addition, the proposed method can overcome the effects of different discharge multipliers, different ambient temperatures, and different aging cycles on SOC estimation. Both theoretical and experimental results show that the IWOA-AdaBoost-Elman algorithm provides a new way for the SOC estimation of Li-ion batteries.