The Research On Harmonic Analysis And Its Application Of The Bi-directional Interactive Energy Measurement

The bi-directional interactive energy measurement can achieve the intelligent power data collection and provides a scientific basis to the analysis and application manage-ment. However, with the widespread proliferation of power electronic loads and other nonlinear loads, the levels of harmonic pollution have been steadily rising toward and even exceeding acceptable limits in the smart grid. In researches on eliminating or at least reducing the impacts of harmonics on power systems, harmonic analysis and har-monic power measurement have been one of the most vital problems that attract most attentions.Conventionally, the fast Fourier transform (FFT) is applied to estimate the fre-quency, phase, and amplitude of the signal of interest for its simplicity and availability, and easiness to implement in embedded systems. However, there are two factors limiting its accuracy, namely the spectral leakage and picket fence effect. Moreover, variations of the fundamental frequency are usually caused by faults in the power systems, which may leads to the uncertainty of the signal harmonics, and the parameters estimation of arbitrary harmonics under frequency variation is still an open problem. Besides, inter-harmonics and noise may exist in the measured signal. Hence, it is of great theoretic and practical importance to study and implement novel methods for accurate and real-time harmonic parameters estimation under nonstationary situations.Firstly, the key technologies of the interactive energy measurement are illustrated for building a bi-directional interactive intelligent power system (BIIPS). The general rules of communication, the structure, and the function module for the optical fiber meter based on the IEC62056standard are discussed. Then the IEC62056-based bi-directional interac-tive communication protocol of the intelligent power system is developed. Combined with the demonstration area of the BIIPS, the interactive SMS platform and the intelligent interactive terminals based on multi-agent technology are designed.Next, according to the high accuracy requirements of frequency measurement and the leakage of FFT due to asynchronous sampling, the reasons and solutions of spectral leakage and picket fence effect are illustrated. Combined with the spectral characteris-tics of classic windows.the design specifications and principles of window functions are analyzed. And an improved FFT method based on the even order cosine window for harmonic parameter estimation is proposed. By searching the local peaks of the discrete spectrum, the maximum and sub-maximum amplitude of spectral linos is determined. and the applicable rectification formulas of the improved FFT are obtained by using the polynomial curve fitting. The proposed method can get exact solutions where frequency deviates from nominal frequency, white noise and interharmonics are present, and voltage fluctuation. The rectification of the fundamental frequency is free of solving high-order equations. Therefore, the proposed method has shorter computation time and the algo-rithm is thus advantageous for efficient applications in embedded systems which requires highly accurate and real-time frequency estimation.Then, a new class of windows, the cosine self-convolution window (CSCW), is pro-posed to solve the problem that the spectral leakage suppression of classic windows con-strained by its side-lobe behaviors when harmonic frequencies are intensive. The main-lobe and side-lobe behaviors of the first to the third order Blackman self-convolution win-dow (BSCW), Rife-Vincent self-convolution window (RSCW), and Nuttall self-convolution window (NSCW), are respectively studied. A CSCW-based improved FFT method for estimating power system signal parameters is given. A CSCW gives a much lower peak side-lobe level along with a higher side-lobe decaying rate compared with classic win-dows. Therefore, leakage; errors and harmonic interferences are thus reduced considerably by weighting samples with the CSCW. Moreover, the adjacent intensive frequency signals can be distinguished easily, and the algorithm is thus advantageous for harmonic analysis of intensive and complicated signal.Harmonic energy can not be measured accurately by the traditional methods, and the interpolation FFT algorithm has computation time for solving high-order equations. Thus, combined with the good side-lobe behaviors of the NSCW, a second order NSCW-based phase difference correction algorithm for harmonic power measurement is proposed. The new algorithm need not to solve high-order equations and has shorter computation time. Therefore, the suppression ability to the spectral leakage of the second order NSCW can be fully applied, as the real-time performance of the algorithm is ensured, and the performance requirements of the embedded system are thus reduced, which made it is easily to implement in DSP.Subsequently, according to the demands of the smart grid and the defects of the traditional ammeter, a novel three-phase multi-function smart meter (TMSM) based on ADS8364+ADSP-BF533+AT91SAM9261is designed. Combined with the proposed har-monic analysis and energy measurement method based on second order NSCW. the im-plementations on accurate bi-directional harmonic energy measurement is given. Then the error models of harmonic energy measurement are established and the error correction methods are proposed. According to the test results, the proposed TMSM can measure both fundamental energy and2nd-21st harmonic energy, respectively. The accuracy of the fundamental energy measurement reaches0.2s class, and the measurements of the harmonic energy accord with the Class A standard requirement of GB/T-14549-1993.