模糊聚类分析在中国火灾危害分类中的应用

Application of Fuzzy Clustering Analysis to Classification of Fire Harm of China

火灾现象具有随机性、模糊性,是个复杂的模糊系统行为。运用模糊聚类分析方法,借助于M ATLAB计算软件,对中国火灾的危害程度进行了分类。根据2000年中国火灾统计资料,把31个省市分为重灾区、较重灾区、一般灾区、轻灾区等4类,与传统的概率统计方法相比,本方法更符合客观实际。城市火灾分类是动态的,重灾区吸取教训,增强消防意识,增大消防投入,可减少火灾隐患;而轻灾区如不注意防范,也会增加火灾隐患。因此不论轻重灾区,均应时时增强防范意识,把火灾危害降至最低。

Every year fires cause enormous losses to people in many countries. It is of practical value to study the rule of fire occurrence and its evolutional trend. Fire is a complex action of fuzzy system with random and fuzzy complexity. In this paper, the fire harm of China is classified by means of fuzzy clustering analysis with calculation software MATLAB. According to the data of fire statistical yearbook of China 2001, the fire harm of 31 regions in 2000 year can be classified into four types ：slightly serious regions, serious regions, more serious regions and most serious regions. The result of fuzzy clustering analysis agree well with the real situations.