Power Quality Disturbances Research Articles

A Generalized Classification Framework for Power Quality Disturbances Based on Synchrosqueezed Wavelet Transform and Convolutional Neural Networks

A Generalized Classification Framework for Power Quality Disturbances Based on Synchrosqueezed Wavelet Transform and Convolutional Neural Networks

PQ Event Detection After Noise Removal Using Fuzzy Transform and Hilbert Spectral Analysis on EMD

In this paper, a noise removal technique from Power Quality (PQ) signals is proposed which is based on a soft thresholding method. Thresholding is accomplished by using Fuzzy Membership Function (MF) and Empirical Mode Decomposition (EMD). To assess the denoising performance, the proposed technique is applied to detect PQ disturbances from the denoised power system signals. The EMD is applied to decompose the signal into different components. Next, the Intrinsic Mode Functions (IMFs) obtained from the EMD are utilized to determine the IMF dependent thresholding parameters, which are used into the modified fuzzy MF for the noise removal. For improved performance, several random versions of the first IMF are created, independently soft thresholded, and averaged to yield the final denoised version of the first IMF. A separate measurement is used to decide how many IMFs are required to be thresholded. For the evaluation of the noise removal performance, 1) the proposed method is compared with a wavelet-transform-based state of art technique, and 2) an application is presented on the denoised signal for the detection of PQ events from real and simulated power system disturbance data. The denoised and processed IMFs are subjected to Hilbert spectral analysis for the identification of the localized features, which are then utilized for the detection of PQ events in smart grid signal.

A Conversion of Signal to Image Method for Two-Dimension Convolutional Neural Networks Implementation in Power Quality Disturbances Identification

The power quality is identified and monitored to prevent the worst effects arise on the electrical devices. These effects can be device failure, performance degradation, and replacement of some device parts. The deep convolutional neural networks (DCNNs) method can extract the complexity of image features. This method is adopted for the power quality disruption identification of the model. However, the power quality signal data is a time series. Therefore, this paper proposes an approach for the conversion of a power quality disturbance signal to an image. This research is conducted in several stages for constructing the approach proposed. Firstly, the size of a matrix is determined based on the sampling frequency values and cycle number of the signal. Secondly, a zero-cross algorithm is adopted to specify the number of signal sample points inserted into rows of the matrix. The matrix is then converted into a grayscale image. Furthermore, the resulting images are fed to the two-dimension (2D) CNNs model for the PQDs feature learning process. When the classification model is fit, then the model is tested for power quality data prediction. Finally, the model performance is evaluated by employing the confusion matrix method. The model testing result exhibits that the parameter values such as accuracy, recall, precision, and f1-score achieve at 99.81%, 98.95%, 98.84, and 98.87 %, respectively. In addition, the proposed method's performance is superior to the previous methods.Â

Identification of Power Quality Disturbances Using S-Transform and Multi-Class Support Vector Machine

Abstract: An essential issue in power quality disturbances is identifying and classifying power quality disturbances from anywhere and at any time. This article proposed a new approach to identify and classify power quality disturbances over the web using S-transform, Multi-Class Support vector machine (SVM), and Matlab framework. S-Transform is used as an extraction feature to obtain the temporal frequency characteristics of power quality events. The development of the multi-class SVM classifier, in which the system classifies various power quality disturbances. Finally, the Matlab framework integrated the graphical and computational processes with remote access via the web. The test result indicated the suggested method's effectiveness and robustness for identifying and classifying power quality disturbances through the web.Abstrak: Masalah penting dalam gangguan kualitas daya adalah mengidentifikasi dan mengklasifikasikan gangguan kualitas daya dari mana saja dan kapan saja. Artikel ini mengusulkan pendekatan baru untuk mengidentifikasi dan mengklasifikasikan gangguan kualitas daya melalui web menggunakan S-transform, Multi-Class Support vector machine (SVM), dan Matlab. S-Transform digunakan sebagai fitur ekstraksi untuk mendapatkan karakteristik frekuensi temporal dari peristiwa kualitas daya. Multi class SVM classifier dikembangkan dimana sistem mengklasifikasikan berbagai gangguan kualitas daya. Akhirnya, Matlab framework mengintegrasikan proses grafis dan komputasi sehingga dapat diakses jarak jauh melalui web. Hasil pengujian menunjukkan efektivitas dan robustnes metode yang usulkan untuk mengidentifikasi dan mengklasifikasikan gangguan kualitas daya melalui web.

Evaluation of wavelet-based feature extraction methods for detection and classification of power quality disturbances

The appearance of Power Quality Disturbances can cause serious damage to the utility grid. Their detection and identification are two of the major problems related to the improvement of Power Quality. This paper presents an evaluation of different combinations of wavelet-based features for the detection and classification of eight types of Single Power Quality Disturbances. A set of disturbances was generated in MATLAB through their mathematical models. The detection stage was performed using Multiresolution Analysis. The extracted features were normalized by Z-score to serve as input to four different classifiers: Multilayer Perceptron, K-Nearest Neighbors, Probabilistic Neural Network, and Decision Tree. The combination of Shannon Entropy and Log-Energy Entropy was found the best with the highest accuracy in all cases. Furthermore, the normalization stage has an impact on classification as it improves accuracy regardless of the classifier used. This fact makes it possible to reduce the computational expense by using only two types of features without compromising the accuracy.

A meta-heuristic optimization-based method for parameter estimation of an electric arc furnace model

Throughout distribution systems, it is usual to find non-linear time-varying loads, such as electric arc furnaces (EAFs), which are widely used in the steel-making industrial sector. Due to the process of melting and refining metals, the EAFs consume large blocks of power (active and reactive power) causing significant power quality disturbances, such as harmonics and voltage fluctuations on distribution networks. Different EAF parametric models have been proposed with the purpose to predict the voltage and current waveforms and then evaluate the performance of the reactive power compensation devices. This paper proposes a novel methodology for the optimal estimation of parameters of an electric arc furnace model, which can achieve lower execution times and error rates compared to some state-of-the-art methods. The methodology was evaluated using three meta-heuristic optimization algorithms such as the particle swarm optimization (PSO) algorithm, the vortex search algorithm (VSA) and the crow search algorithm (CSA); using real and simulated data. From the results, the proposed methodology based on meta-heuristic optimization approaches worked efficiently, allowed estimating the parameters of the electric arc furnace model using a single optimization step, capture the non-sinusoidal, non-linearity and time-varying random behavior that exhibit the real electric arc furnace samples and obtained relative errors of the total harmonic distortion between the measured and estimated voltage and arc current signals around 1.23% and 0.62%, respectively.

All-in-one three-phase smart meter and power quality analyzer with extended IoT capabilities

The traditional power grid is evolving into a new smart grid that requires better coordination of supply and demand, making it necessary to establish precise monitoring strategies in order to determine grid status in real-time. With the aim of providing a low-cost device based on open-hardware and open-source software to the technicians, engineers, and scientists around the world, this paper presents the three-phase openZmeter (3Ph-oZm), an all-in-one device that allows measuring and computing electrical data related to energy and power quality features in three-phase power networks. It has been designed to perform advanced computations for voltage, current, frequency, power, and energy. 3Ph-oZm is able to process high order harmonics, and log power quality disturbance events defined according to the recommendations of some international standards organizations. The data and its associated features are processed on-site using custom software specifically designed and programmed for this purpose that relies on advanced signal analysis techniques. This smart meter significantly improves the capabilities of the single-phase version, and overcomes certain shortcomings of other commercial devices, both in terms of versatility and data acquisition and processing capabilities. The system has been calibrated and validated using laboratory testing set-up and real-world applications, such as long-term photovoltaic power plant metering. The capabilities of 3Ph-oZm can also support a variety of other electrical applications, such as three-phase induction motor health monitoring, energy savings, or microgrid state estimation.

Novel method based on Teager Energy Operator for online tracking of power quality disturbances

• A TEO -based PQ disturbance tracking algorithm is proposed for the voltage fluctuations with sag disturbance. • The inverter-based DG, nonlinear load and fault event influence on the voltage flicker and sag levels. • The new approach can be used to track multiple PQ disturbance with a complex structure. • This method with proper accuracy can track the voltage amplitude variations in the presence of notch and harmonics. This paper presents an optimal algorithm for simultaneous tracking of power quality (PQ) events in distribution systems in the presence of the inverter-based distributed generation and nonlinear load. The algorithm is based on Teager Energy Operator (TEO). The main specification of the Teager Energy Operator, compared with the commonly used methods in the literature, is the simplicity of its mathematical. Using a proposed method, a suitable combination of the TEO and a relevant filter set is defined as a criterion to reduce envelope ripples and improve a suitable tracking for the power quality disturbances. This approach is able to accurately detect the distortions and sudden changes of voltage under different conditions. The paper considers the voltage fluctuations and voltage sag at the point of common coupling (PCC) and nearby points of the PCC as these voltage disturbances. Extensive simulations for the several power quality disturbance signals and the distribution network based on nonlinear elements using the new approach are performed and performance evaluation is clearly distinguished from the results. Simulation results for single and multiple power quality disturbances present that the proposed method is precise and suitable in detecting voltage disturbances.

An image-based deep transfer learning approach to classify power quality disturbances

• Pre-trained deep learning models can be effectively used to detect power quality disturbances. • A new approach is proposed and validated using real-time data. • This approach is based on the use of Transfer Learning and adopts images of voltage waveforms. • The proposed approach allows reducing the complexity of the training process and is suitable for real-time deployment. Power quality disturbances (PQDs) consist in deviation of voltage and current waveforms from the ideal sinusoid at fundamental frequency, and need to be monitored to ensure a reliabile electrical supply. While, traditionally, power quality monitoring has been performed using signal processing techniques, coupled with shallow Machine Learning classifiers or wave change detection methods, more recently, new approaches, based on Deep Learning, have been proposed. These methods have the potential to achieve high classification accuracy and to remove the need of extensive data pre-processing, hence being more suitable for real-time deployments. However, high classification performance has been only demonstrated using synthetically generated data. In order to address limitations related to processing time and accuracy, this paper proposes a novel end-to-end framework for automated detection of PQDs based on Deep Transfer Learning. The proposed approach uses a small set of images of voltage waveforms to train the model and classify different types of PQDs. This method leverages on the high performance of existing pre-trained models for image classification and shows consistent high accuracy for data with varying resolution. The proposed methodology provides a pathway towards effective deployment of Deep Learning in power quality monitoring systems and real-time applications.

Ensemble deep learning for automated classification of power quality disturbances signals

• A deep learning framework for intelligent classification of PQD signals is proposed. • The improved Bagging algorithm has better accuracy and generalization performance. • The active learning strategy selects the most representative samples from a dataset. • The active learning strategy uses less labeled data thus saving manual labeling cost. The automatic classification of power quality disturbances (PQD) is of great significance for solving power quality problems. In this study, we propose an ensemble deep learning framework to realize intelligent classification of PQ disturbances. Specifically, based on the characteristics of the sequence of disturbance signals, the Long Short Term Memory (LSTM) network is used to classify the signals. In addition, the Bagging theory is used to integrate the training results of multiple LSTM networks to improve the generalization of the network. Our contribution lies in the combination of deep learning and ensemble learning to extract the classification representation of PQD signals. In view of the large number of unlabeled power quality disturbance samples in the power grid, the active learning strategy is adopted to select the most representative samples from the data set, which can enhance the model performance with less labeled data. Finally, experiments were conducted in different noise environments. Compared with the existing multi-label learning models, this method achieves better classification performance with good calculation speed. Furthermore, the proposed active learning strategy is able to train the classification model with fewer labeled samples, reducing the manual labeling costs.

Comparison analysis of different classification methods of power quality disturbances

<span lang="EN-US">Good power quality delivery has always been in high demand in power system utilities where different types of power quality disturbances are the main obstacles. As these disturbances have distinct characteristics and even unique mitigation techniques, their detection and classification should be correct and effective. In this study, eight different types of power quality disturbances were synthetically generated, by using a mathematical approach. Then, continuous wavelet transform (CWT) and discrete wavelet transform with multi-resolution analysis (DWT-MRA) were applied, which eight features were then extracted from the synthesized signals. Three classifiers namely, decision tree (DT), support vector machine (SVM) and k-nearest neighbors (KNN) were trained to classify these disturbances. The accuracy of the classifiers was evaluated and analyzed. The best classifier was then integrated with the full model, which the performance of the proposed model was observed with 50 random signals, with and without noise. This study found that wavelet-transform was effective to localize the disturbances at the instant of their occurrence. On the other hand, the SVM classifier is superior to other classifiers with an overall accuracy of 94%. Still, the need for these classifiers to be further optimized is crucial in ensuring a more effective detection and classification system.</span>

Islanding detection in microgrid using deep learning based on 1D CNN and CNN-LSTM networks

Islanding detection is a critical task due to safety hazards and technical issues for the operation of microgrids. Deep learning (DL) has been applied for islanding detection and achieved good results due to the ability of automatic feature learning in recent years. Long short term memory (LSTM) and two dimensional (2D) convolutional neural networks (CNN) based DL techniques are implemented and demonstrated well performance on islanding detection. However, one dimensional (1D) CNN is more suitable for real-time implementations since it has relatively low complexity and cost-effective than 2D CNN. In this paper, for the first time, the 1D CNN and the combination of 1D CNN-LSTM are proposed for islanding detection to better exploit the global information of islanding data samples using the strengths of both networks. The proposed methods utilize only voltage and current harmonic measurements as input at the point of common coupling (PCC). About 4000 cases under the modified CERTS microgrid model are simulated to evaluate the performance of the proposed architectures. The simulation results and the presented analysis show that the proposed networks have achieved the maximum accuracy of 100% on the task of islanding detection; especially the proposed CNN-LSTM model outperforms the other approaches. Furthermore, the robustness of the proposed methods is demonstrated with unseen samples under low none detection zone and the expansion of microgrid topology. • This paper presents a novel 1D CNN-LSTM based deep learning method for islanding detection by using only harmonic measurements at the PCC. • The proposed method is robust to power quality disturbances and sudden changes. • The robustness of the proposed method is verified under the expansion of DGs with the six hundred unseen islanding and non-islanding samples. • The effectiveness of the proposed methods is proven under small power mistmaches, by testing with two hundred unseen events.

Generation of Power Spectrum of Power Quality Disturbances using MATLAB

The aim of this paper is to present power spectrum of electrical power quality disturbances obtained by mathematical modeling. Signals considered for analysis are transients, harmonics, interruption, voltage sag, voltage swell, voltage fluctuations and voltage flicker. These power quality disturbances affect the continuity of supply and results in reduction of stability of the power system. So there is a necessity for an analysis of power quality disturbances in different perspectives. Power spectrum representation of the different disturbance signals is obtained. The values of power in Decibels and frequency in Hertz are obtained which gives more information about the signals in time domain. Power spectrum gives useful information about the variations in signals. Waveforms for transients, interruption, sag, swell, harmonics, voltage fluctuations, and voltage flicker are generated using mathematical equations and their corresponding power spectrum representations are obtained in MATLAB environment.

Adaptive spectral trend based optimized EWT for monitoring the parameters of multiple power quality disturbances

The growing penetration of renewable energy sources leads to power quality disturbances (PQDs) being multiple. The real-time monitoring of the parameters of multiple power quality disturbances (MPQDs) can help electricity distributors better control power quality problems. In this paper, an adaptive spectral trend-based optimized empirical wavelet transform (EWT) is proposed to analyze MPQDs. First, the upper and lower envelopes of the signal frequency spectrum are fitted by piecewise cubic Hermite interpolation. Then the sum of the upper and lower envelopes is halved as a spectral trend. These steps are repeated until all peak-to-peak distances are greater than the key threshold, indicating that the spectral trend has already been optimized. Third, main frequency components are calculated based on peaks of the optimal spectral trend, which can be used as a basis for spectrum segmentation. Eventually, the components decomposed by EWT represent different disturbance elements, whose parameters can be further calculated by Hilbert transform (HT). To verify the effectiveness of the proposed algorithm, two databases are applied for analysis: synthetic signals, and recording signals by the experimental platform. At the same time, four advanced algorithms are used for comparison, which are scale-space optimized EWT (SOEWT), improved EWT (IEWT), variational mode decomposition (VMD) and fast Stockwell transform (FST). The results showed that the proposed method can accurately segment the spectrum and prevent the inappropriate segmentation of the original EWT so that it can monitor the disturbance parameters with high precision. In addition, the proposed algorithm has a simple computation.

Integration of renewable energy sources in the dynamic voltage restorer for improving power quality using ANFIS controller

Nowadays, grid power loss and power quality deterioration are causing many problems, especially for sensitive uses such as data centers, airports, healthcare, banking, insurance, and telecom. These issues must be resolved through energy continuity coverage and optimization. Dynamic voltage restorers (DVR) are generally used to mitigate network voltage disturbances and maintain a constant voltage value between load terminals for a short and limited duration due to limited energy storage facilities. However, this paper's proposed (DVR) system controlled by the adaptive-network fuzzy inference system (ANFIS) controller can compensate for prolonged power quality disturbances by integrating a hybrid renewable energy system (HRES) managed using classical Proportional Integral (PI) controller power management. This paper proposes a new (DVR) topology coupled with a hybrid renewable energy system (HRES) to exploit free and clean energy, consisting of a solar panel, a PEM fuel cell, and a battery storage device connected through DC-DC converters to a DC transmission, the obtained results from the simulation process of the proposed DVR system in the environment of MATLAB / Simulink showed the ability to eliminate sag that exceeded 0.9 pu in a period of more than 3 min and swells that exceeded 1.2 pu in a period that exceeded one minute. In comparison, the load voltage's total harmonic distortion THDv was reduced from 29% to 5%, and the source current total harmonic distortion THDi from 30.25% to 2.79%.

Methodology for the Surveillance the Voltage Supply in Public Buildings Using the ITIC Curve and Python Programming

This paper proposes an easy-to-implement method for detecting and assessing two of the most frequent PQ (Power Quality) problems: voltage sags and swells. These can affect sensitive equipment such as computers, programmable logic controllers, contactors, etc. Therefore, it is of great interest to implement it in any laboratory, not only for protection reasons but also as a safeguard for claims against the supply company. Thanks to the actual context, in which it is possible to manage big volumes of data, connect multiple devices with IoT (Internet of Things), etc., it is feasible and of great interest to monitor the voltage at specific points of the network. This makes it possible to detect voltage sags and swells and diagnose which points are more prone to this type of problems. For the detection of sags and swells, a program written in Python is in charge of crawling all the files in the database and target those RMS values that fall outside the established limits. Compared to LabVIEW, which might have been the most logical alternative, being the acquisition hardware from the same company (National Instruments), Python has a higher computational performance and is also free of charge, unlike LabVIEW. Thanks to the libraries available in Python, it allows a hardware control close to what is possible using LabVIEW. Implemented in MATLAB, the ITIC (Information Technology Industry Council) power acceptability curve reflects the impact of these power quality disturbances in electrical power systems. The results showed that the combined action of Python and MATLAB performed well on a conventional desktop computer.

Fast Support Vector Machine for Power Quality Disturbance Classification

The power quality disturbance (PQD) problem involves problems of voltage swell, voltage sag, power interruption, harmonics and complex events involving multiple PQD problems. The PQD problem attracted considerable attention from utilities, especially when renewable energy is getting a higher penetration. The PQD problem could downgrade the service quality, causing problems of malfunctions and instabilities. This paper proposed a simplified SVM technique to identify the PQD problem including the multiple PQD classification. With the simple structure proposed, the methodology could reduce a great deal of training data; requires much less memory space and saves computing time. An IEEE 14-bus power system was used to show the performance. Many tests were conducted, and the method was compared with an artificial neural network (ANN). Simulation results showed the shortened processing time and the effectiveness of the proposed approach.

Combined ST/MST and radial basis function neural networks for power quality disturbance signal classification

In this paper, the classification of power quality disturbances using combined ST/MST (S-Transform/Modified S-Transform) and Radial Basis Function Neural Network (RBFNN) is proposed. The extraction of significant features from the power quality disturbance signals is one of the challenging tasks in recognizing different disturbances. The Stockwell Transform/Modified Stockwell Transform (ST/MST) based features are distinct, understandable and more immune to noise. The important attributes present in the signals are retrieved from the ST/MST contours, MST 3D plots and MST based statistical curves. The relevant features are also extracted from the statistical curves. The extracted features are given as input to the RBFNN for further classification. This method is evaluated under both noisy and noiseless conditions. The performance of the proposed approach is compared with other conventional approaches in the literature. The simulation results demonstrate that the proposed MST based RFNN technique is more effective for the detection and classification of power quality disturbances.

Analysis, monitoring, and mitigation of power quality disturbances in a distributed generation system

Despite the numerous benefits of distributed generation (DG), such as reliable energy supply and environmental friendliness, there are a number of challenges associated with integrating DG with the grid. In the field of DG, the complication of power quality (PQ) is a major technical challenge. Custom power devices (CPDs) are used in the distribution system to solve complications. This study presents different PQ disturbances (PQDs), monitoring techniques, and fundamental standards. Furthermore, with the widespread literature, the study critically reviews different PQ mitigation techniques such as distribution static compensator (DSTATCOM), dynamic voltage restorer (DVR), unified power quality conditioner (UPQC), and uninterruptable power supply (UPS). The present research work is not only limited to surveying the existing techniques but also analyzing the performance of UPQC using the ANFIS controller. Overall, this study is intended to provide researchers working on improving PQ in the distribution system with a valuable resource that will aid them in enriching their research.

Highly accurate detection of power quality disturbance using segmented and modified S-transform

Power quality (PQ) is important for the normal operation of modern power grids. The accurate detection of PQ disturbances is the basis for dealing with PQ events in power systems. This paper proposes an algorithm based on the segmented and modified S–transform (SMST) to accurately detect PQ disturbances. The SMST uses a modified Gaussian window with two parameters that can be adaptively adjusted according to the intrinsic characteristics of the PQ signals. Furthermore, the idea of segmentation is used in the modified S–transform, whereby the range of frequency of the signals is divided into three parts, i.e., low-, medium-, and high-frequency bands. The two adjustable parameters can then be accordingly selected in different frequency bands. This enables the SMST to adaptively satisfy different requirements of detection at different frequency bands. Finally, the results of detection of the proposed model were compared with those of prevalent algorithms in the area through simulations. They showed that it can significantly increase the time-frequency resolution in different frequency bands, and has superior accuracy of detection.