Focusing on the low efficiency of the rescue robot in the coal mine disaster environment, a new improved boundary exploration algorithm based on multiple Unmanned Aerial Vehicles (multi-UAV) was proposed. Based on the utility value boundary exploration algorithm, the flight angle parameter of UAV was considered, and the distribution function was introduced as a judgment mechanism to construct the objective function. Finally, the ant colony algorithm was used to solve the objective function. Simulation experiments were carried out on a rasterized map with Matlab software. The simulation results show that the improved boundary exploration algorithm can reduce the phenomenon of repeated coverage and crowding, shorten the detection time, meanwhile the energy required by UAV is reduced by about 30%, thus improving the overall exploration efficiency of multi-UAV system.