Magnetorheological damper-based compliance control for hydraulic actuators in Legged Robots
Compliance control is essential for hydraulic legged robots to operate in unstructured environments. However, due to the inherently underdamped nature of valve-controlled cylinders, existing legged robots often struggle with foot impact and vibration. To address these challenges, this study presents a single-leg model and a damping compensation approach based on MRD. A hydraulic damping actuator (HDA) incorporating an MRD is installed at the drive joint to enhance damping performance. To mitigate force and displacement disturbances caused by foot-ground collisions, the proposed method employs an impedance control strategy with a force-based inner loop and a damping compensation strategy with a position-based outer loop. This dual-loop control effectively reduces displacement errors and internal pressure fluctuations. Additionally, a fuzzy controller is designed to realize resistance–impedance coupling between the MRD and the hydraulic actuator. Experimental results demonstrate that under step input conditions, the HDA reduces the maximum overshoot by 74.74% and shortens the settling time by 78.24%. During gait control, the damping compensation strategy significantly suppresses foot vibrations, decreasing the impact velocity by a factor of five. These findings confirm that the proposed HDA system and control algorithm substantially improve compliance control and effectively address the underdamped behavior of hydraulic actuators in legged robots.
Authors
- Qinglong Liao (ORCID: https://orcid.org/0000-0003-1692-0429)
- Baizhou Ma (ORCID: https://orcid.org/0000-0002-2609-7168)
- Heng Du
- Junshou Su
- Xiufang Lin
- Yuzheng Li
- Hui Huang
- Jiayi Zhou
- Yunfeng Fu
Institutions
- Minjiang University (CN)
- Kangmei Pharmaceutical (China) (CN)
- Fuzhou University (CN)
- University of Science and Technology Beijing (CN)
Publication Details
- Journal
- Journal of Intelligent Material Systems and Structures
- Published
- 2026-01-02
- DOI
- https://doi.org/10.1177/1045389x251395460
- Primary Topic
- Robotic Locomotion and Control
- Type
- article
- Field-Weighted Citation Impact
- 0.00