The Development of Dynamic Travel Time Prediction Models for South Jersey Real-Time Motorist Information System

摘要

This study develops a dynamic travel time prediction model for the South Jersey Motorist Real-time Information System. In order to develop an effective and sound predictive system for travel times, the integration of traffic flow theory, management and application of collected data are considered to construct the proposed model, which can generate reliable and accurate prediction results with limited traffic information. A new approach for predicting travel time is proposed considering both the real-time and historical traffic conditions. In this study, a number of sensors are installed at congested places to monitor traffic congestions. A calibrated simulation model is developed to emulate traffic operations for evaluating prediction accuracy. With the travel times collected from the sensors and simulation model, the Kalman filter model is applied to forecast the travel times over different time periods. The results show that the proposed method demonstrates a satisfactory performance.