Location Analysis For Emergency Shelter Of The Mountain Area Utilizing Spatiotemporal Population Estimation

As a mountainous country,mountain disasters occur frequently in China,but the level of disaster prevention and evacuation in mountains areas is relatively backward.As the main form of residential areas in mountainous areas,towns have complex geology,dense population,limited land use and imperfect road system.All of these lead to increased disaster risks and various disaster avoidance problems are more prominent than in plain areas.As an important part of the urban disaster prevention and mitigation system,emergency shelters are of great significance to ensure the public safety and post-disaster operation of cities.Indepth study of the disaster situation in small mountain towns and the development of a set of localized emergency shelter system is an important step to ensure the smooth planning and construction of mountainous areas.This thesis,based on the current standards of shelters in China,sorts out the key points of the spatial layout of shelters suitable for small mountain towns.Taking the perspective of fine-scale population daytime-nighttime distribution as the starting point,the supply-demand relationship of regional disaster prevention planning system is clearly studied.The combination of geographic information technology and management science decision-making is used to solve the optimization problem of emergency shelter layout.Firstly,the statistical data is spatialized to the building scale,and multi-source geographic data such as land use data,building vector data,and Baidu urban population geographic big data are integrated to construct a daytime-nighttime population spatial distribution estimation model based on the law of human spatial displacement.Secondly,from the three aspects of safety,accessibility and effectiveness,the main factors affecting the layout of shelters in mountain areas are proposed,and the population distribution of the obtained building scale is taken as the starting point,and the factors are quantified by GIS technology,and the PSO-AHP-EM method is used to comprehensively determine their weights.In order to further optimize the layout of shelters,the obtained weights are introduced into the multi-objective location-allocation model,and the rationality of shelters,the rapidity of evacuation time and the balance of personnel allocation are comprehensively considered,and the service radius and site capacity are taken as constraints.At the same time,the simulated annealing algorithm improved the particle swarm optimization algorithm is designed to solve it,so as to improve the spatial search ability and solution quality in highdimensional complex problems.Taking Lushan County Seat,a typical mountain disaster-prone area,as an example,the results show that the daytime-nighttime population mobility is small,and the spatial difference is obvious.In order to provide a more refined population evacuation scheme,taking the simulated building distribution as the demand point,the quantification of the influencing factors of the shelter and the optimization process of site selection are more realistic,and the multi-objective planning model comprehensively considers multiple conflicting objectives.It is found that the same shelters are reserved for both day and night after supplementary site optimization,which basically meet the evacuation needs of buildings in the county,and the evacuation efficiency and site utilization rate are well balanced on the basis of cost control.The improved PSO algorithm has a better distribution for solving location planning problems in a large-scale solution scenario.The research conclusion has direct guiding significance for disaster prevention and construction of Lushan County,and also has certain reference significance for emergency evacuation planning of similar mountain towns in southwest China.