Multiclass dynamic emergency traffic collaborative optimization considering multiple solutions with stage-based algorithm

Abstract

Multiclass dynamic emergency traffic consisting of evacuation and rescue vehicles is a common situation on the road network in the response of the sudden local disaster, e.g., fire disaster, hazardous chemical disaster, etc. However, there are several options for the optimal rescue traffic routes sharing the same optimum objective value during the evacuation, and it is not clear how different optimal rescue traffic routes will affect evacuation traffic. Our motivation is based on the possibility for the dynamic rescue traffic optimum to have multiple solutions on the road network shared by multiple vehicle classes, which differ in when and which roads are occupied by how many rescue vehicles during the evacuation. To this end, we formulate multiclass dynamic emergency traffic collaborative optimization problem considering multiple solutions and traffic priority as the MMILP-RRO formulation, which consists of multi-objective mixed integer linear programming formulation (MMILP) and risk-based rescue traffic optimization formulation (RRO), and is solved by a stage-based optimization algorithm. The numerical results are conducted on the Beijing’s Zhongguancun​ road network to provide some insights into the effects of multiple optimal rescue traffic routes on evacuation traffic for multiclass dynamic emergency traffic network flow collaborative management.