Loading...

Fuzzy Control of a Quadrotor

Sheikhpour, Soroush | 2013

699 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 45261 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Bagheri Shouraki, Saeed
  7. Abstract:
  8. In recent years، one of the most popular and wide areas of research have been focused on Unmanned Aerial Vehicles (UAVs). Among them، Quadrotor which has the capability of Vertical Take-off and Landing (VTOL)، hovering، great agility and maneuverability، simple structure، light weight and reasonable size has particular importance. In the industry، we always deal with plants which there are no precise information about their dynamic model، or their model is too complex to identify. So designing controllers with less dependency to system model and no need to exact knowledge of the system is absolutely vital. The arrangement of motors in a Quadrotor provides six Degree Of Freedoms (DOF) for this UAV، but the nonlinear and Multi-Input Multi-Output (MIMO) model of the Quadrotor confronts every control algorithms with fundamental challenges. In this thesis two control algorithms which address the issue of control from two different perspectives are utilized to control the attitude of a Quadrotor. In the first algorithm which is known as Parallel Distributed Compensation fuzzy controller (PDC) the process of designing the controller begins with fuzzy identification of a system and by utilizing of the idea of parallel distribution compensation a fuzzy nonlinear controller is provided، which among the fuzzy controllers is the one which its stability can be proven via Lyapunov’s method. The second algorithm which is known as Self Organizinig Controller (SOC), a nonlinear adaptive fuzzy controller is initiated with just preliminary knowledge of the plant and is trained by assessing its performance over the time. The practical result of the first algorithm implies that it successfully controls a Quadrotor with just eight fuzzy rules. The second algorithm which has no dependency on the exact model of the Quadrotor achieves a fuzzy controller in few minutes with tremendous performance
  9. Keywords:
  10. Quadrotor ; Fuzzy Self-Organizing Controller ; Parallel Distributed Compensation (PDC)Controller ; Fuzzy Control

 Digital Object List

 Bookmark

No TOC