Study on the Forecast of Coalmine Electricity Consumption Based on Holt-Winters Model

The coal industry has a larger demand for electricity consumption, therefore the forecast of coalmine electricity consumption has become an urgent problem for the electric power company and large-scale coal enterprise. Through the analysis of available coalmine electricity consumption data, a weak increasing trend can be observed from the electricity consumption longitudinal analysis curve. A periodic and seasonal pattern can also be shown in the monthly electricity consumption comparative analysis curve. In other words, we find the time series examples of the coalmine electricity consumption show a linear, seasonal and stochastic pattern. In this study, a popular forecasting model based on Holt-Winters method is employed to estimate the trend of coalmine electricity consumption. Meanwhile, two other forecasting models, the classical linear regression (CLR) model and the quadric exponential smoothing (QES) model are utilized in the same data sets. Forecasted results indicate that the Holt-Winters model is outperforms the CLR and the QES models in terms of forecasting evaluation measures. Thus, the Holt-Winters model is an effective and feasible method for the coalmine electricity consumption forecasting.

View article page

Download PDF