Estimation of the state of charge of Ni-MH battery pack based on artificial neural network

摘要

To track the state of charge (SOC) of Ni-MH battery pack at the hybrid electric vehicle, an artificial neural network (ANN) is designed. Current, voltage and the previous SOC are used to inputs of ANN, and output is SOC. The result show that, this artificial neural network can track the state of charge (SOC) of the batteries accurately, in the average tracking error less than 5%; the ANN is in low dependence on the initial SOC, and the output can be achieved target value only in 90 seconds.