Analysis Bases On The Asynchronous Sampling Between Power Steady-State Harmonics And Inter-Harmonics Detection Method

Nowadays’society, with various kinds of electrical and electronic equipment rapidly developing and widely applied, the widespread influence of power system harmonics and inter harmonics and has become more and more serious, the power system harmonics and inter harmonics harm serious gradually with people highly valued. The fast and accurate measurement is the premise to realize the power system harmonics and inter harmonics suppression and to realize the optimal compensation configuration. The existing detection methods have many short comings.In the power signal detection and analysis, the Fourier transform is incorporated into the IEC61000-4-7 standard. The limitation of Fourier transform analysis is synchronous sampling and the length of the sequence is enough to satisfy the frequency resolution. Fourier transform’s, in order to achieve good analysis precision, the computation of the interpolation computation must be high. This chapter describes three-term third derivative Nuttall windowed and four-term fifth derivative Nuttall windowed interpolating Fourier algorithm, and according to the Fourier analysis of non-synchronous sampling frequency spectrum leakage and fence effect principle, eliminating power frequency and second harmonic frequencies of other ingredients leakage and fence effect obtains three-term third derivative Nuttall window and four-term fifth Nuttall window interpolating Fourier algorithm, which improves the accuracy of Nuttall windowed interpolating Fourier transform algorithm for harmonics and inter harmonics analysis.The modern spectrum estimation overcomes the contradiction between the length of the discrete series and the frequency resolution. This chapter describes the relationship between AR model spectrum estimation and signal self-correlation and model parameters, and gives the commonly used method of AR model Burg algorithm and the optimal weighted Burg algorithm. The algorithm principle of modern spectrum estimation is discussed, including the algorithm of MUSIC and ESPRIT. Then, modifying the forward prediction error and backward prediction error of the optimal weighted Burg algorithm further improves the spectral peak shift of the Burg. According to the Root-MUSIC algorithm simplifying the computational complexity of the music algorithm and algorithm PM-MUSIC improving the MUSIC algorithm for frequency analysis of precision, putting forward PM-Root-MUSIC algorithm in a certain extent reduces the computational complexity of the algorithm and improves the analysis accuracy of the algorithm.This chapter briefly introduces the principle of Adaline neuron model, and proposes a complex domain Adaline neural network model and an adaptive convergent learning algorithm, including Newton LMS, variable step size LMS and RLS. An adaptive learning algorithm with good convergence performance is used as the learning criterion for Adaline neural network, which improves the convergence speed and convergence precision of the algorithm.Finally, the addition of window interpolation Fourier and the corresponding improved algorithm, modern spectral estimation and the corresponding improved algorithm are combined with Adaline neural network, and the simulation results verify the effectiveness of the algorithm applied to the steady power harmonics and inter harmonics;This chapter present the hardware and program design of power quality analysis device;With the YXDSP-F28335 digital signal platform, the TMS320F28335 DSP core implements algorithm, which is applied to the steady power harmonics and inter harmonics detection method.