| 来源类型 | Research papers
|
| 规范类型 | 报告
|
| Reliability Analysis of Eenrgy Demand Estimation Based on Energy Input-Output Coefficients |
| J. S. Park
|
| 发表日期 | 2006-12-31
|
| 出版年 | 2006
|
| 语种 | 英语
|
| 摘要 | 1. Research Purpose
An input-output energy analysis method suggests how to use energy input-output table for energy demand forecasting. The method makes it possible to derive the total energy input coefficients from the table and decompose them into structural determinants. An observation of determinants can analyze the cause of changes in total energy input coefficients. Thus, energy input-output analysis helps us systematically understand changes in energy demand and also effectively forecast energy demand with the energy input-output table.
There are RAS method and Lagrange undetermined multiplier approach (Lagrange approach) as commonly used methods for energy demand forecasting. RAS method is an algorithm - which distributes the differences between known product outputs and the sums of the rows of the present intermediate matrix proportionally along the rows and then to distribute the differences between known industry outputs and the sums of the columns proportionally along the columns - to estimate input-output coefficients of current year. This process is carried out continuously until all discrepancies are reduced to an acceptable minimum. Lagrange approach estimates input-output coefficient matrix minimizing squared difference of input-output coefficients between past and current year subject to the given total production, total intermediate demand and input of current year.
These methodologies commonly require current year information obtained from an annual survey which necessarily bears the cost. However, such information may not be available if the survey or information cost is not acceptable. This situation makes it inevitable to use the data of past years. One of demanded solutions to resolve such a situation will be a forecasting method which can enhance efficiency in using information and rule out information costs. The purpose of this research is to contribute to the development of the methodology.
2. Major Results and Limitations
This research first constructs hybrid energy input-output tables, which refer to Shim (2005), to derive total energy input coefficients. The coefficients are decomposed into structural determinants. The determinants of past years are adjusted by four concepts of average including arithmetic, weighted, geometric and harmonic mean to forecast those of current year. Each forecasting method is evaluated by mean absolute percentage error (MAPE). This is called an adjustment process compared to non-adjustment process. An adjustment process and the reliability analysis of energy demand forecasting show a set of analysis steps to reach an efficient method of forecasting and suggest implications to improve it.
A reliability analysis chooses a competitive adjustment process applied to the energy demand forecasting. The results of reliability analysis support that the adjustment process is superior to the non-adjustment process in their reliability. Thereby, it is proved that an adjustment process may possibly be able to improve the reliability of energy demand forecasting.
Some issues have been found in the process of analysis as follows: limitations of two models used in the research, the effect of interaction effect on errors, further decomposed sub-determinants, qualitative analysis, and so forth. In-depth study on these issues are required to shape the methodology.
Literature survey rarely found previous studies matching with this research in purposes and motivations. The main contribution of this research could be to provide a set of analysis steps and primary models for energy demand forecasting under information constraints. I hope that this research will encourage additional research even if this research has various limitations as mentioned in the text and may not be robust as at early stage. |
| URL | http://www.keei.re.kr/web_keei/en_publish.nsf/by_report_year/20486AB6F7CBAFF5492572750007A95D?OpenDocument
|
| 来源智库 | Korea Energy Economics Institute (Republic of Korea)
|
| 资源类型 | 智库出版物
|
| 条目标识符 | http://119.78.100.153/handle/2XGU8XDN/322390
|
推荐引用方式
GB/T 7714 |
J. S. Park. Reliability Analysis of Eenrgy Demand Estimation Based on Energy Input-Output Coefficients. 2006.
|
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。
