APF Method of Dynamic Obstacle Avoidance with Relative Velocity

The Artificial Potential Field (APF) method is a widely adopted approach in autonomous mobile robot path planning due to its mathematical simplicity and computational efficiency. However, its application in dynamic obstacle avoidance remains limited by the static assumptions in its algorithmic logic, leading to challenges in realistic scenarios. This paper addresses these limitations by introducing a novel refinement to the APF method for dynamic obstacle avoidance. Our approach redefines the repulsive field as an adaptive elliptical model, with the obstacle as the focal point and the relative velocity between the robot and the obstacle dictating the major axis. This innovation accounts for realistic motion dynamics, including varying speeds and directions, to enhance safety and path optimization. To validate the effectiveness of the proposed method, we developed new algorithms and tested them in a virtual environment, demonstrating improved adaptability and efficiency compared to traditional APF approaches.