If you have any problems related to the accessibility of any content (or if you want to request that a specific publication be accessible), please contact us at scholarworks@unr.edu.
System Identification and Control Modeling of 6-Rotor Unmanned Air Craft for Wildfire Spread Monitoring
Date
2024Type
ThesisDepartment
Computer Science
Degree Level
Master's Degree
Abstract
Wildfires are natural occurrences that have been significantly acknowledgedas devastating natural disasters globally. The fuel type, temperature, wind and
humidity are some of the elements that directly affect the orientation and propagation
of the fire in the area. For tracking and monitoring purposes, traditional
methods on the ground and in the air constitute great safety risks, especially considering
unpredictable fire behavior. Therefore, there is an increasing need for
novel options for field monitoring approaches to minimize risk to reach real-time
information. The technology to reduce the impacts of wildfires involves the use
of aerial unmanned systems in order to leverage efficient mapping and real-time
surveillance.
This thesis aims to identify the system dynamics and develop a control strategy
for tracking wildfire spread in vast green areas using the DJI Matrice 600 Professional
6-rotor unmanned air vehicle (UAV). Flight data is collected to identify and
generate the mathematical model of the aircraft in each axis motion. Dynamic
identification is conducted through a time-domain dataset in the Matlab toolbox,
driving the Model Predictive Control algorithm. Sensory information is processed
on a high-level onboard computer running a Robot Operating System (ROS).
In this work, we contribute system identification, modeling, and control for
commercial grade drones. MPC is implemented as the preferred model to resolve
proper flight routes for UAV missions. For position control, the linearized dynamics
of the air vehicle has been deployed at the core. Through both experiment
and simulation studies, the proposed autonomous aerial vehicle’s performance has
been demonstrated. Aerial wildfire monitoring in high-risk areas can be carried
out without human intervention using the proposed technique, which is considered
achievable.
Permanent link
http://hdl.handle.net/11714/12705Additional Information
Committee Member | Feil-Seifer, David; Xu, Hao |
---|