To deeply explore the risk factors and evolution paths contained in coal mine accidents and promote the transformation and upgrading as well as high-quality development of the coal industry, this paper adopts text mining and safety risk topic modeling to subdivide the safety risk system of coal mining enterprises into five subsystems: human resource allocation, establishment of scientific research and innovation platforms, technological innovation, policy support, and enterprise system construction. The network analytic hierarchy process and CRITIC weighting method are applied to calculate the comprehensive weights of each risk factor. Taking a certain coal mine in Qinghai as the research object, the system dynamics model is used to construct the causal loop diagram and stock-flow diagram of the safety risk resilience system of coal mining enterprises, and to simulate and analyze the impact of each risk factor on the overall safety risk resilience level of the case coal mine. The key technological innovation factor parameters are regulated to analyze the sensitivity changes of the safety production rate and safety risk resilience level of coal mining enterprises under different scenarios. The results show that: (1) The technological innovation and human resource allocation subsystems have the greatest impact on the safety risk resilience of coal mining enterprises, followed by the enterprise system construction, scientific research and innovation platform establishment, and policy support subsystems. (2) Increasing investment in safety, such as talent cultivation and technological innovation, can enhance the safety risk resilience level and shorten the time required to reach the expected safety value. Conversely, it will reduce the safety risk resilience level and delay the time to reach the expected safety value. (3) Further simulation and evaluation of the technological innovation subsystem reveal that the factors influencing the safety risk resilience of coal mining enterprises in order are construction survey and monitoring, support structure, construction design, mining plan, and engineering technology application. (4) The change trend of the conversion rate of technological innovation investment to engineering technology application and construction survey and monitoring is generally in a logistic “S” shape, mainly due to the time delay and lag of the benefits generated by safety investment.