This paper summary, state and analysis accident data of China coal mine from 1991 to 2008 through a huge amount of groundwork, and discussed the severity of safety situation of China coal mine. And all of this lay the data basis for accident forecasting.According to the complex factors such as limited accident statistics of China coal mine, strong discrete and nonlinear, the author analyzed a variety of forecasting theories and methods, including regression forecasting method, time series forecasting methods and gray-system forecasting method. This paper proposed the approaches of using cubic exponential smoothing and gray-system forecasting methods to accidents forecasting. Based on the stated accidents data, forecasted count number of accident, death tolls and mortality per million tons of China coal mine, by methods of cubic exponential smoothing and gray-system forecasting method. For gray-system forecasting, the data were pretreated by the way of calculating square root and logarithms to improve forecasting accuracy. The precision of cubic exponential smoothing was higher, could be used for short-term forecasts, however, gray-system forecasting method's precision was relatively lower, it could be used in long-term forecasts and reflected the development trend of coal mine safety situation well. This paper combined these two methods and applied to accident prevention and provided support for the decision. And put forward countermeasures and suggestions of preventing and controlling accident, in view of the safety status of China coal mine and combined with the actual conditions, to improve production situation.
|