Study On The Application Of Support Vector Machine With Entropy Method In Coal Mine Safety Evaluation

The coal industry is a specially highly dangerous industry, the situation of safety effects weather the development of coal industry can be health and sustainable or not. In a long time, in our country, coal accidents happened very often, which brought the society a heavy economic lost, and it had become a bottleneck of coal industry. In view of this, studying on comprehensive evaluation of the coal mine's safety to control the risk of coal production effectively has a very important theoretical and realistic meaning. It makes the result more objective and accurate, also improve the generalization ability of this model.On basement of reading lots of documents as well as experts'research achievements at home and abroad, this paper summarizes the evaluate theory of the coal mine's safety theory and current research situation at home and abroad, so as to support the basic theory of support Vector machine classification principle, connecting With the Entropy method to determine the index weight, establish a new model of the evaluation of coal mine's safety and connect with the actual datum to have done some application research.Firstly, this paper analyses the current safety situation of our country's coal mines to lead the thought of necessity to promote the evaluation job of the coal mines safety, summarize the safety evaluation theories and the current technologies development situation at home and abroad, set up the research's meaning, methods and new innovation points. Then, according to the modern accident-causing theory, combined people-machine-ring analysis method with analytic hierarchy process, the paper analyses the elements which effect the safety indicators, and sets up a quite comprehensively coal mine's safety evaluation system including ten first class indicators.40second class indicators, and applies entropy method to determine the index weigh of each indicator. After that, the paper expounds the basic knowledge supporting vector machine, especially based on the theory of vector machine classification principle, according to the feature that support vector machine has an advantage over those traditional ways, and the coal mines safety evaluation system which had set up. using software tool box-CMSVM2.0to collect data from local site, we realized the application of coal mine's safety evaluation research. The evaluation result almost meets with experts'evaluation result, which proves the model's practical applicability. At last is the summary the paper, it shows that the method can improve the scientific and reliability of the coal mine's safety evaluation, which has a broad application, still, there are some problems need to be done some more further up research.Figure16table12reference85...