A New Algorithm of Online Monitoring and Fault Prediction for the Battery Set State

Based on the traditional floating voltage examination method of the VRLA battery, this paper proposed a new online examination method of the VRLA battery interface resistance, and established GM(1,1) forecast model of the battery interface resistance to examine the batteries´ condition and predict the fault. Because many uncertain factors impact the internal resistance, this paper give a real time method and establish a dynamic innovation forecast model of GM(1,1). The forecast precision is improved greatly through experiment, and the feature of internal resistance of battery can be forecast exactly.