Segment Importance Ranking Approach for Traffic Networks Based on Macroscopic Fundamental Diagram

Identifying critical segments in traffic networks is vital to network robustness analysis. Previous methods are difficult to be extensively applied since localized indexes result in localized solutions and single system-wide indexes yield results depending on which index is selected. This paper proposes an approach determining seven system-wide indexes from characteristics of macroscopic fundamental diagram (MFD) to identify critical segments from multiple aspects of network performance. A scalable measure called the average relative change is applied for segment importance identification. Results from simulations under two demand levels and comparison with a single-index approach demonstrate that the proposed approach yields reasonable results from a comprehensive view and involves network performance in different traffic states. The limitation of single-index approaches can be avoided by this approach. In addition, segments can be classified in different importance levels based on the changes in the shape of the MFD, which is significant to network control.