[1]刘学彬,梁智飞,朱卫平,等.基于 EEMD 的固定分段数分段线性表示方法[J].计算机技术与发展,2023,33(11):202-208.[doi:10. 3969 / j. issn. 1673-629X. 2023. 11. 030]
　LIU Xue-bin,LIANG Zhi-fei,ZHU Wei-ping,et al.Piecewise Linear Representation Algorithm of Fixed Section Number Based on EEMD[J].,2023,33(11):202-208.[doi:10. 3969 / j. issn. 1673-629X. 2023. 11. 030]

点击复制

基于 EEMD 的固定分段数分段线性表示方法()

分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:

33

期数:

2023年11期

页码:

202-208

栏目:

新型计算应用系统

出版日期:

2023-11-10

文章信息/Info

Title:

Piecewise Linear Representation Algorithm of Fixed Section Number Based on EEMD

文章编号:

1673-629X(2023)11-0202-07

作者:

刘学彬¹ ; 梁智飞² ; 朱卫平² ; 祝 凯^1*

1. 青岛理工大学 信息与控制工程学院,山东 青岛 266000;

2. 中石油煤层气有限责任公司,北京 102200

Author(s):

LIU Xue-bin1 ; LIANG Zhi-fei2 ; ZHU Wei-ping2 ; ZHU Kai1*

1. School of Information and Control Engineering,Qingdao University of Technology,Qingdao 266000,China;

2. Petrochina Coalbed Methane Company Limited,Beijing 102200,China

关键词:

时间序列; 分段线性表示; 集合经验模态分解; 模态重构; 符号化; 自底向上

Keywords:

time series;  piecewise linear representation;  ensemble empirical mode decomposition;  mode reconstruction;  symbolization; bottom-up

分类号:

TP391

DOI:

10. 3969 / j. issn. 1673-629X. 2023. 11. 030

摘要:

针对采用单一启发式规则的分段线性表示方法存在局部最优化和无法准确预计分段数目的问题,提出了基于集合经验模态分解( EEMD)的固定分段数分段线性表示方法。 该方法通

过将集合经验模态分解和重构思想引入分段线性表示方法研究中,同时将自底向上算法的拟合误差阈值改进为分段数阈值来解决上述两个问题。 首先,通过模态重构思想过滤掉细节信息,提取到全局性分段点;然后,根据各初始分段子序列的波动程度,确定子序列段内分段点数量分布;最后,采用基于分段数阈值的自底向上方法将子序列合并到要求的分段数。 该方法不仅继承了自底向上方法拟合误差小的优点,同时克服了局部最优化以及不能预计分段数的缺点。 通过仿真实验证明了该方法克服了局部性的缺点,并有效减弱了噪声的干扰。 相比现有方法,在压缩率相同的情况下,该方法的拟合误差更小。 最终,在压裂施工时序数据趋势提取的应用中也验证了其有效性。

Abstract:

Aiming at the problems of local optimization and inability to accurately predict the number of segments in the piecewise linearrepresentation method using a single heuristic rule,

a piecewise linear representation method with a fixed number of segments based on Ensemble Empirical Mode Decomposition ( EEMD ) was proposed. This method introduces the idea of ensemble empirical modedecomposition and reconstruction into the research of piecewise linear representation method,and at the same time improves the fittingerror threshold of the bottom-up algorithm to the threshold of piecewise number to solve the above two problems. Firstly,the detail information is filtered out by the idea of modal reconstruction,and the global segmentation point is extracted. Then,the distribution of thenumber of segmentation points in the subsequence is determined according to the fluctuation degree of each initial segmentationsubsequence. Finally,a bottom - up method based on the number of segments threshold is used to merge the subsequences into therequired number of segments. This method not only inherits the advantages of small fitting error of the bottom - up method,but alsoovercomes the shortcomings of local optimization and unpredictable number of segments. The simulation experiment proves that theproposed method overcomes the shortcoming of locality and effectively weakens the interference of noise. Compared with existingmethods,the fitting error of the proposed method is smaller when the compression rate is the same. Finally, its effectiveness is alsoverified in the application of time series data trend extraction of fracturing construction.

常用功能