An Improved Path-planning Method for Multi-agent by Velocity Obstacle Dynamic Window Approach

Dynamic window approach (DWA) is a frequently-used local path planning algorithm for avoiding dynamic obstacles in multi-agent path planning. Velocity obstacle (VO) is a local path planning algorithm that uses the relative velocity of robots to determine whether there will be collisions between robots in the future for collision avoidance. This paper proposes an improved method combining the velocity obstacle and dynamic window approach. The proposed algorithm can solve the problem of path redundancy caused by the influence of robots on each other in the local path planning of multiple robots, solve the problem that the robot can not reach the target point due to the influence of other robots and limitations of robot's kinematics model, reduce the phenomenon of dead cycle around the target point. Finally, the simulation experiment verifies the feasibility and effectiveness of the algorithm proposed in this paper. The simulation result analysis demonstrates the improved algorithm's path planning effect and the reduced running time.