Evaluation of Arterial and Freeway Interaction for Determining the Feasibility of Traffic Diversion

Abstract

Traffic congestion is a common phenomenon in our daily lives that greatly costs society. A better understanding of the interaction between freeways and arterial streets may help traffic engineers and researchers improve the operation of existing facilities and deploy feasible traffic diversion plans to improve the usage of existing road capacity within a traffic network. This paper proposes a novel two-step approach to evaluate the interaction between freeways and arterial streets by comparing their performances. The first step is to identify freeway and arterial travel time pattern similarities via template matching techniques commonly used in computer vision. The interaction is quantified in the second step by using conditional probability theory. The result of the two-step process allows analysts to determine whether traffic diversion is possible or likely between freeways and parallel arterials. The city of Bellevue, Washington was selected as a case study site because of the availability of traffic sensor data. The results demonstrate that the analysis approach allows traffic analysts to more comprehensively observe the interaction between freeway and arterial performance by using existing data collection facilities. The quantitative interaction results provide a better understanding of diversion potential and are useful in incident response, individual route planning, and integrated corridor management. This approach can be applied to any city’s freeway-arterial network if reliable sources of travel time data are available.