Abstract: We analyze the problem that the control of humanizing working face on coal cutter should ensure efficient operation, and control the gas concentration overrun problem. We propose the method based on the gas concentrations advanced prediction method to control the running speed of shearer in advance. Gas concentration in time series is predicted by using ARMA model on the basis of advancing 5 minutes gas concentration range. Using the prediction interval limit according to the statistical characteristics of the gas concentration, the sequence parameters on maximum limit and 95% confidence interval, we comprehensively control coal cutter operation status. For six different conditions, coal cutter speed control methods are proposed respectively. Combined with 030302 working face in a mine of Gansu province, we simulated unmanned work with the method.