Data-based tuning of linear controllers for MIMO twin rotor systems

Abstract

This paper proposes the application of Iterative Feedback Tuning (IFT) as a data-based control technique to parameter tuning of linear controllers for Multi Input-Multi Output (MIMO) twin rotor systems. The azimuth and pitch position control are carried out using a MIMO control system structure with two control loops, one for each position, and two decoupling controllers. The initial controller design is based on zero cross-coupling between the two control loops; therefore, the azimuth and pitch controllers are tuned in terms of the Modulus Optimum method and the initial transfer functions of the two decoupling controllers are zero. An IFT algorithm is applied to tune the four controllers in order to ensure the control system performance improvement by the experiment-based iterative solving of a model reference tracking optimization problem. A set of experimental results is given to show the strong performance improvement after few IFT algorithm iterations.