Application Research of WNN Optimized by GA in VRLA Battary Degradation Prediction

This paper studies the accurate prediction problem of VRLA battery state of charge (SOC) and the remaining capability of the battery, after the comprehensive analysis of the various elements which affect the battery state of charge, we put forward a battery degradation test model based on GA - WNN, and carries on the verification test, in the meantime, we also make it contrast with other algorithms. The results of test show that the model of WNN Optimized by GA has shorter training time and high prediction accuracy, it can predict the battery remaining power more accurately.