Path Following of Cable Suspended Parallel Robots Through Nonlinear Model Predictive Control

Cable-driven parallel robots (CDPRs) are spreading very fast due to their large workspaces and their high payload-to-weight ratio, therefore representing an interesting solution for the next generation of industrial automation. Among the different topologies of CDPRs, the suspended configuration covers a huge interest because it simplifies the robot installation. At the same time, Cable Suspended Parallel Robots (CSPRs) represent a highly challenging configuration from the control design point of view because they lack positive controllability. This paper shows how Nonlinear Model Predictive Control (NMPC) algorithms represent an effective control method to perform path following tasks in the presence of CSPRs; in particular, a two-stage controller is proposed. Firstly, a NMPC algorithm is designed by considering only the dynamical subsystem made of the end-effector, achieving the optimal cable tensions for the execution of the desired path; this procedure easily includes constraints on the feasible tensions. Secondly, the related motor torques are achieved through inversion of the dynamic model of the electric motors. To assess the proposed controller, a 3-dof cable suspended spatial robot, moved by four cables, is considered. Numerical results are reported, showing low contour errors in the execution of the path following tasks and therefore confirming the effectiveness of the proposed method.