Study On Coal Mining Safety Evaluation Based On RS-SVM

Coal is the foundation of the national economy and social development. Coal occupies about 70 percent of the once energy production and consumption structure in our country now, predicting that the percentage will be up to 60 percent in the 2010, the 50 percent in the 2050. Therefore, coal has been being the main energy for a few years. However, the coal production doesn't meet the requirement. Nowadays, the rapidly growing economy calls higher request on the the coal development. For this reason, it is very necessary to guarantee the safe and high-efficiency development for mining production, the same as the safety of coal mines.Based on the study situation of the coal mines safety management method research and safety decision making method, this article analyses the different environment conditions of diverse production stages. Then, construct the safety evaluation index system, at the same time applying rough set theory and RS-SVM predicting modeling. The process is: first, analyse the selected property based on rough set theory to delete the redundant property so as to get the minimal rule set for reducing the number of features and reducing the input dimension of SVM learning samples. Second, establish the evaluation model based on excellent non-linear characteristics and appraise the coal safety production condition. RS-SVM integrated method has advantages at the ways of the Rough Set attributes reduction and SVM small sample and non-linear features. This method improves the system operating speed and prediction precision. In the end, with XueCun corporation as an example, the article investigates the model establishment, function selection and other issues of the support vector machine sample set, in order to improve safety management of coal corporations.