Loading [MathJax]/extensions/MathZoom.js
Forecasting Electricity Outage in KwaZulu-Natal, South Africa using Trend Projection and Artificial Neural Networks Techniques | IEEE Conference Publication | IEEE Xplore

Forecasting Electricity Outage in KwaZulu-Natal, South Africa using Trend Projection and Artificial Neural Networks Techniques


Abstract:

The majority of faults on electric power systems are traceable to distribution networks due to the exposure of the networks to different adverse climatic conditions, amon...Show More

Abstract:

The majority of faults on electric power systems are traceable to distribution networks due to the exposure of the networks to different adverse climatic conditions, among other factors. Power outages due to climatic factors are inevitable since a major part of a distribution network is made up of overhead lines and exposed to different unfavorable weather conditions. Climatic factors need to be fully considered when designing a power system model in order to achieve a meaningful system reliability improvement. The projection of power system outage requires a model with high accuracy, taking climatic conditions into consideration. Conventional models rely only on fault data and do not consider climatic factors, and so they have very low accuracy. This research proposes a computational intelligence model using historical fault and climatic data sets of Newcastle, South Africa. Three models are developed in this paper: the trend projection (TP) model, and two artificial neural network (ANN) models (Models 1 and 2). The performance of the models are examined using statistical parameters. The results of the ANN models were satisfactory unlike the trend projection model; thereby illustrating the efficacy of the computational methods.
Date of Conference: 23-27 August 2021
Date Added to IEEE Xplore: 28 September 2021
ISBN Information:
Conference Location: Nairobi, Kenya

Funding Agency:


I. Introduction

Utilities today put great emphasis on power system reliability because it takes a reliable system to achieve customer satisfaction [1]. Reliability in power systems is the degree to which the utility services can be depended on to be operational and predictable. Consideration must be given to climatic factors that are often responsible for electricity supply disruptions. This would improve distribution system reliability. Distribution networks are made up of overhead lines which are exposed to different weather conditions, often unfavorable conditions. Existing power outage prediction models make use of historical fault data without due consideration of climatic data; hence, the projections have low accuracy, which leads to unreliability.

References

References is not available for this document.