计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (32): 183-185.DOI: 10.3778/j.issn.1002-8331.2009.32.058

• 图形、图像、模式识别 • 上一篇    下一篇

符号分析在混沌序列模式识别中的应用研究

冯明库1,刘炽辉1,刘雄英2   

  1. 1.广东技术师范学院 电子信息工程系,广州 510665
    2.华南理工大学 电子与信息学院,广州 510640
  • 收稿日期:2008-06-25 修回日期:2008-10-08 出版日期:2009-11-11 发布日期:2009-11-11
  • 通讯作者: 冯明库

Research on identifying pattern of chaotic series with symbolic analysis

FENG Ming-ku1,LIU Zhi-hui1,LIU Xiong-ying2   

  1. 1.Department of Electronic Information Engineering,Guangdong Polytechnical Normal University,Guangzhou 510665,China
    2.College of Electronics & Information,South China University of Technology,Guangzhou 510640,China
  • Received:2008-06-25 Revised:2008-10-08 Online:2009-11-11 Published:2009-11-11
  • Contact: FENG Ming-ku

摘要: 针对待测时间序列是来自周期信号源、混沌信号源还是随机信号源的问题,提出一种用符号谱定性判定的方法。先对时间序列量化,然后进行编码,继而绘其符号谱。通过对不同类型时间序列的符号谱的对比分析表明,符号谱比K-S熵和功率谱等其他方法更能简单有效地判定时间序列的确定性来源和模式类型。

关键词: 混沌映射, 模式, 符号谱

Abstract: In view of whether the time series comes from periodic,chaotic or random source,a qualitative method with symbolic spectrum is proposed.In this method,the time series is quantized firstly,then is coded and its symbolic spectrum is plotted.Several symbolic spectrums of different type of time series show that,symbolic spectrum can identify the deterministic source and pattern of time series simply and efficiently,better than K-S entropy and power spectrum.

Key words: chaotic map, pattern, symbolic spectrum

中图分类号: