University of British Columbia
Network and fault parameter estimation and the Cramér-Rao lower bound in power line communications
Abstract
dc:descriptionPower Line Communications (PLC) enables existing power line infrastructure to be used not only for communication, but also as a sensing platform for grid diagnostics. This thesis investigates model-based estimation of network and fault parameters from measured PLC channel transfer functions (CTFs). The theoretical foundations of transmission line (TL) and multi-conductor transmission line (MTL) theory are developed alongside the principles of frequentist and Bayesian estimation theory, numerical optimization, and the Cramér-Rao Lower Bound (CRLB) and Bayesian Cramér-Rao Lower Bound (BCRLB). These techniques are then applied to two models. First, a simple point-to-point two-conductor TL model where we study maximum likelihood estimation of fault and network parameters and investigate the effects of the likelihood landscape, frequency range, cable length, and numerical optimization on estimator performance. Second, a realistic branched low-voltage (LV) network is modeled using MTL theory with frequency-dependent cable and load models. A two-stage Bayesian inference framework based on mean-field stochastic variational inference is developed, in which network parameters are first calibrated under healthy operating conditions and subsequently used to estimate fault location and impedance after fault detection.
Degree
thesis:*- Name thesis:degree_name
- Master of Applied Science - MASc
- Level thesis:degree_level
- master's
- Discipline thesis:degree_discipline
- Electrical and Computer Engineering
- Grantor dc:publisher
- University of British Columbia
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Chambers, Francis
- Contributors dc:contributor
-
- Lampe, Lutz
Rights
dc:rights- Statement dc:rights
-
- Attribution-NonCommercial-NoDerivatives 4.0 International
- Language dc:language
- eng
Identifiers
dc:identifier.*- Handle dc:identifier
- http://hdl.handle.net/2429/95338