The Research On Improved HHT Algorithm In Detection Of The Micro-grid Power Quality

In recent years, as the country propose the development of the smart grid, micro-grid as an important part of the smart grid, not only it has the advantage of benefits of renewable energy generation and other distributed energy system, saving non-renewable energy, improve the environment, and increases the the reliability and availability of the grid. With the application and development of micro-grid, it also has a lot of problems. With a large number of micro-grid electricity power equipment applications and a variety of distributed power, the micro-grid has power quality fluctuations: such power harmonics, voltage fluctuations,voltage sag and other issues.To be effective against a variety of micro-grid power quality fluctuations in effective detection, researchers have proposed a number of methods, such as: wavelet transform analysis, Prony algorithm, short-time Fourier transform analysis. Hilbert- Huang transform analysis method is a new time-frequency analysis method, it has many traditional methods incomparable advantages. HHT method has self-adaptive in analyzing nonlinear signal, Heisenberg uncertainty principle is not restricted, the instantaneous frequency is obtained by the use of derivative. Due to these advantages, HHT algorithm has widespread attention and application in recent years.But HHT method also has some problems, such as the end effect problem, modal aliasing problem, to improve these problems, there are some improvement research. This paper presents a variety of HHT improved algorithms, including: windowing function method, waveform matching method continuation, continuation method mirrors, auto-regressive moving average model method, polynomial fitting, the proportion of continuation method, support vector regression method, artificial neural network, combined with extreme points extension window function method, mirror extension artificial neural network method, these methods to some extent, solve the problem of end effects; intermittent cycle scale method is based on the guidelines, the interference signal extreme value method wavelet pretreatment, integrated EMD method, pseudo-wave method, high-frequency harmonics method, mask signal method, to some extent, these methods can weaken modal aliasing problems.This paper uses mirror extension method to improve end effect, the application of a mask signal method improve modal aliasing.