期刊
MECHANISM AND MACHINE THEORY
卷 168, 期 -, 页码 -出版社
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.mechmachtheory.2021.104573
关键词
Maritime crane operations; Motion compensation; Sensor fusion; Anti-sway; Anti-pendulum; Multi-body dynamics; Marine robotics
资金
- Natural Sciences and Engineering Research Council of Canada (NSERC) [RGPIN-2017-06967]
- Kraken Robotic Systems Inc.
This paper introduces a framework for relative motion compensation, utilizing complementary filters to estimate relative motion by fusing data from Motion Reference Units in real-time. Simulation results demonstrate significant improvements in motion control efficiency and accuracy across various operational scenarios.
In this paper, a framework for relative motion compensation is presented and demonstrated via simulation with kinematic crane-tip control. The method allows for compensation relative to a fixed platform or world-frame (ship-to-shore/platform transfer), a secondary moving body (ship-to-ship transfer), or the host body itself (on-deck operations) without modification. The system utilizes Motion Reference Units (MRUs) which are located on the host ship and the payload/hook. Sensor fusion is performed on the MRU data using real-time complementary filters to estimate the relative motion. The frequency response of the system is investigated, and practical considerations are discussed through a series of case studies, which include time-varying trajectories. Within simulation, the results show that the complementary filter estimates the orientation of the ship and payload under dynamic conditions. The resulting controller provided an improvement of 35 dB attenuation at the natural frequency of the pendulum relative to the uncompensated system. The case study of a pick-and-place operation demonstrates that the proposed system produces an order-of-magnitude reduction in the error metrics for the tracking performance and pendulum suppression. A hardware implementation reduced 83.8% of the motion for the worst-case natural frequency tests and up to 48.3% for the trajectory experiments.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据