Thesis
Vehicle position tracking using visual data in under-tank air slot environment
Washington State University
Master of Science (MS), Washington State University
12/2020
DOI:
https://doi.org/10.7273/000004219
Handle:
https://hdl.handle.net/2376/125039
Abstract
Under-tank inspection is necessary to monitor aging waste-tank infrastructure at the Hanford site near Richland, WA. Integrity inspections inform critical decision making to prioritize waste removal for vitrification. A mother- and child-bot system delivers a nondestructive evaluation (NDE) sensor within the under-tank air-slot system to gather tank bottom integrity data. Position tracking utilizing a cable encoder suffers from slippage and accumulation of error. Further development of independent position tracking techniques is needed to guide sensor placement and determine relative position within ± 0.5 inch such that multiple sensor reading can be combined into a composite map of tank-integrity data. The work in this thesis includes the development, testing and application of position tracking methods for a semi-autonomous vehicle in the under-tank air-slot environment. The method developed combines laser-ranging with front and rear mono-camera setups to calculate a relative change in distance from observed features as the vehicle moves within the highly constrained environment. Laser ranging data is used as a primary distance measurement tool with visual data used to confirm the laser reflection surface. Vehicle behavior and air slot geometry leads to uncertainty in laser ranging data for brief periods of time. A separate visual only method - utilizing edge detection and a perspective correcting algorithm - serves as a secondary position tracking system. This secondary system has potential to suffer from an accumulation of error but provides accurate readings for short distances until laser ranging data is determined to be reliable. Laser ranging and video data are processed and converted into a series of relative position changes that are input into a geometric model. The geometric model is based on construction drawings of the double-shell tank refractory pad and is used for position mapping and location reporting. The model supports the visualization of vehicle position within the air-slot system and provides operators with distance-to-waypoint estimates that can be used to confirm vehicle position within specific segments by comparing to visual data. Considerations for implementing such a system, including automation and adaptation of corrective models based on specific vehicle behaviors, are also included.
Metrics
6 File views/ downloads
24 Record Views
Details
- Title
- Vehicle position tracking using visual data in under-tank air slot environment
- Creators
- Christopher Jon Cree
- Contributors
- John H Miller (Advisor) - Washington State University, School of Engineering and Applied Sciences (TRIC)
- Awarding Institution
- Washington State University
- Academic Unit
- School of Engineering and Applied Sciences (TRIC)
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University
- Identifiers
- 99900896436101842
- Language
- English
- Resource Type
- Thesis