Abstract: In order to solve the problems of poor optimization effect and fuzzy decision-making of safety input in mining enterprises in China, considering the influence of indirect loss on safety management effect, the EAHP method was used to determine the weight of each safety input factors, and the fuzziness and preference difference of experts' opinions were well described. Mine safety input decision model based on dynamic gray area weighted correlation algorithm has been established to reflect the impact of the difference of safety input-output at different time points by asymmetric sine whitening weight function and gray correlation analysis method. The model was applied to the optimization decision analysis of safety input in mining enterprises. The results show that the correlation degree of safety education and safety management in the mine is relatively large, which are 0.188 and 0.170 respectively, indicating that the bottleneck factors of the comprehensive effect of safety operation of the enterprise are safety education and safety management. The safety input optimization decision of the mining enterprise is reasonable and effective, which verifies the applicability of the model, and provides a decision-making reference for enterprises to formulate safety input schemes and improve safety management efficiency.