Abstract:
Electricity supply outages in urban areas in developing countries like Sri Lanka has a significant impact on the social life as well as the country’s economy. Owing to aging of network assets, the frequency of system failures is increasing and therefore, it is a major challenge for utilities to reduce the electricity outages without compromising essential maintenance needs. Therefore, it has become a mandatory requirement to find advanced automated solutions to upgrade traditional procedures in order to reduce the maintenance time.
A deep study about the process flow for planned and un-planned maintenance reveals that inspection is a necessary activity which significantly contributes to the outage time. Apparently, maintenance scheduling also solely depends on inspection results and the quality of results directly influence efficiently of fault clearing or routine maintenance process. Hence, this research was done to develop a solution to reduce inspection time while improving the quality and safety of operation. The scope of the study was identified as the inspection process of short span MV feeders in urban areas and the properties and parameters of this type of feeders were taken to develop the methodology.
As outcome of this research, a quadcopter based aerial observation platform was developed to remotely investigate MV feeders while mitigating the electromagnetic interference. The platform has an imaging unit which transmits images to a ground station and computer based program process images further and identifies defects. In the scope of this research, the MATLAB based algorithm was developed to identify corroded nuts in transformer terminals. The same algorithm can be extended to identify other specific network defects and faults and quadcopter design can be rescaled to operate in a long range as per the desired industrial requirement.
