Robust Performance of Transportation Networks Using Quantile Metrics

The state of the art in transport network modeling is founded on the estimation of travel demands between origins and destinations (O-D matrices) by various techniques. Assuming the estimate matches the true O-D matrix, the analysis culminates with a network assignment procedure that generates the volumes and costs associated with those volumes on the links and paths of the network. However, there are always variations from the truth in such estimates. In this study the authors develop new metrics for network assessment by taking explicit account of such demand variability and uncertainty. The metrics consist of the calculation of quantile network costs. This assessment methodology leads to improved decision-making in transport planning and operations and can be used to develop management and control strategies that result in more robust network performance. The following results were obtained in this study: (1) Characterization of O-D demand variability from field data; (2) Development of p-quantile metrics for network performance; (3) Computational procedures to assess performance using p-quantile metrics; and (4) Calculation of robust controls following accepted norms of behavior.