
改进的R/S方法与中国火灾数据的分析预测
付昱华1、付安捷2
Improved R /S Method and Analysis and Forecast to China Fire Data
Fu Yuhua1、 Fu Anjie2
讨论工程和经济学领域中的R/S分析方法(重标极差方法)的若干改进及应用。对于全国火灾起数的分析,计算赫斯特指数H时应用2种新的数据分组方法;引入赫斯特指数的差值ΔH以利于判断下一年的火灾起数是否会激增;对于已计算出的赫斯特指数H进行R/S分析,得到一组新的赫斯特指数H1,即赫斯特指数的赫斯特指数,以及相应的ΔH1,依此类推可以得到高阶赫斯特指数及其差值H2,ΔH2,H3,ΔH3等;根据1950—1999年全国火灾起数,用R/S方法预测2000年全国火灾起数。
This paper discusses some improvements for the R/S analysis method (rescaled range analysis) in engineering and economics and their application. For the analysis of total fire number in China, two new data grouping methods for calculating the Hurst exponent H are presented. The difference of Hurst exponent AH is introduced to judge the possibility whether or not the next year's fire number will be increased suddenly. For the R/S analysis of the calculated Hurst exponents, a new group of Hurst exponent H1,i. e. , Hurst exponent of Hurst exponent, and related △H1, can be reached, the rest high order Hurst exponent and related difference H2, △H2, H3, △H3 and the like may be deduced by analogy. According to the fire numbers from 1950〜1999, the fire number of year 2000 is forecasted with R/S method.
R/S分析 / 重标极差方法 / 高阶赫斯特指数 / 全国火灾起数 / 预测
R/S analysis / rescaled range analysis / high order Hurst exponent / total fire number of whole country / forecast
/
〈 |
|
〉 |