利用傅里叶谐波分法的时序数据周期迭代辨识算法-黄雄波.pdfVIP

35 6 Vol. 35 No. 6 第 卷第 期 计算机应用研究 优先出版 Application Research of Computers Online Publication 利用傅里叶谐波分析法的时序数据周期迭代辨识算法* 1 2 黄雄波 ，胡永健 (1. 佛山职业技术学院 电子信息系, 广东 佛山 528000; 2. 华南理工大学 电子与信息学院, 广州 510641) 摘 要：针对现有的时序数据其周期辨识算法存在着辨识精度低及计算成本高的问题，在傅里叶谐波分析法的基础上， 提出了一种具有基频迭代机制的周期辨识算法。首先，利用方差分析法从原始序列中析出其周期长度的整型估算值， 然后，以任意小的频率间隔在估算值区间内进行傅里叶谐波的迭代拟合，最后，基于最小拟合残差和的准则来确定最 优的周期成分。实验表明，该算法不仅具有良好的计算效能，而且还能精确地辨识出与序列样本长度无关的周期成分。 关键词：时序数据；周期成分；迭代辨识；傅里叶谐波分析法；方差分析法 中图分类号：TP391 Iterative identification algorithm for periodic component of time series data using fourier analysis method 1 2 Huang Xiongbo , Hu Yongjian (1. Dept of Electronic Information Engineering , Foshan Professional Technical College, Foshan Guangdong 528000, China; 2. School of Electronic Information Engineering , South China University of Technology, Guangzhou 510641, China) Abstract: In view of the existing time series data, the algorithm of periodic identification has the problems of low identification accuracy and high computational cost, on the basis of Fourier harmonic analysis, a periodic identification algorithm with fundamental frequency iteration mechanism is proposed . First of all, using the method of analysis of variance from the original sequence in the precipitation period length integer estimation, then, to the frequency interval arbitrarily small in the e