Estimation of the State of Charge of the Battery Based on Driving Cycles Discriminant

It is the key technical parameter for the battery management system in electric vehicles to estimate the state of charge (SOC) of batteries. It is difficult to establish an accurate mathematical model due to the influence of characteristic of monomer battery, consistency of batteries, and balance control technology. First, the driving cycles of the vehicle are classified by the Bayes classification method; secondly, the SOC prediction model of multi-scale support vector machine based on the driving cycle discrimination is constructed. According to the statistical characteristics of different driving cycles, the model parameters are optimized by Levenberg–Marquardt algorithm to improve the prediction accuracy of SOC. Finally, the rationality and practicability of the proposed method are verified through simulation and analysis.