Overview
This study introduces an optimization-based design framework for modular aerial manipulators. The framework addresses challenges associated with rotor-induced airflow during physical interaction tasks in complex environments. It integrates considerations for task wrench feasibility, end-effector placement, and constraints related to airflow exposure.
Research Context
Aerial manipulation with multirotor platforms enables physical interaction in complex environments. However, rotor-induced airflow poses a critical limitation, particularly for tasks involving airflow-sensitive targets or surroundings. Prior designs for aerial manipulators often assume a fixed end-effector location.
Approach
The research proposes an optimization framework structured around several key components:
- Airflow Tolerance Categorization: A novel categorization of target-side airflow tolerance is introduced. This categorization informs the formulation of corresponding exposure requirements, expressed as geometric constraints.
- Airflow Modeling: To efficiently model rotor-induced airflow, a compact cone-sphere envelope is developed. This envelope approximates the spreading structure of a quadrotor's airflow, while maintaining computational tractability for optimization processes.
- Reconfiguration Optimization: Building on the airflow formulation and categorization, a reconfiguration optimization is proposed. This optimization adapts a modular aerial manipulator to diverse task wrench requirements. The optimization process enforces both target-side airflow exposure and intra-platform airflow interference constraints.
- Integrated End-Effector Optimization: The framework optimizes the end-effector placement concurrently with the overall platform configuration, departing from prior designs that typically assume a static end-effector location.
The effectiveness of the proposed framework was evaluated through scalability experiments and ablation studies.
Findings
- The introduced categorization of target-side airflow tolerance allows for the formulation of airflow exposure requirements as geometric constraints.
- The compact cone-sphere envelope provides an efficient and tractable model for approximating quadrotor airflow in optimization.
- The proposed reconfiguration optimization method can adapt modular aerial manipulators to varying task wrench requirements.
- The framework successfully enforces both target-side airflow exposure and intra-platform airflow interference constraints during optimization.
- Simultaneous optimization of end-effector placement and platform configuration is achieved.
- Scalability experiments and ablation studies validated the framework's effectiveness.
Why This Matters
The ability to design aerial manipulators that account for and manage rotor-induced airflow is critical for their application in environments where targets or surroundings are sensitive to airflow. By optimizing end-effector placement and platform configuration dynamically, the framework expands the range of tasks aerial manipulators can perform while mitigating adverse airflow effects.