Wyniki wyszukiwania - BazTech

1

Optimized Approach of Feature Selection Based on Binary Genetic Algorithm in Classification of Induction Motor Faults

Le Truong-An

Annals of Computer Science and Information Systems

|

2022

|

Vol. 33

145--150

EN

In this paper, an effective model for detection and classification of multiple faults in induction motors is presented. It used S-transform method is used to analyze current signals measured from four different motors including a healthy motor, broken rotor bars, bearing damage, stator winding short-circuits fault. The feature set is extracted based on signal spectrum. With strong exploration capabilities in the search space, binary genetic algorithm (BGA) is proposed to select the optimal feature subset. As the classifier, the backpropagation neural network and support vector machine are used. The simulation results showed that the average accuracy of 100 trails is 98.3\% and the optimal feature subset equal to 36\% of total original features. That means the number of redundant features removed is 64\%. In conclusion, the proposed model combined with BGA reached highly effective in the classification of induction motor.

2

Improving the resolution of seismic data based on S transform and modified variational mode decomposition, an application to Songliao Basin, Northeast China

Liu Siqin, Zhou Zhusheng, Peng Sanxi, Yang Yang, Zeng Weizu, Chen Keyang

Acta Geophysica

|

2022

|

Vol. 70, no 3

1103--1113

EN

The sandstone of the lithologic reservoir in Songliao Basin is characterized by thin vertical thickness and rapid lateral pinch outs, which make it difficult to identify and describe the sandstone due to insufficient resolution of existing seismic data. To solve this problem, a method based on S-transform and modified variational mode decomposition is proposed to improve the resolution in the time–frequency domain by cepstrum deconvolution. Firstly, the time–frequency spectrum of seismic records is obtained by generalized S-transform; then it is transformed into cepstrum; and the wavelet amplitude spectrum is extracted by modified variational mode decomposition with permutation entropy, to realize the time–frequency domain deconvolution of cepstrum. After improving the resolution of seismic data in the study area, the frequency band of the data is broadened by more than 20%. After frequency expansion, the seismic refection structure reflects the sedimentary characteristics of the actual sandstone in the area. The seismic refection has a good correspondence with the well point sandstone, and the thin sandstone is clearly depicted. It is fully verified that this method can effectively improve the resolution of seismic data and has the characteristics of relative amplitude preservation.

3

SpEC: A system for patient specific ECG beat classification using deep residual network

Allam Jaya Prakash, Samantray Saunak, Ari Samit

Biocybernetics and Biomedical Engineering

|

2020

|

Vol. 40, no. 4

1446--1457

EN

Electrocardiogram (ECG) is a non-invasive technique used to detect various cardiac disorders. One of the major causes of cardiac arrest is an arrhythmia. Furthermore, ECG beat classification is essential to detect life-threatening cardiac arrhythmias. The major limitations of the traditional ECG beat classification systems are the requirement of an extensive training dataset to train the model and inconsistent performance for the detection of ventricular and supraventricular ectopic (V and S) beats. To overcome these limitations, a system denoted as SpEC is proposed in this work based on Stockwell transform (ST) and two-dimensional residual network (2D-ResNet) for improvement of ECG beat classification technique with a limited amount of training data. ST, which is used to represent the ECG signal into a time-frequency domain, provides frequency invariant amplitude response and dynamic resolution. The resultant ST images are applied as input to the proposed 2D-ResNet to classify five different types of ECG beats in a patient-specific way as recommended by the Association for the Advancement of Medical Instrumentation (AAMI). The proposed SpEC system achieved an overall accuracy (Acc) of 99.73%, sensitivity (Sen) = 98.84%, Specificity (Spe) = 99.50%, Positive predictivity (Ppr) = 98.20% on MIT-BIH arrhythmia database, and shows an overall Acc of 89.87% on real-time acquired ECG dataset with classification time of single ECG beat image = 0.2365 (s) in detecting of five arrhythmia classes. The proposed method shows better performance on both the database compared to the earlier reported state-of-art techniques.

4

Cauchy-Stieltjes Families with Polynomial Variance Functions and Generalized Orthogonality

Bryc Włodzimierz, Fakhfakh Raouf, Młotkowski Wojciech

Probability and Mathematical Statistics

|

2019

|

Vol. 39, Fasc. 2

237--258

EN

This paper studies variance functions of Cauchy-Stieltjes Kernel (CSK) families generated by compactly supported centered probability measures. We describe several operations that allow us to construct additional variance functions from known ones. We construct a class of examples which exhausts all cubic variance functions, and provide examples of polynomial variance functions of arbitrary degree. We also relate CSK families with polynomial variance functions to generalized orthogonality. Our main results are stated solely in terms of classical probability; some proofs rely on analytic machinery of free probability.

5

Tandem repeats detection in DNA sequences using Kaiser window based adaptive S-transform

Sharma S. D., Saxena R., Sharma S. N.

Bio-Algorithms and Med-Systems

|

2017

|

Vol. 13, no. 3

167--173

EN

In computational biology the development of algorithms for the identification of tandem repeats in DNA sequences is a challenging problem. Tandem repeats identification is helpful in gene annotation, forensics, and the study of human evolution. In this work a signal processing algorithm based on adaptive S-transform, with Kaiser window, has been proposed for the exact and approximate tandem repeats detection. Usage of Kaiser window helped in identifying short as well as long tandem repeats. Thus, the limitation of earlier S-transform based algorithm that identified only microsatellites has been alleviated by this more versatile algorithm. The superiority of this algorithm has been established by comparative simulation studies with other reported methods.

6

Dispersion calculation method based on S-transform and coordinate rotation for Love channel waves with two components

Feng L., Zhang Y.

Acta Geophysica

|

2017

|

Vol. 65, no. 4

757--764

EN

Dispersion analysis is an important part of in-seam seismic data processing, and the calculation accuracy of the dispersion curve directly influences pickup errors of channel wave travel time. To extract an accurate channel wave dispersion curve from in-seam seismic two-component signals, we proposed a time–frequency analysis method based on single-trace signal processing; in addition, we formulated a dispersion calculation equation, based on S-transform, with a freely adjusted filter window width. To unify the azimuth of seismic wave propagation received by a two-component geophone, the original in-seam seismic data undergoes coordinate rotation. The rotation angle can be calculated based on P-wave characteristics, with high energy in the wave propagation direction and weak energy in the vertical direction. With this angle acquisition, a two-component signal can be converted to horizontal and vertical directions. Because Love channel waves have a particle vibration track perpendicular to the wave propagation direction, the signal in the horizontal and vertical directions is mainly Love channel waves. More accurate dispersion characters of Love channel waves can be extracted after the coordinate rotation of two-component signals.

7

Event-based S-transform approach for nonintrusive load monitoring

Khalid K., Mohamed A., Shareef H., Sabeeh M.

Przegląd Elektrotechniczny

|

2016

|

R. 92, nr 5

194--198

EN

In this study, a nonintrusive load monitoring system is developed by analyzing the power signal obtained from a single point of power meter installation to detect ON/OFF load activities. A mathematically designed model with backpropagation neural network is utilized in load pattern recognition to decompose the load operation. Leveraging its unique load signature profile, the S-transform approach is employed to extract the features from the aggregate power signal and analyze the detection of load start-up transient from signal processing. To improve the accuracy of load identification for unknown data, the power factor is used as an additive feature with 99.32% load recognition accuracy.

PL

W artykule analizowany jest system monitorowania obciążenia sieci. Wykorzystano sieć neuronową do rozpoznawania rodzaju Transformata S jest użyta do ekstrakcji danych z sygnału mocy. Dodatkowo do identyfikacji obciążenia użyto współczynnik mocy.

8

Application of S-transform to signal analysis

Szymczyk P., Szymczyk M.

Bio-Algorithms and Med-Systems

|

2015

|

Vol. 11, no. 4

223--229

EN

This paper presents opportunities of signal analysis using S-transform. Images of the module and the angle of S-transform show the results. These images may be used as models to compare with unknown signals and to detect patterns and anomalies, even in very sophisticated signals.

9

Wyznaczanie atrybutów sygnału EEG w oparciu o transformatę Stockwella

Rutkowski G., Patan K.

Pomiary Automatyka Kontrola

|

2013

|

R. 59, nr 3

208--211

PL

W nowoczesnych systemach medycznych w dziedzinie elektroencefalografii coraz bardziej zwiększa się nacisk na udoskonalanie aparatury pomiarowej. Nieustannie poszukuje się rozwiązań poprawiających niedoskonałości sprzętowe, a trudności dotyczą zarówno sfery konstrukcyjnej, jak i zaimplementowanych algorytmów. Problemy dotyczą eliminacji artefaktów i samej charakterystyki sygnałów EEG. Proponowane rozwiązania począwszy od metod klasycznych, a skończywszy na metodach opartych na sztucznej inteligencji ciągle ewoluują i pozwalają na wdrażanie coraz to nowszych rozwiązań na potrzeby kliniczne. Techniki oparte na analizie widmowej pozwalają wspomóc pracę lekarzy specjalistów w procesie diagnostycznym dla poszczególnych dysfunkcji o podłożu neurologicznym. Jednym ze stosowanych rozwiązań jest dynamicznie rozwijająca się metodologia oparta na zaawansowanych narzędziach analizy widmowej. Transformata S pozwala na wprowadzenie i zastosowanie funkcji okna o zmiennej szerokości zależnej od częstotliwości. Uzyskane informacje pozwalają zarówno określić rozdzielczość zależną od częstotliwości, jak i wyznaczyć widmo. W artykule opisano eksperyment na próbkach rzeczywistych pomiarów EEG zgromadzonych przy ścisłej współpracy z Oddziałem Neurologii i Udarów Szpitala Wojewódzkiego w Zielonej Górze. Zaprezentowano wyniki przy użyciu transformaty S w ekstrakcji cech i klasyfikacji zaburzeń neurologicznych dla przypadków napadów epileptycznych.

EN

In modern medical systems more and more emphasis is put on improvement of the measuring equipment. We are constantly looking for solutions to improve both hardware and software. The main problems relate to the elimination of artifacts and the analysis of EEG signal characteristics. To date elaborated solutions still evolve and allow for the implementation of still newer and newer solutions for clinical needs. One of possible solutions is to use a dynamically developing methodology based on advanced spectral analysis tools. First, the Fourier Transform was used, but it turned out to be effective only for stationary signals. The Fourier Transform allows extracting information about the signal spectrum components, without providing the information about the component occurrence time of the component. Unfortunately, EEG signals are non-stationary in nature. The solution may be S Transform, which can be viewed as an extension of the popular Short-Time Fourier Transform and wavelet transform. S Transform allows for the introduction and application of window functions with a variable width frequency dependent. The resulting information helps us to determine attributes of EEG signals needed for classification. The paper deals with experiments carried out using EEG samples collected in close collaboration with the Ward of Neurology and Strokes of Provincial Hospital of Zielona Góra. EEG signals were recorded using 16-channel equipment under the supervision of experts in neurology practices. In result, 1154 sequences were acquired including both dysfunctions (586 epileptic seizures) and normal records (568). EEG sequences were analyzed using S-transform to extract signal features. The last step was classification of EEG signals performed using a nearest neighbor classifier. The presented results are very promising and may have an impact on the improvement and refinement of medical diagnostic tools.

10

Zwektoryzowane algorytmy obliczania Transformaty S dla środowiska wieloprocesorowego

Tariov A., Gliszczyński M.

Pomiary Automatyka Kontrola

|

2011

|

R. 57, nr 11

1401-1403

PL

W pracy został zaprezentowany wektoryzowany algorytm obliczania transformaty S w dwóch wariantach - w postaci sekwencyjno-równoległej pozwalającej na oszczędzenie zasobów sprzętowych oraz w postaci równoległej pozwalającej wykorzystać, nowoczesne wielordzeniowe platformy obliczeniowe. W drugim przypadku możliwa jest znaczna redukcja czasu trwania algorytmu. Obie metody mogą znaleźć zastosowanie praktyczne zależnie od oczekiwanej dokładności (rozdzielczości) i szybkości działania jak też możliwości platformy obliczeniowej.

EN

In the paper the algorithm for calculating N by N-point S Transform is presented. In a sequential, recursive option hardware resources saving is available, while on the other hand, a parallel version of the algorithm allows increasing the accuracy and reducing the time when using multi-core platforms. Two of these approaches can be implemented in practical use depending on the expected accuracy, speed and power of the hardware platform. At the beginning of the paper uses of S Transform with other similar solutions are described. Advantages and disadvantages of S Transform, which are good properties of the time-frequency analysis of non-stationary signals thanks to a movable, different sized Gaussian window, but at the same time a long computation time of the standard, sequential method, are considered. Next, the theoretical, continuous form of the transform and the discrete form with the sequential algorithm are presented. Later The main part of the work deals with synthesis of the sequential and parallel version of the algorithm in the matrix-vector form. The data flow in the algorithms in space and time is shown in Figs. 1 and 2 (for sequential and parallel approach). Finally, the computation times of two versions are compared. The advantage of the two presented approaches is simple and understandable tensor product representation which makes the implementation easy. The sequential algorithm can be used for slower platforms, where the real time analysis is not necessary, while the parallel version offers quick computation on multi-core processors.

11

Time-frequency Representation to the Analysis of Partial Discharges in Power Cables Using Novel Transform Method

Guo C., Huang C., Qian Y., Jiang X.

Przegląd Elektrotechniczny

|

2010

|

R. 86, nr 6

288-291

EN

Partial discharge is induced by the defect or failure inside high voltage power cables. The non-stable property of pulse signals requires both time and frequency domain information to indicate the characteristics. This paper introduces the application of the newly developed timefrequency representation, i.e. S-transform, to the analysis of partial discharge signals. Through simulation and on-site experiment, S-transform presents its effectivity of extracting partial discharge signal information.

PL

Wyładowanie niezupełne może być powodowane przez defekty lub uszkodzenia wewnątrz kabli wysokiego napięcia. Analiza sygnałów impulsowych wymaga zastosowania analizy zarówno czasowej jak i częstotliwościowej aby można było uzyskać pełną informację o zjawisku. W artykule przedstawiono nowe metody takiej analizy, jak na przykład transformatę S. Na podstawie symulacji i eksperymentów można stwierdzić że transformata S jest efektywnym narzędziem do wydobycia informacji o wyładowaniach.

12

Power quality diagnosis in distribution networks using support vector regression based S-transform technique

Faisal M., Mohamed A., Shareef H.

Przegląd Elektrotechniczny

|

2010

|

R. 86, nr 11a

38-42

EN

This paper presents a novel method for performing automatic power quality diagnosis to identify the causes of short duration voltage disturbances such as voltage sags and swells. Such voltage disturbances can be caused by permanent or non permanent faults. A permanent fault causes permanent damage and power interruption to the customers whereas a non permanent fault can be categorized as either transient or incipient faults. In the proposed power quality diagnosis method, a time frequency analysis technique called as the S-transform is used to analyse and extract features of voltage disturbances recorded from the power quality monitoring system. The support vector regression which is an intelligent technique is then used identify whether the voltage disturbances are caused by permanent, non permanent, transient or incipient faults. Test results proved that the proposed power quality diagnosis method can provide accurate diagnosis on the causes of short duration voltage disturbances.

PL

W artykule przedstawiono nową metodę przeprowadzania automatycznej diagnostyki jakości energii elektrycznej do identyfikacji przyczyn krótkoczasowych zakłóceń napięcia, takich jak zapady napięcia. Zaburzenia napięcia mogą być spowodowane przez długotrwałe lub chwilowe awarie. W proponowanej metodzie diagnozowania jakości zasilania, zastosowano transformatę S do ekstrakcji charakterystyk zarejestrowanych przebiegów z systemu monitoringu. Zastosowano regresję SVR jako technike inteligentna, pozwalajaca na rozróżnienie pomiędzy typami awarii. Wyniki badań wykazały, że proponowana metoda diagnozowania jakości zasilania może zapewnić dokładną diagnozę na temat przyczyn zaburzeń napięcia o krótkim czasie trwania.