Non-linear System Identification for UAS Adaptive Control

Fast aircraft prototyping, fault detection, morphing surfaces, and real-time generation of dynamic models are just some of the advantages of a model identification adaptive controller (MIAC). The research presented in this paper introduces a MIAC architecture and validates a novel data-driven algorithm to be used for online system identification of unmanned aerial system (UAS). The simulation results illustrate the effects and the limits of short training time and sensor noise on the identified model.