Path planning plays a crucial role in multi-robot systems, and its effectiveness directly impacts the system’s performance. A multi-robot conflict-elimination path planning method (CEPP) for dynamic environments is proposed. The method fuses the adaptive dynamic-window algorithm (ADWA) with the Repulsive function-based optimized A* algorithm (R–A*) to deal with multi-robot path planning (MRPP) and introduces a safe area radius and priority strategy to solve the multi-robot collision conflict problem. Among them, ADWA first adds the time cost and target point distance evaluation function to the original evaluation function and introduces adaptive weights to accelerate the efficiency of the robot in finding the target point. Then a target point detection waiting mechanism is introduced to solve the problem that the robot cannot find the endpoint. Finally, the effectiveness of the CEPP algorithm for MRPP in dynamic environments is verified by simulation. Meanwhile, the CEPP algorithm is compared and analyzed with the traditional fusion algorithm (A*-DWA), and the simulation results show that the average running time and path length of this method are better than the A*-DWA algorithm.