Developing an Improved Grey Prediction Model for Application to Electricity Consumption Prediction: Toward Enhanced Model Accuracy

In the economy, one of the crucial trends or processes that play an important role involves electricity consumption prediction. Indeed, forecasting the consumption of electricity with accuracy and precision paves the way for relevant policy makers to establish strategies for electricity supply. Despite this promising and beneficial effect of accurate forecasting, limited variables and data are unlikely to offer adequate data through which satisfactory prediction accuracy might be gained. Due to the need to address this dilemma, this study developed a novel model as an improvement of the grey forecasting model. The proposed framework combined the background value’s interpolation optimization (for the GM model) and the original data sequence’s data transformation. Also, cases studies were conducted to discern the proposed model’s prediction performance. From the findings, the proposed model outperformed most of the other grey-linked frameworks relative to the parameter of forecasting accuracy. Apart from forecasting accuracy, another parameter on which the proposed model exhibited superior results compared to grey modification frameworks and the traditional GM model involves the total electricity consumption. The implication is that the findings were informative and of practical importance whereby they would allow relevant agencies in the electricity sector to develop short-term plans or strategies (due to its electricity consumption prediction accuracy), even in situations where the source data is limited.