The Research Of Power Quality Disturb Signal Denoising

As the high duty electronic equipment has been commonly utilized and increasingly popular nowadays, power quality issue within power system area has brought lots of attentions. In order to revolve power quality issues and purify power environment, the technical data reflected by the various power quality issues is necessary to be precisely measured, however, is largely influenced by noise signal factors such as electromagnetic interference and nonlinearity of detective sensors. Consequently, the method for removing the noise signal becomes the key factors of resolving the power quality problems.Signal processing is the major field of study in the development of information science. Signal extraction, signal transformation, the analysis and synthesis signal processes etc. are collectively called signal processing, generally including signal denoising, character abstraeting and border abstraeting. The method of de nosing can be varied, traditionally includes: Linear filtering denoising, Nonlinear filtering denoising and Fourier transform denoising. The newly discovered approaches includes Wavelet transform denoising, Wavelet packet analysis denoising, EMD decomposition denoising, Gauss filter denoising, the Singular value decomposition denoising and the Wavelet neural network denoising. The techniques mentioned above are various from the de noise effects and their pros and cons.The overall objective of this thesis is researching the noise signal within the topic of power quality. The signals in power quality can be categorized as two divisions: one division covers the signal voltage swell, sag, interruption and harmonic voltage fluctuation in the short period of time; the other one contains transient voltage fluctuation signals such as once mutation and oscillation mutation signal.Firstly, this thesis explores the techniques for the analysis of the wavelet transform denoising, followed by the detailed discussion of wavelet transform denoising properties. With respect to wavelet transform denoising, three solutions are covered in this thesis: decomposition, the selection of wavelet base and the selection of threshold value which is the significant limited factor the wavelet transform denoising. With an example of the voltage fluctuation signal, this thesis briefly explains the identification of decomposition level, the selection of wavelet base function according to wavelet basis; and how to determine the threshold values.In subsequent sections of this thesis deals with wavelet packet analysis denoising and EMD decomposition denoising according to the voltage fluctuation signal short-time Fourier transform for denoising, wavelet denoising. These methods above are applied to the voltage fluctuation by MATLAB software simulation, with the combination of the modeling and programming, power quality signals are analyzed and compared by evaluating the de noise effects.Last of section of this thesis is conducting wavelet basis selection method as well as four types of de nosing, and the conclusion is drawn according to the comparison and analysis of transit power quality signal.A summary of the work has been carried out in this thesis, make an outlook about some works and research direction on the next step.