VSPNet: A vehicle speed prediction model incorporating transformer and BiLSTM

Abstract In recent years, with the increasing adoption of hybrid vehicles, energy management strategies 
have become a prominent research focus. Accurate Vehicle Speed Prediction (VSP) is a critical 
prerequisite for achieving optimal results in predictive energy management strategies. However, 
existing speed prediction algorithms fail to fully leverage vehicle data to enhance prediction 
accuracy. Therefore, a novel Vehicle Speed Prediction Net (VSPNet) is proposed in this study. 
Firstly, we constructed a combined cycle condition for model training through comprehensive 
analysis and analysed the vehicle feature parameters through the Random Forest (RF) algorithm 
and Pearson correlation analysis to select the best input feature parameters. Then a VSPNet 
speed prediction model is proposed based on the Transformer model. In the encoder part, firstly, 
by assigning weights to the input feature parameters and incorporating the temporal attention 
mechanism, the model is made to make better use of the input features from two dimensions, 
and at the same time the Transformer model's encoder based on positional coding combined 
with Bi-directional Long Short-Term Memory (BiLSTM) belonging to Recurrent Neural 
Networks(RNN), which is used as a decoder to better catch and handle long-term dependencies 
in sequence data. Finally, a comparative experiment between VSPNet and the classical speed 
prediction models was carried out. The proposed VSPNet model reduces the RMSE by 37%, 
22%, and 20% and MAE by 39%, 25, and 24% compared to the LSTM model for the prediction 
time horizons of 3s, 5s, and 8s. The RMSE is reduced by 47%, 28%, and 7%, and the MAE is 
reduced by 47%, 30, and 9% compared to the Transformer model for the prediction time 
horizons of 3s, 5s, and 8s. The experimental results demonstrate the superiority of this speed 
prediction model.