| 一种基于数据融合的林火火线动态蔓延预测方法的研究 |
| |
| 引用本文： | 翟春婕,张思玉,张怀兵,闫德民,彭徐剑,曹兆楼. 一种基于数据融合的林火火线动态蔓延预测方法的研究[J]. 火灾科学, 2018, 27(3): 141-147 |
| |
| 作者姓名： | 翟春婕 张思玉 张怀兵 闫德民 彭徐剑 曹兆楼 |
| |
| 作者单位： | 南京森林警察学院森林消防学院，南京，210023；中国科学技术大学火灾科学国家重点实验室，合肥，230026;南京森林警察学院森林消防学院，南京，210023;南京森林警察学院森林消防学院，南京，210023;南京森林警察学院森林消防学院，南京，210023;南京森林警察学院森林消防学院，南京，210023;南京信息工程大学光电工程系，南京，210044) |
| |
| 基金项目： | 中央高校基本科研业务费专项资金项目（LGYB201615）；江苏省自然科学基金（BK20150929 , BK20140501）；国家自然科学基金（61605081） |
| |
| 摘 要： | 准确预测林火蔓延对于有效防治野火危害具有重要意义。传统林火蔓延速度场经验模型需要可燃物特性、坡度、温度及湿度等众多实测参数,由于参数均具有不同的时空分布,限制了模型的实际应用。因此提出速度场实时测量及水平集法模拟林火蔓延两种技术相结合的方法,通过测定火蔓延锋面的位置获取当前时刻锋面处的速度场分布并对此速度场在未燃烧区域进行延拓,结合水平集法预测林火蔓延,并使用数值模拟及实验数据验证了该方法能够有效预测短期内林火蔓延趋势,得出基于速度场的数据融合方法可满足实际林火蔓延建模的需求。
|
| 关 键 词： | 数据融合；速度场；水平集；林火蔓延预测 |
| 收稿时间： | 2017-06-25 |
| 修稿时间： | 2017-09-12 |
Data-assimilation-based prediction of wildland fire propagation |
| |
| ZHAI Chunjie,ZHANG Siyu,ZHANG Huaibing,YAN Demin,PENG Xujian and CAO Zhaolou. Data-assimilation-based prediction of wildland fire propagation[J]. Fire Safety Science, 2018, 27(3): 141-147 |
| |
| Authors: | ZHAI Chunjie ZHANG Siyu ZHANG Huaibing YAN Demin PENG Xujian CAO Zhaolou |
| |
| Affiliation: | College of Forest Fire Protection, Nanjing Forest Police College, Nanjing 210023, China;State Key Laboratory of Fire Science, University of Science and Technology of China, Hefei 230026, China;College of Forest Fire Protection, Nanjing Forest Police College, Nanjing 210023, China;College of Forest Fire Protection, Nanjing Forest Police College, Nanjing 210023, China;College of Forest Fire Protection, Nanjing Forest Police College, Nanjing 210023, China;College of Forest Fire Protection, Nanjing Forest Police College, Nanjing 210023, China;Department of Optical Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China |
| |
| Abstract: | Accurate prediction of wildland fire is important for effective protection of wildland fire. Conventional empirical model of rate of spread involves the parameters such as slope, moisture and temperature, which vary over time or space. This limits the model application. In this paper, a method based on data assimilation is developed to overcome this problem. The rate of spread is retrieved with the measured position of fire perimeter and extended in the unburnt area. With the extended rate of spread, the fire perimeter is evolved by level-set method. Numerical results and experimental data demonstrate that the proposed method can effectively predict short-time fire perimeter evolution, which provides basis for further application of data assimilation in fire protection. |
| |
| Keywords: | Data assimilation Rate of spread Level-set method Prediction of wildland fire propagation |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《火灾科学》浏览原始摘要信息 |
|
点击此处可从《火灾科学》下载全文 |
|
