用于处理电能质量信号的最优低通差分器的研究

在某些电能质量检测和分析计算中,有时需采用差分方法计算信号的变化率,常见的差分方法有基本中心差分、Newton-Cores差分、Lanczos差分等.采样得到的电能质量信号一般含有较多的噪声,如差分算法选择不当,噪声将被显著地放大,影响后续信号的分析、处理和评估.因此需寻找一种低通差分滤波器算法,在保留有效扰动信号的同时,能抑制噪声对数值差分的影响.文章在逼近理想差分滤波器幅频特性的基础上,通过带约束条件的最小化方法建模求解差分滤波器的最优系数.仿真计算表明,最优低通差分滤波器(OLPDF)能准确地完成数值微分计算及有效地抑制噪声.OLPDF的带通频率可根据应用体条件选取,且可根据噪声情况设计其阻带性能.研究表明,最优低通差分滤波器是一种分析电能质量干扰信号的比较有效的工具.

OPTIMAL DIGITAL LOW-PASS DIFFERENTIATOR FOR POWER QUALITY DISTURBANCES SIGNAL PROCESSING

In the certain kinds of detection and analysis of power quality sometimes the difference method should be used to calculate the change rate of signals, the common difference methods are basic central difference, Newton-Cores difference, Lanczos difference, etc. In general, the collected power quality signals are polluted by high frequency noises, if the method of difference has not been properly chosen the noises will be significantly amplified, so the follow-up analysis, processing and evaluation of the signals may be affected. Therefore, it is necessary to find a low-pass differentiation algorithm by which the affection of noises on numerical difference can be restrained while the effective interference can be effectively reserved. Here, on the basis of approximating the amplitude-frequency characteristic of ideal difference filter, the optimal coefficients of difference filter are obtained through the modeling by minimized method with constraint conditions. Simulation results from two calculation examples show that using the proposed optimal low-pass difference filter (OLPDF) the numerical differentiation can be carried out more accurately and the noises can be effectively suppressed. The band-pass frequency of OLPDF can be chosen according to the application conditions and its band-stop performance can be designed according to the noise situation. It is indicated in the research results that OLPDF is an effective tool in the analysis of power quality disturbance signals.