Abstract: By establishing the evaluation system of emergency management ability of coal mine enterprises, we can find out the problems and shortcomings in emergency management of coal mines, and improve the emergency management ability of coal mines.Based on theoretical analysis and field investigation, this paper systematically considers and extracts 26 factors that affect the emergency management ability of coal mine enterprises, and uses the trained convolutional neural network model as an effective evaluation tool to accurately and effectively predict the satisfaction of samples outside the training sample set, thus providing suggestions for improving the emergency management of coal mines.The trained model is used for empirical prediction and analysis of Xiaoyun Coal Mine.The results show that the overall emergency management ability of the coal mine is Grade Ⅱ,which is at a good level.The evaluation results are consistent with the actual situation.