Adaptive Model-based Compensation with Robust Tuning for Real-time Hybrid Simulation
Abstract
In this study, an adaptive compensation method is utilized to design a controller capable to compen- sate for different experimental substructures. The initial conditions of the controller are defined using a model of the transfer system without considering the interaction with any experimental substruc- ture. Then, adaptive compensation is implemented to update the control parameters during the test, maintaining excellent compensation when the experimental substructure interacts with the actuator. The proposed methodology is validated through numerical simulations in a virtual RTHS benchmark problem, but incorporating more complex scenarios such as different experimental substructures for the same controller and non-linear experimental substructures. The results show excellent performance in terms of accuracy and robustness. Notice that a controller independent from the experimental sub- structure allows to avoid the previous test of the physical specimen to design a controller, and the adaptation capacity allows to keep a good compensation in presence of time-varying properties of the experimental substructure.
