Demand Estimation And Dispatch Of Urban Flood Emergency Materials Based On Baidu Heatmap

In recent years,under the background of continuous deterioration of global climate conditions,severe environmental damage and rapid urbanization,floods have occurred frequently in many large and medium cities in China,causing serious social problems and economic losses.Flood disasters are one of the most widespread,longest-lasting,and most harmful disasters in China.When the disaster occurs,it is necessary to efficiently and accurately deliver materials to the disaster-stricken area to protect the needs of the people in the disaster area and prevent the disaster from further expanding.When urban floods occur,reasonable,effective,and accurate estimation of emergency materials needs has become an urgent problem.The estimation of emergency material demand is mainly based on the number of people affected by the disaster.At present,the number of people affected is mainly from static census data.If real-time population data can be used,it can provide more reliable and accurate quantities of emergency materials for the disaster area.Real-time information is conducive to scientific and reasonable emergency material dispatching.When the demand for emergency materials is determined,it needs to be quickly and accurately transported to the disaster site to provide sufficient material security for the people in the disaster area.Urban floods usually result in multiple disaster-stricken sites,and the geographical locations between the disaster-stricken sites are relatively close.They involve common emergency supplies and rescue forces,and require the joint support of multiple reserves.How to reasonably allocate the materials of multiple emergency material storage points to the disaster-stricken points and how to make the transportation route the shortest has also become a problem to be solved,which has practical significance and theoretical significance.This article mainly researches from two aspects: the estimation of urban flood emergency materials demand,material allocation scheduling and path planning.On the one hand,a method of calculating population dynamics based on Baidu’s heat map was proposed,and an extreme learning machine method was used to construct a flood disaster emergency material demand estimation model to estimate the quantity of different types of material demand.On the other hand,for multiple disaster-stricken points,multiple reserve points,and multiple types of emergency supplies,a genetic algorithm is designed to propose a reasonable solution for the shortest delivery path of emergency supplies dispatch.(1)Based on the heat map data in Baidu’s big data,the information of the dynamic spatiotemporal distribution of the affected population is extracted to obtain the dynamic data of the population at different times,which solves the problem that the traditional census cannot reflect the dynamic population and spatial distribution.Synthesize relevant data after multiple periods of flood disasters,select the affected population,area coefficient,seasonal coefficient as input indicators,tents,quilts,and drinking water as output indicators,and test with root mean square error,model validity coefficient,and determination coefficient It shows that the simulation accuracy is high,and it can be used to accurately estimate the amount of emergency materials,and provide a scientific decision basis for government departments’ post-disaster rescue and emergency material reserves.(2)Emergency material dispatch is an important guarantee for rescue work in natural disasters and emergencies.Scientifically transporting and distributing emergency materials is the key to achieving efficient rescue.The material storage points are generally divided into government and social reserves.Different types of materials correspond to the types of storage points.The storage points that meet the needs are selected.Emergency supplies from multiple storage points are transported to multiple disaster-stricken points.The objective function is the shortest transportation path.,Design genetic algorithm to realize emergency material allocation and transportation route decision.The heat map based on Baidu big data can monitor regional population movement trends in real time,accurately sense changes in urban population,and lay a data foundation for studying real-time population.Taking the statistics of the real-time population and regional and seasonal coefficients as input indicators,and tents,quilts,and drinking water as output indicators in emergency supplies,the extreme learning machine method is used to build a model of emergency supplies demand and perform precision tests.It shows that the model is feasible to estimate the demand for tents,quilts and drinking water separately.Aiming at multiple disaster-stricken points and multiple reserve points,a genetic algorithm was designed to implement emergency material allocation and transportation routes,and a minimum path plan for emergency material dispatch was obtained.