Nonlinear Observer-Based Tension Distribution for Cable-Driven Parallel Robots

This paper addresses a new methodology for positive tension distribution in cable-driven parallel robots (CDPRs) employing a nonlinear disturbance observer. Cables can merely apply positive tensile force; thus, a CDPR’s control scheme should keep the cable tension positive. This is generally provided with a redundancy resolution (RR) scheme, which maps the control effort into a positive range without changing the motion characteristics of the robot. RR methods are Jacobian-dependent and prone to kinematic uncertainties. Thus, such uncertainties induce extra errors in the computation of the cable tension, apart from the control scheme sensitivity to dynamic uncertainties and external disturbances. Additionally, RR demands high-performance computing due to the optimization-based nature of the approach. The proposed control strategy eliminates the complexity of having RR in the control architecture by incorporating a positive saturation-type function to map the control effort into a positive range and a nonlinear disturbance observer to compensate for the saturation effects. The stability of the proposed control scheme is investigated through Lyapunov’s second method in detail. Furthermore, its performance is compared with a conventional RR-based method for a spatial six-degree of freedom CDPR, considering uncertainties .