Development of neural network-based computer vision system for automated grading operation of a hydraulic excavator

Show simple item record

dc.contributor Williams, Keith A.
dc.contributor Shen, Xiangrong
dc.contributor Vikas, Vishesh
dc.contributor Li, Shuhui
dc.contributor.advisor Yoon, Hwan-Sik
dc.contributor.author Xu, Jiaqi
dc.date.accessioned 2018-12-14T18:12:00Z
dc.date.available 2018-12-14T18:12:00Z
dc.date.issued 2018
dc.identifier.other u0015_0000001_0003103
dc.identifier.other Xu_alatus_0004D_13601
dc.identifier.uri http://ir.ua.edu/handle/123456789/5235
dc.description Electronic Thesis or Dissertation
dc.description.abstract Hydraulic excavators are widely used in the construction and mining industries. While conventional hydraulic excavators have been manually operated by a human operator, automated systems are being developed as an effective alternative to manual operations for common tasks that excavators routinely perform in typical work sites. An example is an automated ground grading system that can enhance the productivity of an excavator by assisting an operator to perform ground grading in a fast and accurate manner. For this purpose, a sliding mode controller is developed for automated grading of a hydraulic excavator in this research. First, an excavator manipulator model is developed in Simulink by using SimMechanics and SimHydraulics toolboxes. Then, the sliding mode controller is used to control the manipulator to trace a predefined trajectory for a grading task. The simulation results show that the sliding mode controller can control the grading operation with less tracking errors than a PI controller. As an alternative to conventional displacement sensors in an automated excavation system, a novel approach to estimate the position of a hydraulic manipulator using a neural network-based vision system is also studied in this research. A webcam is used to capture images of a moving manipulator, and the captured images are used to train neural networks. Then, the trained neural networks can be used to estimate the position of the excavator manipulator for a feedback control system. A case study was conducted to investigate the factors that affect the performance of the neural networks. A simulation study shows a stable grading performance when a PI controller is used to control the manipulator based on the estimated manipulator position.
dc.format.extent 403 p.
dc.format.medium electronic
dc.format.mimetype application/pdf
dc.language English
dc.language.iso en_US
dc.publisher University of Alabama Libraries
dc.relation.ispartof The University of Alabama Electronic Theses and Dissertations
dc.relation.ispartof The University of Alabama Libraries Digital Collections
dc.relation.hasversion born digital
dc.rights All rights reserved by the author unless otherwise indicated.
dc.subject.other Mechanical engineering
dc.subject.other Computer science
dc.title Development of neural network-based computer vision system for automated grading operation of a hydraulic excavator
dc.type thesis
dc.type text
etdms.degree.department University of Alabama. Department of Mechanical Engineering
etdms.degree.discipline Mechanical Engineering
etdms.degree.grantor The University of Alabama
etdms.degree.level doctoral
etdms.degree.name Ph.D.


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search DSpace


Browse

My Account