Research and Design of Pre-warning System for Coal and Gas Outburst Based on Data Mining

摘要

Coal and gas outburst is one of the most important natural disasters in our country. Thus pre-warning has the great significance to production safety for coal mine. In view of the facts of current pre-warning technologies, such as single pre-warning indicators, poor performance under interference and low real-time capability, some new pre-warning models are presented. The models include G line, moving average line and so on. Based on data mining theory, the pre-warning system works on the time series of gas emission concentration (which is called G series), and the pre-warning system for outburst is designed as well. "4.19" coal and gas outburst in Dashu Village Coal Mine of Hebei Fengfeng Coal Mining Group is studied based on pre-warning system. And the result is in accord with accident investigation. The analysis shows that the pre-warning system can mine the outburst developing process information of coal-rock dynamical system from the time series of gas concentration and realize outburst accidents pre-warning.