Monash University

Embargoed and Restricted Access

Reason: Under embargo until June 2018. After this date a copy can be supplied under Section 51(2) of the Australian Copyright Act 1968 by submitting a document delivery request through your library

Radiometric detection of partial discharge sources using smart antennas

posted on 2017-01-23, 22:32 authored by Bandara, Uditha Wijethilaka
Radiometric partial discharge (PD) detection is an attractive technique with potential for condition monitoring of live assets in high voltage power networks. Reported developments in this domain have resulted in handheld PD probes and time of arrival-based source localization using high speed digital sampling oscilloscopes. However, due to the operator involvement in using handheld probes and the cost associated with high-speed digital sampling oscilloscopes, the practical applicability of radiometric PD detection has been limited. A review of the reported literature shows that these systems make use of basic antenna structures for the reception of radiated PD signals. Investigations on a purpose-built automated antenna system for PD detection have not been reported in open literature. This research project investigated the development of an automated condition monitoring system using a smart antenna. The smart antenna not only facilitates detection of PD, but also provides location information of the radiation source within the switchyard. The developed smart antenna system consists of a phased array antenna and an automated PD detector. The phased array antenna is designed with high gain and agile beam for operation in a frequency band where significant power from the radiated PD signal is concentrated. Signals detected by this antenna are fed to a digitizer for automatic identification of PD occurrences. Antenna components,including broadband antenna elements, power dividers, and a digital beamforming network for electronic control, have been developed from scratch. A low-cost digital phase shifter for use in this beamforming network has been designed, using LC-compensated band switching PIN diodes. A frequency down-converter has been designed to facilitate PD detection using a low-cost waveform digitizer. The digitized signal is processed to automatically identify the presence of PD signals. The location of the PD source is estimated by combining this PD detector result and orientation of the antenna main-lobe. A novel post-processing technique based on sensor fusion has been developed to improve the location identification of radiating sources using the smart antenna.


Campus location


Principal supervisor

Nemai Chandra Karmakar

Year of Award


Department, School or Centre

Electrical and Computer Systems Engineering


Doctor of Philosophy

Degree Type



Faculty of Engineering

Usage metrics

    Faculty of Engineering Theses


    No categories selected


    Ref. manager