Shelter Allocation and Location Model for Evacuation Plan of Coastal Cities During Floods (multi-level model)

Disasters are classified into two groups: man-made disasters and natural disasters. One of the main goals of crisis management is to reduce the damage and minimize personal suffering and improve the situation as much as possible. This study examines the evacuation planning before a flood occurs and the design of suitable shelter locations for flood evacuation. The issue involves the leadership (authorities) determining the locations of the shelter to minimize evacuation time and a few followers (migrants) who are able to choose the destination (shelter) and the route leading to it. In this study, in order to investigate the evacuation planning before the flood disaster, a location allocation model for the implementation of the flood evacuation plan is proposed. The proposed model is formulated as a two-tier programming problem using a genetic algorithm. The high-level problem is the selection of shelters with minimum evacuation time and the low-level problem is a combination of distribution and allocation (CDA) problems. In order to solve the two-level program problem, GA-based problem solving method is designed. The results showed that planning is one of the most important tools used in crisis and disaster management. The choice of shelter location and the effects of capacity range on the evacuation plan have the particular importance. As the capacity of the existing shelter increases, the number of selected shelters and the total evacuation time will decrease. Also, any increase in the occupancy rate of the vehicle will lead to a reduction in the total unloading time.