Study on large-scale crowd evacuation plan for typhoon disaster
Typhoon is one of the most destructive and serious natural disasters.Since the generation of typhoon disasters is predictable,it is necessary to evacuate and relocate people in the risk area in advance to effectively avoid casualties and protect the safety of the public.Historical practices and previous studies show that organized crowd evacuation based on public transport has become the main way to deal with typhoon disasters.Therefore,we developed a vehicle routing model based on the minimum evacuation completion time.To make the model more suitable for the actual scenario,multiple demand points and evacuation shelters were added to the evacuation network.Besides,considering the limited resources of vehicles and the capacity of shelters,"multi-points round trip"was employed in the model.That is,the vehicle can pick up people from any given demand point and send people to any shelter during each evacuation trip.Next,we introduced the concept of virtual demand points to simplify the model calculation.Moreover,since the vehicle routing problem in this paper is a typical NP-hard problem,it cannot be solved using the exact algorithm.A genetic algorithm based on a two-dimensional real number encoding was then employed to solve the above model,which was then validated by an example.Finally,a case study was conducted and analyzed.Specifically,21 demand points and 10 evacuation shelters were determined according to the actual situation,and different numbers of service vehicles were analyzed.Our proposed model is compared with the traditional vehicle routing model,and results show that our model can effectively improve evacuation efficiency.In addition,our model is more advantageous in areas with scarce transportation resources.Our proposed model in turn is proven to be credible and effective,which can contribute to the formulation of a large-scale crowd evacuation plan.
public safetytyphoon disastercrowd evacuationvehicle routemulti-point round trip
NETL
NSTL
万方数据
维普
