Waco, Texas
March 24, 2021
March 24, 2021
March 26, 2021
23
10.18260/1-2--36412
https://peer.asee.org/36412
597
Andres Enriquez Fernandez was born in Ciudad Juarez, Chihuahua, Mexico. He earned his bachelor’s degree in Mechanical Engineering at the University of Texas at El Paso. After graduating in Spring of 2006, he started working full-time at an automotive company’s technical center as a product development engineer in Ciudad Juarez.
While working full-time in Fall of 2017, Andres returned to The University of Texas at El Paso to start the master’s graduate school program in Mechanical Engineering. While obtaining his master’s degree, he moved to Milwaukee, Wisconsin in 2020 to work full-time at a motorcycle’s company development center as a CAE Engineer.
Dr. Everett is the MacGuire Distinguished Professor of Mechanical Engineering at the University of Texas El Paso. Dr. Everett's current research is in the areas of Mechatronics, Freshman Programs and Student Engagement. Having multiple years of experience in several National Laboratories and Industries large and small, his teaching brings real world experiences to students. As a former NSF Program Director he works regularly helping faculty develop strong education proposals.
Miguel Cedeno is Adjunct Professor at The University of Texas at El Paso. He received his B.S. in Mechanical Engineering from ESPOCH, his M.S. and Ph.D. in Petroleum Engineering at Missouri University of Science and Technology. His research areas include artificial intelligence, machine learning applied to aerospace and mechanical engineering. He works with CFD applied to refinery equipment design for oil and gas industry. He lectures Thermodynamics, Heat Transfer, Fluid Mechanics, Thermal System Design (Heat Exchanger Design) and VBA Applications for Mechanical Engineers. He is a member of the American Society of Mechanical Engineers (ASME) and the Society of Petroleum Engineers (SPE).
Parameter identification of Unmanned Aerial Vehicles (UAV) is very helpful for understanding cause-effect relationships of physical phenomenon, investigating system performance and characteristics, fault diagnostics, control development/tuning, and more. Traditional ways of performing parameter identification involve establishing a mathematical model that describes the system’s behavior. The equations in the model contain parameters that are estimated indirectly from measured flight data. This parameter identification process requires knowledge of the physics involved. Also, it necessitates a careful consideration of the aircraft instrumentation for accurate measurements. It also requires careful design of the flight maneuvers to ensure thorough excitation of the flight dynamics involved. Finally, one must select a suitable identification method. The purpose of this paper is to show the application of machine learning for parameter identification of a UAV model. The machine learning algorithm does not require developing parameterized models; hence it is an equation-less identification method. To provide input to the system, a simulation model of the aircraft is generated. The parameters of the model can be modified in the simulation. The aircraft flight measurement data is obtained directly from the model as simulation outputs from a predetermined flight path. The data is submitted to a machine learning algorithm that is able to read and recognize the data. The machine learning algorithm is trained with a set of flight data that incorporates variations in the parameters to be identified. Finally, the algorithm is tested by feeding unknown flight data to predict the output. To achieve autonomous and consistent flights, a Software-In-the-Loop (SIL) simulation is constructed between X-Plane and Mission Planner. X-Plane is a realistic flight simulator where the UAV model is created, and flight physics are modeled. Mission Planner is the Ground Control Station (GCS) that generates and sends the flight commands to be executed in X-Plane. Several machine learning regression models are explored including linear regression, regression trees, Gaussian process regression, support vector machines and ensembles of regression trees.
Enriquez Fernandez, A., & Everett, L. J., & Cedeno, M. (2021, March), UAV PARAMETER ESTIMATION THROUGH MACHINE LEARNING Paper presented at ASEE 2021 Gulf-Southwest Annual Conference, Waco, Texas. 10.18260/1-2--36412
ASEE holds the copyright on this document. It may be read by the public free of charge. Authors may archive their work on personal websites or in institutional repositories with the following citation: © 2021 American Society for Engineering Education. Other scholars may excerpt or quote from these materials with the same citation. When excerpting or quoting from Conference Proceedings, authors should, in addition to noting the ASEE copyright, list all the original authors and their institutions and name the host city of the conference. - Last updated April 1, 2015