Evaluation Study On Coal Mine Emergency Rescue Ability With Imbalance Support Vector Machine

Coal as main energy in China, will be in a major position for a long time in future. Coal mine safety situation in our country has improved in recent years, but still has some problems. We can not completely avoid coal mine accidents with our current technology. However, effective coal mine emergency rescue can reduce casualties and property losses, so research on coal mine emergency rescue capability evaluation has important scientific significance and broad application prospect.On the basis of research on emergency rescue system home and broad, the paper expands into the field of coal mine emergency rescue capability evaluation theory. By analyzing the problems and making full use of existing research achievements, the paper develops an evaluation model to provide powerful decision support for coal mine emergency rescue capability evaluation practice.Firstly, the paper analyzes the domestic and foreign research achievements in the field of emergency rescue. By analyzing limitations of traditional evaluation method, we discuss the research significance and main thought of the paper. Secondly, the paper analyzes comprehensive causes that lead to coal mine accidents. We construct the coal mine emergency rescue capability evaluation indicators that consist of5secondary indicators and18tertiary indicators. Here, we make use of the analytic hierarchy process (AHP) method to select indexes, and then introduce the entropy weight method to calculate entropy weight of each index. According to these weights, we further screen indexes by sorting, to get indexes suitable for support vector machine (SVM) learning. Then, the paper introduces the normal support vector machine theory knowledge, and brings in the imbalance adjustment factor after considering characteristics of coal mine emergency rescue capability evaluation samples. So the imbalance SVM classification model based on coal mine emergency rescue capability evaluation is established. At last, the paper collects1set training sample and1set prediction sample for calculation by using the SVM packages of MATLAB2010b calculus. By compassion among results of imbalanced SVM evaluation, fuzzy comprehensive evaluation and expert evaluation, we demonstrate that, based on coal mine emergency rescue capability evaluation, the imbalance SVM evaluation model has a good performance and batter classification accuracy than normal SVM model and so meet the requirements of coal mine emergency rescue capability evaluation.