Enhancing Urban Traffic Efficiency Through Traffic Flow Prediction using Long Short-Term Memory Neural Networks

Abstract

Traffic congestion is a widespread issue affecting urban areas worldwide, leading to significant economic and environmental costs. Predicting traffic flow accurately is crucial for effective traffic management and planning. This study aims to develop a robust traffic flow prediction model that leverages the capabilities of Long Short-Term Memory (LSTM) neural networks in handling time series data. Suggested models were trained and tested on a comprehensive dataset, which included various traffic parameters provided by The Luxembourg administration of Ponts et Chaussées. The models achieved high accuracy in forecasting the average speed and flow rate in a studied location. So, the outputs can be used in an assistance system to help humane operators adjust traffic signal timings based on the predicted traffic conditions, reducing congestion and improving flow.