摘要
将排列熵与滑动窗口和滑动移除数据技术相结合,比较滑动排列熵和滑动移除排列熵方法在非线性时间序列动力学结构突变检测中的性能。发现滑动排列熵方法依赖于滑动窗口的选取,得到的突变点位置误差比较大,不利于实际应用;而滑动移除排列熵方法的检测结果几乎不依赖于滑动移除窗口的大小,能够更为精准地得出时间序列的突变位置,明显优于滑动排列熵方法,具有良好的应用前景。
Performance of dynamic structure mutational detection in nonlinear time series is compared with moving permutation entropy and moving cut data-permutation entropy.The analysis shows that the moving permutation entropy method depends heavily on the selection of sliding window,and its results has big position errors of the mutational point,which is unfavorable to the practical application.Moving cut data-permutation entropy method does not depend on the scale of sliding remove window to give results,and is capable of more accurate drawing of the mutational position of sequence.The moving cut data-permutation entropy has a good application prospect,and is significantly better than moving permutation entropy method.
作者
罗文相
万丽
赖斯敏
LUO Wenxiang;WAN Li;LAI Simin(School of Mathematics and Information Science,Guangzhou University,Guangzhou 510006,China;Key Laboratory of Mathematics and Interdisciplinary Sciences of Guangdong Higher Education Institutes,Guangzhou University,Guangzhou 510006,China)
出处
《厦门理工学院学报》
2018年第3期91-96,共6页
Journal of Xiamen University of Technology
基金
国家自然科学基金项目(41172295)
广州大学研究生创新研究资助计划(2017GDJC-M32)
关键词
突变检测
排列熵
滑动排列熵
滑动移除排列熵
mutational detection
permutation entropy
moving permutation entropy
moving cut data permutation entropy
