A Layered Neural Network Competitive Algorithm for Short-Term Traffic Forecasting

A reliable and accurate short-term traffic forecasting system is crucial for the successful deployment of any intelligent transportation system. To address the complexity of real-world traffic forecasting conditions, this paper presents a layered traffic forecasting algorithm, which is implemented by a clustering neural network, Kohonen self-organizing map (KSOM) and four neural network paradigms. In system training stage, KSOM is first trained and tested using historical traffic data to obtain an optimal forecasting scheme. In system online operation stage, real-time traffic forecasting is made according to the system optimal forecasting scheme. Case studies are carried out using real-time traffic data. The obtained results demonstrated the superiorities of the proposed algorithm to existing forecasting models.