欢迎访问林业科学,今天是

林业科学 ›› 2006, Vol. 42 ›› Issue (7): 73-77.doi: 10.11707/j.1001-7488.20060713

• 论文及研究报告 • 上一篇    下一篇

基于计算机图形技术的森林火灾模拟蔓延模型

秦向东1,3 阿布里提2   

  1. 1.中国科学院地理科学与资源研究所,北京100101;2.青森工业综合研究所,日本青森030-0113;3.中国科学院研究生院北京100039
  • 收稿日期:2004-06-26 修回日期:1900-01-01 出版日期:2006-07-25 发布日期:2006-07-25

A Spread Model for Simulation of Forest Fire Based on Computer Graphics Technique

Qin Xiangdong1,3,Abuliti2   

  1. 1. Institute of Geographical Sciences and Natural Resources Research, Chinese Academy of Sciences Beijing 100101;2. Aomori Industrial Research Center Aomori 030-0113, Japan; 3. Graduate School of Chinese Academy of Sciences Beijing 100039
  • Received:2004-06-26 Revised:1900-01-01 Online:2006-07-25 Published:2006-07-25

摘要:

利用计算机图形技术,以波动传播模型为基础,吸收了邻接单元模型的优点,设计基于计算机图形技术的林火模拟蔓延模型。这种模型用预定义的灰度图表示蔓延状态,将图元集中存储的过火区形状按一定的像素栅格运算规则写入灰度图中,来模拟火灾的波动蔓延过程。模型还利用漫染填充算法来处理具有复杂拓扑关系的地理要素附近的蔓延,并将过火区形状的投影变换转化为图元的旋转和错切变换。分析这种模型性能上的优势和局限,包括对局部失真和畸变的影响。从给出的算例和初步的理论分析可见,新模型可以将林火蔓延模拟的计算速度提高1个数量级以上,其他计算性能也有大幅度提高。模型可以为森林火灾的扑救提供精确、实时、全局的火场发展预测图,并且不受火行为的复杂性和地理要素的复杂性的限制。

关键词: 森林火灾, 计算机模拟, 蔓延模型, 图形技术

Abstract:

Two categories of spread models used for forest fire simulation, namely contiguous cellular model and wave propagation model, were strongly complemented each other in their advantages and limitations, which was utilized in the paper to design a spread model based on computer graphic techniques. In the new model, states of fire spread are expressed as a predefined gray level image. A series of fire shapes stored in metagraphic set are written to the gray level image according to given rules of raster operation while simulating the wave propagation of fire spread. Scan-line flood-fill algorithm is used to deal with the spread near some geographical elements with complex topological relations. Projection of fire shapes is translated into the combined transformation of graphic rotation and shear. The advantages and deficiencies of the model are discussed, including its effect on local distortion and aberration. By primary theoretical analysis and a calculation case, it is deduced that the new model can accomplish the simulation for forest fire faster than the previous models by more than one order of magnitude, while other performances of simulation are highly improved as well. Therefore, it can provide forest fire fighting with precise and overall predicted diagram of fire development in real time, not restricted by complexities of fire behaviors or geographical elements.

Key words: forest fire, computer simulation, spread model, graphics technique