Forecasting energy consumption in Anhui province of China through two Box-Cox transformation quantile regression probability density methods
| Autor: | Yaya Zheng, Qifa Xu, Yaoyao He |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Consumption (economics)
education.field_of_study Applied Mathematics 020208 electrical & electronic engineering 010401 analytical chemistry Population 02 engineering and technology Energy consumption Power transform Condensed Matter Physics 01 natural sciences Gross domestic product 0104 chemical sciences Quantile regression Normal distribution 0202 electrical engineering electronic engineering information engineering Economics Econometrics Gamma distribution Electrical and Electronic Engineering education Instrumentation |
| Zdroj: | Measurement. 136:579-593 |
| ISSN: | 0263-2241 |
| DOI: | 10.1016/j.measurement.2019.01.008 |
| Popis: | Energy consumption can be taken as one of crucial indicators for economic development in any regions. Rapid development of economic, population and urbanization in Anhui province of China has led energy consumption to grow rapidly. The considerable demand for energy has become an issue of great concern. For proper policy formulation, it is necessary to have reliable forecasts for energy consumption. However, energy consumption forecasting is affected by some potential factors, including historical energy consumption, economic activities, population and weather. These uncertain factors put forward higher requirements for energy consumption forecasting methods. Considering these problems, a reasonable feature selection method called stepwise regression is adopted to extract important variables before prediction. Historical energy consumption, average annual GDP growth rate, and total GDP are identified as key factors on annual energy consumption. Then, two probability density forecasting methods based on Box-Cox transformation quantile regression via normal distribution (N-BCQR) and gamma distribution(G-BCQR) are proposed to measure the uncertainty and estimate the future demand of energy in Anhui province of China. The comparatives results show that the N-BCQR outperforms G-BCQR and other existing methods in terms of point forecasts and interval predictions. In addition, on the strength of predefined assumptions regarding the average annual gross domestic product (GDP) growth rate, the energy consumption of Anhui province through 2023 is forecasted. The results indicate that total energy consumption of Anhui is an important component determining local economic growth. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect