The Improved HHT Algorithm And Its Application In The Power Quality Detection

With the continuous acceleration of global industrialization process, human activities have produced tremendous effect on the earth’s environment. Therefore, it is an urgent and arduous task for people all over the word to protect environment and realize sustainable development. The intelligentization of electric network not only can promote the development of renewable energy, but also an important means to realize the energy saving and emission reduction, and it may become the important direction for the development of electric power industry. Along with the gradual construction of smart grid, the domestic power equipment manufacturers will have a more broad market space, the requirement of power quality from consumers will become much higher due to the development of a large number of intelligent equipments, and the electric power made to measure the sensitive power consumers will become a new direction. But because of the widely used of nonlinear and impulsive loads, the power quality is deteriorating, and it is imperative to take measures to govern the power quality problems. The accurate and real-time detection of power quality disturbances is the premise of government, and also the important premise for the stable operation of electric power system. Under the factors of non-stationary, randomness and distributivity, the traditional power quality detection methods are not applicable any more, seeking a new method that is more suitable for the analysis of nonlinear, non-stationary signal is particularly important.This paper firstly analyzes the overseas and domestic research status of traditional power quality detection methods, and then leads to the Hilbert-Huang transform method. Secondly, based on the in-depth analysis of the empirical mode decomposition method and Hilbert transform, the concept of Hilbert spectrum is introduced. For the end effect of Hilbert-Huang transform, this paper proposes to use the combination of least squares support vector regression machine and the mirror extension method to extend the power quality disturbance signal; for the mode mixing problem, this paper puts forward using the wavelet transform to pretreat the signal. Thirdly, by solving the end effect problem and the mode mixing problem of Hilbert-Huang transform, this paper presents the improved HHT method and then applies this method to power quality disturbance detection (transient power quality, harmonic, voltage fluctuation and flicker). Finally, apply the improved HHT to detect the harmonic generated by PV grid-connected system which belongs to State Key Laboratory of Alternate Electrical Power System with Renewable Energy Sources. The simulation results show that the improved HHT method can detect the variation of harmonic real-timely and achieves very good results.