Quaternion Nonlinear Dynamic Inversion Autonomous Flight Controller for Supermaneuverable Aircraft

  • 康 心奕

Student thesis: Master's Thesis


The control of aerial vehicle is always one of the major challenges in aerospace technology development High agility maneuvers on fighter jets are still one of the most seriously treated properties of the modern fighter design The ability to make aggressive maneuvers often means first the vehicle has unstable dynamics and second the vehicle would constantly go through extreme dynamic conditions such as excessive angle of attack (AOA) or high rotation rate The instability of such vehicle will require a computer assisted control system the high ? and body rate condition mean the control system would have to deal with the nonlinearity of the vehicle dynamics In this thesis we utilize the control method of Nonlinear Dynamic Inversion (NDI) The NDI is sometimes known as nonlinear feedback linearization Basically with full knowledge of vehicle dynamics the controller would try to predict and cancel the nonlinear behavior of the system With complete nonlinear aerodynamic and kinematic models of the high angle-of-attack research vehicle (HARV) Nonlinear Dynamic Inversion (NDI) control equations are derived Another aspect when dealing with such agile vehicles is their rapid changing omni-directional nature We need some kind of mathematics to avoid the problems of using traditional aircraft Euler-angle expressions The quaternion is the suitable solution Under the rules of quaternion calculation not only the attitude of the aircraft can be represented important states of aerodynamic angles can also be calculated via the quaternion method Last but not least the control development wouldn’t be complete without the verification of experiments In addition to the numerical simulation of the aircraft model and control scenes a micro aerial vehicle system is designed and built The scaled down model fully mimics all aspects of the full-scale vehicle from aerodynamic configuration to actuators This aircraft model is fitted with powerful commercially available flight controller and sensors State estimation algorithms using quaternion for such MAV flight controller are implemented into the firmware to provide constant and stable execution of the program Dozens of experiments are carried out along the process of developing the MAV system Assisting equipment and procedures are invented to guide the process Control algorithms and practical logic sequences designed for the experiment are developed and implemented Experiences methods and solutions to constructing the aircraft build up along the path The complete autonomous aerial vehicle system is presented in this thesis
Date of Award2016 Aug 2
Original languageEnglish
SupervisorJiun-Haur Tarn (Supervisor)

Cite this