Autonomous Vehicle Traffic Rate Prediction in Dense Networks: A Deep Learning Approach
Abstract
Autonomous Vehicles (AVs) are expected to transform transportation, although their effective implementation depends on precise and prompt traffic forecasting. To tackle this difficulty, we examine the utilization of deep learning models, including Bidirectional Long Short-Term Memory (BiLSTM), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU), for predicting traffic flow in congested autonomous vehicle networks. These approaches are adept at capturing the intricate temporal interdependence and spatial correlations of traffic data. Our findings demonstrate that BiLSTM consistently outperforms other models, even under variable traffic volumes and conditions. This research propels the development of intelligent transportation systems and improves the optimization of autonomous vehicle operations.