R/S Analysis for Coal Mine Different Grade Accidents Based on Fractal Theory

In order to master the evolution characteristics and laws of different rank coal mine accidents in China, the R/S analysis method of the fractal theory is used in this paper to build three time series based on the number of deaths of different rank coal mine accidents in China from 2005 to 2014, and its fractal characteristics are researched. Linear regression correlation coefficient R of its rescaled range distributes from 0.981 66 to 0.994 83, and its linear regression equation is reliable by unilateral t test on the basis of the significance level α=0.05; Hurst index H of three time series distributes from 0.701 9 to 0.940 95; fractal dimension D distributes from 1.059 05 to 1.298 1, three time series present distinct Hurst phenomenon and striking Joseph effect. The research results show that the tendency of coal mine production safety accidents reducing and production safety condition improving is strong, and R/S analysis method can better reflect complicated non-linear system dynamics characteristics of coal mine production safety.