CSI Researchers have developed a groundbreaking surrogate model to identify critical bridges in traffic networks during earthquakes, significantly accelerating seismic risk assessments. The model eliminates the need for time-intensive traffic simulations by using Markov random walk and random forest techniques, reducing computational time by 98%. Tested on the San Francisco Bay Area road network, the approach also introduces a new ranking method to prioritize bridge upgrades, ensuring efficient resource allocation and improved network resilience with minimal data requirements.
Figure 1. Traffic zones and road network of the Bay Area in California, USA.
Surrogate modeling for identifying critical bridges in traffic networks under earthquake conditions
Pengshun Li, Ziqi Wang, Bingyu Zhao, Tracy Becker, Kenichi Soga