Artificial Potential Field for Dynamic Obstacle Avoidance with MPC-Based Trajectory Tracking for Multiple Quadrotors

Abstract

This paper’s objective is to navigate multiple quadrotors using Artificial Potential Field (APF) technique for obstacle avoidance in a known environment. The idea is to use the dynamic potential field concept that does not depend on the quadrotor’s dynamics to ensure avoiding collisions among quadrotors. Each quadrotor navigates with repulsion force surrounding it to avoid obstacles including other quadrotors. The quadrotor follows its free-collision path using Model Predictive Control (MPC) as the trajectory tracker. The MPC controller task is to control each quadrotor independently to follow the planned path.