Safety evaluation of human accidents in coal mine based on ant colony optimization and SVM

The man-made causes have manifested great significance in the safety evaluation of coal mine. Because factors which have been affecting human accidents are quite various and, their interaction are so highly complicated that statistical analysis can hardly be carried out, this paper has made elaborate analysis of the leading factors through four aspects: employees’ qualities, organization management, facilities factors and working environment. Based upon those specific analysis, by converting feature selection of intricate multi-factors to parallel combinatorial optimization and global optimization, evaluative indices of human accident at coal mines and relevant feature extraction method have been come up with in this paper. After effective feature selection, a safety evaluation model of human accidents based upon SVM is established which can availably tackle the problem of “hard to evaluate accurately with limited samples of human accidents in coal mine”. Evaluation and analysis on human accidents data of ventilating system of Caitun Mine proves that the proposed approach is effective and practicable.