With the development and application of electric vehicles, the significance of power battery SOC estimation is becoming more and more important. In order to improve the accuracy of SOC estimation, this paper proposes to use the backtracking search algorithm to improve the RBF neural network model based on the standard RBF network model. By optimizing and solving the objective function in the lithium battery model, the SOC estimation accuracy of the RBF network model is improved by finding the best objective weights and thresholds. Finally, an experimental simulation platform was built to compare and analyze the algorithm SOC estimation before and after the improvement. The experimental results prove that the improved RBF network has higher SOC estimation accuracy than the standard RBF network algorithm, and reduces the estimation error to less than 2%. Lithium-ion batteries have good estimation accuracy and have certain theoretical research significance.
