Friction identification in suspended cable-driven robots using current-controlled actuation and force sensing

This paper presents an experimental framework for identifying friction parameters in cable-actuated systems representative of suspended Cable-Driven Parallel Robots (CDPRs). The proposed method combines current-controlled actuation with dual force sensing—using both a dynamic load cell and a static strain gauge—to estimate motor inertia, viscous friction, static friction, and suspended mass. A linear dynamic model is identified via regression from time-resolved current and velocity data. Experimental validation under quasi-static and dynamic conditions demonstrates repeatability, with mass estimates accurate to within 1% of ground truth. Observations reveal tension asymmetries and hysteresis effects induced by friction, which explain the residual platform sway often observed under conventional position control. These findings underscore the need for friction-aware modeling and real-time tension feedback in control strategies. The proposed identification approach provides a scalable method for characterizing frictional dynamics in underactuated or suspended CDPR configurations.