Nonlinear MPC of Quadrotors for Path Planning and Dynamic Obstacle Avoidance
Yu Du,Junxiang Kang,Xiaojing Tian,Minghao Wang,Dong Liu,Ming Cong
DOI: https://doi.org/10.1109/cyber63482.2024.10749189
2024-01-01
Abstract:The quadrotor aircraft holds extensive prospects for applications in various fields, including transportation, military operations, rescue, and search missions. These domains present challenging demands, offering significant potential for utilization. In dynamic environments, trajectory tracking and obstacle avoidance become particularly crucial for quadrotor aircraft. This study proposes a nonlinear model predictive control (NMPC) approach to address challenges posed by environments with dynamic and static obstacles, aiming to achieve tasks such as pinpoint takeoff, obstacle avoidance, trajectory tracking, and landing of quadrotor aircraft. The formulated obstacle avoidance constraints involve nonlinear and complex issues, rendering the optimization problem more intricate. To tackle this challenge, this study adopts a strategy of transforming obstacle objects into state constraints and utilizes the ACADO open-source toolkit for efficient solving, aiming to accomplish collision avoidance and trajectory optimization goals. Through simulation validation, the NMPC method proposed in this study demonstrates feasibility and practicality in complex obstacle environments.