【正文】 號(hào)周期交替變化引起的背景梯度，提出了基于TopHat形態(tài)變換的抑制方法，針對(duì)噪聲引起的背景梯度，構(gòu)造了合適的形態(tài)濾波器，提高了擾動(dòng)檢測(cè)的靈敏度。7）提出了一種改進(jìn)的單相電壓暫降特征參數(shù)實(shí)時(shí)檢測(cè)算法。分析了單相電壓暫降檢測(cè)時(shí)“異動(dòng)”現(xiàn)象的產(chǎn)生原因，提出了基于形態(tài)運(yùn)算的解決方法。針對(duì)三相不平衡電壓暫降，通過(guò)正序dq變換和負(fù)序dq變換，經(jīng)由形態(tài)低通濾波得到三相電壓的基波正序、負(fù)序分量，采用改進(jìn)的單相電壓暫降檢測(cè)算法得到零序分量，進(jìn)而得到各相電壓暫降的特征參數(shù)。本文提出的電能質(zhì)量參數(shù)估計(jì)理論和系列算法，為電能質(zhì)量實(shí)時(shí)檢測(cè)和離線分析提供了有效的工具。關(guān)鍵詞：電能質(zhì)量；參數(shù)估計(jì)；數(shù)字信號(hào)處理；檢測(cè)；算法AbstractWith rapid development of economy and science technology, power quality (PQ) problems have increasingly bee a substantial concern in recent years. Completely and accurately assessing power quality is the first step in solving and mitigating the problems, fast and advanced algorithms are the basis for effective measurement and evaluation.Due to the plexity of power system signal and the variety of power quality event, it is a challenge to detect power quality indices pletely, accurately and promptly.In accordance with national and international power quality standard, a systematic and indepth research about power quality major indices evaluation is carried out based on analyzing domestic and overseas research situations. By use of digital signalprocessing methods as well as power system automation techniques, the research works pay more attention to the measurement theories and estimation algorithms of important indices. The main contributions of this dissertation are listed as follows:1）A new digital algorithm based on adaptive ShortTime Fourier Transform (STFT) is presented to estimate some basic power quality indices, such as voltage deviation, frequency deviation, voltage unbalance degree and power factor of fundamental ponent. The algorithm employs fixed sampling period although the frequency of the signal varies. No recursive calculation is needed to get evaluation values of above mentioned parameters. As a result, the algorithm is easy to implement with less putation burden and short response time. To restrain the spectrum leakage of harmonics, a novel kind of window called Rectangle SelfConvolution Window (RSCW) is adopted and the window length is adjusted according to the frequency variation of the signal. It is thus suitable for realtime measuring with high presicion.2）According to the spectrum characteristics of harmonics/interharmonics, an approach is proposed to pute harmonics/interharmonics in real time. Considering the characteristic of sine/cosine function, the harmonic ponent can be transformed into DC ponent, the rms value and initial phase angle of harmonic then can be got after lowpass filtering. As to interharmonics, it is generally considered to be a challenge because the interharmonics only have small ratio, it is hard to estimate them accurately. By subtracting fundamental and harmonic ponents from the original signal, maximum spectrum searching technique is used to pute frequency, rms value and initial phase angle of interharmonic. The proposed approach is easy to implement with high precision and good dynamic tracking feature, it suits to estimate harmonic/interharmonic parameters of voltage or current signals with discrete spectrum.3）It is known that flicker phenomena are closely related with harmonics and interharmonics, but at present, the detection methods of harmonics/interharmonics and flicker are pletely different. Based on deducing the equivalent relation of interharmonic and amplitude modulation wave to flicker, the detection methods of interharmonic and flicker are bined. Therefore, according to national and International Electrotechnical Commission (IEC) standard, a new flicker evaluation algorithm is presented based on interharmonic estimation. The algorithm can not only pute instantaneous visual acuity, shortterm flicker severity index and longterm flicker severity index, but also provide frequency and amplitude imformation of voltage fluctuation, thus, it benefits to analyze and manage voltage fluctuation and flicker. Furthermore, it provides a novel implemental model for flicker detection.4）To detect and locate nonsteadystate disturbances, digital image processing theory and image edge morphology detecting method are introduced. The outstanding features of the method are easy implementation, less puting plexity, high time resolution and low demand of sampling frequency, therefore, it can be used to effectively detect mon power disturbances such as voltage sag, swell, shorttime interruption, shorttime high frequency oscillatory, notching , etc.5）Deep analysis reveals that restrain background gradient is the key step for disturbing detection. Based on examining the origin of background gradient, two different methods for restrain background gradient are proposed. One uses TopHat morphology transform to restrain the background gradient produced by normal periodic alternation of analyzed signal, the other constructs suitable morphology filters to restrain the background gradient introduced by noises. Thus, the detecting sensitivity can be improved obviously.6）A new disturbing detection method based on order morphology and TopHat transform is then presented. It can not only restrain the background gradient produced by steady variation of analyzed signal, but also resist noises while keep the partial details of signal’s abrupt change. Therefore, it ensures the sensitivity and time locating precision of distrurbing detection. 7）An improved single phase voltage sag dececting algorithm is put forward. A new matrix, which is designed to decrease the phase delay that is unavoidable in traditional fictitious three phase systems, is constructed. The improved algorithm can almost instantaneously quantify the magnitude, duration and angel jump for voltage sag in single phase systems. Since ‘a(chǎn)bnormal variation’ phenomenon presents in the detecting process, a corresponding solving measure is carried out based on morphology calculation.8）As to three phase unbalanced voltage sag, an improved detection method is also presented. After carrying out positive and negative dq transform, morphology low pass filter wh