AL-TUNE: A Family of Methods to Effectively Tune UAV Controllers in In-flight Conditions

Abstract In the paper, a family of novel real-time tuning methods for an unmanned aerial vehicle (UAV) altitude controller in in-flight conditions. The allow controller’s gains to be adapted only on basis measurements from basic sensory equipment and by constructing optimization cost function on-line fashion with virtually no impeding computational complexity; case as this paper hexacopter, were used only. are not dependent measurement level, present approach generally applicable form arbitrary controllers low number parameters. Real-world experimental flights, preceded simulation tests, have shown which method should behave best noisy environment when e.g. wind disturbances act UAV while it is autonomous flight. As can potentially extended other control loops or types, making versatile, rapid-tuning tool. It has been that well-tuned using proposed AL-TUNE scheme outperforms tuned just stabilize system. provides new way UAVs terms adaptivity changing their dynamic properties deployed rapidly. This enables applications extends usability fully UAVs, unlike methods, basically require availability model. core difference respect research field authors either use model optimize analytically machine learning techniques, what increases time consumption, whereas presented offer rapid tune controllers, reliable way, deterministic requirements.