A priori model inclusion in the multisine maneuver design

The aim of the project was to determine if including a-priori information in the system identification multisine maneuver design process will improve the quality of the estimated data in the lateral motion. During this study we used a linear aircraft model which was containing a-priori information in terms of aerodynamic derivatives. A-priori knowledge allowed us to obtain a non-uniform power spectrum of the signals that excited the ailerons and rudder. Three cases were analyzed for comparison. The first design was obtained when no a-priori information was available. The second case was designed on the basis of the a-priori model state matrix. We found dominant frequencies of the system and assigned to them more power. In the third case the non-uniform power spectrum was optimized. A genetic algorithm was used for that purpose. The efficiency of the design was optimized thought out Relative Peak Factor. The a-priori information was introduced by using Marchand methodology. All the obtained maneuvers were used to excite the aircraft model. The response of the system was registered. Recorded data was used later to perform a parameter estimation for each case. In the end we compared the results from the identification process in terms of the accuracy.