Forecasting Flicker Severity by Grey Predictor

Rapid voltage fluctuations in an electric power network may produce significant levels of flickers, which have negative impacts on human eyes and power system components. This paper proposes a grey predictor model for the forecast of the flicker severity level associated with operating a large electric arc furnace load. Actual measured flicker index data are adopted to implement the predictor model. Test results based on the proposed model are compared with another neural-network-based method. It shows that a more accurate forecast is achieved by using the proposed grey predictor model.