Research on an Algorithm for Prediction of Lead-Acid Battery Charging Time

Mastering the charging time of lead-acid batteries is particularly important for the use and management of batteries. Considering this, the lead-acid battery charging-time prediction model based on MIV-AdaBoost algorithm is established. The input variables of the prediction model are evaluated and filtered by the Mean Influence Value (MIV) algorithm, and on this basis, a strong predictor is constructed by the AdaBoost algorithm. The prediction errors of charging time are compared between the BP neural network and the MIV-AdaBoost model. The verification of the testing samples shows that the prediction accuracy of MIV-AdaBoost model is better, and the Charging time of lead-acid battery can be predicted more accurately based on the MIV-AdaBoost model.