Abstract: With the continuous increase of the depth and intensity of coal mining in China, the problem of underground dust pollution is becoming more and more serious, which has become one of the key factors restricting the safe production and sustainable development of coal mines. High concentration of dust will not only lead to the occurrence of occupational diseases such as pneumoconiosis, but also seriously damage the health of underground workers. It is more likely to cause coal dust explosion accidents and pose a major threat to mine safety. In order to realize the accurate prediction of dust concentration and promote the intrinsic safety and sustainable development of coal mine production, a prediction method of dust concentration in coal mining face based on ARIMA-LSTM combined model is proposed. Taking the 1606 coal mining face of a coal mine in Inner Mongolia as the research object, the autoregressive integrated moving average model（ARIMA） is used to generate the initial prediction value of dust concentration. Based on the field measured data and the initial prediction value, the residual sequence is constructed, and the long short-term memory network model （LSTM） is further used to correct the residual. The prediction results of the combined model are compared with the measured data. The results show that the prediction accuracy of the ARIMA-LSTM combined model is significantly better than that of the single model. The average absolute error, root mean square error and maximum relative error are greatly reduced, and the key error index is reduced by up to 22.058 715 mg/m³. Especially in the stage of violent fluctuation of dust concentration, the combined model can still maintain stable prediction performance, and the prediction curve is in good agreement with the measured data, showing stronger adaptability and robustness. The combined model can effectively improve the accuracy of dust concentration prediction, which not only provides a more scientific technical means for mine dust control, but also provides an important support for the construction of mine intelligent monitoring system, and plays a positive role in promoting the fine and intelligent management of coal mine safety production.