Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.581903
Title: Positioning for mobile NDE inspection robots
Author: Summan, Rahul
Awarding Body: University of Strathclyde
Current Institution: University of Strathclyde
Date of Award: 2012
Availability of Full Text:
Access through EThOS:
Access through Institution:
Abstract:
Ageing infrastructure worldwide requires periodic inspection, often in-situ, in order to ensure continued safe and economic operations as well as adherence to stringent quality and performance requirements. In service automated Nondestructive Evaluation, where feasible, is highly attractive, and potentially allows inspection of operational plant. The use of such technology is very attractive in terms of safety, cost and the potential for minimal disruption to the inspection site especially if plant operations can remain online. Automated Nondestructive Evaluation in the form of remotely operated robotic vehicles is an active area of research. Knowledge of position relative to a frame of reference is a key aspect for a robotic Nondestructive Evaluation system in order to associate sensor measurements with locations on the structure being investigated. This thesis investigates relative and absolute positioning techniques for a single robot. The accuracy and repeatability of a photogra mmetry system is characterised over a large volume using a high accuracy metrology instrument. It was found that the photogrammtery system was most accurate in the centre of the volume and least accurate at the edges. This photogrammetry system is then used to evaluate the performance of algorithms developed in subsequent research. An image based positioning system is implemented which extracts motion information from a camera carried onboard a robot. The system is evaluated on surfaces typically found in industrial environments. Ultrasonic ranging techniques are investigated for robot positioning. In particular a low cost, modular, ultrasonic positioning system is characterised and calibrated. Bayesian filtering in the form of an Extended Kalman and Particle Filter are implemented to fuse noisy optical encoders estimates available at 100 Hz and the ultrasonic positioning measurements available at 3 Hz. The extended Kalman Filter, at lower computational cost, was found to produce the lowest error.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.581903  DOI: Not available
Share: