CRITICAL LINK DETECTION AND RANKING IN DEGRADABLE TRANSPORTATION NETWORKS

Critical component analysis for a degradable transportation network presents numerous additional challenges for capturing the impacts of uncertain events. In the state-of-practice, the implementation of critical component analysis in transportation planning usually employs scenario analysis techniques which theoretically assumes links have a binary status (fail or not). This study proposes an analytical methodology and an efficient solution procedure to evaluate link criticality considering impacts of uncertain events via capacity uncertainty at the network-scale. The road capacity is considered as a random variable with certain distribution based on day-to-day roadway traffic conditions. Sample-Average Approximation (SAA) is employed to generate plausible realizations of link capacity values and then solve the resulting stochastic model. The results are analyzed using several statistical measures including rank correlation coefficients (RCCs), standardized rank regression coefficients (SRRCs) and partial rank correlation coefficients (PRCCs). Finally, a case study based on a portion of New Jersey roadway network is presented.