OKID-based identification and control of unmanned ducted-fan helicopter

This paper presents system identification and control of unmanned helicopter(UH) with ducted fan. The plant is an unstable multi-input multi-output (MIMO) system. Four identification methods have been considered in this work. Two of these methods studied the identification of the system as a set of transfer functions relating different inputs and outputs namely least square with QR-factorization (LS-QR) and recursive least square method (RLS). On the other hand, subspace method and observer/Kalman filter identification (OKID) methods have been used to identify the system in state space representation. The results of the identifications are presented and benchmarked based on degree of fitness. Comparison between output and methods are conducted and analyzed. OKID proves to be the best method for such system and subsequently serves as the basis for the design of LQGI-based velocities tracking controllers.