Study On The Problem Of Tourism Emergency Shelter Location Based On Two-stage Stochastic Optimization

The frequent occurrence of natural disasters has caused massive casualties and economic losses to the booming tourism industry.In order to deal with the negative impacts,disaster management of tourism is required.As an important means of tourism disaster management,the construction of emergency shelters can effectively reduce casualties and property losses.In addition,the uncertainty of disasters leads to the uncertainty of demand.At the same time,most studies on shelter location focus on vehicle evacuation,ignoring the influence of helicopter evacuation on shelter location and material pre-positioning.Considering the uncertainty of demand and the particularity of helicopter evacuation,this paper constructs two tourism emergency shelter location models based on two-stage stochastic optimization,and uses a chance constraint to balance rescue cost and reliability level.Since secondary disasters may occur in the disaster-stricken scenic spots,this paper selects safe zones away from the center of the scenic area as potential candidates.Considering the shelter capacity,the maximum rescue time,the minimum shelter utilization,etc.,two-stage stochastic models with and without the chance constraint are established to minimize the total cost.In the first stage,decisions on shelter location and material pre-positioning should be made.In the second stage,the evacuation plan is optimized.This paper uses the sample average approximation(SAA)method to transform the two-stage stochastic optimization models that are difficult to solve directly.Several simulation experiments are designed to explore the performance of the models.Based on the experimental results,the feasibility of the stochastic model with the chance constraint is analyzed.Finally,the Jiuzhaigou scenic spot is selected as the empirical research area to verify the optimization effect of the stochastic model with the chance constraint.