Thesis Examination Committee
Prof Chih-Chen CHANG, CIVL/HKUST (Chairperson)
Prof Ross MURCH, ECE/HKUST (Thesis Supervisor)
Prof Qing Huo LIU, Department of Electrical and Computer Engineering, Duke University (External Examiner)
Prof Matthew MCKAY, ECE/HKUST
Prof Chi Ying TSUI, ECE/HKUST
Prof Andrew Brian HORNER, CSE/HKUST
Wired, optical and pipeline networks support a plethora of critical systems and services in our society today. These include the distribution of electricity, signals, gas and water in diverse and critical infrastructure such as buildings, transport systems, urban water systems, automobiles, airplanes, electrical grids and wireless communications. Sustainable, safe and reliable operation of these utility networks is therefore critical and requires the availability of techniques for detecting and locating faults that will, or have already, occurred.
In this thesis a common framework to model utility networks is presented and applied to specific domains of application. Specific domains of interest include the investigation of diagnostic techniques for transmission line cable networks using electromagnetic wave propagation and Urban Water Supply Systems using acoustic wave propagation. Contributions to three areas of one-dimensional direct and inverse problems are presented. The first contribution is for the direct problem of characterizing the acoustic waveguide channel when filled with gas or water. Specific findings include the demonstration that for water pipes elastic boundary conditions must be considered and that channel attenuation depends critically on whether propagation is primarily occurring in the pipe wall or in the waveguide medium. The second contribution includes the development of an analytical formulation, based on the Born approximation, for predicting impedance faults in lossy transmission lines and pipelines. Experimental and simulation results show that surprisingly accurate results can be obtained for both transmission lines and water pipelines. The third contribution is the extension of these one-dimensional inverse problem techniques to the estimation of multiple distributed parameters in lossy transmission lines and pipelines. These techniques make use of S-parameters from both ends of the transmission line or pipeline and analytical formula are again provided. Simulation and experimental results are also presented for transmission lines and they demonstrate that both distributed impedance and shunt conductance can be accurately estimated simultaneously when 2-port S-parameters are available. Finally conclusions are provided and future research directions are suggested.