An exploratory analysis on half-hourly electricity load patterns leading to higher performances in neural network predictions

Accurate prediction of electricity demand can bring extensive benefits to any country as the forecasted values help the relevant authorities to take decisions regarding electricity generation, transmission and distribution appropriately. The literature reveals that, when compared to conventional time series techniques, the improved artificial intelligent approaches provide better prediction accuracies. However, the accuracy of predictions using intelligent approaches like neural networks are strongly influenced by the correct selection of inputs and the number of neuro-forecasters used for prediction. Deshani, Hansen, Attygalle, & Karunarathne (2014) suggested that a cluster analysis could be performed to group similar day types, which contribute towards selecting a better set of neuro-forecasters in neural networks. The cluster analysis was based on the daily total electricity demands as their target was to predict the daily total demands using neural networks. However, predicting half-hourly demand seems more appropriate due to the considerable changes of electricity demand observed during a particular day. As such clusters are identified considering half-hourly data within the daily load distribution curves. Thus, this paper is an improvement to Deshani et. al. (2014), which illustrates how the half hourly demand distribution within a day, is incorporated when selecting the inputs for the neuro-forecasters.