Battery state of charge estimation for electric vehicle based on neural network

Prediction of battery´s remaining capacity is always a significant issue to which electric vehicle researchers paid close attention. Batteries of different types or the same type batteries of different model varies in prediction model of remaining capacity, the expert´s advice obtained from experiment is not so universal that it is significant to build and improve the prediction model of remaining capacity for batteries of different types. This article takes iron phosphate Li-ion battery as the object of study, based on charge-discharge performance test of iron phosphate Li-ion battery, introduces neural network method to build prediction model for remaining capacity of battery and verify the model with test data in the end.