Abstract: An increasing number of robot-based application require the use of manipulators with small weight and limited inertia. Unfortunately these devices are also prone to undesirable vibration effects, which are due to the flexibility of their links. In order to reduce the unwanted vibrations in mechanisms retaining their high-speed motion particular control techniques must be employed. For this reason in this paper an innovative controller for flexible-links mechanisms based on MPC (Model Predictive Control) with constraints is proposed. So far this kind of controller has been employed almost exclusively for controlling slow processes, like chemical plants, but the authors’ aim
is to show that this approach can be successfully adapted to plants whose dynamical behavior is both nonlinear and fast changing. The effectiveness of this control system will be compared to the
performance obtained with a standard control strategy employed in industrial applications. The reference mechanism chosen to evaluate the performance of this control strategy is a single-link planar mechanism laying on the vertical plane driven by a torque-controlled electric actuator. The control strategies for vibration reduction are quite hard to test, since flexible-links mechanisms are quite prone to mechanical failure. In case of a non well-done tuning of the control system
the links are exposed to strong strains, especially when dealing with closed-loop structures. During these tests a frequent replacement of the mechanism links is required, and potential safety risks are encountered. A good solution to those problems can be found in the use of Hardware-In-the-Loop- Simulation (HILS) technology. This emerging technology, which is used mainly to design and test control systems, is based upon the interaction between a real hardware and a virtual system (i.e. a simulation based mathematical model) that emulates and physically replace a real system or one of his components. The main advantage of this approach is that it allows the use of a virtual model directly inside of the control loop: in this way a fine and accurate tuning of the control system parameters can be provided without involving the flexible-link mechanism. For this purpose a simulator named FLiMHILS (Flexible Link Mechanisms HIL Simulator) has been developed,
and will be employed in this paper to test and tune the innovative constrained MPC controller. The results show the optimal performances of control system and the capability of the HIL approach.