Dogan Gökhan Ece

Power-quality event analysis using higher order cumulants and quadratic classifiers

In this paper, we present a novel power-quality (PQ) event detection and classification method using higher order cumulants as the feature parameter, and quadratic classifiers as the classification method. We have observed that local higher order statistical parameters that are estimated from short segments of 50-Hz notch-filtered voltage waveform data carry discriminative features for PQ events analyzed herein. A vector with six parameters consisting of local minimas and maximas of higher order central cumulants starting from the second (variance) up to the fourth cumulant is used as the feature vector. Local vector magnitudes and simple thresholding provide an immediate event detection criterion. After the detection of a PQ event, local maxima and minima of the cumulants around the event instant are used for the event-type classification. We have observed that the minima and maxima for each statistical order produces clusters in the feature space. These clusters were observed to exhibit noncircular topology; hence, quadratic-type classifiers that require the Mahalanobis distance metric are proposed. The events investigated and presented are line-to-ground arcing faults and voltage sags due to the induction motor starting. Detection and classification results obtained from an experimentally staged PQ event data set are presented.

2-D analysis and compression of power-quality event data

This paper introduces a novel two-dimensional (2-D) representation of the power quality event data. 2-D discrete-time wavelet transform is applied to the 2-D representation of real-life event data. The proposed representation and the transform is tested in terms of both event analysis and data compression. The experimental results indicate that the 2-D transform of the event data outperforms the results obtained by conventional one-dimensional (1-D) wavelet transform-based methods.

Covariance analysis of voltage waveform signature for power-quality event classification

In this paper, covariance behavior of several features (signature identifiers) that are determined from the voltage waveform within a time window for power-quality (PQ) event detection and classification is analyzed. A feature vector using selected signature identifiers such as local wavelet transform extrema at various decomposition levels, spectral harmonic ratios, and local extrema of higher order statistical parameters, is constructed. It is observed that the feature vectors corresponding to power quality event instances can be efficiently classified according to the event type using a covariance based classifier known as the common vector classifier. Arcing fault (high impedance fault) type events are successfully classified and distinguished from motor startup events under various load conditions. It is also observed that the proposed approach is even able to discriminate the loading conditions within the same class of events at a success rate of 70%. In addition, the common vector approach provides a redundancy and usefulness information about the feature vector elements. Implication of this information is experimentally justified with the fact that some of the signature identifiers are more important than others for the discrimination of PQ event types

Condition monitoring of speed controlled induction motors using wavelet packets and discriminant analysis

This work presents an intelligent method for the condition monitoring of induction motors supplied with adjustable speed drives (ASD). Most of the previous work in this area concentrated on the fault detection and classification of induction motors supplied directly from an a.c. line. However, ASD driven induction motors are widely used in industrial processes and therefore obtaining an intelligent tool for the condition monitoring of these motors is necessary in terms of preventive maintenance and reducing down time due to motor faults. Here 3-phase supply side current of the ASD driving an induction motor is used to extract statistical features of wavelet packet decomposition coefficients within a frequency range of interest. This way, the information regarding the output frequency of the ASD and hence the motor speed is not required. Six identical three-phase induction motors were used for the experimental verification of the proposed method. One healthy machine was used as a reference, while other five with various synthetic faults were used for condition detection and classification. Extracted features obtained from decomposition coefficients of different wavelet filter types for all motors were employed in three different and popular classifiers. The proposed method and the performance of the features used for fault detection and classification are examined at various motor loads and speed levels, and it is shown that a successful condition monitoring system for induction motors supplied with ASDs is developed. The effect of selected filter type in wavelet decomposition to the condition monitoring process is analyzed and presented.

Modeling of hysteresis and power losses in transformer laminations

In this paper, a method for the representation of hysteresis and power losses in the laminations of power transformers is proposed. The developed model is based on data supplied from a steel manufacturer and able to predict hysteresis and eddy current losses.

An adaptive statistical method for power quality analysis

Digital fault recorders installed for monitoring current and/or voltage waveforms acquire and store vast amount of waveform data for post processing. Because of this, effective offline automated event detection from acquired data is necessary. In this work, we propose a new automatic event detection method which takes the acquired data and produces event flags at instances of events. The method is based on the statistical analysis of adaptive decomposition signals. The combination of an adaptive prediction filter-based subband decomposition structure with a rule-based histogram analysis block produced successful detection and localization results on our real-life power system transient data.

An analysis of transformer excitation current under nonsinusoidal supply voltage

In this paper, the effect of nonsinusoidal supply voltage on the core loss and the transformer excitation current is investigated. The time-domain waveform of the transformer excitation current is calculated for a worst-case harmonic composition of supply voltage.

Analysis of voltage flicker using Hilbert-Huang Transform

Hilbert Huang Transform (HHT), which was proposed by Huang and developed by Flandrin and his group, is a new signal processing method that can be used in the analysis of nonlinear and nonstationary signals. This study suggests an approach of using Hilbert Huang Transform to measure the voltage flicker in power systems. In the suggested method, voltage signal is decomposed into Emprical Mode Decomposition EMD and Intrinsic Mode Function IMF components. These components are used in the calculation of the frequency and amplitude of voltage flicker. The clear success of EMD in depicting envelope variations of a sinusoidal waveform has been the main motivation for the adoption of HHT in flicker analysis. Simulations are done by using input signal modulated with single flicker frequencies and multi flicker frequencies and input signals which include harmonic. Simulations show that the method may be used in voltage flicker analyse and gives good results.

Hilbert–Huang Transform Based Approach for Measurement of Voltage Flicker Magnitude and Frequency

Abstract Voltage flicker is a non-stationary waveform for which direct spectral analysis is not appropriate. To overcome this difficulty, a Hilbert–Huang transform based technique is proposed here. Hilbert–Huang transform is a new signal processing method that can be used in the analysis of non-linear and non-stationary signals. In the suggested method, the recorded voltage signal is decomposed into Hilbert–Huang transform components, namely the empirical mode decomposition and intrinsic mode function components. These components are used in the calculation of the frequency and amplitude of voltage flicker. The clear success of empirical mode decomposition in depicting envelope variations of a sinusoidal waveform has been the main motivation for the adoption of Hilbert–Huang transform in flicker analysis. Simulations are performed over waveforms, including single- and multiple-flicker frequencies and flicker with harmonic, voltage sag, and voltage swell. The waveforms are selected as pure sinusoids, as well as harmonically rich voltage waveforms. Simulation results show that the proposed methodology constitutes a plausible way to analyze voltage flickers, making it an alternative to the available flicker analysis tools.AbstractVoltage flicker is a non-stationary waveform for which direct spectral analysis is not appropriate. To overcome this difficulty, a Hilbert–Huang transform based technique is proposed here. Hilbert–Huang transform is a new signal processing method that can be used in the analysis of non-linear and non-stationary signals. In the suggested method, the recorded voltage signal is decomposed into Hilbert–Huang transform components, namely the empirical mode decomposition and intrinsic mode function components. These components are used in the calculation of the frequency and amplitude of voltage flicker. The clear success of empirical mode decomposition in depicting envelope variations of a sinusoidal waveform has been the main motivation for the adoption of Hilbert–Huang transform in flicker analysis. Simulations are performed over waveforms, including single- and multiple-flicker frequencies and flicker with harmonic, voltage sag, and voltage swell. The waveforms are selected as pure sinusoids, as wel...

A 2D representation for analysis and coding of power quality events

In this work, we demonstrate examples about uses of practical image processing techniques over a new interpretation of power quality event data. Power quality event data are 1D data obtained from a real life system with three-phase RL loads, induction motors and varying mechanical loads. First, the new 2D representation is presented. Next, 2D wavelet transform based analysis and compression methods are presented. 2D wavelet transform of the 2D representation enable us to clearly detect power quality events that perturbs the normal operation waveform. Furthermore, the transform coefficients are observed to be more suitable for compression than the conventional 1D wavelet based results that could be found in the literature. Simulations are presented.