Short term power demand forecasting in light- and heavy-duty electric vehicles through linear prediction method

In this paper a novel method based on linear prediction technique is proposed for short term power demand forecasting in light and heavy-duty electric vehicles for improvement in the overall efficiency of the vehicle. The paper also utilizes filtering of unnecessary information which would have been a major bottleneck in improving the method's accuracy. The predicted demand function is fed to a wavelet function, which apportions the share between the battery and the ultracapacitor of the considered energy management system. The proposed method is validated with empirical power demand data obtained from on road tests of both light and heavy-duty electric vehicles through numerical investigations.