Forecasting For Power Battery Remaining Capacity And Driving Range Of Electric Vehicle

With the continuous consumption of energy and increasing environmentalpollution, the development of new energy vehicles, especially pure electric vehiclewith zero emissions is necessary. In recent years governments and enterprises areactively promoting the development of electric vehicles. The battery managementsystem and the prediction of remaining battery capacity is a core technology inelectric vehicles (EV).The topic is under the background of "863" joint commitment by our schooland an Automobile Company cooperation, forecasting for battery remainingcapacity and driving ranges of an electric vehicle. Firstly, on the basis of analyzingthe characteristics of power battery, do a lot of charge-discharge experiments for aparticular battery pack(including discharging under different current rate、differentSOC and cycle mode) which lay foundation for forecasting remaining capacity ofbattery pack.After anglicizing and comprising many methods of remaining battery capacityprediction, select BP neural network to predict battery remaining capacity.Combined with the battery experimental data, established the BP neural networkmodel for forecasting remaining capacity of battery. Using this model to forecastthe remaining capacity of Lithium iron Phosphate Battery pack, and then combiningwith the result of battery test bed, indicate that the prediction accuracy is in thedesigned range (5%).On the basis of theoretical analysis of EV driving ranges, using cruisesoftware to simulate driving range of battery model established with BP neuralnetwork on an EV. Through comparison with actual value, the error is withindesigned range (5%).