地标性建筑群火灾蔓延风险空间感知算法

A Spatial Perception Algorithm for Fire Spreading Risk of Landmark Buildings

为了改善火灾蔓延风险空间感知水平较低的问题,研究了地标性建筑群火灾蔓延风险空间感知算法。采集地标性建筑物数字高程数据、建筑物地面矢量数据以及建筑物表面纹理信息,通过SketchUp软件建立地标性建筑群的3维模型,分割地标性建筑群空间区域,优化元胞自动机的元胞边长、火灾蔓延速度、元胞状态以及元胞着火概率,将优化后的元胞自动机应用于建筑群3维模型中,获取火灾蔓延风险空间感知结果。实验结果表明:该算法在不同风向下感知的火灾蔓延风险空间均未超过建筑群边界,具有可行性;风险空间建模均方误差低于0.16,能够精确感知地标性建筑群火灾蔓延风险空间,提高了风险空间感知水平,为地标性建筑群的优化设计提供依据。

In order to improve the low level of spatial perception of fire spread risk,the spatial perception algorithm of fire spread risk of landmark buildings was studied.The digital elevation data,ground vector data and surface texture information of landmark buildings were collected,and the 3D model of landmark building complex was established through SketchUp software,the spatial area of landmark building group was divided,the cell side length,fire spread speed,cell state and cell fire probability of the cell automata were optimized,and the optimized cellular automata was applied to the 3D model of the building complex to obtain the spatial perception results of fire spread risk.The experimental results show that the fire spread risk space never exceeds the building complex boundary,and the mean square error of risk space modeling is less than 0.16,increased level of risk spatial perception,which can accurately perceive the fire spread risk space of landmark buildings and provide a basis for the optimal design of landmark buildings.