Prediction of lead-acid storage battery´s remaining capacity based on LM-BP neural network

With the rapid development of renewable energy, energy storage technology is widly used in the grid system. Stable working state of storage battery is an important guarantee of reliable and safe operation for renewable energy generation connecting to the grid. The remaining capacity or state of capacity (SOC) is an important parameter reflecting the performance of storage battery. In this paper, on the base of analysing traditional predicting methods, a method of predicting the lead-acid storage battery´s remaining capacity based on LM-BP neural network was proposed. Specific model of storage battery was not necessary in this method, which avoided the complex calculation of the model´s mathematical expression. Terminal voltage and discharge current were used as the input sample of the training model, because both of them were easy to get by experiment, which simplified the operation of data acquisition and reduced the damage to the battery´s service life. The prediction methods of support vector machine (SVM) and LM-BP neural network were also contrasted in this paper. The prediction results show that in the aspect of predicting SOC of storage battery, the prediction accuracy of LM-BP neural network is higher than that of SVM and the training time is also shorter, which is more appropriate for practical application.