Power System Harmonic Detection Method

Due to the widespread proliferation of nonlinear distorting loads nowadays, harmonic distortions in a power network becomes a great concern. Besides conventional solutions such as passive filters and changed system configurations to harmonic problems, the use of active power filters (APF) to eliminate harmonic currents injected by nonlinear loads has attracted much attention. The premise of realizing APF is detecting the harmonic currents quickly, accurately. So, many scholars aim to design a harmonic detecting method with good real time, high precision and wide applicability.This paper analyzes and studies all kinds of harmonic detecting methods in the world, gives emphasis to introduce the method based on i_p - i_q theory, the adaptive filter and the predictive harmonic detecting method, summarizes the advantages and disadvantages of every harmonic detecting method,, Choosing three representative and wide used harmonic detecting methods, the method based on RDFT, i_p-i_q method and the method based on Caiman filter, with dSpace and Matlab tool, we compare the real time implementation, precision, adaptive ability,computing complexity of them.Recently, the wavelet transform has attracted more and more attentions by electric engineer because of its excellent time-frequency localization property, the harmonic detection with the wavelet transform has been a hotspot。 This paper studies the wavelet transform deeply, introduces the relative theory, the simulation results verify, the method based on the wavelet transform can detect the fundamental frequency signal accurately and quickly, it is suitable for the detection and time-frequency analysis of nonstationary and time-varying signals, especially for the signal with the voltage flickero However, the traditional dyadic-orthonormal wavelet transform only the low frequency bands are subdivided stepwise that results in high resolution in low frequency bands, whereas low resolution in high frequency bands, it can't detect the frequency components near the cut-off frequency efficiently, so it can't detect all neighboring harmonics in harmonic detection。 Based on the conventional wavelet transform, this paper proposes a kind of improved wavelet transform, it subdivides both the low frequency bands and the high frequency bands to ensure the uniform resolution in all frequency bands. The improved wavelet transform improve the ability of decomposing the high frequency, furthermore it strengthens the ability of detecting the abrupt signal, simulation results verifies, the improved wavelettransform method has good real-time ability ^ high precision! and can be used widely0 Based on the wavelet transform, the harmonic monitor is developed.. It can show harmonic content, harmonic ration, total harmonic distortion, online diagnosiso It is an integrated harmonic detection system with monitoring, analyzing, anticipating, alarming, protecting function, providing an effective method for monitoring the harmonic o...