Research On The Location-allocation Model For Urban Fire Stations Under Fire Risk

Since the beginning of the 21st century,China’s urbanization has developed rapidly.However,fire protection planning in most cities lags behind the urbanization process at different degrees.Multiple factors such as a wide variety of urban fire hazards,uncertain fire protection needs,and complex and changeable road traffic conditions make the the layout of some urban fire stations is not reasonable enough.It can be seen that studying the layout planning of urban fire stations and optimizing the allocation of fire resources has important theoretical significance and practical value for ensuring urban fire safety.Based on the three types of modeling ideas of deterministic planning,collaborative planning and distributed robust planning,this paper deeply studies the pain points and difficult problems in fire planning from different perspectives:balancing the fairness and efficiency of fire site selection and layout;dividing the collaborative fire service areas under practical fire protection planning;and the numerical uncertainty of key parameters faced in fire planning,so as to provide a theoretical basis for improving the scientificity of urban fire layout planning.The main contents and research results are as follows:Firstly,a multi-level gradual coverage fire station location model based on fire risk is constructed using the deterministic modeling idea.Historical fire data and variable data such as meteorology,roads,and population density are firstly collected,and a random forest model is used to predict the distribution of urban fire risk.In addition,a genetic algorithm is designed to solve the model on a city-scale case.The research results show that,compared with the traditional P-median model(efficiency priority)and the maximum coverage model(fairness priority),the model proposed in this paper takes into account the fire risk level coverage and area coverage in fire protection services.Particularly,when the ideal coverage radius(R1,R2,R3)is set as(2km,2.5km,3km),the fire rescue performance is the best in the medium risk area of the study area,and the fire service performance in the high risk area and low risk area is located at between the optimal placement of the P-median model and the maximum coverage model.The analysis results of taking Hefei City as the research area show that,based on the fire response time,relocating 30 fire stations can achieve the same service performance as the existing 44 fire stations,and achieve a fire response time of 5 minutes in 75%of the area.The above examples show that the model in this paper can effectively optimize the location planning and layout of the fire station.Secondly,a fire station location-allocation model based on collaborative fire service is built.The goal of the model is to maximize the collaborative fire service performance in the study area,and it comprehensively considers a variety of key factors in fire planning,including traffic conditions in multiple time periods,rescue dispatches with different types and numbers of fire vehicles,types of fire stations,fire response time,and reliability.The above factors are integrated by designing two types of coverage criteria:one is vehicle coverage,which characterizes fire service reliability,and the other is time coverage,which describes fire service accessibility.By further integrating the two criterias,a collaborative fire service standard that takes into account both the coverage of fire rescue vehicles and the coverage of fire response time is designed.Taking Hefei City as the research area,a sensitivity analysis experiment is designed to illustrate the relationship between fire budget and other factirs,such as vehicle coverage,time coverage,the number of fire trucks,and the number of fire stations.The distribution of optimal fire stations provides timely and reliable fire-fighting vehicle services to the study area,indicating that the model can optimize the layout of fire stations under collaborative rescue.An example of scenario analysis with Hefei City as the research area shows that when the cooperative firefighting service of rescue time response and firefighting vehicle response is comprehensively considered,compared with the existing fire station layout(44),which covers 50%of the area within the 5minute collaborative fire response time,the 5-minute collaborative fire coverage of the relocated 42 fire stations has increased to 75%.The results show that the collaborative fire protection coverage model can optimize and improve the service performance of the fire rescue system.Thirdly,this paper chooses the distributed robust optimization modeling idea to construct the fire station location-allocation model under the uncertainty of fire demand and fire response time.The objective of the model is to minimize the expected total cost of random variables under the worst distribution,including the construction cost of fire station,the purchase cost of fire fighting vehicle,the transportation cost,and the penalty cost for unsatisfied demands and overtime.Sample mean absolute deviation and sample average are used to design fuzzy set for describing the probability distribution of random fire protection demand and random fire response time.Subsequently,the model is solved using the Lagrangian duality-based cutting plane method.Since it is difficult to solve large-scale cases,two approximation algorithms are introduced,namely the Lagrangian duality-based linear decision approximation algorithm and the three-point discrete approximation algorithm.Taking Hefei City as an example,the sensitivity experiment analysis shows that the heuristic algorithm can effectively solve large-scale problems.In addition,by designing internal and external sample experiments,comparing the performance of the optimal solution of the distributed robust optimization model and the stochastic programming model on external samples,it is found that the former is 19.2%lower in rescue cost when the random distribution of fire demand is in the worst case.The results show that the distributed robust optimization model can enhance the robustness of the fire rescue service system.