CAJ | 학술논문

针对扰动检测过程中存在的脉冲噪声和白噪声影响暂态电能质量信息准确提取的问题，文章设计了一种基于LMS的多结构复合形态滤波算法，该算法具有在保留信号的有效成分的基础上最大限度地滤除噪声的功能。文章使用电压暂降、电压暂升、电压中断等对所提方法进行了仿真实验。仿真结果表明，基于LMS的多结构复合滤波算法得到的信噪比是多结构自适应复合滤波算法的1．01倍左右，是小波变换的1．1倍左右，即该算法能有效地滤除暂态电能质量扰动分析中的随机噪声和脉冲噪声；而且，该算法可较好地保持扰动信号的形状和特征。
침대우동검측과정중존재적맥충조성화백조성영향잠태전능질량신식준학제취적문제，문장설계료일충기우LMS적다결구복합형태려파산법，해산법구유재보류신호적유효성분적기출상최대한도지려제조성적공능。문장사용전압잠강、전압잠승、전압중단등대소제방법진행료방진실험。방진결과표명，기우LMS적다결구복합려파산법득도적신조비시다결구자괄응복합려파산법적1．01배좌우，시소파변환적1．1배좌우，즉해산법능유효지려제잠태전능질량우동분석중적수궤조성화맥충조성；이차，해산법가교호지보지우동신호적형상화특정。
Aiming to the problem that pulse noise and white noise have an influence on accurate extraction of transient power quality information, so an effective filter algorithm is proposed to filter out noise to great extend on the basis of keeping effective components of signal.This algorithm is based on multiple structure elements composite filtering of morphology LMS.This paper utilizes voltage sag, voltage surge and voltage interrupt and so on to carry out simulation experiment.The simulation results show that the SNR which the algorithm based on multiple structure elements com-posite filtering of morphology LMS gains is about 1 .01 times than that of multiple structure elements self-adaptive and composite algorithm and about 1.1 times than that of wavelet transform.In other words, this algorithm not only effec-tively filters out noise of transient power quality signals, but also remains characters and shape of disturbance signals.