Integrated System Design for Intelligent Subterranean Exploration using Experimental UAVs

• Author / Creator

  Allan, Shane

• Unique challenges arise when attempting to navigate unknown GPS-denied subterranean environments. To successfully explore these environments a vehicle much be able to reconcile the nonlinear dynamics of motion, accommodate for diverse terrain, adjust to static and moving obstacles, combat degraded sensor conditions, and transmit high quality information. Each of these fields have been investigated independently, however within the scope of this research these challenges are addressed collectively by a prototype unmanned aerial vehicle (UAV) which utilizes a sophisticated motion planner, customized hardware, and artificial intelligence-based image enhancement.
  A motion planner suitable for the described challenges addresses a four-part problem. Generation of a high-fidelity map, exploratory motivation, a global navigational strategy, and an optimized and adaptable local trajectory planner. Each of these sub-tasks are addressed by implementing the Fast Autonomous UAV Exploration (FUEL) algorithm into the prototype platform.
  Limitations to hardware systems present unique boundaries to efficient flight performance. Consumers typically address these issues through costly performance up- grades of existing components. However, when the flight terrain is predetermined, slight geometric modifications to the airframe can be implemented to optimize efficiency. Within this research, the vehicle uses custom-designed twisted rotor arms. This strategy is used to redirect a portion of the vertical thrust to produce greater yaw authority.
  Arguably, the main value of a UAV is measured by the information it transmits to remote operators. Motivated by this, five convolutional neural networks tasked with enhancing degraded low light images are implemented onboard the vehicle. Each of these networks are tested in real-time experiments to quantify their performance and their net benefits to the operator.
  The ultimate goal of this research is to experimentally validate all presented design selections, specifically to understand the limitations and real-world performance of the systems and algorithms used.

• Subjects / Keywords

  • UAV
  • Path Planning
  • GPS-Denied
  • Autonomous
  • CNN
  • Low Light
  • Hardware
  • Tilted
  • Subterranean

• Graduation date

  Fall 2023

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/r3-fab6-x726

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.