Nonlinear Model Predictive Control for Pneumatic Driven Quasi Continuum Manipulators

With the ongoing digitalization of manufacturing, an increasing amount of production steps is carried out by robots. Typically they work autonomously, however so-called soft robots allow to further increase their interactivity with human workers. Yet, the modeling and control of such soft robots is still a relatively new research field. In particular, the accuracy and speed of the existing controllers is not adequate for industrial use. Therefore, this paper proposes the application of nonlinear model predictive control (NMPC) for a continuum manipulator as a special case of soft robotics. Successful experimental results are presented using the sequential quadratic programming real-time iteration (SQP-RTI) scheme for embedded optimization. The controller performance is superior to a nonlinear PD controller, which is used as a benchmark. In addition, the proposed NMPC allows to explicitly define state and input constraints on the manipulator, which can be exploited for further controller improvement in future work.