Adaptive Quaternion Control for a Miniature Tailsitter UAV

摘要: ADAPTIVE QUATERNION CONTROL OF A MINIATURE TAILSITTER UAV Nathan B. Knoebel Department of Mechanical Engineering Master Science The miniature tailsitter is a unique aircraft with inherent advantages over typical unmanned aerial vehicles. With the capabilities both hover and level flight, these small, portable systems can produce efficient maneuvers for enhanced surveillance autonomy little threat to surroundings system itself. Such vehicles are accompanied control challenges due two different flight regimes. Problems conventional attitude representation arise in estimation as departs from conditions. Furthermore, changing dynamics limitations modeling sensing give rise significant design challenges. Restrictions computation also result limited size weight capacity airframe. In this research, discussed above addressed computationally adaptive quaternion algorithm. backstepping method model cancellation consistent tracking reference derived. This used conjunction algorithms designed identification parameters. For metric baseline performance, gain-scheduled feedback developed. regularized data-weighting recursive least-squares parameter algorithm, controller shown be better than simulation hardware results. universal performance all three surface actuators (aileron, elevator, rudder) barring saturation assuming accurate identification. Testing requires development quaternionbased navigational estimation. novel technique north/east position Also, altitude hover, given an inconsistent thrust system, original on-line throttle Means quaternion-based produced adaptations made existing techniques employed Brigham Young University Multi-Agent Coordination Control Lab. Also generated simple trajectories transitions between modes. developed, which uses multiple sensors combined filtering similar fixed-gain Kalman filter. Simulation results methods presented concept validation. discussion production testing means (a environment test system) provided. culmination, fully autonomous demonstrating its various capabilities.