A Hybrid Method Combining Data-Driven and Model-Based Algorithms for External Force-Sensing and Haptics Control of Cable-Pulley-Driven Surgical Robotic Manipulator

The implementation of the high-precision tracking control and external force-sensing ability of a manipulator is important for achieving refined surgical robot operation. In this paper, a hybrid method based on data-driven and model-based algorithms is proposed for the manipulator of a cable-pulley-driven surgical robot. This method integrates an artificial neural network and a dynamic model rotation angle estimation, and a full closed-loop control architecture is further constructed. The algorithm compensates for the hysteresis of the joint angle and effectively improves the tracking control precision. Based on the architecture, the external force estimator (EFE) using a joint torque disturbance observer and the force interaction teleoperation control strategy using a direct force feedback framework (DFF) are implemented. In the force loading experiment, it was shown that the EFE performs well for static and dynamic force estimation, and the teleoperated haptic control experiment showed that the DFF-EFE-based system has a high position-tracking accuracy with real-time external force-sensing ability.