| 摘要: |
| 能源是促进社会经济发展、保证人民生产生活质量的重要基础。能源价格变化与世界经济走势密切相关,与国际关系、各国政策也有紧密联系。但能源价格近年来波动剧烈,加大了价格序列的分析复杂度,因此分解重构方法在能源价格序列分析预测中得到越来越广泛的应用。以原油期货价格为例,应用四种常见的分解方法(小波变换、奇异谱分析、经验模态分解、变分模态分解)进行分析与对比,实证表明,四种分解方法可以得到相似的分解量,并且分解方法可以有效地分离价格序列的多种波动特征,降低序列分析复杂度。除此之外四种分解方法也存在不同的优势:小波变换选择合适的基函数可以获得良好的正交性,奇异谱分析可以有效提取信号的主要成分,经验模态分解算法实现快速简单且无需设置参数,变分模态分解选择合适的分解数量可以有效避免模态混叠现象。结果表明,针对数据特点和分析目的选择合适的分解方法可以更有效地对能源价格进行分析。 |
| 关键词: 能源价格 原油期货 分解 重构 |
| DOI：10.13216/j.cnki.upcjess.2019.04.0001 |
| 分类号: |
| 基金项目:国家自然科学基金面上项目(71771208,3) |
|
| Construction Study on Application of Decomposition and Reconstruction for Energy Price Analysis |
|
LI Xiang1,2, WANG Jue1,2, WANG Zhen1,2, ZHOU Hao1,2
|
|
(1. CFS, MDIS, Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing 100190, China;2. School of Economics and Management, University of Chinese Academy of Sciences, Beijing 100190, China)
|
| Abstract: |
| Energy is an important foundation in ensuring people 's living standards and developing national economy. Changes in energy prices are closely related to the development of global economy, international relations and national policies. In recent years, the dramatic fluctuations of energy prices have increased the complexity of analyzing the price series. Therefore, the decomposition and reconstruction method has been widely used in energy price series analysis and prediction. This paper uses oil futures price as the representative of energy and applies four common decomposition methods (wavelet transform, singular spectrum analysis, empirical mode decomposition, variational mode decomposition) on oil price for analysis and comparison. Empirical studies show that the four decomposition methods can obtain similar decompositions. The decomposition method can effectively separate multiple fluctuation characteristics of price series and reduce the complexity of price analysis. Furthermore, the wavelet transform has orthogonality with appropriate basis functions; the singular spectrum analysis can extract the principal components of the signal; the empirical mode decomposition algorithm is fast and easy to implement but sometimes suffered from mode mixing problem; variational mode decomposition can effectively avoid mode mixing. It will be helpful to choose a suitable decomposition method according to characteristics of energy price series and analysis purposes. |
| Key words: energy price oil futures decomposition reconstruction |
