Gas accidents are the main type of accidents affecting the safety production of coal mines.In order to reduce the risk of gas accidents,an innovative and practical coal mine gas risk level evaluation method is proposed to provide assistance for the prevention and control of coal mine gas accidents.The study mainly includes three steps:first,the real data of coal mine gas accidents are collected;second,due to the large number of attribute features,the dataset is too high-dimensional,large-scale,and high-complexity structural characteristics,thus the t-distributed random domain embedding(t-SNE)meth-od is used to process the complex high-dimensional gas accident data;and finally,the genetic algorithm(GA)is adopted to optimize the support vector machine(SVM)to predict the severity of the coal mine gas accident.The results show that by comparing the performance of prediction effect,error distribution,and time cost,the t-SNE-introduced evaluation model can accurately predict 89％of accidents,saving about 60％of the time cost.
关键词
风险评估/煤矿瓦斯事故/t-分布随机邻域/遗传算法(GA)/支持向量机(SVM)
Key words
risk assessment/coal mine gas accident/t-distributed random domain/genetic algorithm(GA)/support vector machine(SVM)
维普
万方数据