T.K. Abdel-Galil

Power quality disturbance classification using the inductive inference approach

This paper presents a novel approach for the classification of power quality disturbances. The approach is based on inductive learning by using decision trees. The wavelet transform is utilized to produce representative feature vectors that can accurately capture the unique and salient characteristics of each disturbance. In the training phase, a decision tree is developed for the power quality disturbances. The decision tree is obtained based on the features produced by the wavelet analysis through inductive inference. During testing, the signal is recognized using the rules extracted from the decision tree. The classification accuracy of the decision tree is not only comparable with the classification accuracy of artificial Neural networks, but also accounts for the explanation of the disturbance classification via the produced if... then rules.

Disturbance classification utilizing dynamic time warping classifier

The application of deregulation policies in electric power systems results in the absolute necessity to quantify power quality. This fact highlights the need for a new classification strategy which is capable of tracking, detecting, and classifying power-quality events. In this paper, a new classification approach that is based on the dynamic time warping (DTW) algorithm is proposed. The new algorithm is supported by the vector quantization (VQ) and the fast match (FM) techniques to speed up the classification process. The Walsh transform (WT) and the fast Fourier transform (FFT) are adopted as feature extraction tools. The application of the combined fast match-dynamic time warping (FM-DTW) algorithms provides superior results in speed and accuracy compared to the traditional artificial neural networks and fuzzy logic classifiers. Moreover, the proposed classifier proves to have a very low sensitivity to noise levels.

Online tracking of voltage flicker utilizing energy operator and Hilbert transform

The Teager energy operator (TEO) and the Hilbert transform (HT) are introduced in this paper as effective approaches for tracking the voltage flicker levels in distribution systems. The mathematical simplicity of the proposed techniques, compared with the commonly used algorithms in the literature, renders them competitive candidates for the online tracking of voltage flicker levels. Moreover, the TEO and the HT are capable of tracking the amplitude variations of the voltage flicker and supply frequency in industrial systems with only a 3% margin of error. Such accurate tracking facilitates the implementation of the control of flicker mitigation devices. A detailed comparison of the two proposed approaches that profile the different factors affecting tracking accuracy is presented. The results are provided to verify the tracking capabilities of both HT and TEO and to indicate the superior performance of the TEO and the HT in tracking voltage flicker.

Hilbert transform based control algorithm of the DG interface for voltage flicker mitigation

Distributed Generation (DG) is used widely in the modern distribution systems. This paper proposes a novel functionality of the interface between DG and the utility network to mitigate the voltage flicker and to regulate the voltage at the Point of Common Coupling (PCC) in addition to its main function of controlling the power flow. A new control algorithm for the DG interface based on the Hilbert transform (HT) is presented. The HT is employed as an effective technique for tracking the voltage flicker levels in distribution systems. The mathematical simplicity of the proposed technique, compared with the commonly used algorithms in the literature, renders them competitive candidates for the on-line tracking of voltage flicker. The accurate tracking of the HT facilitates its implementation for the control of flicker mitigation devices. Simulation results are provided to verify the tracking capabilities of the HT and to evaluate the performance of the proposed DG interface for multifunction operation.

Power quality event detection using Adaline

Abstract Vast spread of sensitive loads in power systems results in increasing susceptibility to power quality problems, which makes fast detection and classification algorithms a necessity. A new approach for power quality event detection is presented in this paper. This approach utilizes linear adaptive neuron, which is called Adaline. Unlike the previously proposed detection methods, which are based on wavelet analysis that encounter mathematical calculations burden, Adaline is fast due to its simple construction, which makes it more suitable for on-line applications.

Partial discharge pattern classification using the fuzzy decision tree approach

Partial discharge (PD) measurement is a proven flaw detection technique for finding cavities that are defects in the insulating material. In this paper, a novel approach for the classification of cavity sizes, based on their maximum PD charge transfer-applied voltage (/spl Delta/Q-V) characteristics using a fuzzy decision tree system, is proposed. The (/spl Delta/Q-V) partial discharge patterns for different cavity sizes are represented by features extracted from their pulse shapes, and the classification rules are directly extracted from the data using the decision tree. The decision rules obtained from the decision tree are then converted to the fuzzy IF-then rules, and the back-propagation algorithm is utilized to tune the parameters of the membership functions employed in the fuzzy classifier. The neuro-fuzzy classification technique is shown to provide successful classification of void sizes in an easily interpretive fashion.

SVM classification of contaminating particles in liquid dielectrics using higher order statistics of electrical and acoustic PD measurements

Electrical and acoustic partial discharge (PD) measurement and pattern recognition procedures are described for detecting and identifying contaminating particles in transformer mineral oils. This work introduces the use of support vector machines (SVM), a nonlinear non-parametric automatable machine learning algorithm, for the purpose of classifying the size and composition of such particles. The training and validation of acoustic and electrical PD measurement data, which are contaminated by time varying noise, are first filtered adaptively using wavelet decomposition. Statistics of a particles impact upon collision with the walls of a tank, containing the electrode test assembly and the inter arrival time between collisions constitute the features for the SVM classifier. These statistics include higher order moments and the entropy of the estimated density function of the features. Results based on experimental training and testing data indicate that fusing of the acoustic and electric PD information at the features level provides a nearly perfect classification success rate. These observations demonstrate that, while electrical and acoustic PD data are correlated, they contain individually independent and complementary information regarding the state and condition of transformer type mineral oils.

Partial discharge pulse pattern recognition using an inductive inference algorithm

This paper presents a novel approach in the area of time dependent partial discharge (PD) pulse pattern recognition, to applications based on the inductive learning (decision tree) approach. Different attributes based on pulse shape analysis are used as representative feature vectors that can accurately capture the unique and salient characteristics of the PD pulse shape. In the training phase, a decision tree is developed to relate the pulse shape with the cavity size by using inductive machine learning. The C4.5 machine learning algorithm is deployed to realize the tree using the training data, since it has the capability of inferring the rules and to produce the tree in terms of continuous features. During testing, the cavity size is recognized by means of the rules extracted from the decision tree. The dependency between the features and the classes are examined using the mutual information approach. The proposed algorithm possesses the inherent advantage of explaining the result via the self-created rule base as demonstrated by the results obtained. Those self-created rules can be employed as the basis for applying a fuzzy expert system for the classification of void sizes in an easily interpreted fashion.

Partial discharge pulse pattern recognition using Hidden Markov Models

An approach for the classification of cavity sizes based on their maximum charge transfer characteristics, applied voltage partial discharge pattern using Hidden Markov Models, is described. In these models, the partial discharge patterns for different cavity sizes are represented by a sequence of events rather than by the actual curves. In the training phase, each cavity size represents a unique class, which emits its own eigen sequence. Vector Quantization is deployed to assign labels for this particular sequence of events. A Hidden Markov Model is trained for each class, using a set of training patterns consisting of the labels produced by Vector Quantization. During testing, the sequence of events to be recognized is quantized and then matched against all the developed models. The best-matched model pinpoints the cavity size class. Experimental results demonstrate the remarkable capability of the proposed algorithm.

Particle identification in terms of acoustic partial discharge measurements in transformer oils

Acoustic measurements of partial discharge (PD) are employed to classify particles in transformer mineral oil according to material and size. Wavelet multi-resolution analysis data of the acoustic signals together with higher order statistics of the particle intercollision times and magnitudes comprise the input features to a Support Vector Machine (SVM) classifier. The training and validation measurement data, which are contaminated by time varying noise, are first filtered using wavelet decomposition. Results indicate that the SVM algorithm with the selected features provides a remarkably high success rate when classifying particles by size and material type. A potentially significant conclusion is that acoustic measurements alone are by themselves effective in classifying discharged particles in terms of the foregoing selected features. The proposed algorithm can be employed to enhance quality control procedures based on acoustic measurements of partial discharge.