摘要
提出了基于最小二乘法原理与FFT相结合的改进型谐波分析方法。基本原理是:提供一组采集信号,利用FFT先计算出谐波的初始参数,然后把初始参数代入模型波形成为其初始参数,计算模型波形和实际波形之间的均方差,如若均方差不满足条件,则进行参数的修正。当均方差最小时,模型波形的参数可以代表实际波形的参数。为避免分析出的谐波次数不准确而出现无效参数,把实际采样数据分成训练组和测试组。在训练组中用最快下降梯度查询学习策略的迭代循环修正参数,在测试组中检测谐波次数的正确性,获得准确的分析结果,实现对次谐波和频率相隔很小的谐波的同步跟踪与分析。通过仿真实验证明了该方法的有效性和准确性。
An improved FFT algorithm for harmonic analysis method based on least square method is presented. The keystone is.. a set of sampled signals is provided, and it makes use of FFT to work out the initial parameters of harmonics which turns into initial parameters of model waveform. The mean square deviation between the waveform of the model and actual waveform is counted, if it does not satisfy the condition, the parameters are modified. While the mean square deviation is minimal, the parameters of the model can represent the parameters of actual waveform. To avoid invalid parameters caused by inaccurate order of harmonics being analyzed, the actual sampled data are divided into training group and testing group. The correction parameters of iterative loops for search strategy can be inquired about in train- ing group, arid the correctness of harmonic order can be detec- ted to obtain effective results of harmonic analysis in testing group, in order to realize the synchronous tracking of non-integer harmonics and harmonics with small frequency interval. Simulation results show that the proposed method is effective.
出处
《现代电力》
2008年第2期24-28,共5页
Modern Electric Power
关键词
谐波分析
FFT
最小二乘法
参数修正
仿真
harmonic analysis. FFT： least square method： parametric modification： simulation
