The Prediction of Freeway Traffic Conditions for Logistics Systems

Abstract

With a steady increase in the number of vehicles predicted, traffic congestion has become a significant logistical challenge. The increase in traffic not only results in pollution and traffic congestion, but also leads to increased travel time and productivity loss. Thus, traffic prediction has become an important research topic in the academia. In fact, logistics managers are more concerned about predicting short-term traffic conditions than the accuracy of prediction. Therefore, this study used a discrete-time Markov chain and online traffic monitoring data to predict the probability of traffic congestion and identify the freeway bottlenecks. The findings of the study revealed the high probability of National Freeway 3's northern section being non-congested during the morning and afternoon rush hours. However, several bottlenecks were found in the links to nearby urban areas. The results of this study can not only facilitate logistics managers to optimize vehicle routes but can also support transportation control centers with regulating traffic flow in freeways during peak periods.

Keywords

• Discrete-time Markov chain
• freeway traffic congestion
• logistics management
• short-term traffic prediction