An effective wavelet-based feature extraction method for classification of power quality disturbance signals

https://doi.org/10.1016/j.epsr.2008.03.002Get rights and content

Abstract

This paper presents a wavelet norm entropy-based effective feature extraction method for power quality (PQ) disturbance classification problem. The disturbance classification schema is performed with wavelet-neural network (WNN). It performs a feature extraction and a classification algorithm composed of a wavelet feature extractor based on norm entropy and a classifier based on a multi-layer perceptron. The PQ signals used in this study are seven types. The performance of this classifier is evaluated by using total 2800 PQ disturbance signals which are generated the based model. The classification performance of different wavelet family for the proposed algorithm is tested. Sensitivity of WNN under different noise conditions which are different levels of noises with the signal to noise ratio is investigated. The rate of average correct classification is about 92.5% for the different PQ disturbance signals under noise conditions.

Introduction

As it is well known, an ideal three-phase ac supply consists of three-phase voltages that are 120° out of phase and have identical magnitudes. Above all, these voltages should be sinusoidal and should be available continuously. Any diversion from these requirements is considered as poor quality [1].

Poor quality of electric power is normally caused by power-line disturbances, such as impulses, notches, glitches, momentary interruptions, wavefaults, voltage sag/swell, harmonic distortion, and flicker, resulting in misoperation or failure of end-use equipment. In order to improve power quality (PQ), the sources and causes of such disturbances must be known before appropriate mitigating actions can be taken. A feasible approach to achieve this goal is to incorporate detection capabilities into monitoring equipment so that events of interest will be recognized, captured, and classified automatically. Hence, good performance monitoring equipment must have functions which involve the detection, localization, and classification of transient events. In particular, when the disturbance type has been classified accurately, the PQ engineers can define the major effects of the disturbance at the load and analyze the source of the disturbances so that an appropriate solution can be formulated [2].

Artificial neural network (ANN) systems can provide an effective method to cope with such problems [8], [10]. However, the complexity of the classifier structure may depend on the choice of the feature parameters (magnitude, duration, frequency component, or waveform shape) as well building the ANN system. Therefore, an effective signal processing technique must be offer for analyzing PQ related problems.

In literature, the signal processing techniques are available for analyzing PQ disturbance. Some examples are fast Fourier transform method [3], fractal-based method [4], S-transform method [5], time–frequency ambiguity plane method [6], short time power and correlation transform method [7], wavelet transform method [8].

The major problem of the traditional analyzing methods is that it is not provide sufficient information on the time domain. One technique emerged to overcome the above-mentioned problem is by using wavelet transform whose strength is on handling signals on short time intervals for high frequency components and long time intervals for low frequency components. By means of the strength, wavelet transform is considered suitable for analyzing signals with localized impulses and oscillations particularly for those commonly present in fundamental and low order harmonics.

In 1994, the usage of wavelet transform was proposed to study power systems non-stationary harmonic distortion [9]. In 1996, this technique was started to be use as a power tool that has ability to analysis in both time and frequency domains for detect and localize PQ problems [8]. After these papers, many studies were done on automatic classification of PQ disturbances. Gaouda et al. proposed the wavelet-multiresolution analysis technique to detect, localize, and classify different PQ problems. A new feature extraction method based the standard deviation at different resolution levels was applied as inputs to neural network to classify PQ disturbances type [10]. Elmitwally et al. proposed a new classifier using neuro-fuzzy network. A PQ disturbance recognition system using wavelet statistical features is employed [11]. A new classification methodology that is based on machine inductive learning implemented using the C4.5 algorithm, decomposed with wavelet transform of original signals, is proposed in this paper [12]. It is showed that typical PQ disturbances are correctly classified. A novel classifier is performed by using probabilistic neural network (PNN) for PQ application [2]. In this scheme, energy distribution features from wavelet analysis and Parseval's theorem are used to PQ disturbance recognition and classification. He and Starzyk proposed a novel approach for PQ disturbances classification based on the wavelet transform and self organizing learning array (SOLAR) system in the paper [13]. Energy value at each decomposition level using multiresolution analysis is applied to SOLAR. The performance of this system tested under different noise conditions is investigated and it can be correctly classified PQ disturbances.

To utilize a reliable classifier, it is essential to extract a useful feature vector that can reduce data size as well indicating and recognizing the main characteristics of signal. Hence, in this paper, a wavelet feature extraction technique based on norm entropy is proposed for automatic PQ disturbances classification. The disturbance classification schema is performed with wavelet-neural network (WNN). WNN realizes a feature extraction and a classification algorithm composed of a wavelet feature extractor based on norm entropy and a multi-layer perceptron classifier. WNN is applied on a set of different PQ disturbances, such as pure sine (normal), sag, swell, outage, harmonics, and sag with harmonic and swell with harmonic. Then, the effect on the classification success of different wavelet family is investigated. Sensitivity to noise of the proposed schema is tested under different noise conditions. Finally, the performance comparison between the proposed method and previous literature reports are presented for a better validation. The result shows that WNN could analyze the PQ signal efficiently.

The novelty presented in this paper can summarized as follows.

A norm entropy-based effective feature extraction method that is reduced size of the feature vector from the wavelet decomposition and multi resolution analysis is proposed. Using this method, the classification accuracy percentage of the PQ disturbances can be increased by an easier disturbance classifier based on a multi-layer perceptron.

The paper is organized as follow. In Section 2, it is given a preliminary for wavelet transform, multiresolution analysis and WNNs. Several brief definitions can be seen at this section. In Section 3, the methodology and the implementation of the proposed process is given. In Section 4, simulation and analysis study is introduced and the classification results are shown. Finally, conclusions are discussed in Section 5.

Section snippets

Wavelet transform

Wavelet transform has been proven to very efficient in signal analysis. It finds applications in different areas of engineering due to its ability to analyze the local discontinuities of signals. The main advantages of wavelets is that they have a varying window size, being wide for slow frequencies and narrow for the fast ones, thus leading to an optimal time–frequency resolution in all the frequency ranges. Furthermore, owing to the fact that windows are adapted to the transients of each

Methodology

The WNN structure is consisted of two stages. These are feature extraction stage, classification stage.

Data generation

Data generation by parametric equations for classifiers tests has advantageous in some ways. It was possible to change training and testing signal parameters in a wide range and in a controlled manner. The signals simulated with this way were very close to the real situation. On the other hand, different signals belonging to the same class gave possibility to estimate generalization ability of classifiers based on neural networks [25].

The input data to the WNN based on PQ disturbances

Conclusion

In this paper, a wavelet norm entropy-based effective feature extraction method is proposed for the automatic PQ disturbance classification. The disturbance classification schema is performed with wavelet-neural network (WNN). It performs a feature extraction and a classification algorithm composed of a wavelet feature extractor based on norm entropy and a classifier based on a multi-layer perceptron. In addition, the comparison of wavelet families (Daubechies, coiflets, and biorthogonals) for

References (28)

  • S.J. Huang et al.

    Feasibility of fractal-based methods for visualization of power system disturbances

    Int. J. Elect. Power Energy Syst.

    (2001)
  • A. Sengur et al.

    Wavelet packet neural networks for texture classification

    Expert Syst. Appl.

    (2007)
  • A.M. Gargoom et al.

    A comparative study on effective signal processing tools for optimum feature selection in automatic power quality events clustering

  • Z.L. Gaing

    Wavelet-based neural network for power disturbance recognition and classification

    IEEE Trans. Power Deliv.

    (2004)
  • G.T. Heydt et al.

    Applications of the windowed FFT to electric power quality assessment

    IEEE Trans. Power Deliv.

    (1999)
  • P.K. Dash et al.

    Power quality analysis using S-transform

    IEEE Trans. Power Deliv.

    (2003)
  • M. Wang et al.

    A classification of power quality disturbances using time-frequency ambiguity plane and neural networks

    IEEE Power Eng. Soc. Summer Meet.

    (2001)
  • J. Wen et al.

    A method for detection and classification of power quality disturbances

    Autom. Elect. Power Syst. China

    (2002)
  • S. Santoso et al.

    Power quality assessment via wavelet transform analysis

    IEEE Trans. Power Deliv.

    (1996)
  • P. Ribeiro

    Wavelet transform: an advanced tool for analyzing non-stationary harmonic distortion in power systems

  • A.M. Gaouda et al.

    Power quality detection and classification using wavelet-multiresolution signal decomposition

    IEEE Trans. Power Deliv.

    (1999)
  • A. Elmitwally et al.

    Proposed wavelet-neurofuzzy combined system for power quality violations detection and diagnosis

    IEE Proc.: Gen. Transm. Distrib.

    (2001)
  • T.K. Abdel-Galil et al.

    Power quality disturbance classification using the inductive inference approach

    IEEE Trans. Power Deliv.

    (2004)
  • H. He et al.

    A self-organizing learning array system for power quality classification based on wavelet transform

    IEEE Trans. Power Deliv.

    (2006)
  • Cited by (225)

    • Wavelet based real-time monitoring of electrical signals in Distributed Generation (DG) integrated system

      2021, Engineering Science and Technology, an International Journal
      Citation Excerpt :

      Discrete Wavelet Transform (DWT) DWT is a method that disintegrates a discrete signal into different resolution levels [9]. It deals with two phases.

    View all citing articles on Scopus
    View full text