首页 | 本学科首页   官方微博 | 高级检索  
     检索      

火灾预测的模糊马尔柯夫模型
引用本文:姜学鹏,徐志胜,冷彬.火灾预测的模糊马尔柯夫模型[J].灾害学,2006,21(3):27-32.
作者姓名:姜学鹏  徐志胜  冷彬
作者单位:1. 中南大学,防灾科学与安全技术研究所,湖南,长沙,410075
2. 机械工业出版社,高等教育分社,北京,100037
摘    要:火灾发生并不是一个严格马尔柯夫随机过程,而是一个“近似具有马尔柯夫性”的模糊马尔柯夫过程。本文在对我国55年火灾统计资料“相对化”处理的基础上,根据模糊马尔柯夫理论和方法,建立我国火灾相对变动模糊马尔可夫预测模型。根据隶属度最大原则,确定所属状态,进行火灾预测。2005年我国火灾预测值与实际值的相对误差为0.0148,表明模型具有相当高的预测精度。对我国2006年进行预测,得到火灾相对变动状态等级为微降,我国2006年的火灾发生率与2005年相比降低了0~0.1,火灾频数约为212347~235941起,预测结果可为有关部门的决策提供一定的依据。

关 键 词:模糊马尔可夫模型  相对化处理  火灾  预测
文章编号:1000-811X(2006)03-0027-06
收稿时间:2006-03-19
修稿时间:2006-03-19

Application of Fuzzy-Markov Method in Fire Forecast
JIANG Xue-peng,XU Zhi-sheng,LENG Bin.Application of Fuzzy-Markov Method in Fire Forecast[J].Journal of Catastrophology,2006,21(3):27-32.
Authors:JIANG Xue-peng  XU Zhi-sheng  LENG Bin
Institution:1. Disaster Prevention Science and Safety Technology Institute, Central South University, Changsha 410075, China; 2. Department of Higher Education, Mechanical Industry Press, Beijing 100037, China
Abstract:Fire occurrence is not a strict Markov process, but an approximate Fuzzy-Markov process. By relative processing the fire statistic data in China from 1950 to 2004, a relative Fuzzy-Markov fire forecasting model is worked out by using Fuzzy-Markov forecasting method and theory. According to its maxim principle, the fire relative fluctuating state in future is forecasted. The relative error of forecasted value and the real value in 2005 is 0. 00148, which shows that the model has a high forecasting precision. The forecast shows that fire frequency in 2006 will be lower than that in 2005 by 0-0. 1, it will be between 212347 and 235941. The forecast results will be of a certain guiding significance to the related departments.
Keywords:Fuzzy-Markov model  relative processing  fire  forecast
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号