Research on a Force-Sensorless Control Method for an Agricultural Manipulator Based on Fuzzy Sliding Mode Control and a Disturbance Observer
Agricultural manipulators are vulnerable to disturbances induced by compliant and hard contact, variations in object stiffness, and sudden constraint release during branch severing in fruit harvesting and pruning. External force/torque sensors, however, are costly, difficult to install, and poorly suited to agricultural environments. This study proposes a force-sensorless compliant control method integrating fuzzy sliding mode control and a disturbance observer (FSMC-DOB). Experiments are conducted on an RML63 6-DOF manipulator. Fuzzy sliding mode control enhances joint trajectory-tracking robustness, while fuzzy adaptation mitigates the chattering inherent in conventional sliding mode control. A disturbance observer estimates external disturbances at the wrist joints, and bias learning, smooth saturation, and task-specific compliance compensation enable contact disturbance estimation and control adjustment without external force sensing. A dual-timescale disturbance feature-extraction and contact-confidence mapping scheme suppresses false contact triggering caused by short-duration disturbance transients and improves the robustness of contact detection. In addition, a snap/release event detector with short-term impact suppression reduces end-effector rebound and torque shocks after sudden constraint removal during pruning. MATLAB simulations show satisfactory free-motion tracking. Under soft-contact harvesting, the maximum and RMS compression deflections are reduced to 0.090 rad and 0.046 rad, respectively, outperforming PID and conventional sliding mode control. Under hard-contact pruning, the global RMSE and post-snap peak error are 0.1491 rad and 0.3722 rad. ROS–Gazebo simulations and physical experiments further confirm motion continuity, reliable contact identification, and engineering feasibility, demonstrating improved stability, safety, and adaptability in complex agricultural contact operations.
Authors
- Wenshuo Gao (ORCID: https://orcid.org/0000-0001-9197-959X)
- Kun Jiang (ORCID: https://orcid.org/0000-0002-2912-5136)
- Bao-guo Yao (ORCID: https://orcid.org/0009-0001-4855-8076)
- Xifeng Liang
Institutions
- China Jiliang University (CN)
Publication Details
- Journal
- Agronomy
- Published
- 2026-10-04
- DOI
- https://doi.org/10.3390/agronomy16191940
- Primary Topic
- Adaptive Control of Nonlinear Systems
- Type
- article
- Field-Weighted Citation Impact
- 0.00