Whole-Body Integrated Motion Planning for Aerial Manipulators
Weiliang Deng, Hongming Chen, Biyu Ye, Haoran Chen, Ziliang Li, Ximin Lyu
AI summary
Problem
Existing planners are often task-specific or decoupled, struggling with high-dimensional state spaces and complex intermediate constraints required for versatile, aggressive aerial manipulation.
Approach
The method formulates a spatio-temporal optimization problem that jointly plans quadrotor and delta-arm trajectories using flexible partial waypoint constraints, guided by an imitation learning network to avoid poor local optima during dynamic maneuvers.
Key results
- Simultaneous quadrotor and manipulator trajectory optimization
- Flexible partial waypoint constraints for selective position, velocity, and orientation control
- Imitation learning-guided two-stage optimization to escape local optima
- Validated across nine manipulation skills in simulation and real-world experiments
Why it matters
Provides a versatile, collision-free planning foundation for aerial robots to perform complex, dynamic manipulation tasks in unstructured environments.
Abstract
Expressive motion planning for Aerial Manipulators (AMs) is essential for tackling complex manipulation tasks, yet achieving coupled trajectory planning adaptive to various tasks remains challenging, especially for those requiring ag- gressive maneuvers. In this work, we propose a novel whole- body integrated motion planning framework for quadrotor-based AMs that leverages flexible waypoint constraints to achieve versatile manipulation capabilities. These waypoint constraints enable the specification of individual position requirements for either the quadrotor or end-effector, while also accommodating higher-order velocity and orientation constraints for complex manipulation tasks. To implement our framework, we exploit spatio-temporal trajectory characteristics and formulate an op- timization problem to generate feasible trajectories for both the quadrotor and manipulator while ensuring collision avoidance considering varying robot configurations, dynamic feasibility, and kinematic feasibility. Furthermore, to enhance the ma- neuverability for specific tasks, we employ Imitation Learning (IL) to facilitate the optimization process to avoid poor local optima. The effectiveness of our framework is validated through comprehensive simulations and real-world experiments, where we successfully demonstrate nine fundamental manipulation skills across various environments.