中文摘要：

针对煤矿瓦斯爆炸灾害子系统的复杂非线性动力学特性,以南方某省1958—2007年的煤矿瓦斯爆炸事故发生数构建时间序列,应用霍斯特时间序列数据的标度行为,提出瓦斯爆炸时间序列的重标极差分析（R/S分析）方法。结果表明：瓦斯爆炸时间序列的霍斯特指数为0.550 2,分形维数为1.449 8,其分式布朗运动轨迹表现出一定的持久性和非高斯性。为诊断瓦斯爆炸事故时间序列变异年份,对时间序列进行分段R/S分析,获得变异年份为1974年,以1974年为分界点构成2个时间序列,霍斯特指数分别为0.589 1和0.697 5,分形维数为1.410 9和1.302 5,与煤矿瓦斯爆炸事故时间序列的持久性趋势一致。煤矿瓦斯爆炸事故时间序列看似无序却蕴藏规律,反映了系统的动力学变化特征,R/S分析方法可以描述煤矿瓦斯爆炸事故时间序列的动力学特征。

英文摘要：

In view of the complex nonlinear dynamics characteristics of gas explosion subsystems in coal mine, according to the number of gas accidents from 1958 to 2007 in one of southern province in China, a time series of gas explosions domain rescaled method was proposed based on the data scaling behavior of Hurst time series. By R/S （ Rescaled Range） analysis and double logarithmic linear fitting, the Hurst index of the time series was 0.550 2 and its fractal dimension was 1. 449 8, which indicates the trajectories of fractional Brownian motion show some persistent and non - Gaussian. For diagnosing the variance years of gas explosion accidents in this time series, the time series was divided into two parts for R/S analysis, and a variance year1974 was obtained. So, two times series were established with 1974 as the cutoff point. The Horst index of the two times series was 0.589 1 and 0.697 5, respectively, and their fractal dimensions were 1. 410 9 and 1. 302 5, respectively. These results accord with the actual situation, which indicates that the seemingly chaotic time series encompasses regularity and reflects the characteristics of the system dynamics. R/S analysis can be described as the dynamic characteristics of time series of gas explosions in coal mines.