The Vehicle’s Velocity Prediction Methods Based on RNN and LSTM Neural Network

Abstract

The electrification of fuel vehicles has become a trend all around the world. For HEV and PHEV, the prediction of the future velocity affects the energy distribution and energy management directly. The vehicle's velocity prediction models based on Recurrent Neural Networks (RNN) and Long Short Term Memory (LSTM) Neural Network have been established in this study, to predict the future vehicle velocity. Through the test data validation, both models can improve the accuracy of vehicle velocity prediction. But as the prediction length increases, the root mean square error (RMSE) of the LSTM prediction model is smaller than the RNN model, which means the LSTM prediction model has better performance in vehicle velocity prediction.