Dynamic tube model predictive control for a class of soft manipulators with fluidic actuation

Abstract

Controlling soft continuum manipulators is a challenging task due to the structural compliance of these systems which results in underactuation. While a variety of control approaches have been proposed for regulation tasks, tracking control requires further investigation. To this end, a new dynamic tube-Model Predictive Control formulation is presented, which includes an ancillary control law constructed analytically with an energy-shaping approach. Two different models are considered: (i) a rigid-link model with direct-torque actuation, which is representative of miniature soft continuum manipulators supplied by digital pressure regulators; (ii) an extended model that includes the pressure dynamics of the fluid, which is representative of fluidic actuation by means of a variable flow rate. The effectiveness of the proposed algorithm is demonstrated with numerical simulations considering bounded external disturbances, either constant or time-varying.