Study On Mine Safety Evaluation Based On Support Vector Maching With Entropy Cross Validation

The coal industry is a high-risk industry, coal mine safety accident frequency, and mine workers have a considerable number of people suffering from occupational disease; Not only cause huge economic damage to property, but also cause social unrest, it is not conducive to the healthy development of coal industry and the social harmony and progress. So the research of coal mine safety evaluation has very important theoretical value and practical significance.Based on the function of support vector machine classification, by identifying the risk factors of coal mine safety production system, combined with the entropy weight method, man-machine analysis, chromatographic analysis, constructs the comprehensive evaluation index system of mine safety including 11 first-level indicators,90 secondary index. Set up a support vector machine (SVM) based on entropy weight of mine safety evaluation model by the data collect from local site, after comparing with using heuristic algorithm model, determine the advantages of the model in this paper, and completed the application instance on the 1930 mine in xinjiang coking coal group, for instance, the evaluation results are basically identical with the results of expert evaluation, proves the correctness and practicability of the model. Then by analyzing the level of safety evaluation indexes to determine the weak link in coal mine production, guiding the direction of the coal mine safety work, rapidly improve the level of coal mine safety production. The results show that this method can make accurate classification to the safety status of coal mine, provides a new method for safety evaluation of coal mines.