Non-linear modelling of a one-degree-of-freedom twin-rotor multi-input multi-output system using radial basis function networks

Modelling of innovative aircraft such as unmanned air vehicles (UAVs), X-wing, tilt body and delta-wing aircraft is not easy. It is argued in this paper that non-linear system identification is suitable for modelling air vehicles of complex configuration. This approach is demonstrated through a laboratory test rig. Extensive time and frequency-domain model-validation tests are employed in order to instil confidence in the estimated model. The estimated model has a good predictive capability and can be utilized for non-linear flight simulation studies. Some aspects of the modelling approach presented may be relevant to flight mechanics modelling of new generations of air vehicle.