Path Planning of Multi-robots in Confined Spaces

Mobile robots can navigate in various environments if a suitable path-planning technique is provided. The importance of the right path-planning strategy for a given environment is emphasized in this paper. We elaborate on how the structure and complexity of the path can affect the performance of path-planning algorithms. Rapidly-exploring Random Tree (RRT), RRT*, and HybridA*, which perform perfectly in wide spaces, are implemented in confined spaces. We simulated the environment of a confined space based on actual air-conditioning (AC) duct dimensions. We focused on the issue of path planning in narrow T-junctions as it delimits the robot’s movements. By adjusting the minimum turning radius and primitive length, the algorithms are enabled to generate trajectories in a narrow path. Algorithms’ performance is evaluated based on the distance they cover to reach the goal and the smoothness of the path. The simulation results verified the prominent difference in execution of the same algorithm in different environments.