AI-based vehicular traffic flow forecast for advanced transportation systems

Abstract

For traffic control systems, the intricate nature and quantity of vehicular traffic have recently risen significantly. For transportation systems to be optimized and overall efficiency to be increased, accurate prediction of traffic is essential. To anticipate vehicular traffic flow, we introduce a unique reptile-ant optimized bidirectional gated recurrent unit (RAO-BiGRU) method in this study. Ant colony optimization (ACO) and reptile search optimization (RSO) methods are combined in the RAO. The Bi-GRU model's performance is improved by using the RAO to better choose the input features. Extensive experiments are run utilizing the traffic data database to assess the efficacy of the suggested approach. The outcomes show that in terms of predicting accuracy and computational effectiveness, the suggested RAO-BiGRU strategy performs better than the conventional forecasting strategies.