Development of a deep learning software for visual analysis of high voltage insulators

Savva, Timotheos and Albano, Maurizio ORCID: https://orcid.org/0000-0002-5486-4299 2020. Development of a deep learning software for visual analysis of high voltage insulators. Presented at: 55th International Universities Power Engineering Conference (UPEC 2020), Virtual - Torino, Italy, 1-4 September 2020. 2020 55th International Universities Power Engineering Conference (UPEC). IEEE, pp. 1-6. 10.1109/UPEC49904.2020.9209804

Full text not available from this repository.

Abstract

Silicone rubber insulators have been favoured over conventional insulators to be adopted on high voltage power systems due to their hydrophobic properties and better performance in polluted environments. However, dry-bands and discharges can still be initiated on the polluted surfaces of polymeric insulators causing surface degradation and lifetime reduction. This paper presents a new procedure that identifies and assesses location and frequency of electrical stresses such as discharges and partial arcs on high voltage insulator surface analyzing visual recordings. The procedure is based on image analysis and deep learning techniques to be fully automatic and to minimize user intervention. Also, a MATLAB GUI was developed to provide a fast and user-friendly control of all steps of the analysis and its results.

Item Type:	Conference or Workshop Item (Paper)
Date Type:	Published Online
Status:	Published
Schools:	Engineering
Publisher:	IEEE
ISBN:	9781728110783
Last Modified:	09 Nov 2022 09:40
URI:	https://orca.cardiff.ac.uk/id/eprint/136575

Actions (repository staff only)

Edit Item