Higher-Order Model-Free Control Tuned byFictitious Reference Iterative Tuning for 3D Cranes

The current paper proposes to mix two data-driven algorithms, i.e. Model-Free Control (MFC) and Fictitious Reference Iterative Tuning (FRIT). The MFC algorithm has the main advantage of not using the precise process model, and the FRIT algorithm has the main benefit of determining the optimal parameters of the controller using a set of initial input/output data. Therefore, the mix between these two algorithms improves the performance of the overall control loop with MFC algorithm from one iteration to the next since the optimal parameters of the MFC will be tuned via FRIT. The benefit of this mix is that the parameters of the MFC algorithm are optimally computed using FRIT. An empirical sensitivity study is performed, where higher-order MFC algorithms are tuned via FRIT. The study aims to determine how a higher-order MFC improves the performance of the control loop. The resulting MFC-FRIT algorithms are experimentally validated on the 3D crane laboratory equipment