불균형 패널모형을 이용한 지역별 전력소비량 및 최대전력 예측

초록

We propose a model for forecasting long-run regional electricity consumption and regional peak electricity demand. We first estimate regional electricity consumption, which varies over time and with regional income, using the unbalanced panel model presented by Chang et al. (2021). By analyzing the errors in the estimated model, we assess the impact of changes in regional industrial structure on electricity consumption. For regional peak demand forecasting, we adjust the existing incomplete regional load data and verify the robustness of the adjusted data. Using this newly adjusted dataset, we construct a model for forecasting regional peak electricity demand and apply it long-run peak demand forecasts. The forecasting performance of our regional models has proven to be stable over a 1- to 5-year time span, with an average MAPE(Mean Absolute Percentage Error of less than 5%.

키워드