Safety of users in road evacuation: Algorithms for path design of emergency vehicles

Abstract

In this paper an advancement on the SICURO Research Project relating to thedesign of path choice models for emergency vehicles is proposed. In particular, with respect to the previous work in which the procedures to be planned and activated in emergency conditions were defined in order to allow the evacuation of weak users (the disabled, senior citizens, etc.) from the area affected by a disaster, some advanced instruments and scenarios for designing the optimal path for emergency vehicles to reduce evacuation times are introduced. The problem of an emergency vehicle that has to pick up some users at fixed points of the network and to take them to the refuge area is schematized (as in a previous work) with two different approaches: as a shortest-path problem (one to one) and as a vehicle routing problem (many to one). In the first exact approach the k shortest paths algorithm is applied to obtain the best k paths that satisfy specific choice criteria; in the second approach a metaheuristic procedure (genetic algorithm) which allows route optimization of a fleet of emergency vehicles is proposed. Some new experimental results obtained by applying the proposed advanced model to a real road transport network during a simulation of evacuation at urban scale in the context of the SICURO Project are also reported and compared with those obtained previously.