The estimation of the capacity of lead-acid storage battery using artificial neural networks

The capacity of lead-acid storage battery for communication system has been long estimated by constant current discharge method in the past. It spends a lot of time and labor and wastes more energy. This paper proposes a new method combining the measured data of battery discharge and the back-propagation neural network. After they are trained and learned, the back-propagation neural network can estimate the capacity of lead-acid storage battery after half hour discharge test. Therefore, the advantages of this paper are less discharge time of storage battery, less working hour and saving energy. The practical results show that the method has good performances.