To improve the path optimization effect and search efficiency of ant colony algorithms, an improved ant colony algorithm for the path planning of mobile robots under the grid map environment is proposed. In traditional ant colony algorithms, ants generally search 4 directions, 4 neighborhoods or 8 directions, 8 neighborhoods. Based on that, this paper proposes am improved search method for ants to search 16 directions and 24 neighborhoods, and also shows the mobile rules of ants. For heuristic information, we combine the idea of vector angle to design two methods to calculate heuristic information, and also analyze the application characteristics of those two methods through experiments. In the transfer probability part, we introduce the transfer probability control parameters. The search range of the algorithm can be adjusted by adjusting the transfer probability control parameters. Finally, the simulation experiments under grid map environments with different scales verify the effectiveness of the improved ant colony algorithm.