WO2022119426A1 - Intelligent system for immediate detection and notification of disturbances in electrical signal quality - Google Patents
Intelligent system for immediate detection and notification of disturbances in electrical signal quality Download PDFInfo
- Publication number
- WO2022119426A1 WO2022119426A1 PCT/MA2020/050006 MA2020050006W WO2022119426A1 WO 2022119426 A1 WO2022119426 A1 WO 2022119426A1 MA 2020050006 W MA2020050006 W MA 2020050006W WO 2022119426 A1 WO2022119426 A1 WO 2022119426A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- electrical signal
- notification
- quality
- processing unit
- faults
- Prior art date
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 14
- 230000007547 defect Effects 0.000 claims abstract description 19
- 238000012545 processing Methods 0.000 claims abstract description 15
- 238000013145 classification model Methods 0.000 claims abstract description 12
- 238000007781 pre-processing Methods 0.000 claims abstract description 11
- 230000005540 biological transmission Effects 0.000 claims abstract description 6
- 238000013135 deep learning Methods 0.000 claims abstract description 6
- 230000004075 alteration Effects 0.000 claims abstract description 4
- 230000002123 temporal effect Effects 0.000 claims description 5
- 238000000034 method Methods 0.000 abstract description 8
- 238000013473 artificial intelligence Methods 0.000 abstract 1
- 238000011179 visual inspection Methods 0.000 abstract 1
- 230000001052 transient effect Effects 0.000 description 12
- 230000003534 oscillatory effect Effects 0.000 description 11
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/28—Testing of electronic circuits, e.g. by signal tracer
- G01R31/2832—Specific tests of electronic circuits not provided for elsewhere
- G01R31/2836—Fault-finding or characterising
- G01R31/2846—Fault-finding or characterising using hard- or software simulation or using knowledge-based systems, e.g. expert systems, artificial intelligence or interactive algorithms
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/28—Testing of electronic circuits, e.g. by signal tracer
- G01R31/2832—Specific tests of electronic circuits not provided for elsewhere
- G01R31/2836—Fault-finding or characterising
- G01R31/2837—Characterising or performance testing, e.g. of frequency response
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Definitions
- the study of signal quality relates to the study of electromagnetic occurrences that deflect electrical signals.
- the detection of these defects is essential to improve the quality of the signal.
- a signal processing technique e.g., Fourier transform[1], [2], wavelet transform[3]-[5], WVD[ 6], S-transform[7], etc.
- the second step is the classification of these defects by a specific method (e.g., Support Vector Machine (SVM)[8], Artificial Neural Network (ANN)[7][9], Decision trees[10], etc.).
- SVM Support Vector Machine
- ANN Artificial Neural Network
- CNN Convolution Neural Network
- the deep learning approach unlike standard approaches, does not require manual feature extraction from each defect, which minimizes human interaction in extracting and choosing the best defect features by allowing the model to create its own features.
- the present invention therefore aims to:
- the present invention relates to an intelligent system for the detection and instantaneous notification of faults in the quality of the electrical signal characterized by:
- Said processing unit manages the execution and storage of the results of the pre-processing of the electrical signal and of the classification model.
- Said classification model takes said image to identify the existence of defects using a deep learning architecture based on a multi-label classification.
- An intelligent system for instantaneous detection and notification of faults in the quality of the electrical signal characterized by:
- Said processing unit manages the execution and storage of the results of the pre-processing of the electrical signal and of the classification model.
- Said classification model takes said image to identify the existence of defects using a deep learning architecture based on a multi-label classification.
- ResNet-18 we used several types of the ResNet model: ResNet-18, ResNet-34, ResNet-50, ResNet-101 for multi-label signal quality classification.
- the difference between these models is their depth of 18 to 101 layers.
- the learning is done on 500 images per class (a total of 14500 images), using the cross-validation method of 5 stratified groups which reorganizes these images into 5 groups in a random way in order to guarantee a better representation of each class in each group.
- These models are tested on a dataset of 100 images per class (a total of 2900 images) never seen during training, to quantify their performance and degree of generalization.
- the element (1) is said system
- the element (2) is said connectors
- the element (3) is said electromagnetic adapter
- the element (4) is said unit of processing which can take a memory card (5) as input
- the element (6) is said wireless transmission system.
- the development process of said classification model is done in three phases: the generation of the signal dataset, the preprocessing of the signals, and the learning and selection of the best performing model.
- the generation of signals is done through several mathematical models that are part of a public model developed under MATLAB [13], and defining several classes
- Class 1 Pure sinusoidal • Class 13 - Sag with Oscillatory transient
- Class 2 Sag • Class 14 - Swell with Oscillatory transient
- Class 9 Harmonics with Swell • Class 20 - Sag with Harmonics with Flicker
- Class 12 Flicker with Swell • Class 22 - Sag with Harmonics with • Class 26 - Harmonics with Sag with Flicker Oscillatory transient with Oscillatory transient
- the preprocessing of the signals transforms each into an image of size 128xl28pixels (named FTS image) which contains the superposition of its temporal representation and its frequency representation described in three distinct gray levels:
- the detection of defects existing in the signal is done by injecting the image resulting from the preprocessing to said learned classification model.
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Image Analysis (AREA)
Abstract
The present invention relates to an intelligent system for immediate detection and notification of defects in electrical signal quality, and is characterised by: • a system comprising input connectors for electrical cables, an electromagnetic adapter, a processing unit and a wireless transmission system. • The electromagnetic adapter adapts, without alteration, the input voltage to the input voltage of the processing unit. • The processing unit manages the execution and storage of the results of the preprocessing of the electrical signal and the classification model. • The preprocessing of the electrical signal makes it possible to convert it into an image in which the time representation and the frequency representation of the signal overlap one another. • The classification model takes the image to identify the existence of defects using a deep learning architecture based on a multi-label classification. This system makes it possible to detect a plurality of single defects or a combination of defects in the electrical signals in an immediate, intelligent and intuitive manner. This product is based on an artificial intelligence method, in particular a deep learning method, which applies the concept that a visual inspection by a person of the time and frequency domain of the signal enables the identification of a defect or defects.
Description
Système intelligent pour détection et notification instantanée des perturbations de la qualité du signal électrique. Intelligent system for instant detection and notification of electrical signal quality disturbances.
Description Description
L'étude de la qualité du signal se rapporte à l'étude des occurrences électromagnétiques déviant les signaux électriques. Ainsi, la détection de ces défauts est essentielle pour améliorer la qualité du signal. Plusieurs efforts on étaient faites pour l'analyse et la classification de ces défauts. En fait, le processus d'identification des défauts de la qualité de la puissance s'effectue par deux étapes principales. La premier étape est l'extraction des caractéristiques des défauts par le biais d'une technique du traitement du signal (e.g., Transformée de fourrier[l], [2], transformée d'ondelette[3]-[5], WVD[6], S- transformée[7], etc.). La deuxième étape est la classification de ces défauts par une méthode spécifique (e.g., Support Vector Machine (SVM)[8], Artificial Neural Network (ANN)[7][9], Arbres de décisions[10], etc.). The study of signal quality relates to the study of electromagnetic occurrences that deflect electrical signals. Thus, the detection of these defects is essential to improve the quality of the signal. Several efforts were made for the analysis and classification of these defects. In fact, the process of identifying power quality faults involves two main steps. The first step is the extraction of the features of the defects through a signal processing technique (e.g., Fourier transform[1], [2], wavelet transform[3]-[5], WVD[ 6], S-transform[7], etc.). The second step is the classification of these defects by a specific method (e.g., Support Vector Machine (SVM)[8], Artificial Neural Network (ANN)[7][9], Decision trees[10], etc.).
La classification se basant sur des modèles de deep learning devient très courante dans plusieurs domaines et application[ll], [12]. Les auteurs dans [6] ont entraîné un modèle de Convolution Neural Network (CNN) pour classifier six défauts seules et trois de leurs combinaisons avec des images de taille 200x200pixels générées par l'algorithme Wigner-Ville distribution appliqué sur chaque signal de défaut. Classification based on deep learning models is becoming very common in several fields and applications[ll], [12]. The authors in [6] trained a Convolution Neural Network (CNN) model to classify six single defects and three of their combinations with images of size 200x200pixels generated by the Wigner-Ville distribution algorithm applied to each defect signal.
L'approche deep learning, contrairement aux approches standards, ne nécessite pas une extraction manuelle des caractéristiques de chaque défaut, ce qui minimise l'interaction de l'humain à extraire et choisir les meilleures caractéristiques des défauts en permettant au modèle de créer ses propres caractéristiques. The deep learning approach, unlike standard approaches, does not require manual feature extraction from each defect, which minimizes human interaction in extracting and choosing the best defect features by allowing the model to create its own features.
La présente invention vise donc à : The present invention therefore aims to:
• Couvrir un plus grand nombre de défauts. • Cover more defects.
• Développer une nouvelle représentation du signal en image, tout en maintenant l'automatisation de l'extraction de ses caractéristiques. • Develop a new representation of the signal in image, while maintaining the automation of the extraction of its characteristics.
• Minimiser la taille de l'image d'entrée et ainsi la complexité et le temps de calcul nécessaire pour les générer et les traiter. • Minimize the size of the input image and thus the complexity and computation time required to generate and process them.
Plus particulièrement, la présente invention porte sur un système intelligent pour détection et notification instantanée des défauts de la qualité du signal électrique caractérisé par : More particularly, the present invention relates to an intelligent system for the detection and instantaneous notification of faults in the quality of the electrical signal characterized by:
• Un système incorporant des connecteurs d'entrée pour câbles électriques, un adaptateur électromagnétique, une unité de traitement et un système de transmission sans fil. • A system incorporating input connectors for electrical cables, an electromagnetic adapter, a processing unit and a wireless transmission system.
• Ledit adaptateur électromagnétique adapte sans altération la tension d'entrée à la tension d'entrée de ladite unité de traitement. • Said electromagnetic adapter adapts without alteration the input voltage to the input voltage of said processing unit.
• Ladite unité de traitement gère l'exécution et le stockage des résultats du pré-traitement du signal électrique et du modèle de classification. • Said processing unit manages the execution and storage of the results of the pre-processing of the electrical signal and of the classification model.
• Le prétraitement dudit signal électrique permet de le convertir en une image superposant la représentation temporelle et la représentation fréquentielle du signal. • The pre-processing of said electrical signal makes it possible to convert it into an image superimposing the temporal representation and the frequency representation of the signal.
• Ledit modèle de classification prend ladite image pour identifier l'existence des défauts en utilisant une architecture de deep learning se basant sur une classification multi-label.
Un système intelligent pour détection et notification instantanée des défauts de la qualité du signal électrique caractérisé par : • Said classification model takes said image to identify the existence of defects using a deep learning architecture based on a multi-label classification. An intelligent system for instantaneous detection and notification of faults in the quality of the electrical signal characterized by:
• Un système incorporant des connecteurs d'entrée pour câbles électriques, un adaptateur électromagnétique, une unité de traitement et un système de transmission sans fil. • A system incorporating input connectors for electrical cables, an electromagnetic adapter, a processing unit and a wireless transmission system.
• Ledit adaptateur électromagnétique adapte sans altération la tension d'entrée à la tension d'entrée de ladite unité de traitement. • Said electromagnetic adapter adapts without alteration the input voltage to the input voltage of said processing unit.
• Ladite unité de traitement gère l'exécution et le stockage des résultats du pré-traitement du signal électrique et du modèle de classification. • Said processing unit manages the execution and storage of the results of the pre-processing of the electrical signal and of the classification model.
• Le prétraitement dudit signal électrique permet de le convertir en une image superposant la représentation temporelle et la représentation fréquentielle du signal. • The pre-processing of said electrical signal makes it possible to convert it into an image superimposing the temporal representation and the frequency representation of the signal.
• Ledit modèle de classification prend ladite image pour identifier l'existence des défauts en utilisant une architecture de deep learning se basant sur une classification multi-label. • Said classification model takes said image to identify the existence of defects using a deep learning architecture based on a multi-label classification.
Nous avons utilisé plusieurs types du modèle ResNet : ResNet-18, ResNet-34, ResNet-50, ResNet-101 pour une classification multi-label de la qualité du signal. La différence entre ces modèles est leur profondeur de 18 à 101 couches. We used several types of the ResNet model: ResNet-18, ResNet-34, ResNet-50, ResNet-101 for multi-label signal quality classification. The difference between these models is their depth of 18 to 101 layers.
L'apprentissage est fait sur 500 images par classe (un total de 14500 images), en utilisant la méthode cross-validation de 5 groupes stratifiés qui réorganise ces images en 5 groupes d'une façon aléatoire afin de garantir une meilleure représentation de chaque classe dans chaque groupe. Lesdits modèles sont testés sur une dataset de 100 images par classe (un total de 2900 images) jamais vues durant l'apprentissage, pour quantifier leur performance et dégrée de généralisation. The learning is done on 500 images per class (a total of 14500 images), using the cross-validation method of 5 stratified groups which reorganizes these images into 5 groups in a random way in order to guarantee a better representation of each class in each group. These models are tested on a dataset of 100 images per class (a total of 2900 images) never seen during training, to quantify their performance and degree of generalization.
Descriptions des figures Figure Descriptions
-Tel que représenté sur la figure 1, l'élément (1) est ledit système, l'élément (2) est lesdits connecteurs, l'élément (3) est ledit adaptateur électromagnétique, l'élément (4) est ladite unité de traitement qui peut prendre en entrée une carte mémoire (5), et l'élément (6) est ledit système de transmission sans fil. -As shown in figure 1, the element (1) is said system, the element (2) is said connectors, the element (3) is said electromagnetic adapter, the element (4) is said unit of processing which can take a memory card (5) as input, and the element (6) is said wireless transmission system.
-En référence à la figure 2, le processus de développement dudit modèle de classification se fait en trois phases : la génération de la dataset des signaux, le prétraitement des signaux, et l'apprentissage et la sélection du meilleur modèle performant. La génération des signaux se fait par le biais de plusieurs modèles mathématiques faisant parties d'un modèle publique développé sous MATLAB [13], et définissant plusieurs classes -Referring to Figure 2, the development process of said classification model is done in three phases: the generation of the signal dataset, the preprocessing of the signals, and the learning and selection of the best performing model. The generation of signals is done through several mathematical models that are part of a public model developed under MATLAB [13], and defining several classes
Classe 1 - Pure sinusoidal • Classe 13 - Sag with Oscillatory transientClass 1 - Pure sinusoidal • Class 13 - Sag with Oscillatory transient
Classe 2 - Sag • Classe 14 - Swell with Oscillatory transientClass 2 - Sag • Class 14 - Swell with Oscillatory transient
Classe 3 - Swell • Classe 15 - Sag with HarmonicsClass 3 - Swell • Class 15 - Sag with Harmonics
Classe 4 - Interruption • Classe 16 - Swell with HarmonicsClass 4 - Interruption • Class 16 - Swell with Harmonics
Classe 5 - Transient/lmpulse/Spike • Classe 17 - Notch Class 5 - Transient/impulse/Spike • Class 17 - Notch
Classe 6 - Oscillatory transient • Classe 18 - Harmonics with Sag with FlickerClass 6 - Oscillatory transient • Class 18 - Harmonics with Sag with Flicker
Classe 7 - Harmonics • Classe 19 - Harmonics with Swell withClass 7 - Harmonics • Class 19 - Harmonics with Swell with
Classe 8 - Harmonics with Sag Flicker Class 8 - Harmonics with Sag Flicker
Classe 9 - Harmonics with Swell • Classe 20 - Sag with Harmonics with FlickerClass 9 - Harmonics with Swell • Class 20 - Sag with Harmonics with Flicker
Classe 10 - Flicker • Classe 21 - Swell with Harmonics withClass 10 - Flicker • Class 21 - Swell with Harmonics with
Classe 11 - Flicker with Sag Flicker Class 11 - Flicker with Sag Flicker
Classe 12 - Flicker with Swell
• Classe 22 - Sag with Harmonics with • Classe 26 - Harmonics with Sag with Flicker Oscillatory transient with Oscillatory transient Class 12 - Flicker with Swell • Class 22 - Sag with Harmonics with • Class 26 - Harmonics with Sag with Flicker Oscillatory transient with Oscillatory transient
• Classe 23 - Swell with Harmonics with • Classe 27 - Harmonics with Swell with Oscillatory transient Flicker with Oscillatory transient • Class 23 - Swell with Harmonics with • Class 27 - Harmonics with Swell with Oscillatory transient Flicker with Oscillatory transient
• Classe 24 - Harmonics with Sag with • Classe 28 - Sag with Harmonics with Flicker Oscillatory transient with Oscillatory transient • Class 24 - Harmonics with Sag with • Class 28 - Sag with Harmonics with Flicker Oscillatory transient with Oscillatory transient
• Classe 25 - Harmonics with Swell with • Classe 29 - Swell with Harmonics with Oscillatory transient Flicker with Oscillatory transient • Class 25 - Harmonics with Swell with • Class 29 - Swell with Harmonics with Oscillatory transient Flicker with Oscillatory transient
-En référence à la figure 3, le prétraitement des signaux transforme chacun en une image de taille 128xl28pixels (nommée image FTS) qui contient la superposition de sa représentation temporelle et sa représentation fréquentielle décrit en trois niveaux de gris distincts : -Referring to Figure 3, the preprocessing of the signals transforms each into an image of size 128xl28pixels (named FTS image) which contains the superposition of its temporal representation and its frequency representation described in three distinct gray levels:
• Un niveau de gris pour les pixels de la représentation temporel uniquement, • A level of gray for the pixels of the temporal representation only,
• Un niveau de gris pour les pixels de représentation fréquentiel uniquement, • A gray level for the frequency representation pixels only,
• Un niveau de gris pour les pixels d'intensité mutuelles appartenant aux deux représentations en même temps. • A level of gray for the pixels of mutual intensity belonging to the two representations at the same time.
-En référence à la figure 4, la détection des défauts existants au signal se fait en injectant l'image issue du prétraitement à ledit modèle de classification appris. -Referring to Figure 4, the detection of defects existing in the signal is done by injecting the image resulting from the preprocessing to said learned classification model.
Bibliography : Bibliography:
[1] G. T. Heydt, P. S. Fjeld, C. C. Liu, D. Pierce, L. Tu, and G. Hensley, "Applications of the windowed FFT to electric power quality assessment," IEEE Trans. Power Deliv., vol. 14, no. 4, pp. 1411-1416, Oct. 1999. [1] G. T. Heydt, P. S. Fjeld, C. C. Liu, D. Pierce, L. Tu, and G. Hensley, "Applications of the windowed FFT to electric power quality assessment," IEEE Trans. Power Deliver., vol. 14, no. 4, p. 1411-1416, Oct. 1999.
[2] Y. Gu and M. H. J. Bollen, "Time-frequency and time-scale domain analysis of voltage disturbances," IEEE Trans. Power Deliv., vol. 15, no. 4, pp. 1279-1284, Oct. 2000. [2] Y. Gu and M. H. J. Bollen, "Time-frequency and time-scale domain analysis of voltage disturbances," IEEE Trans. Power Deliver., vol. 15, no. 4, p. 1279-1284, Oct. 2000.
[3] F. Ucar, O. F. Alcin, B. Dandil, and F. Ata, "Power quality event detection using a fast extreme learning machine," Energies, vol. 11, no. 1, Jan. 2018. [3] F. Ucar, O. F. Alcin, B. Dandil, and F. Ata, "Power quality event detection using a fast extreme learning machine," Energies, vol. 11, no. 1, Jan. 2018.
[4] F. Jandan, S. Khokhar, S. Abid, A. Shaha, and F. Abbasi, "Recognition and Classification of Power Quality Disturbances by DWT-MRA and SVM Classifier," 2019. [4] F. Jandan, S. Khokhar, S. Abid, A. Shaha, and F. Abbasi, "Recognition and Classification of Power Quality Disturbances by DWT-MRA and SVM Classifier," 2019.
[5] A. M. Gaouda, M. M. A. Salama, M. R. Sultan, and A. Y. Chikhani, "Power quality detection and classification using wavelet-multiresolution signal decomposition," IEEE Trans. Power Deliv., vol. 14, no. 4, pp. 1469-1476, Oct. 1999. [5] A. M. Gaouda, M. M. A. Salama, M. R. Sultan, and A. Y. Chikhani, "Power quality detection and classification using wavelet-multiresolution signal decomposition," IEEE Trans. Power Deliver., vol. 14, no. 4, p. 1469-1476, Oct. 1999.
[6] K. Cai, W. Cao, L. Aarniovuori, H. Pang, Y. Lin, and G. Li, "Classification of Power Quality Disturbances Using Wigner-Ville Distribution and Deep Convolutional Neural Networks," IEEE Access, vol. 7, pp. 119099-119109, Aug. 2019. [6] K. Cai, W. Cao, L. Aarniovuori, H. Pang, Y. Lin, and G. Li, "Classification of Power Quality Disturbances Using Wigner-Ville Distribution and Deep Convolutional Neural Networks," IEEE Access, vol . 7, p. 119099-119109, Aug. 2019.
[7] K. Daud, A. F. Abidin, A. P. Ismail, M. D. A. Hasan, M. A. Shafie, and A. Ismail, "Evaluating windowing-based continuous S-transform with neural network classifier for detecting and classifying power quality disturbances," Indones. J. Electr. Eng. Comput. Sci., vol. 13, no. 3, pp. 1136-1142, Mar. 2019.
[8] Z. Liu, Y. Cui, and W. Li, "A Classification Method for Complex Power Quality Disturbances Using EEMD and Rank Wavelet SVM," IEEE Trans. Smart Grid, vol. 6, no. 4, pp. 1678-1685, Jul. 2015. [7] K. Daud, AF Abidin, AP Ismail, MDA Hasan, MA Shafie, and A. Ismail, "Evaluating windowing-based continuous S-transform with neural network classifier for detecting and classifying power quality disturbances," Indones. J. Electr. Eng. Computer. Science, vol. 13, no. 3, p. 1136-1142, Mar. 2019. [8] Z. Liu, Y. Cui, and W. Li, "A Classification Method for Complex Power Quality Disturbances Using EEMD and Rank Wavelet SVM," IEEE Trans. Smart Grid, vol. 6, no. 4, p. 1678-1685, Jul. 2015.
[9] R. Kumar, B. Singh, D. T. Shahani, A. Chandra, and K. Al-Haddad, "Recognition of Power-Quality Disturbances Using S-Transform-Based ANN Classifier and Rule-Based Decision Tree," IEEE Trans. Ind. Appl., vol. 51, no. 2, pp. 1249-1258, Mar. 2015. [9] R. Kumar, B. Singh, D. T. Shahani, A. Chandra, and K. Al-Haddad, "Recognition of Power-Quality Disturbances Using S-Transform-Based ANN Classifier and Rule-Based Decision Tree," IEEE Trans . Ind. Appl., vol. 51, no. 2, p. 1249-1258, Mar. 2015.
[10] M. Biswal and P. K. Dash, "Measurement and classification of simultaneous power signal patterns with an s-transform variant and fuzzy decision tree," IEEE Trans. Ind. Informatics, vol. 9, no. 4, pp. 1819-1827, 2013. [10] M. Biswal and P. K. Dash, "Measurement and classification of simultaneous power signal patterns with an s-transform variant and fuzzy decision tree," IEEE Trans. Ind. Informatics, vol. 9, no. 4, p. 1819-1827, 2013.
[11] S. M. Islam, H. Mahmood, A. A. Al-Jumaily, and S. Claxton, "Deep learning of facial depth maps for obstructive sleep apnea prediction," in Proceedings - International Conference on Machine Learning and Data Engineering, iCMLDE 2018, 2019, pp. 154-157. [11] S. M. Islam, H. Mahmood, A. A. Al-Jumaily, and S. Claxton, "Deep learning of facial depth maps for obstructive sleep apnea prediction," in Proceedings - International Conference on Machine Learning and Data Engineering, iCMLDE 2018, 2019 , p. 154-157.
[12] A. Sevik, P. Erdogmus, and E. Yalein, "Font and Turkish Letter Recognition in Images with Deep Learning," in International Congress on Big Data, Deep Learning and Fighting Cyber Terrorism, IBIGDELFT 2018 - Proceedings, 2019, pp. 61-64. [12] A. Sevik, P. Erdogmus, and E. Yalein, "Font and Turkish Letter Recognition in Images with Deep Learning," in International Congress on Big Data, Deep Learning and Fighting Cyber Terrorism, IBIGDELFT 2018 - Proceedings, 2019, p.p. 61-64.
[13] R. Igual, C. Medrano, F. J. Arcega, and G. Mantescu, "Integral mathematical model of power quality disturbances," in Proceedings of International Conference on Harmonics and Quality of Power, ICHQP, 2018, vol. 2018-May, pp. 1-6.
[13] R. Igual, C. Medrano, F. J. Arcega, and G. Mantescu, "Integral mathematical model of power quality disturbances," in Proceedings of International Conference on Harmonics and Quality of Power, ICHQP, 2018, vol. 2018-May, pp. 1-6.
Claims
5 5
Revendications Claims
1. Un système intelligent pour détection et notification instantanée des défauts de la qualité du signal électrique caractérisé par : o Un système incorporant des connecteurs d'entrée pour câbles électriques, un adaptateur électromagnétique, une unité de traitement et un système de transmission sans fil. o Ledit adaptateur électromagnétique adapte sans altération la tension d'entrée à la tension d'entrée de ladite unité de traitement. o Ladite unité de traitement gère l'exécution et le stockage des résultats du pré-traitement du signal électrique et du modèle de classification. o Le prétraitement dudit signal électrique permet de le convertir en une image superposant la représentation temporelle et la représentation fréquentielle du signal. o Ledit modèle de classification prend ladite image pour identifier l'existence des défauts en utilisant une architecture de deep Learning se basant sur une classification multi-label. 1. An intelligent system for the detection and instantaneous notification of faults in the quality of the electrical signal characterized by: o A system incorporating input connectors for electrical cables, an electromagnetic adapter, a processing unit and a wireless transmission system. o Said electromagnetic adapter adapts without alteration the input voltage to the input voltage of said processing unit. o Said processing unit manages the execution and storage of the results of the pre-processing of the electrical signal and of the classification model. o The preprocessing of said electrical signal makes it possible to convert it into an image superimposing the temporal representation and the frequency representation of the signal. o Said classification model takes said image to identify the existence of defects using a deep learning architecture based on a multi-label classification.
2. Un système de détection et notification instantanée des défauts de la qualité du signal électrique selon la revendication 1 caractérisé en ce que ladite image à trois niveaux de gris distinguant les pixels appartenant uniquement à la représentation temporelle, des pixels appartenant uniquement à la représentation fréquentielle, et des pixels à mutuelles intensités. 2. A system for instantaneous detection and notification of faults in the quality of the electrical signal according to claim 1, characterized in that said three-level gray image distinguishing the pixels belonging only to the time representation, from the pixels belonging only to the frequency representation , and pixels with mutual intensities.
3. Un système de détection et notification instantanée des défauts de la qualité du signal électrique selon la revendication 1 caractérisé en ce que ledit modèle de classification a appris à détecter, en plus de l'état sain, différents défauts de qualité du signal électrique. . Un système de détection et notification instantanée des défauts de la qualité du signal électrique selon la revendication 1 et 3, caractérisé en ce que ledit modèle de classification est capable de classifier des combinaisons de défauts. 3. A system for instantaneous detection and notification of faults in the quality of the electrical signal according to claim 1, characterized in that said classification model has learned to detect, in addition to the healthy state, various faults in the quality of the electrical signal. . An instant electrical signal quality fault detection and notification system according to claim 1 and 3, characterized in that said classification model is capable of classifying combinations of faults.
5. Un système de détection et notification instantanée des défauts de la qualité du signal électrique selon la revendication 1 caractérisé en ce que ladite unité de traitement stocke les instants de détection des défauts et leurs labels dans un fichier en mémoire interne et optionnellement dans ladite carte mémoire. 5. A system for the instantaneous detection and notification of faults in the quality of the electrical signal according to claim 1, characterized in that said processing unit stores the fault detection times and their labels in a file in internal memory and optionally in said card memory.
6. Un système de détection et notification instantanée des défauts de la qualité du signal électrique selon la revendication 1 caractérisé en ce que ledit système de transmission sans fil envoie des notifications et reçoit des requêtes.
6. An instant detection and notification system for faults in the quality of the electrical signal according to claim 1, characterized in that said wireless transmission system sends notifications and receives requests.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/MA2020/050006 WO2022119426A1 (en) | 2020-12-01 | 2020-12-01 | Intelligent system for immediate detection and notification of disturbances in electrical signal quality |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/MA2020/050006 WO2022119426A1 (en) | 2020-12-01 | 2020-12-01 | Intelligent system for immediate detection and notification of disturbances in electrical signal quality |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022119426A1 true WO2022119426A1 (en) | 2022-06-09 |
Family
ID=74191828
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/MA2020/050006 WO2022119426A1 (en) | 2020-12-01 | 2020-12-01 | Intelligent system for immediate detection and notification of disturbances in electrical signal quality |
Country Status (1)
Country | Link |
---|---|
WO (1) | WO2022119426A1 (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115796668A (en) * | 2022-11-28 | 2023-03-14 | 广东新亚光电缆股份有限公司 | Signal quality management system of control cable |
CN116184124A (en) * | 2023-04-26 | 2023-05-30 | 华东交通大学 | Power distribution network fault type identification method |
Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106443379A (en) * | 2016-09-22 | 2017-02-22 | 广州供电局有限公司 | Transformer partial discharge fault type identifying method and transformer partial discharge fault type identifying device |
CN108664936A (en) * | 2018-05-14 | 2018-10-16 | 浙江师范大学 | A kind of diagnostic method and system based on mechanical disorder |
CN110307982A (en) * | 2019-06-19 | 2019-10-08 | 电子科技大学 | Bearing fault classification method based on CNN and Adaboost |
CN110363130A (en) * | 2019-07-08 | 2019-10-22 | 国网四川省电力公司电力科学研究院 | Voltage sag source discrimination method and device for identifying based on variation mode decomposition |
CN110399796A (en) * | 2019-09-02 | 2019-11-01 | 国网上海市电力公司 | A kind of electrical energy power quality disturbance recognition methods based on improvement deep learning algorithm |
US20200209109A1 (en) * | 2018-12-28 | 2020-07-02 | Shanghai United Imaging Intelligence Co., Ltd. | Systems and methods for fault diagnosis |
CN111541511A (en) * | 2020-04-20 | 2020-08-14 | 中国人民解放军海军工程大学 | Communication interference signal identification method based on target detection in complex electromagnetic environment |
CN111800811A (en) * | 2020-09-08 | 2020-10-20 | 中国人民解放军国防科技大学 | Unsupervised detection method, unsupervised detection device, unsupervised detection equipment and storage medium for frequency spectrum abnormality |
CN111789629A (en) * | 2020-06-29 | 2020-10-20 | 中南大学 | Breath sound intelligent diagnosis and treatment system and method based on deep learning |
KR102180748B1 (en) * | 2019-08-21 | 2020-11-19 | 연세대학교 산학협력단 | Multi-core cable fault classification system and method |
KR20200130528A (en) * | 2019-05-08 | 2020-11-19 | (주)제이에스 시스템 | Automatic equipment failure prediction system in smart factory |
-
2020
- 2020-12-01 WO PCT/MA2020/050006 patent/WO2022119426A1/en active Application Filing
Patent Citations (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106443379A (en) * | 2016-09-22 | 2017-02-22 | 广州供电局有限公司 | Transformer partial discharge fault type identifying method and transformer partial discharge fault type identifying device |
CN108664936A (en) * | 2018-05-14 | 2018-10-16 | 浙江师范大学 | A kind of diagnostic method and system based on mechanical disorder |
US20200209109A1 (en) * | 2018-12-28 | 2020-07-02 | Shanghai United Imaging Intelligence Co., Ltd. | Systems and methods for fault diagnosis |
KR20200130528A (en) * | 2019-05-08 | 2020-11-19 | (주)제이에스 시스템 | Automatic equipment failure prediction system in smart factory |
CN110307982A (en) * | 2019-06-19 | 2019-10-08 | 电子科技大学 | Bearing fault classification method based on CNN and Adaboost |
CN110363130A (en) * | 2019-07-08 | 2019-10-22 | 国网四川省电力公司电力科学研究院 | Voltage sag source discrimination method and device for identifying based on variation mode decomposition |
KR102180748B1 (en) * | 2019-08-21 | 2020-11-19 | 연세대학교 산학협력단 | Multi-core cable fault classification system and method |
CN110399796A (en) * | 2019-09-02 | 2019-11-01 | 国网上海市电力公司 | A kind of electrical energy power quality disturbance recognition methods based on improvement deep learning algorithm |
CN111541511A (en) * | 2020-04-20 | 2020-08-14 | 中国人民解放军海军工程大学 | Communication interference signal identification method based on target detection in complex electromagnetic environment |
CN111789629A (en) * | 2020-06-29 | 2020-10-20 | 中南大学 | Breath sound intelligent diagnosis and treatment system and method based on deep learning |
CN111800811A (en) * | 2020-09-08 | 2020-10-20 | 中国人民解放军国防科技大学 | Unsupervised detection method, unsupervised detection device, unsupervised detection equipment and storage medium for frequency spectrum abnormality |
Non-Patent Citations (11)
Title |
---|
A. M. GAOUDAM. M. A. SALAMAM. R. SULTANA. Y. CHIKHANI: "Power quality détection and classification using wavelet-multiresolution signal decomposition", IEEE TRANS. POWER DELIV., vol. 14, no. 4, October 1999 (1999-10-01), pages 1469 - 1476, XP011089188, DOI: 10.1109/61.796242 |
F. JANDANS. KHOKHARS. ABIDA. SHAHAF. ABBASI, RECOGNITION AND CLASSIFICATION OF POWER QUALITY DISTURBANCES BY DWT-MRA AND SVM CLASSIFIER, 2019 |
F. UCARO. F. ALCINB. DANDILF. ATA: "Power quality event détection using a fast extreme learning machine", ENERGIES, vol. 11, no. 1, January 2018 (2018-01-01) |
G. T. HEYDTP. S. FJELDC. C. LIUD. PIERCEL. TUG. HENSLEY: "Applications of the windowed FFT to electric power quality assessment", IEEE TRANS. POWER DELIV., vol. 14, no. 4, October 1999 (1999-10-01), pages 1411 - 1416, XP011089181, DOI: 10.1109/61.796235 |
GUO MOU-FA ET AL: "Deep-Learning-Based Fault Classification Using Hilbert-Huang Transform and Convolutional Neural Network in Power Distribution Systems", IEEE SENSORS JOURNAL, IEEE SERVICE CENTER, NEW YORK, NY, US, vol. 19, no. 16, 15 August 2019 (2019-08-15), pages 6905 - 6913, XP011735595, ISSN: 1530-437X, [retrieved on 20190716], DOI: 10.1109/JSEN.2019.2913006 * |
K. CAIW. CAOL. AARNIOVUORIH. PANGY. LING. LI: "Classification of Power Quality Disturbances Using Wigner-Ville Distribution and Deep Convolutional Neural Networks", IEEE ACCESS, vol. 7, August 2019 (2019-08-01), pages 119099 - 119109, XP011743955, DOI: 10.1109/ACCESS.2019.2937193 |
K. DAUDA. F. ABIDINA. P. ISMAILM. D. A. HASANM. A. SHAFIEA. ISMAIL: "Evaluating windowing-based continuous S-transform with neural network classifier for detecting and classifying power quality disturbances", INDONES. J. ELECTR. ENG. COMPUT. SCI., vol. 13, no. 3, March 2019 (2019-03-01), pages 1136 - 1142 |
LEE CHUN-KWON ET AL: "Detection and Assessment of I&C Cable Faults Using Time-Frequency R-CNN-Based Reflectometry", IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, IEEE SERVICE CENTER, PISCATAWAY, NJ, USA, vol. 68, no. 2, 5 February 2020 (2020-02-05), pages 1581 - 1590, XP011817889, ISSN: 0278-0046, [retrieved on 20201103], DOI: 10.1109/TIE.2020.2970677 * |
LI CHUNLIN ET AL: "Fault Diagnosis Method Based on Encoding Time Series and Convolutional Neural Network", IEEE ACCESS, IEEE, USA, vol. 8, 1 September 2020 (2020-09-01), pages 165232 - 165246, XP011809731, DOI: 10.1109/ACCESS.2020.3021007 * |
Y. GUM. H. J. BOLLEN: "Time-frequency and time-scale domain analysis of voltage disturbances", IEEE TRANS. POWER DELIV., vol. 15, no. 4, October 2000 (2000-10-01), pages 1279 - 1284, XP011049937 |
YANPING CAI ET AL: "Application of Image Recognition Technology based on Fractal Dimension for Diesel Engine Fault Diagnosis", YOUNG COMPUTER SCIENTISTS, 2008. ICYCS 2008. THE 9TH INTERNATIONAL CONFERENCE FOR, IEEE, PISCATAWAY, NJ, USA, 18 November 2008 (2008-11-18), pages 2948 - 2953, XP031373649, ISBN: 978-0-7695-3398-8 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115796668A (en) * | 2022-11-28 | 2023-03-14 | 广东新亚光电缆股份有限公司 | Signal quality management system of control cable |
CN115796668B (en) * | 2022-11-28 | 2023-09-19 | 广东新亚光电缆股份有限公司 | Signal quality management system of control cable |
CN116184124A (en) * | 2023-04-26 | 2023-05-30 | 华东交通大学 | Power distribution network fault type identification method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Chen et al. | Detection and classification of transmission line faults based on unsupervised feature learning and convolutional sparse autoencoder | |
Chen et al. | Environmental sound classification with dilated convolutions | |
Kolagati et al. | Exposing deepfakes using a deep multilayer perceptron–convolutional neural network model | |
WO2022119426A1 (en) | Intelligent system for immediate detection and notification of disturbances in electrical signal quality | |
Ilyas et al. | AVFakeNet: A unified end-to-end Dense Swin Transformer deep learning model for audio–visual deepfakes detection | |
Fan et al. | Semi-supervised time series classification by temporal relation prediction | |
Senthilkumar et al. | Speech emotion recognition based on Bi-directional LSTM architecture and deep belief networks | |
Sunaryono et al. | Gradient boosting machines fusion for automatic epilepsy detection from EEG signals based on wavelet features | |
Uckol et al. | Workmanship defect classification in medium voltage cable terminations with convolutional neural network | |
WO2018104682A1 (en) | Detection of nerves in a series of echographic images | |
Yuan et al. | Faulty feeder detection based on image recognition of current waveform superposition in distribution networks | |
He | Facial expression recognition using multi-branch attention convolutional neural network | |
Gómez-Tapia et al. | On the minimal amount of EEG data required for learning distinctive human features for task-dependent biometric applications | |
Cheng et al. | Deeply explain CNN via hierarchical decomposition | |
Mishra et al. | EEG-based image feature extraction for visual classification using deep learning | |
Vesperini et al. | A capsule neural networks based approach for bird audio detection | |
Saealal et al. | Three-Dimensional Convolutional Approaches for the Verification of Deepfake Videos: The Effect of Image Depth Size on Authentication Performance | |
Lu et al. | A novel approach of facial expression recognition based on shearlet transform | |
Nazira et al. | Depression detection using convolutional neural networks | |
Rajalaxmi et al. | Deepfake Detection using Inception-ResNet-V2 Network | |
Biswas et al. | Classification of various colon diseases in Colonoscopy video using Cross-Wavelet features | |
Resmi et al. | Person Identification by Models Trained Using Left and Right Ear Images Independently | |
Jamunadevi et al. | Performance Analysis of Random Forest Classifier in Extracting Features from the EEG signal | |
Miakshyn et al. | Face Recognition Technology Improving Using Convolutional Neural Networks | |
Daoud et al. | Generative adversarial network based semi-supervised learning for epileptic focus localization |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 20842740 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 20842740 Country of ref document: EP Kind code of ref document: A1 |