A group of students from Dr. Andrea L’afflitto’s Flight Controls class created the following video:
According to Dr. L’afflitto, this project consisted of designing an autopilot for a quadrotor using some modern, very aggressive control techniques. The purpose of this video is to show the results achieved graphically, however, the mathematical models, the control design problem and the numerical simulations have very deep roots.
“I am extremely proud of their work because these are all undergraduate students, but the quality and the mathematical complexity is the one of a graduate project,” said Dr. L’afflitto. “We all can imagine the impact of the development of such technology, considering the growing attention that OU is putting on the UAS technology.”
This video shows the result of a students’’ project developed as part of the AME 4513/5513 “Flight Controls” course at the University of Oklahoma in Fall 2016. A DJI F450 will inspect some buildings of OU’s main campus. The drone’s autopilot implements an algorithm based on Model Reference Adaptive Control.
An important feature of this simulation is that the quadrotor dynamics is not captured by a set of nonlinear differential equations, but it is deduced from a SimMechanics model of a DJI F450. This guarantees high accuracy of the results presented.
The adaptive control technology allows precise, aggressive maneuvers in the vicinity of obstacles, such as buildings.
Next, we compare the performance of a quadrotor (in white) implementing an adaptive control law and a quadrotor (in black) implementing a classic PID controller.
Created by: Blake Anderson
Kevin Murray Jr.