Resilience analysis of highway network under rainfall using a data-driven percolation theory-based method

This paper proposes a data-driven approach using percolation theory to analyze the resilience of highway networks under rainfall conditions. The proposed approach's main contribution is integrating real-world traffic data with percolation theory to evaluate the impact of rainfall on traffic flow and identify the critical links of highway networks. The resilience indicators, accounting for network topology and functionality, were formulated. To calculate these indicators under various rainfall intensities, the traffic flow fundamental diagrams were established using empirical rainfall and traffic data, and a probabilistic rainfall simulation model was developed. A case study of the East Midlands, UK highway network under a heavy rainfall event on September 27, 2019, validated the approach's feasibility. Furthermore, control experiments showed that the critical links identified by the proposed method enhance highway network resilience more effectively than traditional methods, thus validated the novelty of our approach.