Sensitivity Analysis-Based Multi-Modal Transportation Network Vulnerability Assessment with Weibit Choice Models

Abstract

This study proposes a sensitivity analysis-based multi-modal network vulnerability analysis through the variation in utility-based accessibility measures associated with mode and route choice dimensions. The weibit expected travel disutility is used to assess network accessibility, which is the composite travel cost derived from the weibit travel choice model. Benefiting from the properties of weibit model, the proposed accessibility measures can accurately reflect proportional variation in network performance and account for heterogenous network scales. The equilibrium mode and route choice patterns are reproduced via a weibit-based combined modal split and traffic assignment (CMSTA) model while specifically considering the mode similarity and route overlapping in travel choice modeling. The sensitivity analysis of the weibit-based CMSTA model with respect to inputs from both sides of supply and demand is conducted for vulnerability assessment and critical link identification. The sensitivity analysis-based method can reduce the computational burden of repetitively solving the CMSTA model required by commonly used enumeration- or scenario-based approaches. The proposed vulnerability measures and analysis method are demonstrated via numerical experiments on a multi-modal transportation network.