基于VMD-GRGC-FFT的BDI指数周期特性研究

发布时间：2018-04-19 16:33

本文选题：BDI指数 + 周期特性　；参考：《管理评论》2017年04期

【摘要】：本文对BDI指数周期特性进行了深入探讨,主要创新点有:一是提出了BDI指数周期划分和形成机制的理论分析框架,分析了需求侧、供给侧和非经济三个维度的影响因素综合促成的BDI指数周期波动。二是构建了基于VMD-GRGC-FFT的BDI指数长、中和短周期测算分析框架。借助变分模态分解法(VMD)对BDI指数进行模态分量分解,采用灰色关联度聚类分析法(GRGC)将模态分量重构合成低频、中频和高频三个新分量,以及通过快速傅立叶变换(FFT)模型全面测算BDI指数新分量对应的平均周期。三是采集1985-2015年BDI指数频度日/周/月的数据进行实证分析,结果表明:BDI指数的长周期约16年、中周期2-4年、短周期0.4-0.6年;在2016年至未来几年的时期内,BDI指数仍将处于长周期的萧条阶段和中周期的回复阶段。
[Abstract]:In this paper, the characteristics of BDI exponent cycle are deeply discussed. The main innovations are as follows: first, the theoretical analysis framework for the division and formation mechanism of BDI exponent cycle is proposed, and the demand side is analyzed.Supply-side and non-economic dimensions of the impact of factors combined to promote the BDI index cycle fluctuations.Secondly, the analysis frame of long, medium and short period of BDI index based on VMD-GRGC-FFT is constructed.The modal components of BDI exponent are decomposed by the variational mode decomposition method, and the modal components are reconstructed into three new components: low frequency, middle frequency and high frequency by grey correlation degree clustering analysis.The mean period corresponding to the new component of BDI exponent is calculated by fast Fourier transform (FFT) model.Thirdly, the data of daily / week / month frequency of BDI index from 1985 to 2015 were collected for empirical analysis. The results show that the long period of BDI index is about 16 years, the middle period is 2-4 years, and the short period is 0.4-0.6 years.The BDI index will remain in a long period of depression and a mid-cycle recovery period in the period from 2016 to the next few years.
【作者单位】：大连海事大学综合交通运输协同创新中心;
【基金】：国家自然科学基金项目(71672016) 长江学者和创新团队发展计划资助项目(IRT13048) 河北省交通运输厅重点项目(ZJT2015037)
【分类号】：F551

【相似文献】