基于改进混合遗传算法的教学楼火灾逃离路径优选研究  

Study on Optimum Evacuation Path of Teaching Building Fire Based on Improved Hybrid Genetic Algorithms

作　　者：吴元君[1] 徐玉林[2] WU yuanjun;XU YuLin(Anhui Finance&Trade Vocational College,Hefei 230601,China;Hefei Normal University,Hefei 230601,China)

机构地区：[1]安徽财贸职业学院云桂信息学院,安徽合肥230601 [2]合肥师范学院,安徽合肥230601

出　　处：《灾害学》2020年第2期75-79,共5页Journal of Catastrophology

基　　金：国家自然科学基金项目(31401657)。

摘　　要：为解决教学楼火灾疏散问题,更好地服务于校园火灾应急管理,研究基于改进混合遗传算法的教学楼火灾逃离路径优选。教学楼火灾发生时,确定影响火灾逃离路径优选的权值因子,如楼梯或消防通道长和宽、人流密度、逃离速度、烟雾扩散范围等,动态分析教学楼火灾发展情况,建立教学楼内部空间对象三维网络数据集,有效分析教学楼各楼层空间实体间的顺序、度量、拓扑空间关系,通过图像细化和图像特征点提取算法建立完整的教学楼三维路径模型。采用引入模拟退火拉伸方法的改进混合遗传算法,针对所建立三维路径模型,展开教学楼火灾逃离路径优选。模拟实验研究发现,该方法能从火灾条件、早期生长阶段和火灾发生三个阶段得到最佳的火灾逃生途径。In order to solve the problem of fire evacuation of teaching building and better serve the campus fire emergency management, the optimization of fire escape path of teaching building based on improved hybrid genetic algorithm is studied.The building when the fire broke out, to determine the weights of factors that affects the fire escape route optimization, such as: the stairs or fire long and wide, the density of stream of people, escape velocity, smoke diffusion range, etc., the dynamic analysis of the building fire development situation, to establish teaching building internal space object 3 D network dataset, effective analysis of the building floors space sequence, measurement, topological spatial relations between entities, through the image and image feature points extraction algorithm is used to build complete teaching building 3 D model of the path.An improved hybrid genetic algorithm based on simulated annealing tensile method is used to optimize the fire escape path of teaching building.It is found in the simulation experiment that this method can obtain the best escape route of the fire in the teaching building in three stages: fire condition, early growth stage and spread over a large area.

关 键 词：混合 遗传算法 教学楼 火灾 逃离 路径优选 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] X45[自动化与计算机技术—计算机科学与技术]