森林火灾蔓延模拟及应急路径规划优选  

Simulation of Forest Fire Spread and Optimization of Forest Fire Emergency Prevention Planning

作　　者：陈卓[1] 刘海洋[1] 黄全义[1] 郑明章[2] 栗健 Chen Zhuo;Liu Haiyang;Huang Quanyi;Zheng Mingzhang;Li Jian(Institute for Public Safety Research,Tsinghua University,Beijing 100084;Jilin Province Forestry Surveydesign&Research Institute,Changchun 130022)

机构地区：[1]清华大学公共安全研究院,北京100084 [2]吉林省林业勘察设计研究院,长春130022

出　　处：《林业科学》2024年第4期52-61,共10页Scientia Silvae Sinicae

基　　金：国家自然科学青年科学基金项目(42007420);“十四五”国家重点研发计划(2022YFC3105100);“十三五”国家重点研发计划(2017YFC1405300)。

摘　　要：【目的】利用云计算手段模拟林火蔓延过程,规划火灾应急路径,以促进救援力量快速、高效到达指定火灾地点,为布置林火阻隔措施预留充足时间。研究也为吉林省通化市东昌区的林火应急响应和预防规划提出建设性意见。【方法】以吉林省通化市东昌区为研究区域进行实地调研和模型推演。着眼于林火点单源和多源差异性,利用应急路径规划模型、Dijkstra算法模拟计算单源林火情况下最短救援路径,并纳入多源林火优先级理论,实现救援资源利用最大化。采用FARSITE模型对区域实例森林火灾进行蔓延速率模拟,获取林火发生后各阶段实时数据,综合考虑森林火点附近资源布局情况以及道路交通能力,进行救援路径的推算和防火道路的布局,以验证方法的可行性,并对森林火灾应急响应救援路径进行智能优选。【结果】在一般森林火灾发生时单个着火点的公式基础上,为2个及以上着火点引入优先级理论,再根据火情严重程度通过Dijkstra算法进行依次推演。此外,增加μ作为通行能力系数,以考虑路况、道路属性对救援通行时间的影响,得到新的路径公式。通过FARSITE模型,对2010年“5.3”一般森林火灾(20100503-220502-01)进行蔓延推演,在火源点起火5 h后,如果没有得到有效的人为干预,火情将演变为重大森林火灾,并蔓延到人居地区。由于风向影响,火源点东北侧的火情优先级高于北侧,若有条件同时进行救援,应急资源应按优先级倾斜。【结论】模拟结果表明,通过应急路径规划模型、Dijkstra算法和优先级理论,可以提高救援路径的获取速率,提前进行资源调配和人员安排,FARSITE模拟结合防火道路规划,联系火灾发生后各时间段的蔓延情况,能有效提升森林火灾决策效率,事半功倍地形成林火应急响应和预防规划方案。【Objective】In this study,cloud computing method was used to simulate the spread process of forest fire and plan the fire emergency path,so as to promote the rescue forces to quickly and efficiently reach the designated place,and to reserve enough time for arranging forest fire prevention measures.This study also puts forward constructive suggestions for forest fire emergency response and prevention planning in Dongchang District,Tonghua City,Jilin Province.【Methods】An on-site investigation was conducted in Dongchang District,Tonghua City,Jilin Province,and then the model deduction was carried out.With a view to the differences between single-source and multi-source forest fires,the emergency path planning model and Dijkstra algorithm were used to simulate and calculate the shortest rescue path in the case of single-source forest fires,and the multi-source forest fire priority theory was incorporated to maximize the utilization of rescue resources.The FARSITE model was used to simulate the spread rate of forest fires in regional examples,and real-time data of each stage after forest fires occurrence were obtained.In considering the resource layout near forest fires and road traffic capacity comprehensively,the rescue path was calculated and the fire prevention road layout was carried out,to verify the feasibility of the method,and the rescue path of forest fires emergency response was intelligently optimized.【Result】Based on the formula of a single fire point in a general forest fire occurrence,the priority theory was introduced for two or more fire points,and then Dijkstra algorithm was used to deduce in turn according to the severity of the fire. In addition, μ, as a capacity coefficient, was added to consider the influence of road conditions and road attributes on the rescue transit time, and obtain a new path formula. By using the FARSITE model, the spread rate of the "5.3" general forest fire (20100503-220502-01) in 2010 was simulated. In five hours after the fire started, if there was no effecti

关 键 词：智慧林业 应急预防规划 森林火灾决策 云计算 FARSITE DIJKSTRA 

分 类 号：S762[农业科学—森林保护学]