Hybrid Tension-Position Control of Cable-Driven Serial Manipulator to Improve Accuracy and Avoid Cable Slack and Breakage

Cable-driven serial manipulators have significant potential for applications in confined spaces. However, due to the unidirectional driving force and hysteresis characteristics of cables, nonsynchronous motion between motors and tracking errors on each axis can cause cables to become too loose or too tight, leading to cable slack or breakage. To address this issue, we propose a hybrid tension-position control (HTPC) method to achieve high position accuracy while maintaining cable tension within a safe range. First, we establish multiple-space kinematic and static models to reflect the motion and force mapping between motors, cables, joints, and the end-effector. Second, cable terminal tensions are predicted by combining artificial neural networks with friction models and measuring cable tension at the motor side. Based on risk assessment, the desired tensions are determined according to force balance constraints to ensure safety. Third, we present a two-level closed-loop control method composed of resolved motion position control and hybrid tension-position control. Each driving group consists of three driving cables and three motors. The cable with the highest risk is controlled in hybrid tension-position mode to maintain its tension, while the other two cables are controlled in position mode to achieve the desired joint positions. Finally, we conducted several hardware experiments to validate the proposed method. The results show that the HTPC can reduce motion errors by more than half compared to traditional methods while simultaneously avoiding cable slack and breakage.