Study Of Mine Gas Monitoring Data Analysis And Its Predictive Application Based On Rbfnn

Gas disaster, as one of the most mainly hazards, is always puzzling the mine safeproduction. Based on the study of gas monitoring method and pre-warning technology, thispaper takes the real-time monitoring data and outburst prevention data (that can reflect gasemission features) the research object. The relative feature that different data reveals amongthe gas concentration, real-time gas monitoring data and outburst prevention data, is describedby Neural Network to research the gas concentration pre-warning and outburst risk predictionpre-warning. And the main works as follow:Firstly, introduce the basic theory and the prediction method of Radial Basis FunctionNeural Network. Then analyze the feasibility of the mine real-time monitoring dataapplication and the gas prediction basic principles.Secondly, aiming at the mine factual monitoring data features, use interpolation methodto have data processing and build up the fully-mechanized face gas concentration predictionand pre-warning model based on the real-time gas monitoring data, which is used to realizegas real-time prediction and pre-warning.Thirdly, according to the monitoring data, extract the statistical feature parameters of gasreal-time monitoring data and build up heading face gas outburst risk prediction andpre-warning model, which is to realize the outburst risk prediction and pre-warning based onthe data fusion analysis.Finally, make the prediction and pre-warning model apply to the example mine and havea analysis and verification.Above all the gas prediction and pre-warning technology, combined to the real-timemonitoring data application, it shows a great applicability, which can be extended as a effective method of mine safe monitoring system function and have a application value.