Abstract: As the coal mine robot's explosion-proof safety level has not yet reached the level of ia Intrinsic safety, When performing environmental detection and rescue missions the Coal Mine Robot can not enter the gas dangerous area to avoid causing disasters, Therefore, the coal mine robot needs to avoid the dangerous area of gas. In this paper, the path planning method of the coal mine robot to avoid obstacles in the regional distribution of gas is proposed, which Using MAKLINK method to characterize the working environment of coal mine robot and Ant colony optimization Dijkstra algorithm is used to calculate the coal mine robot operation path, the experimental study on the path planning is carried out respectively in the two cases, which are only the regional distribution of gas and the regional distribution of the gas dangerous area and the obstacle. Experimental results show that the method proposed in this paper can achieve the optimal path planning under the condition of satisfying the safe distance and convergence speed. This method has important significance for the coal mine robot to work quickly and to improve the intelligent of the coal mine.