Genetic RBF Neural Network for Estimating State-of-Charge of Lithium-Ion Batteries in a Pure Electric Vehicle

Considering the complexity of lithium-ion batteries in series and the advantages of radial basis function neural network (RBF NN) in solving nonlinear problems, RBF NN was used to set up a model of the SOC estimation and proposed a genetic RBF NN method for estimating the SOC of the lithium-ion batteries in a pure electric vehicle. The practical data obtained from the pure electric buses running during 2010 Shanghai World Expo was used to train the genetic RBF NN and to do the experiments of the SOC estimation. The experimental results show that the root mean square error of the SOC estimation is 0.002 4 and the estimation precision is improved.