Research on Prediction of Electric Taxi Travel Patterns Based on Deep Learning

With the popularization of electric vehicles, electric taxis, as an important part of residents’ transportation, and their travel behavior characteristics can lead to traffic congestion, unbalanced utilization of charging facilities, and a series of other problems. This paper processes the relevant data obtained and builds a CNN-LSTM-Attention prediction model based on PSA optimization. PSA is used to optimize the parameters in the CNN-LSTM-Attention model, and the performance of this model is compared with three traditional models, showing that this model has better prediction performance. Finally, this model is used to predict the travel patterns of electric taxis, and the travel behavior characteristics of electric taxis under two travel patterns, namely working days and holidays, are obtained.