Loading...

Modeling and Flight Controller Design of an Insect-Like Flapping Wing in Hover

Khodabakhsh, Hamid Reza | 2011

671 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 41900 (45)
  4. University: Sharif University of Technology
  5. Department: Aerospace Engineering
  6. Advisor(s): Banazadeh, Afshin
  7. Abstract:
  8. There are many applications for micro aerial vehicles (MAV) and in nature they are flying as birds and insects. It can be said that a MAV that can mimic flying capabilities of birds and insects can be used as a solution for flying in low Reynolds number regimes. It seems that Flapping Wing MAVs have better performance than fixed and rotary wings MAV in low Reynolds number regimes. So far there haven’t been many theories and methods introduced for controlling flapping wing MAVs and many researches failed to provide a convenient way for designing an effective controller. On the other hand, models presented in literature for aerodynamics, dynamics and kinematics of flapping wing are not accurate and are oversimplified. As a result a complete model representing the true dynamics of a flapping wing doesn’t exist. The current research attempts to present a more accurate aerodynamic and dynamic model based on the today’s best understanding of insect’s flight. Then kinematic parameters are selected so that we can control the flapping wing with active control of only one degree of freedom of the wing. Because of underactuated nature of the system on one hand, and heavy coupling of aerodynamic forces on the other hand, designing an effective controller is a very hard task. To address this problem the optimal discrete linear feedback matrix is found via Island Parallelization of Non-Dominated Sorting Genetic Algorithm II around various initial conditions around hovering point. Because of highly nonlinear and time varying nature of the system and low performance of discrete linear feedback controller in different flight conditions, input output pairs of discrete linear feedback controller is used to train a neuro-fuzzy system. Finally it has been shown that trained neuro-fuzzy controller has the ability to stabilize and control flapping wing MAV in a wide range of conditions. Also it has been shown that the designed controller is very robust to external disturbance and parameter variations while the discrete linear feedback controller is sensitive to disturbance
  9. Keywords:
  10. Ornithopter ; Hover Flight ; Genetic Algorithm ; Adaptive Neuro-Fuzzy Inference System (ANFIS) ; Micro Air Vehicle (MAV)

 Digital Object List

 Bookmark

No TOC