Research On Complex Electrical Signal Phasor Measurement Algorithm

Electrical signal phasor measurement provides the basic electrical signal parameterinformation for real-time monitoring of the grid，which has been widely applied in the areaof synchronized phasor measurement、state estimation、static stability monitoring、transientstability prediction and control，as well as error analysis and other fields in powersystem．The existing phasor measurement algorithms have high accuracy when measuringsimple electrical signals；but for those complex electrical signals whose frequency deviatesfrom rating，or whose amplitude、frequency changes with time，because of being affectedby signal’s dynamic characteristics， the accuracy of algorithms will be seriouslyaffected．This paper does research on the measurement algorithm of complex electricalsignals，for improving steady-state measurement accuracy and dynamic tracking speed ofthe measurement algorithm，and has great significance in dynamic monitoring、safe andstable operation of the power system．This paper classifies the electrical signals and builds their models， and thensummarizes and generalizes the commonly used parameter estimation algorithm contrary todifferent types of electrical signals，on the basis this paper researches and improves severalphasor measurement algorithms．First of all，this paper does research on the most commonly used algorithm—discreteFourier transform (DFT) algorithm which is used to estimate electrical parameters，deepanalyzes its measurement error and the cause of it， and gives an improved DFTmeasurement algorithm based on fixed frequency sampling．In the case of that signal’sfrequency and amplitude change with time，simulation tests the capability when thealgorithm is used to measure electrical parameters．Secondly，when the electrical signal contains decaying DC component，this papercompares and analyzes the least squares algorithm，three-point algorithm and half-waveFourier algorithm，and proposes an improved full-wave Fourier algorithm．The algorithmutilizes the principle that the sum of fundamental signal’s two sampled values which isseparated by half a cycle is zero to do Fourier transform twice，obtaining an expression toeliminate decaying DC component．The simulation results show the effectiveness of theimproved algorithm．In addition，for dynamic signal this paper proposes a method toeliminate decaying DC component based on adaptive filtering and designs correspondingadaptive filtering algorithm． Thirdly，this paper analyzes the spectral properties of several commonly used windowfunction，and gives an improved windowed interpolation algorithm to measure fundamentaland harmonic’s parameters when frequency of electrical signal deviates from ratedfrequency．This paper presents calculation formula of the algorithm in detail，and verifiesthe accuracy of the algorithm by simulation．For time-varying dynamic electrical signal，this paper researches fundamental and harmonic parameters estimation algorithm based onKalman filter．At last，this paper proposes a harmonic and interharmonic analysis method based onS-transform and TT-transform．S-transform is a reversible time-frequency analysis methodthrough windowed Fourier transform，and has good time-frequency characteristics；TT-transform is an inverse Fourier transform of S-transform，and provides a representationof an one-dimensional time series into a two-dimensional time-time series．The methoddetects steady harmonics and interharmonics by using TT-transform’s amplification to highfrequency component of the signal and interpolation algorithm，and solves the problem thatthe high frequency harmonics and interharmonics in power system is too less to detect．Themethod detects frequency and amplitude parameters of time-varying harmonics in signal byusing high time resolution of TT-transform and good time-frequency analysis capability ofS-transform， and determines the beginning and ending time of each frequencycomponent．Simulation results show that the method has high accuracy and is suitable foranalysis of steady-state and dynamic harmonic signals．...