Limit Cycle-Based Artificial Fields for Obstacle Avoidance in Robot Path Planning

This paper introduces a novel limit cycle-based approach for obstacle avoidance and compares its performance against the conventional Attractive-Repulsive method, a widely used artificial potential field (APF) technique. While traditional APF methods model obstacles as repulsive fields and the goal as an attractive field, the proposed method incorporates a virtual limit cycle around each obstacle, enabling the robot to engage tangentially with the obstacle boundary and navigate around it smoothly. The limit cycle field is designed to drive the robot along a near-optimal path without the need for complex computations or manual switching strategies. Simulation results demonstrate that the proposed method consistently yields shorter and smoother trajectories compared to the AttractiveRepulsive approach, while maintaining low computational overhead. Although experiments were conducted on a two-degree-of-freedom cable-driven robot in a vertical plane, the methodology can be generalized to a broad range of robotic systems with higher degree-of-freedom.