Abstract
This paper presents a new estimation approach of residual available capacity for lead acid batteries in electric vehicles (EVs). The essence of this approach is to model lead acid batteries in EVs by using a neural network (NN) with the specially defined output and the proposed inputs. The inputs are the battery surface temperature and the discharged and regenerative capacity distribution to describe the discharge current profiles of lead acid batteries during EV operations. The output is the state of available capacity (SOAC) representing battery residual available capacity (BRAC). Then, SOACs of lead-acid batteries in EVs are experimentally investigated under different EV discharge current profiles in the presence of various battery surface temperatures. The corresponding data are recorded to train and verify the NN. The results indicate that the proposed NN can provide accurate and effective BRAC estimation for lead-acid batteries in EVs.
