WO2024020706A1 - Multisensory method and system for detecting surface discharges in insulators - Google Patents

Multisensory method and system for detecting surface discharges in insulators Download PDF

Info

Publication number
WO2024020706A1
WO2024020706A1 PCT/CL2023/050060 CL2023050060W WO2024020706A1 WO 2024020706 A1 WO2024020706 A1 WO 2024020706A1 CL 2023050060 W CL2023050060 W CL 2023050060W WO 2024020706 A1 WO2024020706 A1 WO 2024020706A1
Authority
WO
WIPO (PCT)
Prior art keywords
sensors
antenna
sensor
signals
danger
Prior art date
Application number
PCT/CL2023/050060
Other languages
Spanish (es)
French (fr)
Inventor
Jorge Alfredo ARDILA REY
Rodrigo Andrés ROZAS VALDERRAMA
Luis Alberto ORELLANA GONZÁLEZ
Original Assignee
Universidad Técnica Federico Santa María
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Universidad Técnica Federico Santa María filed Critical Universidad Técnica Federico Santa María
Publication of WO2024020706A1 publication Critical patent/WO2024020706A1/en

Links

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
    • G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models

Definitions

  • the present invention relates to the field of measurements and testing of electrical variables, specifically to the field of arrangements for testing electrical properties, locating electrical faults, and arrangements for electrical testing.
  • a multisensory system for early detection of surface discharges in primary equipment insulators is provided.
  • said invention aims to provide a method and a device for diagnosing an anomaly of an electrical power installation by using a complex diagnostic system, which can perform an ultrasonic diagnosis at the same time, in which it processes the data collected by multiple diagnostic facilities for multiple data to perform self-diagnosis and analyze it with the aim of dramatically improving the accuracy of judgment for the diagnosis of abnormal deterioration of power facilities.
  • Said invention is also directed to a diagnostic method, where the purpose is to find abnormal deterioration of electrical energy installations by detecting light waves, sound, heat and electromagnetic waves generated by these electrical energy installations at a given time. through the use of equipment.
  • a diagnostic judgment system of the artificial intelligent electrical system comprising: a light source for emitting light (ultraviolet rays, ultrasonic waves), a detection unit including an independent ultraviolet detection device, a device ultrasonic detection device, a thermal image detection device and a high frequency detection device that perform different functions to simultaneously detect the intensity of an electromagnetic wave (high frequency); a signal processor to convert each of the detected abnormal signals into a digital signal, filtering and amplifying the digital signal. Ultrasonic waves are analyzed by analyzing an anomaly in the signal provided by the signal processing unit.
  • document CN10591 1438 refers to a method and a risk assessment system for a local energy source of the GIS (Gas /nsu ate S /to/?gear) substation type, capable of detecting a signal based on local partial discharges, with a high sensitivity detection and high diagnostic accuracy.
  • GIS Global System for Interoperability
  • This system is used for a GIS and for a 66 kV or higher substation tank.
  • Major equipment such as circuit breaker performs partial discharge load test to assess system risk to guide maintenance decision.
  • the GIS risk assessment method based on partial discharge electrification detection comprises collecting partial discharge signals, then performs partial discharge type analysis and energy localization in partial discharge signals, and subsequently analyzes the partial discharge severity depending on the type of discharge; to finally, according to the type of local discharge, the source of local discharge and the severity of partial discharge are found, combined with GIS operational and maintenance data, the risk level of GIS is evaluated.
  • the background of the state of the art focuses on the monitoring and maintenance of insulators in high voltage power electrical systems that contain a plurality of sensors, and only in one of the cases do they use data analysis to make a diagnosis. from the application of artificial intelligence in the measurements made by said sensors. Furthermore, this background does not provide an alert system that allows knowing the current state of the device and does not describe an analysis by groups of signals obtained by the multiple sensors, where each of these analyzes separately provides as a result an indicator that can complement the independently the danger information in which the insulator is located.
  • the present invention provides a multisensory system for early detection of surface discharges in primary equipment insulators, which is characterized in that it comprises: sensors to measure electromagnetic radiation at a distance, where one of said sensors is an antenna-type electromagnetic signal sensor and the other is an acoustic sensor; Said system also includes: a filter that limits the bandwidth allowed by the sensors; a data acquisition and analysis system obtained by the sensors, through artificial intelligence, where said data analysis system gives a forecast of the probability of occurrence of a surface failure; and a continuous voltage source to power the sensors.
  • the multisensory early detection system is characterized in that the antenna-type electromagnetic signal sensor measures in the giga Hertz range and the acoustic sensor measures mechanical waves.
  • the multisensory early detection system is characterized in that the acoustic sensor is an ultrasonic acoustic type sensor that measures ultrasonic frequency signals.
  • the multisensory early detection system is characterized in that the system comprises a plurality of acoustic sensors.
  • the system of the claim is characterized in that the system comprises a plurality of ultrasonic acoustic sensors.
  • the multisensory early detection system is characterized in that said antenna-type electromagnetic signal sensor is an open slot antenna.
  • the multisensory early detection system is characterized in that a filter that limits the bandwidth allowed by the sensors includes a frequency range between 500-3000MHz.
  • the present invention also provides a method for the early detection of surface discharges in primary equipment insulators, which are characterized because it comprises: acquiring and measuring, using an antenna, signals corresponding to the electromagnetic radiation emitted by a primary equipment isolator to obtain data from these signals; acquiring and measuring, using an ultrasonic acoustic sensor, signals corresponding to the acoustic waves emitted by a primary equipment isolator to obtain data from these signals; process the data obtained by the ultrasonic acoustic sensor for the evaluation of mechanical vibrations resulting from mechanical alterations emitted by the insulator; analyze the data obtained by the antenna and the ultrasonic acoustic sensor using an artificial intelligence algorithm; provide a numerical indicator that represents the danger of the insulator; separate the danger indicator by ranges to determine a normal activity, an intermediate risk activity, and a dangerous situation in the insulator; compare the danger indicators to give a final diagnosis of the condition of the insulator; and analyze these danger indicators through hysteresis cycles that allow
  • the method for early detection of downloads is characterized in that the artificial intelligence algorithm is through neural networks.
  • the method for the early detection of discharges is characterized in that the signals measured by the antenna and the ultrasonic acoustic sensor are analyzed separately using neural networks.
  • the method for the early detection of discharges is characterized in that the processing of the data coming from the antenna and the ultrasonic sensor before it enters the neural network uses Hilbert transforms and the Fourier transform.
  • the method for early detection of discharges is characterized in that the acquisition of antenna signals is carried out in groups of 5 and the signals acquired by the acoustic sensor are carried out in groups of 10.
  • the method for early discharge detection is characterized in that the artificial intelligence algorithms are trained based on 5 series of experimental measurements.
  • the method for early detection of downloads is characterized in that the artificial intelligence algorithms deliver a numerical value between 0 and 1 as a result, which is interpreted as a dangerousness index.
  • the method for early detection of discharges is characterized in that said values that are interpreted as a danger index include a range between 0 and 0.35, the value of which is considered normal activity, a range between 0.35 and 0. .8 whose value is considered an activity that should be paid attention to, and a value greater than 0.8 is considered an alarm or danger situation.
  • the method for early detection of discharges is characterized in that the hysteresis consists of waiting 20 successive decisions of the type of normal activity to descend to this state from a state of an activity to which attention must be paid; 20 successive decisions of the type of an activity that must be paid attention to in order to descend to this state from an alarm or danger situation; and 40 successive decisions of the type of normal activity to descend to this state from an alarm or danger situation.
  • FIG. 1 is a photograph of an ultrasonic acoustic probe type sensor for the detection of partial discharges through mechanical vibrations.
  • FIG. 2 illustrates the open slot antenna in two parts, in (a) it shows a photo of the acoustic sensor, and in (b) it shows the frequency spectrum of the reflection coefficient of said sensor.
  • FIG. 3 schematically illustrates using blocks the assembly used to be able to perform measurements with the sensors of the present invention.
  • FIG. 4 schematically illustrates the assembly circuit used to perform measurements with the sensors of the present invention. DETAILED DESCRIPTION OF THE INVENTION
  • the present invention details a system and a method that offers a technological tool that comprises a system that performs measurements of electrical variables using multiple sensors, focusing its use on measurements on insulators used in primary high voltage equipment for the prevention of future failures or discharges. superficial in this type of systems.
  • the multisensor system for early detection of surface discharges in primary equipment insulators of the invention is characterized in that it comprises: a system using sensors that uses the approach based on remotely measuring electromagnetic radiation; Said system contains an antenna-type electromagnetic signal sensor that measures frequencies in the giga Hertz range and an acoustic sensor; a filter that limits the bandwidth allowed by the sensors; a system for acquiring and analyzing data obtained by sensors, through artificial intelligence; where said data analysis system gives a forecast of the probability of occurrence of surface failure; and a continuous voltage source to power the sensors.
  • the multisensor early detection system of the invention is characterized in that the antenna it uses is an open slot so that it is highly directional.
  • the sensors of the system correspond to a directional antenna for the transmission and reception of broadband directional radio signals such as, for example, Deepace KC R102 and an ultrasonic acoustic probe type sensor that can measure acoustic or mechanical waves that come from an activity of download such as AA Ultrasonic PD Sensor.
  • the antenna corresponds to an open slot type antenna, for example a Deepace KC R102 antenna, which makes it highly directional, this is more efficient in detecting electromagnetic radiation depending on the direction it is pointed.
  • Figure 2 shows this antenna and its frequency response to the reflection coefficient, called S1 1, which accounts for which frequencies the antenna measures more efficiently, that is, with greater amplitude.
  • S1 1 the reflection coefficient
  • This frequency range for which the antenna is adapted is between 1500 and 2500 MHz, which corresponds to the ultra high frequency (UHF) radio frequency range.
  • a bandwidth capable of observing the associated electrical transients is required, which is why the measurement with this sensor is through an oscilloscope that has a wide of at least 3 GHz band and 5 GHz sampling frequency using one of the 50Q BNC type coaxial cables. Therefore, before connecting the antenna to the oscilloscope, a high-pass filter (HPF) is attached in the range 500-3000 MHz, to reduce the influence of low-frequency phenomena on the measurement.
  • HPF high-pass filter
  • the acoustic sensor is designed by the manufacturer to detect the phenomenon of partial discharges through the mechanical vibrations they produce. This sensor works in the ultrasound range, specifically at 40 kHz. Requires a 12 V direct voltage source before being connected to the data acquisition system. As the phenomenon of mechanical vibrations is detected on the scale of milliseconds, it is not possible to adequately measure it simultaneously with the antenna, which in turn detects its signals on the scale of tens and hundreds of nanoseconds. It is for this reason that the acoustic sensor signals are acquired by an acquisition card such as an oscilloscope such as, for example, the NI USB-5133 card with 50 MHz bandwidth and sampling. maximum of 100 MHz, through one of the coaxial cables, which has lower bandwidth and sampling requirements.
  • an acquisition card such as an oscilloscope such as, for example, the NI USB-5133 card with 50 MHz bandwidth and sampling. maximum of 100 MHz, through one of the coaxial cables, which has lower bandwidth and sampling requirements.
  • FIG 3 shows schematically using blocks the assembly or system used to perform measurements with the sensors of the present invention, in which:
  • the circumference at the top of the diagram represents the insulator to be monitored, which is located at a distance of 50cm from each sensor.
  • the acquisition of signals from both sensors, the antenna and the acoustic sensor is directed through a computer with an operating system that supports Python and Labview, using the latter as a graphical interface.
  • the graphical interface in turn calls for the programming of the artificial intelligence model developed in Python, which analyzes the partial discharge signals, classifies them according to risk to produce a contour, and makes a decision to issue alarms.
  • flashover will be understood when the current discharge occurs along the surface or a surface failure of a solid insulation placed within a gaseous insulation or dielectric liquids
  • the present invention also provides a method for the early detection of surface discharges in primary equipment insulators, which is characterized in that it comprises: measuring by means of an antenna the electromagnetic radiation emitted by a primary equipment insulator; measuring data from acoustic waves emitted by a primary equipment isolator using a plurality of ultrasonic sensors; process the data obtained by the acoustic sensor to obtain the global behavior of mechanical vibrations; analyze the data obtained by said sensors using a neural network algorithm; and deliver a numerical indicator that represents the danger of the insulator; separates the danger indicator by ranges to indicate, such as: normal activity, intermediate risk activity, and a dangerous situation; compare the indicators to give a final diagnosis of the condition of the insulator; and analyze the danger indicators using hysteresis cycles.
  • the method for early detection of surface discharges in primary equipment insulators is characterized in that the signals measured by said sensors are analyzed separately using neural networks.
  • the operation of the partial discharge detection method to issue the bypass alert is carried out in three stages:
  • Stage 1 It consists of measuring signals through the antenna and the acoustic sensor.
  • the neural network model that processes these signals requires that the antenna signals be acquired in groups of 5 and the acoustic sensor signals in groups of 10.
  • the acoustic sensor signals are preprocessed before entering the neural network, using the calculation of the envelope of each signal using the Hilbert transform and the subsequent calculation of the Fourier transform of this envelope.
  • the information from the acoustic sensor that is delivered to the neural network corresponds to the global behavior of the mechanical vibrations from partial discharges.
  • This approach for the acoustic sensor is based on previous studies (S. Polisetty, A. El-Hag, and S. Jayram, “Classification of common discharges in outdoor insulation using acoustic signals and artificial neural network,” High Voltage, vol. 4 , no. 4).
  • Stage 2 In this stage, the measurements of both sensors, the antenna and the acoustic sensor, are analyzed separately using two artificial intelligence models, with the same architecture, but optimized to analyze their corresponding signals. These models were trained based on 5 experimental measurement signals that allowed the partial discharge activity to be measured from a low-risk condition to a high-risk condition, reaching contour. Artificial intelligence models deliver a real number as a result that can be between 0 and 1, which can be interpreted as a danger index.
  • Stage 3 In this final stage the results of both hazard indices are used to determine the general, and therefore definitive, decision of the system. Each hazard index, from the antenna and the acoustic sensor, is compared to decision thresholds specified by the user depending on how rigid the alarm system is desired. For the purposes of showing results, the interval between 0 and 0.35 has been chosen as the one where the danger indices are considered normal activity, that is, without danger. Between 0.35 and 0.8 is considered an activity that should be paid attention to, indicating an intermediate risk. If it is greater than 0.8, it is an alarm or danger situation.
  • the hysteresis used to obtain results includes waiting 20 successive decisions of the normal type to descend to this state from one of attention; 20 successive decisions of the attention type to go down to this state from one of alarm; and 40 successive decisions of the normal type to go down to this state from the alarm state.
  • contamination will be understood as any object other than the electronics of the device that rests on the surface of the insulator, which may mechanically or electronically interfere with its normal operation, and affect the danger index detected in it.
  • the final decision to timely intervene in the isolator is a combination of both states determined by the antenna and the acoustic sensor. If the determination of both sensors is normal or at least one of them is in attention, the overall decision of the system is a “normal” state, which is indicated on the interface with a green color. If both sensor determinations reveal that they are in a state of attention or at least one of them is in an alarm state, that is, the acoustic sensor in alarm and the antenna in attention, then the general decision is an “attention” state. which is indicated with yellow color. In any other case, the general decision is considered to be the “alarm” state, which is indicated with a red color.
  • FIG. 4 The circuit used for the experimental tests is shown in Figure 4, where a high voltage source appears that was achieved through a 400V/150kV high voltage test transformer connected to the 220 V 50 Hz industrial network controlled by a regulator transformer. Also, there is a capacitive voltage divider in parallel to the test object that was used to measure the voltage applied to the object.
  • the object of the test was a transformer bushing type insulator, this is a primary equipment insulator and the nominal voltage on the insulator is 34.5 kV.
  • salt contamination NaCl was dissolved in water at a concentration of 167 g/L, and said solution was spread over the insulator, and subsequently dried with an industrial dryer to produce an initial deposit of salt on its surface. .
  • the nominal voltage of the insulator was maintained at 34.5 kV throughout the experiment to simulate its operation in a contaminated condition.
  • the saline solution was additionally spread periodically, in series every 3 minutes and in others every 5 minutes with the objective of simulating the humidity phenomenon, which is highly random in the field. Spraying this contamination while the insulator was energized was done using a hand fumigation pump. To contain salt contamination, the test object was placed inside an acrylic box.
  • a Keysight Infiniium DSOS804a oscilloscope was used to measure the antenna signals using a time window of 1 ps and a sampling rate of 5 GHz.
  • the acoustic sensor signals were acquired by the NI USB-5133 card, using a time window of 100 ms and sampling frequency of 200 MHz.
  • the Keysight oscilloscope used corresponds to the integration of an oscilloscope and a computer with an operating system that supports labview 2018 and Python 3.6, so in these experimental series these two components, schematized in Figure 3, were integrated into only one.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Molecular Biology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Testing Relating To Insulation (AREA)

Abstract

The present invention relates to the field of electronics, specifically to measurements of electrical variables. In particular, the invention provides a multisensory system for detecting surface discharges in insulators of primary equipment, which is characterised in that it comprises: a sensor-based system that uses remote measurement of electromagnetic radiation, wherein the system contains an antenna-type sensor of electromagnetic signals that measures frequencies in the range of gigahertz, and an acoustic sensor; a filter that limits the bandwidth permitted by the sensors; a system for acquiring and analysing data obtained by the sensors, by means of artificial intelligence; and a constant voltage source for powering the sensors, wherein the data analysis system generates a final diagnosis of the status of the insulator.

Description

SISTEMA Y MÉTODO MULTISENSORIAL DE DETECCIÓN DE DESCARGAS SUPERFICIALES EN AISLADORES MULTI-SENSORY SYSTEM AND METHOD FOR DETECTION OF SURFACE DISCHARGES IN INSULATORS
CAMPO TÉCNICO DE LA INVENCIÓN TECHNICAL FIELD OF THE INVENTION
La presente invención se relaciona con el campo de las mediciones y pruebas de variables eléctricas, específicamente con el campo de los arreglos para probar propiedades eléctricas, localizar fallas eléctricas y disposiciones para pruebas eléctricas. En particular, se proporciona un sistema multisensorial de detección temprana de descargas superficiales en aisladores de equipos primarios. The present invention relates to the field of measurements and testing of electrical variables, specifically to the field of arrangements for testing electrical properties, locating electrical faults, and arrangements for electrical testing. In particular, a multisensory system for early detection of surface discharges in primary equipment insulators is provided.
ANTECEDENTES DE LA INVENCIÓN BACKGROUND OF THE INVENTION
Dentro del campo de la electrónica, una parte importante es la medición de variables eléctricas y magnéticas, y la realización de pruebas sobre equipos electrónicos. Estas mediciones o pruebas sirven para prevenir futuros desperfectos en los equipos, y por otro lado, también sirven para realizar futuras mejoras en dichos equipos electrónicos. Los métodos y sistemas que se desarrollan con estos fines permiten prevenir eventuales fallas en estos equipos, de manera que estas no escalen a problemas mayores, y así ayudar a abaratar costos al evitar reemplazos en los mismos que se pudieran anticipar. Within the field of electronics, an important part is the measurement of electrical and magnetic variables, and carrying out tests on electronic equipment. These measurements or tests serve to prevent future damage to the equipment, and on the other hand, they also serve to make future improvements to said electronic equipment. The methods and systems that are developed for these purposes make it possible to prevent eventual failures in this equipment, so that they do not escalate to major problems, and thus help reduce costs by avoiding replacements that could be anticipated.
Actualmente, existen diversos sistemas y métodos de mediciones de variables eléctricas para la prevención de fallas en dispositivos electrónicos, de manera particular las mediciones de descargas eléctricas superficiales sobre aisladores para los equipos de alta tensión, ya que es conocido que dichos aisladores se caracterizan por estar en ambientes al aire libre, por lo que estos aisladores están expuestos a la contaminación. Por consiguiente, la evaluación de la contaminación en los aisladores es crítica para prevenir un contorneo repentino que podría impedir la operación permanente de la red de transmisión y distribución de energía. Dichos sistemas de mediciones sobre estos equipos electrónicos tienen diferentes configuraciones mecánicas de diferentes geometrías, y eléctricas con diversos sensores, para la medición de estas variables eléctricas sobre los aisladores. Un ejemplo de este tipo de sistemas que realizan mediciones electrónicas sobre aisladores en sistemas de alta tensión se describe en el documento KR101235777, el cual se refiere a un sistema de procesamiento de datos que detecta simultáneamente la luz del efecto corona, el calor y las ondas electromagnéticas generadas por una instalación eléctrica. Además, dicha invención tiene el objeto de proporcionar un método y un dispositivo para diagnosticar una anomalía de una instalación de energía eléctrica mediante el uso de un sistema de diagnóstico complejo, que puede realizar un diagnóstico ultrasónico a la vez, en el que procesa los datos recopilados por múltiples instalaciones de diagnóstico para múltiples datos para realizar un autodiagnóstico y analizarlo con el objetivo de mejorar drásticamente la precisión del juicio para el diagnóstico de deterioro anormal de las instalaciones de energía. Dicha invención está dirigida también a un método de diagnóstico, donde cuyo propósito es encontrar el deterioro anormal de las instalaciones de energía eléctrica mediante la detección de ondas de luz, sonido, calor y ondas electromagnéticas generadas por estas instalaciones de energía eléctrica en un momento dado mediante el uso del equipo. Currently, there are various systems and methods for measuring electrical variables to prevent failures in electronic devices, particularly measurements of surface electric discharges on insulators for high voltage equipment, since it is known that said insulators are characterized by being in outdoor environments, so these insulators are exposed to contamination. Therefore, assessment of contamination on insulators is critical to prevent sudden bypass that could prevent permanent operation of the power transmission and distribution network. Said measurement systems on these electronic equipment have different mechanical configurations of different geometries, and electrical configurations with various sensors, for the measurement of these electrical variables on the insulators. An example of this type of system that performs electronic measurements on insulators in high voltage systems is described in document KR101235777, which refers to a data processing system that simultaneously detects corona light, heat and waves. electromagnetic waves generated by an electrical installation. Furthermore, said invention aims to provide a method and a device for diagnosing an anomaly of an electrical power installation by using a complex diagnostic system, which can perform an ultrasonic diagnosis at the same time, in which it processes the data collected by multiple diagnostic facilities for multiple data to perform self-diagnosis and analyze it with the aim of dramatically improving the accuracy of judgment for the diagnosis of abnormal deterioration of power facilities. Said invention is also directed to a diagnostic method, where the purpose is to find abnormal deterioration of electrical energy installations by detecting light waves, sound, heat and electromagnetic waves generated by these electrical energy installations at a given time. through the use of equipment.
En esa invención, de acuerdo con un sistema de juicio de diagnóstico del sistema eléctrico inteligente artificial comprende: una fuente de luz para emitir luz (rayos ultravioleta, ondas ultrasónicas), una unidad de detección que incluye un dispositivo de detección ultravioleta independiente, un dispositivo de detección ultrasónico, un dispositivo de detección de imagen térmica y un dispositivo de detección de alta frecuencia que realizan diferentes funciones para detectar simultáneamente la intensidad de una onda electromagnética (alta frecuencia); un procesador de señal para convertir cada una de las señales anormales detectadas en una señal digital, filtrando y amplificando la señal digital. Las ondas ultrasónicas se analizan analizando una anomalía en la señal proporcionada por la unidad de procesamiento de señal. In that invention, according to a diagnostic judgment system of the artificial intelligent electrical system comprising: a light source for emitting light (ultraviolet rays, ultrasonic waves), a detection unit including an independent ultraviolet detection device, a device ultrasonic detection device, a thermal image detection device and a high frequency detection device that perform different functions to simultaneously detect the intensity of an electromagnetic wave (high frequency); a signal processor to convert each of the detected abnormal signals into a digital signal, filtering and amplifying the digital signal. Ultrasonic waves are analyzed by analyzing an anomaly in the signal provided by the signal processing unit.
Por otro lado, el documento CN10591 1438 se refiere a un método y un sistema de evaluación de riesgos de una fuente de energía local del tipo subestación con aislamiento de gas GIS (Gas /nsu ate S /to/?gear), capaz de detectar una señal basada en descargas parciales locales, con una alta sensibilidad de detección y una alta precisión de diagnóstico. Este sistema se utiliza para un GIS y para un tanque de una subestación de 66 kV o superior. El equipo principal, como el interruptor automático, realiza una prueba de carga de descarga parcial para evaluar el riesgo del sistema para guiar la decisión de mantenimiento. On the other hand, document CN10591 1438 refers to a method and a risk assessment system for a local energy source of the GIS (Gas /nsu ate S /to/?gear) substation type, capable of detecting a signal based on local partial discharges, with a high sensitivity detection and high diagnostic accuracy. This system is used for a GIS and for a 66 kV or higher substation tank. Major equipment such as circuit breaker performs partial discharge load test to assess system risk to guide maintenance decision.
El método de evaluación de riesgos GIS basado en la detección de electrificación de descargas parciales, comprende la recolección señales de descargas parciales, luego se realiza un análisis de tipo de descarga parcial y la localización de energía en señales de descarga parcial, y posteriormente analiza la severidad de la descarga parcial según el tipo de descarga; para finalmente, de acuerdo con el tipo de descarga local, se encuentra la fuente de la descarga local y la gravedad de la descarga parcial, combinados con los datos operativos y de mantenimiento de GIS, se evalúa el nivel de riesgo de GIS. The GIS risk assessment method based on partial discharge electrification detection, comprises collecting partial discharge signals, then performs partial discharge type analysis and energy localization in partial discharge signals, and subsequently analyzes the partial discharge severity depending on the type of discharge; to finally, according to the type of local discharge, the source of local discharge and the severity of partial discharge are found, combined with GIS operational and maintenance data, the risk level of GIS is evaluated.
Sin embargo, los antecedentes del estado de la técnica se enfocan en el monitoreo y mantención de aisladores en sistemas eléctricos de potencia de alto voltaje que contienen una pluralidad de sensores, y solo en uno de los casos utilizan el análisis de datos para realizar un diagnóstico a partir de la aplicación de inteligencia artificial en las mediciones realizadas por dichos sensores. Además, estos antecedentes no proveen un sistema de alerta que permita saber el estado actual del dispositivo y tampoco describen un análisis por grupos de señales obtenidas por los múltiples sensores, donde cada uno de estos análisis por separado proporcione como resultado un indicador que pueda complementar de forma independiente la información de peligrosidad en la que se encuentra el aislador. However, the background of the state of the art focuses on the monitoring and maintenance of insulators in high voltage power electrical systems that contain a plurality of sensors, and only in one of the cases do they use data analysis to make a diagnosis. from the application of artificial intelligence in the measurements made by said sensors. Furthermore, this background does not provide an alert system that allows knowing the current state of the device and does not describe an analysis by groups of signals obtained by the multiple sensors, where each of these analyzes separately provides as a result an indicator that can complement the independently the danger information in which the insulator is located.
En consecuencia, existe la necesidad de crear un sistema que ayude a medir este tipo de aisladores en sistemas de alto voltaje y analice los resultados de estas mediciones mediante sistemas de inteligencia artificial, de manera que indique en qué estado de desperfecto se encuentra dicho dispositivo, considerando la información por separado de cada uno de los sensores utilizados en el mismo. SUMARIO DE LA INVENCIÓN Consequently, there is a need to create a system that helps measure this type of insulators in high voltage systems and analyzes the results of these measurements using artificial intelligence systems, so as to indicate in what state of damage said device is. considering the information separately from each of the sensors used in it. SUMMARY OF THE INVENTION
La presente invención proporciona un sistema multisensorial de detección temprana de descargas superficiales en aisladores de equipos primarios, que se caracteriza porque comprende: sensores para medir radiación electromagnética a distancia, donde uno de dichos sensores es un sensor de señales electromagnéticas tipo antena y el otro es un sensor acústico; dicho sistema incluye, además: un filtro que limita el ancho de banda permitido por los sensores; un sistema de adquisición y análisis de datos obtenidos por los sensores, por medio de inteligencia artificial, donde dicho sistema de análisis de datos da un pronóstico de probabilidad de ocurrencia de una falla superficial; y una fuente de voltaje continuo para la alimentación de los sensores. The present invention provides a multisensory system for early detection of surface discharges in primary equipment insulators, which is characterized in that it comprises: sensors to measure electromagnetic radiation at a distance, where one of said sensors is an antenna-type electromagnetic signal sensor and the other is an acoustic sensor; Said system also includes: a filter that limits the bandwidth allowed by the sensors; a data acquisition and analysis system obtained by the sensors, through artificial intelligence, where said data analysis system gives a forecast of the probability of occurrence of a surface failure; and a continuous voltage source to power the sensors.
En una realización preferida, el sistema multisensorial de detección temprana se caracteriza porque el sensor de señales electromagnéticas tipo antena mide en el rango de los giga Hertz y el sensor acústico mide ondas mecánicas. In a preferred embodiment, the multisensory early detection system is characterized in that the antenna-type electromagnetic signal sensor measures in the giga Hertz range and the acoustic sensor measures mechanical waves.
En otra realización preferida, el sistema multisensorial de detección temprana se caracteriza porque el sensor acústico es un sensor del tipo acústico ultrasónico que mide señales de frecuencias ultrasónicas. In another preferred embodiment, the multisensory early detection system is characterized in that the acoustic sensor is an ultrasonic acoustic type sensor that measures ultrasonic frequency signals.
En una realización preferida adicional, el sistema multisensorial de detección temprana se caracteriza porque el sistema comprende una pluralidad de sensores acústicos. In a further preferred embodiment, the multisensory early detection system is characterized in that the system comprises a plurality of acoustic sensors.
En una realización más preferida, el sistema de la reivindicación se caracteriza porque el sistema comprende una pluralidad de sensores acústicos ultrasónicos. In a more preferred embodiment, the system of the claim is characterized in that the system comprises a plurality of ultrasonic acoustic sensors.
En otra realización preferida, el sistema multisensorial de detección temprana se caracteriza porque dicho sensor de señales electromagnéticas tipo antena es de una antena de ranura abierta. In another preferred embodiment, the multisensory early detection system is characterized in that said antenna-type electromagnetic signal sensor is an open slot antenna.
En una realización preferida, el sistema multisensorial de detección temprana se caracteriza porque filtro que limita el ancho de banda permitido por los sensores, incluye un rango de frecuencias entre 500-3000MHz. In a preferred embodiment, the multisensory early detection system is characterized in that a filter that limits the bandwidth allowed by the sensors includes a frequency range between 500-3000MHz.
La presente invención proporciona también un método para la detección temprana de descargas superficiales en aisladores de equipos primarios, , que se caracteriza porque comprende: adquirir y medir, mediante una antena, señales correspondientes a la radiación electromagnética que emite un aislador de equipo primario para obtener datos a partir de estas señales; adquirir y medir, mediante un sensor acústico ultrasónico, señales correspondientes a las ondas acústicas que emite un aislador de equipo primario para obtener datos a partir de estas señales; procesar los datos obtenidos por el sensor acústico ultrasónico para la evaluación de las vibraciones mecánicas resultantes de alteraciones mecánicas que emite el aislador; analizar los datos obtenidos por la antena y el sensor acústico ultrasónico mediante un algoritmo de inteligencia artificial; entregar un indicador numérico que representa la peligrosidad del aislador; separar por rangos el indicador de peligrosidad para determinar una actividad normal, una actividad de riesgo intermedio, y una situación de peligro en el aislador; comparar los indicadores de peligrosidad para dar un diagnóstico final del estado del aislador; y analizar mediante ciclos de histéresis dichos indicadores de peligrosidad que permiten la detección temprana de descargas superficiales en los aisladores. The present invention also provides a method for the early detection of surface discharges in primary equipment insulators, which are characterized because it comprises: acquiring and measuring, using an antenna, signals corresponding to the electromagnetic radiation emitted by a primary equipment isolator to obtain data from these signals; acquiring and measuring, using an ultrasonic acoustic sensor, signals corresponding to the acoustic waves emitted by a primary equipment isolator to obtain data from these signals; process the data obtained by the ultrasonic acoustic sensor for the evaluation of mechanical vibrations resulting from mechanical alterations emitted by the insulator; analyze the data obtained by the antenna and the ultrasonic acoustic sensor using an artificial intelligence algorithm; provide a numerical indicator that represents the danger of the insulator; separate the danger indicator by ranges to determine a normal activity, an intermediate risk activity, and a dangerous situation in the insulator; compare the danger indicators to give a final diagnosis of the condition of the insulator; and analyze these danger indicators through hysteresis cycles that allow early detection of surface discharges in the insulators.
En una realización preferida, el método para la detección temprana de descargas se caracteriza porque el algoritmo de inteligencia artificial es mediante redes neuronales. In a preferred embodiment, the method for early detection of downloads is characterized in that the artificial intelligence algorithm is through neural networks.
En otra realización preferida, el método para la detección temprana de descargas se caracteriza porque mediante las redes neuronales se analizan por separado las señales medidas por la antena y el sensor acústico ultrasónico. In another preferred embodiment, the method for the early detection of discharges is characterized in that the signals measured by the antenna and the ultrasonic acoustic sensor are analyzed separately using neural networks.
En otra realización preferida adicional, el método para la detección temprana de descargas se caracteriza porque el procesamiento de los datos provenientes de la antena y del sensor ultrasónico antes que ingrese a la red neuronal, utiliza las transformadas de Hilbert y la transformada de Fourier. In another additional preferred embodiment, the method for the early detection of discharges is characterized in that the processing of the data coming from the antenna and the ultrasonic sensor before it enters the neural network uses Hilbert transforms and the Fourier transform.
En una realización preferida, el método para la detección temprana de descargas se caracteriza porque la adquisición de señales de la antena se realiza en grupos de 5 y las señales adquiridas por el sensor acústico se realiza en grupos de a 10. In a preferred embodiment, the method for early detection of discharges is characterized in that the acquisition of antenna signals is carried out in groups of 5 and the signals acquired by the acoustic sensor are carried out in groups of 10.
En una realización más preferida, el método para la detección temprana de descargas se caracteriza porque los algoritmos de inteligencia artificial son entrenados en base a 5 series de mediciones experimentales. En otra realización preferida, el método para la detección temprana de descargas se caracteriza porque los algoritmos de inteligencia artificial entregan como resultado un valor numérico entre 0 y 1 , que se interpreta como un índice de peligrosidad. In a more preferred embodiment, the method for early discharge detection is characterized in that the artificial intelligence algorithms are trained based on 5 series of experimental measurements. In another preferred embodiment, the method for early detection of downloads is characterized in that the artificial intelligence algorithms deliver a numerical value between 0 and 1 as a result, which is interpreted as a dangerousness index.
En otra realización preferida, el método para la detección temprana de descargas se caracteriza porque dichos valores que se interpretan como índice de peligrosidad incluyen un rango entre 0 y 0,35 cuyo valor se considera como actividad normal, un rango entre 0,35 y 0,8 cuyo valor se considera como una actividad a la que se debe prestar atención, y un valor superior a 0,8 que se considera como una situación de alarma o peligro. In another preferred embodiment, the method for early detection of discharges is characterized in that said values that are interpreted as a danger index include a range between 0 and 0.35, the value of which is considered normal activity, a range between 0.35 and 0. .8 whose value is considered an activity that should be paid attention to, and a value greater than 0.8 is considered an alarm or danger situation.
En otra realización preferida, el método para la detección temprana de descargas se caracteriza porque la histéresis consiste en esperar 20 decisiones sucesivas del tipo de actividad normal para bajar a este estado desde un estado de una actividad a la que se le debe prestar atención; 20 decisiones sucesivas del tipo de una actividad a la que se debe prestar atención para bajar a este estado desde una situación de alarma o peligro; y 40 decisiones sucesivas del tipo de actividad normal para bajar a este estado desde una situación de alarma o peligro. In another preferred embodiment, the method for early detection of discharges is characterized in that the hysteresis consists of waiting 20 successive decisions of the type of normal activity to descend to this state from a state of an activity to which attention must be paid; 20 successive decisions of the type of an activity that must be paid attention to in order to descend to this state from an alarm or danger situation; and 40 successive decisions of the type of normal activity to descend to this state from an alarm or danger situation.
BREVE DESCRIPCIÓN DE LAS FIGURAS BRIEF DESCRIPTION OF THE FIGURES
La FIG. 1 es una fotografía de un sensor tipo sonda acústica ultrasónica para la detección de descargas parciales mediante vibraciones mecánicas. FIG. 1 is a photograph of an ultrasonic acoustic probe type sensor for the detection of partial discharges through mechanical vibrations.
La FIG. 2 ¡lustra en dos partes la antena de ranura abierta, en (a) muestra foto del sensor acústico, y en (b) muestra el espectro de frecuencia del coeficiente de reflexión de dicho sensor. FIG. 2 illustrates the open slot antenna in two parts, in (a) it shows a photo of the acoustic sensor, and in (b) it shows the frequency spectrum of the reflection coefficient of said sensor.
La FIG. 3 ¡lustra esquemáticamente mediante bloques el montaje utilizado para poder realizar mediciones con los sensores de la presente invención. FIG. 3 schematically illustrates using blocks the assembly used to be able to perform measurements with the sensors of the present invention.
La FIG. 4 ¡lustra esquemáticamente el circuito del montaje utilizado para poder realizar mediciones con los sensores de la presente invención. DESCRIPCIÓN DETALLADA DE LA INVENCIÓN FIG. 4 schematically illustrates the assembly circuit used to perform measurements with the sensors of the present invention. DETAILED DESCRIPTION OF THE INVENTION
La presente invención detalla un sistema y un método que ofrece una herramienta tecnológica que comprende un sistema que realiza mediciones de variables eléctricas mediante múltiples sensores, enfocado su uso en mediciones sobre aisladores utilizados en equipos primarios de alta tensión para la prevención de futuras fallas o descargas superficiales en este tipo de sistemas. The present invention details a system and a method that offers a technological tool that comprises a system that performs measurements of electrical variables using multiple sensors, focusing its use on measurements on insulators used in primary high voltage equipment for the prevention of future failures or discharges. superficial in this type of systems.
Como resultado de estudios anteriores, hay evidencia que el comportamiento de una onda de corriente de fuga es no lineal (Y. Liu , M. Farzaneh, “Nonlinear characteristics of leakage current for flashover monitoring of ice-covered suspension insulators”, IEEE Transactions on Dielectrics and Electrical Insulation, vol. 23, no. 3) a medida que el voltaje aumenta constantemente en un aislador cubierto de hielo, reportando que, de hecho, la forma de la onda varía en diferentes etapas previas al contorneo. El presente sistema y método de la invención ofrece la posibilidad de medir continuamente y predecir un quiebre del funcionamiento típico, cuando hay presencia de contaminación o algún agente que altere el funcionamiento normal de dichos equipos primarios. As a result of previous studies, there is evidence that the behavior of a leakage current wave is nonlinear (Y. Liu, M. Farzaneh, “Nonlinear characteristics of leakage current for flashover monitoring of ice-covered suspension insulators”, IEEE Transactions on Dielectrics and Electrical Insulation, vol. 23, no. 3) as the voltage increases steadily in an ice-covered insulator, reporting that, in fact, the shape of the wave varies at different stages prior to flashover. The present system and method of the invention offers the possibility of continuously measuring and predicting a break in typical operation, when there is the presence of contamination or some agent that alters the normal operation of said primary equipment.
De manera particular, el sistema multisensohal de detección temprana de descargas superficiales en aisladores de equipos primarios de la invención se caracteriza porque comprende: un sistema mediante sensores que utiliza el enfoque basado en medir a distancia la radiación electromagnética; dicho sistema contiene un sensor de señales electromagnéticas tipo antena que mide frecuencias en el rango de los giga Hertz y un sensor acústico; un filtro que limita el ancho de banda permitido por los sensores; un sistema de adquisición y análisis de datos obtenidos por los sensores, por medio de inteligencia artificial; donde dicho sistema de análisis de datos da un pronóstico de probabilidad de ocurrencia de falla superficial; y una fuente de voltaje continuo para la alimentación de los sensores. In particular, the multisensor system for early detection of surface discharges in primary equipment insulators of the invention is characterized in that it comprises: a system using sensors that uses the approach based on remotely measuring electromagnetic radiation; Said system contains an antenna-type electromagnetic signal sensor that measures frequencies in the giga Hertz range and an acoustic sensor; a filter that limits the bandwidth allowed by the sensors; a system for acquiring and analyzing data obtained by sensors, through artificial intelligence; where said data analysis system gives a forecast of the probability of occurrence of surface failure; and a continuous voltage source to power the sensors.
En el contexto de la presente invención, y a modo de aclaración general a lo largo de toda la descripción, se entenderá como equipos primarios a todos aquellos equipos que están conectados directamente a un circuito de alta tensión. In the context of the present invention, and as a general clarification throughout the entire description, primary equipment will be understood as all those equipment that is directly connected to a high voltage circuit.
En una realización preferida, el sistema multisensohal de detección temprana de la invención se caracteriza porque la antena que utiliza es una ranura abierta para que sea altamente direccional. Los sensores del sistema corresponden a una antena direccional para la transmisión y recepción de señales de radio direccionales de banda ancha como, por ejemplo, Deepace KC R102 y un sensor tipo sonda acústica ultrasónica que pueda medir ondas acústicas o mecánicas que provenga de una actividad de descarga como, por ejemplo, AA Ultrasonic PD Sensor. In a preferred embodiment, the multisensor early detection system of the invention is characterized in that the antenna it uses is an open slot so that it is highly directional. The sensors of the system correspond to a directional antenna for the transmission and reception of broadband directional radio signals such as, for example, Deepace KC R102 and an ultrasonic acoustic probe type sensor that can measure acoustic or mechanical waves that come from an activity of download such as AA Ultrasonic PD Sensor.
La antena corresponde a una antena de tipo ranura abierta, por ejemplo, una antena Deepace KC R102, lo que la hace altamente direccional, esto es más eficiente en la detección de radiación electromagnética según la dirección hacia donde se apunte. La Figura 2 muestra esta antena y su respuesta en frecuencia al coeficiente de reflexión, denominado S1 1 , el cual da cuenta a cuáles frecuencias la antena mide más eficientemente, es decir con mayor amplitud. Este rango de frecuencia para cual la antena está adaptada, está entre los 1500 hasta los 2500 MHz, lo que corresponde al rango de radio frecuencia ultra high frequency (UHF). Como la emisión de radiación electromagnética desde descargas parciales está en el rango sobre 100 MHz, se requiere un ancho de banda capaz de observar los transitorios eléctricos asociados, es por esta razón que la medición con este sensor es a través de un osciloscopio que disponga ancho de banda de al menos 3 GHz y frecuencia de muestreo 5 GHz mediante uno de los cables coaxiales tipo BNC de 50Q. Por ello, antes de conectar la antena al osciloscopio, se acopla un filtro paso alto (HPF) en el rango 500-3000 MHz, para disminuir la influencia de fenómenos de baja frecuencia en la medición. The antenna corresponds to an open slot type antenna, for example a Deepace KC R102 antenna, which makes it highly directional, this is more efficient in detecting electromagnetic radiation depending on the direction it is pointed. Figure 2 shows this antenna and its frequency response to the reflection coefficient, called S1 1, which accounts for which frequencies the antenna measures more efficiently, that is, with greater amplitude. This frequency range for which the antenna is adapted is between 1500 and 2500 MHz, which corresponds to the ultra high frequency (UHF) radio frequency range. As the emission of electromagnetic radiation from partial discharges is in the range over 100 MHz, a bandwidth capable of observing the associated electrical transients is required, which is why the measurement with this sensor is through an oscilloscope that has a wide of at least 3 GHz band and 5 GHz sampling frequency using one of the 50Q BNC type coaxial cables. Therefore, before connecting the antenna to the oscilloscope, a high-pass filter (HPF) is attached in the range 500-3000 MHz, to reduce the influence of low-frequency phenomena on the measurement.
El sensor acústico está diseñado por el fabricante para detectar el fenómeno de descargas parciales mediante las vibraciones mecánicas que estas producen. Este sensor trabaja en el rango de ultrasonido, específicamente a 40 kHz. Requiere de una fuente de tensión continua de 12 V antes de ser conectado al sistema de adquisición de datos. Como el fenómeno de vibraciones mecánicas es detectado en la escala de los milisegundos, no es posible de medir adecuadamente de forma simultánea junto con la antena, que a su vez sus señales se detectan en la escala de las decenas y cientos de nanosegundos. Es por esta razón que las señales del sensor acústico son adquiridas por una tarjeta de adquisición como un osciloscopio como, por ejemplo, la tarjeta NI USB-5133 con ancho de banda 50 MHz y muestreo máximo de 100 MHz, a través de uno de los cables coaxiales, que tiene menores requerimientos de ancho de banda y muestreo. The acoustic sensor is designed by the manufacturer to detect the phenomenon of partial discharges through the mechanical vibrations they produce. This sensor works in the ultrasound range, specifically at 40 kHz. Requires a 12 V direct voltage source before being connected to the data acquisition system. As the phenomenon of mechanical vibrations is detected on the scale of milliseconds, it is not possible to adequately measure it simultaneously with the antenna, which in turn detects its signals on the scale of tens and hundreds of nanoseconds. It is for this reason that the acoustic sensor signals are acquired by an acquisition card such as an oscilloscope such as, for example, the NI USB-5133 card with 50 MHz bandwidth and sampling. maximum of 100 MHz, through one of the coaxial cables, which has lower bandwidth and sampling requirements.
En la Figura 3 se muestra esquemáticamente mediante bloques el montaje o sistema utilizado para poder realizar mediciones con los sensores de la presente invención, en la cual: Figure 3 shows schematically using blocks the assembly or system used to perform measurements with the sensors of the present invention, in which:
1 . Representa la antena direccional para la transmisión y recepción de señales Deepace KC R102. 1 . It represents the directional antenna for the transmission and reception of Deepace KC R102 signals.
2. Representa el sistema de sensado por el sensor acústico ultrasónico AA Ultrasonic PD Sensor. 2. Represents the sensing system by the AA Ultrasonic PD Sensor ultrasonic acoustic sensor.
3. Representa el filtro pasa altos en el rango 500-3000 MHz que se encuentra interiorizado entre en el sistema de adquisición. 3. Represents the high-pass filter in the 500-3000 MHz range that is internalized in the acquisition system.
4. Representa el sistema de adquisición de señales, donde en este caso particular esquematiza un osciloscopio que disponga ancho de banda de al menos 3 GHz y frecuencia de muestreo 5 GHz. 4. Represents the signal acquisition system, where in this particular case it schematizes an oscilloscope that has a bandwidth of at least 3 GHz and a sampling frequency of 5 GHz.
5. Representa el computador con sistema operativo que soporte software de control y automatización del sistema. 5. Represents the computer with an operating system that supports system control and automation software.
6. Representa una tarjeta de adquisición NI USB-5133 con un ancho de banda 50 MHz y muestreo máximo de 100 MHz. 6. Represents an NI USB-5133 acquisition card with a bandwidth of 50 MHz and maximum sampling of 100 MHz.
7. Representa una fuente de tensión continua de 12 V. 7. Represents a 12 V direct voltage source.
8. Representa 2 Cables coaxiales tipo BNC de 50Q que conecta al sensor acústico y la antena con el sistema de adquisición. 8. Represents 2 50Q BNC type coaxial cables that connect the acoustic sensor and the antenna with the acquisition system.
Además, la circunferencia en la parte superior del esquema representa el aislador a monitorear, que se encuentra a una distancia de 50cm de cada sensor. Furthermore, the circumference at the top of the diagram represents the insulator to be monitored, which is located at a distance of 50cm from each sensor.
En la presente invención la adquisición de señales desde ambos sensores, la antena y el sensor acústico, es dirigida a través de un computador con sistema operativo que soporte Python y Labview, utilizando este último como una interfaz gráfica. La interfaz gráfica a su vez llama a la programación del modelo de inteligencia artificial desarrollada en Python que es la que analiza las señales de las descargas parciales, las clasifica según riesgo para producir un contorneo y toma una decisión para la emisión de alarmas. En el contexto de la presente invención, y a modo de aclaración general a lo largo de toda la descripción, se entenderá como contorneo cuando la descarga de corriente se produce a lo largo de la superficie o una falla superficial de un aislamiento sólido colocado dentro de un aislamiento gaseoso o líquidos dieléctricos In the present invention, the acquisition of signals from both sensors, the antenna and the acoustic sensor, is directed through a computer with an operating system that supports Python and Labview, using the latter as a graphical interface. The graphical interface in turn calls for the programming of the artificial intelligence model developed in Python, which analyzes the partial discharge signals, classifies them according to risk to produce a contour, and makes a decision to issue alarms. In the context of the present invention, and by way of general clarification throughout the entire description, flashover will be understood when the current discharge occurs along the surface or a surface failure of a solid insulation placed within a gaseous insulation or dielectric liquids
La presente invención proporciona también un método para la detección temprana de descargas superficiales en aisladores de equipos primarios, que se caracteriza porque comprende: medir mediante una antena la radiación electromagnética que emite un aislador de equipo primario; medir datos de las ondas acústicas que emite un aislador de equipo primario mediante una pluralidad de sensores ultrasónicos; procesar los datos obtenidos por el sensor acústico para la obtención del comportamiento global de las vibraciones mecánicas; analizar los datos obtenidos por dichos sensores mediante un algoritmo de redes neuronales; y entregar un indicador numérico que representa la peligrosidad del aislador; separa por rangos el indicador de peligrosidad para indicar, como: actividad normal, actividad de riesgo intermedio, y una situación de peligro; comparar los indicadores para dar un diagnóstico final del estado del aislador; y analizar mediante ciclos de histéresis los indicadores de peligrosidad. The present invention also provides a method for the early detection of surface discharges in primary equipment insulators, which is characterized in that it comprises: measuring by means of an antenna the electromagnetic radiation emitted by a primary equipment insulator; measuring data from acoustic waves emitted by a primary equipment isolator using a plurality of ultrasonic sensors; process the data obtained by the acoustic sensor to obtain the global behavior of mechanical vibrations; analyze the data obtained by said sensors using a neural network algorithm; and deliver a numerical indicator that represents the danger of the insulator; separates the danger indicator by ranges to indicate, such as: normal activity, intermediate risk activity, and a dangerous situation; compare the indicators to give a final diagnosis of the condition of the insulator; and analyze the danger indicators using hysteresis cycles.
En una realización preferida, el método para la detección temprana de descargas superficiales en aisladores de equipos primarios se caracteriza porque mediante redes neuronales se analizan por separado las señales medidas por dichos sensores. In a preferred embodiment, the method for early detection of surface discharges in primary equipment insulators is characterized in that the signals measured by said sensors are analyzed separately using neural networks.
El funcionamiento del método de detección de descargas parciales para emitir la alerta de contorneo se realiza en tres etapas: The operation of the partial discharge detection method to issue the bypass alert is carried out in three stages:
• Etapa 1 : Consiste en la medición de señales a través de la antena y el sensor acústico. El modelo de red neuronal que procesa estas señales requiere que las señales de la antena sean adquiridas en grupos de 5 y las señales del sensor acústico en grupos de a 10. Las señales del sensor acústico son preprocesadas antes de entrar a la red neuronal, utilizando el cálculo de la envolvente de cada señal mediante transformada de Hilbert y el posterior cálculo de la transformada de Fourier de esta envolvente. De esta forma la información del sensor acústico que se entrega a la red neuronal corresponde al comportamiento global de las vibraciones mecánicas provenientes de las descargas parciales. Este enfoque para el sensor acústico está basado en estudios anteriores (S. Polisetty, A. El-Hag, and S. Jayram, “Classification of common discharges in outdoor insulation using acoustic signals and artificial neural network,” High Voltage, vol. 4, no. 4). • Stage 1: It consists of measuring signals through the antenna and the acoustic sensor. The neural network model that processes these signals requires that the antenna signals be acquired in groups of 5 and the acoustic sensor signals in groups of 10. The acoustic sensor signals are preprocessed before entering the neural network, using the calculation of the envelope of each signal using the Hilbert transform and the subsequent calculation of the Fourier transform of this envelope. In this way, the information from the acoustic sensor that is delivered to the neural network corresponds to the global behavior of the mechanical vibrations from partial discharges. This approach for the acoustic sensor is based on previous studies (S. Polisetty, A. El-Hag, and S. Jayram, “Classification of common discharges in outdoor insulation using acoustic signals and artificial neural network,” High Voltage, vol. 4 , no. 4).
• Etapa 2: En esta etapa se analizan por separado las mediciones de ambos sensores, la antena y el sensor acústico, mediante dos modelos de inteligencia artificial, de igual arquitectura, pero optimizados para analizar sus correspondientes señales. Estos modelos fueron entrenados en base a 5 señes de mediciones experimentales que permitieron medir la actividad de descargas parciales desde una condición de poca peligrosidad hacia una de alta peligrosidad, alcanzando el contorneo. Los modelos de inteligencia artificial entregan como resultado un número real que puede valer entre 0 y 1 , que es interpretable como un índice de peligrosidad. • Stage 2: In this stage, the measurements of both sensors, the antenna and the acoustic sensor, are analyzed separately using two artificial intelligence models, with the same architecture, but optimized to analyze their corresponding signals. These models were trained based on 5 experimental measurement signals that allowed the partial discharge activity to be measured from a low-risk condition to a high-risk condition, reaching contour. Artificial intelligence models deliver a real number as a result that can be between 0 and 1, which can be interpreted as a danger index.
• Etapa 3: En esta etapa final los resultados de ambos índices de peligrosidad son usados para determinar la decisión general del sistema, y por lo tanto definitiva. Cada índice de peligrosidad, de la antena y del sensor acústico, es comparado respecto a umbrales de decisión especificados por el usuario dependiendo de que tan rígido desea el sistema de alarma. Para efectos de mostrar resultados, se ha escogido el intervalo entre 0 y 0,35 como aquel donde los índices de peligrosidad son considerados como de actividad normal, esto es sin peligro. Entre 0,35 y 0,8 se considera actividad que se debe prestar atención, indicando un riesgo intermedio. Si es superior a 0,8 es una situación de alarma o peligro. Considerando que la actividad de descargas parciales es variable según el nivel de contaminación, que en la realidad puede ir variando, por lo que se incluye una histéresis que permita a estos índices devolverse de estos estados de alarma. La histéresis empleada para obtener resultados comprende esperar 20 decisiones sucesivas del tipo normal para bajar a este estado desde uno de atención; 20 decisiones sucesivas del tipo atención para bajar a este estado desde uno de alarma; y 40 decisiones sucesivas del tipo normal para bajar a este estado desde el estado de alarma. • Stage 3: In this final stage the results of both hazard indices are used to determine the general, and therefore definitive, decision of the system. Each hazard index, from the antenna and the acoustic sensor, is compared to decision thresholds specified by the user depending on how rigid the alarm system is desired. For the purposes of showing results, the interval between 0 and 0.35 has been chosen as the one where the danger indices are considered normal activity, that is, without danger. Between 0.35 and 0.8 is considered an activity that should be paid attention to, indicating an intermediate risk. If it is greater than 0.8, it is an alarm or danger situation. Considering that the activity of partial discharges is variable depending on the level of contamination, which in reality can vary, so a hysteresis is included that allows these indices to return from these alarm states. The hysteresis used to obtain results includes waiting 20 successive decisions of the normal type to descend to this state from one of attention; 20 successive decisions of the attention type to go down to this state from one of alarm; and 40 successive decisions of the normal type to go down to this state from the alarm state.
En el contexto de la presente invención, y a modo de aclaración general a lo largo de toda la descripción, se entenderá como contaminación a todo objeto ajeno a la electrónica del dispositivo que se posa sobre la superficie del aislador, que pueda interferir de manera mecánica o electrónica en su funcionamiento normal, e incidir en el índice de peligrosidad que se detecte en el mismo. In the context of the present invention, and as a general clarification throughout the entire description, contamination will be understood as any object other than the electronics of the device that rests on the surface of the insulator, which may mechanically or electronically interfere with its normal operation, and affect the danger index detected in it.
La decisión final para intervenir oportunamente en el aislador es una combinación de ambos estados determinados por la antena y el sensor acústico. Si la determinación de ambos sensores es normal o al menos uno de ellos está en atención, la decisión general del sistema es un estado “normal”, lo que se indica en la interfaz con un color verde. Si ambas determinaciones de los sensores revelan que están en estado de atención o al menos uno de ellos está en estado de alarma, es decir el sensor acústico en alarma y la antena en atención, entonces la decisión general es un estado de “atención”, lo que se indica con color amarillo. En cualquier otro caso se considera que la decisión general es el estado de “alarma”, lo que se indica con un color rojo. The final decision to timely intervene in the isolator is a combination of both states determined by the antenna and the acoustic sensor. If the determination of both sensors is normal or at least one of them is in attention, the overall decision of the system is a “normal” state, which is indicated on the interface with a green color. If both sensor determinations reveal that they are in a state of attention or at least one of them is in an alarm state, that is, the acoustic sensor in alarm and the antenna in attention, then the general decision is an “attention” state. which is indicated with yellow color. In any other case, the general decision is considered to be the “alarm” state, which is indicated with a red color.
EJEMPLO DE REALIZACIÓN EXAMPLE OF REALIZATION
Circuito eléctrico para la validación de la configuración experimental Electrical circuit for validation of the experimental setup
Se recreó la situación de un aislador contaminado con sal, donde su humedad va incrementándose periódicamente hasta producir el contorneo o descarga total a lo largo de la superficie del aislador. En esta validación es el contorneo el que se quiere detectar en etapas tempranas a través del comportamiento de descargas parciales previas. The situation of an insulator contaminated with salt was recreated, where its humidity increases periodically until it produces total flashover or discharge along the surface of the insulator. In this validation, it is the contouring that we want to detect in early stages through the behavior of previous partial discharges.
El circuito utilizado para las pruebas experimentales se muestra en la Figura 4, donde aparece una fuente de alta tensión que se logró a través de un transformador de ensayo para alta tensión 400V/150kV conectado a la red industrial de 220 V 50 Hz controlado mediante un transformador regulador. También, existe un divisor de tensión capacitivo en paralelo al objeto de ensayo se utilizó para medir la tensión aplicada al objeto. El objeto del ensayo fue un aislador tipo bushing de transformador, esto es un aislador de equipo primario y la tensión nominal sobre el aislador es de 34.5 kV. Por otro lado, para la contaminación salina, se disolvió NaCI en agua a una concentración de 167 g/L, y dicha solución se espació sobre el aislador, y posteriormente se secó con un secador industrial para producir un depósito inicial de sal en su superficie. En cada serie experimental se mantuvo la tensión nominal del aislador a 34,5 kV durante todo el experimento para simular su operación en condición de contaminación. La solución salina se esparció adicionalmente de forma periódica, en series cada 3 minutos y en otras cada 5 minutos con el objetivo de simular el fenómeno de humedad, el cual es altamente aleatorio en terreno. La aspersión de esta contaminación mientras el aislador estaba energizado se hizo mediante una bomba manual de fumigación. Para contener la contaminación salina el objeto de ensayo fue puesto dentro de una caja de acrílico. The circuit used for the experimental tests is shown in Figure 4, where a high voltage source appears that was achieved through a 400V/150kV high voltage test transformer connected to the 220 V 50 Hz industrial network controlled by a regulator transformer. Also, there is a capacitive voltage divider in parallel to the test object that was used to measure the voltage applied to the object. The object of the test was a transformer bushing type insulator, this is a primary equipment insulator and the nominal voltage on the insulator is 34.5 kV. On the other hand, for salt contamination, NaCl was dissolved in water at a concentration of 167 g/L, and said solution was spread over the insulator, and subsequently dried with an industrial dryer to produce an initial deposit of salt on its surface. . In each experimental series, the nominal voltage of the insulator was maintained at 34.5 kV throughout the experiment to simulate its operation in a contaminated condition. The saline solution was additionally spread periodically, in series every 3 minutes and in others every 5 minutes with the objective of simulating the humidity phenomenon, which is highly random in the field. Spraying this contamination while the insulator was energized was done using a hand fumigation pump. To contain salt contamination, the test object was placed inside an acrylic box.
Cabe mencionar que otros medios del tipo electrónico fueron descartados dado que introducían ruido eléctrico a la medición realizada por la antena. En la parte superior de la caja se dispuso un foco de luz con el objetivo de simular el efecto del sol que tiende a secar la humedad del aislador, por lo tanto, se hizo variable el comportamiento de las descargas parciales. It is worth mentioning that other electronic means were discarded since they introduced electrical noise to the measurement carried out by the antenna. A light source was placed in the upper part of the box with the aim of simulating the effect of the sun that tends to dry out the moisture in the insulator, therefore, the behavior of the partial discharges was made variable.
Para las series experimentales se utilizó un osciloscopio Keysight Infiniium DSOS804a para medir las señales de antena utilizando una ventana temporal de 1 ps y una frecuencia de muestreo de 5 GHz. Las señales del sensor acústico fueron adquiridas por la tarjeta NI USB-5133, utilizando una ventana temporal de 100 ms y frecuencia de muestreo de 200 MHz. El propio osciloscopio Keysight utilizado corresponde a la integración de un osciloscopio y un computador con sistema operativo que soporta labview 2018 y Python 3.6, por lo que en estas series experimentales estos dos componentes, esquematizados en la Figura 3, estuvieron integrados en solo uno. For the experimental series, a Keysight Infiniium DSOS804a oscilloscope was used to measure the antenna signals using a time window of 1 ps and a sampling rate of 5 GHz. The acoustic sensor signals were acquired by the NI USB-5133 card, using a time window of 100 ms and sampling frequency of 200 MHz. The Keysight oscilloscope used corresponds to the integration of an oscilloscope and a computer with an operating system that supports labview 2018 and Python 3.6, so in these experimental series these two components, schematized in Figure 3, were integrated into only one.

Claims

REIVINDICACIONES
1 . Un sistema multisensorial de detección temprana de descargas superficiales en aisladores de equipos primarios, CARACTERIZADO porque comprende: sensores para medir radiación electromagnética a distancia, donde uno de dichos sensores es un sensor de señales electromagnéticas tipo antena y el otro es un sensor acústico; dicho sistema incluye, además: un filtro que limita el ancho de banda permitido por los sensores; un sistema de adquisición y análisis de datos obtenidos por los sensores, por medio de inteligencia artificial, donde dicho sistema de análisis de datos da un pronóstico de probabilidad de ocurrencia de una falla superficial; y una fuente de voltaje continuo para la alimentación de los sensores. 1 . A multisensory system for early detection of surface discharges in primary equipment insulators, CHARACTERIZED because it comprises: sensors to measure electromagnetic radiation at a distance, where one of said sensors is an antenna-type electromagnetic signal sensor and the other is an acoustic sensor; Said system also includes: a filter that limits the bandwidth allowed by the sensors; a data acquisition and analysis system obtained by the sensors, through artificial intelligence, where said data analysis system gives a forecast of the probability of occurrence of a surface failure; and a continuous voltage source to power the sensors.
2. El sistema de la reivindicación 1 , CARACTERIZADO porque el sensor de señales electromagnéticas tipo antena mide en el rango de los giga Hertz y el sensor acústico mide ondas mecánicas. 2. The system of claim 1, CHARACTERIZED because the antenna-type electromagnetic signal sensor measures in the giga Hertz range and the acoustic sensor measures mechanical waves.
3. El sistema de la reivindicación 1 , CARACTERIZADO porque el sensor acústico es un sensor del tipo acústico ultrasónico que mide señales de frecuencias ultrasónicas. 3. The system of claim 1, CHARACTERIZED in that the acoustic sensor is an ultrasonic acoustic type sensor that measures ultrasonic frequency signals.
4. El sistema de la reivindicación 1 , CARACTERIZADO porque el sistema comprende una pluralidad de sensores acústicos. 4. The system of claim 1, CHARACTERIZED in that the system comprises a plurality of acoustic sensors.
5. El sistema de la reivindicación 4, CARACTERIZADO porque el sistema comprende una pluralidad de sensores acústicos ultrasónicos. 5. The system of claim 4, CHARACTERIZED in that the system comprises a plurality of ultrasonic acoustic sensors.
6. El sistema de la reivindicación 1 , CARACTERIZADO porque dicho sensor de señales electromagnéticas tipo antena es de una antena de ranura abierta. El sistema de la reivindicación 1 , CARACTERIZADO porque filtro que limita el ancho de banda permitido por los sensores, incluye un rango de frecuencias entre 500-3000MHz. Un método para la detección temprana de descargas superficiales en aisladores de equipos primarios, CARACTERIZADO porque comprende: adquirir y medir, mediante una antena, señales correspondientes a la radiación electromagnética que emite un aislador de equipo primario para obtener datos a partir de estas señales; adquirir y medir, mediante un sensor acústico ultrasónico, señales correspondientes a las ondas acústicas que emite un aislador de equipo primario para obtener datos a partir de estas señales; procesar los datos obtenidos por el sensor acústico ultrasónico para la evaluación de las vibraciones mecánicas resultantes de alteraciones mecánicas que emite el aislador; analizar los datos obtenidos por la antena y el sensor acústico ultrasónico mediante un algoritmo de inteligencia artificial; entregar un indicador numérico que representa la peligrosidad del aislador; separar por rangos el indicador de peligrosidad para determinar una actividad normal, una actividad de riesgo intermedio, y una situación de peligro en el aislador; comparar los indicadores de peligrosidad para dar un diagnóstico final del estado del aislador; y analizar mediante ciclos de histéresis dichos indicadores de peligrosidad que permiten la detección temprana de descargas superficiales en los aisladores. El método de la reivindicación 8, CARACTERIZADO porque el algoritmo de inteligencia artificial es mediante redes neuronales. 6. The system of claim 1, CHARACTERIZED in that said antenna-type electromagnetic signal sensor is an open slot antenna. The system of claim 1, CHARACTERIZED because a filter that limits the bandwidth allowed by the sensors, includes a frequency range between 500-3000MHz. A method for the early detection of surface discharges in primary equipment insulators, CHARACTERIZED because it comprises: acquiring and measuring, by means of an antenna, signals corresponding to the electromagnetic radiation emitted by a primary equipment insulator to obtain data from these signals; acquiring and measuring, using an ultrasonic acoustic sensor, signals corresponding to the acoustic waves emitted by a primary equipment isolator to obtain data from these signals; process the data obtained by the ultrasonic acoustic sensor for the evaluation of mechanical vibrations resulting from mechanical alterations emitted by the insulator; analyze the data obtained by the antenna and the ultrasonic acoustic sensor using an artificial intelligence algorithm; provide a numerical indicator that represents the danger of the insulator; separate the danger indicator by ranges to determine a normal activity, an intermediate risk activity, and a dangerous situation in the insulator; compare the danger indicators to give a final diagnosis of the condition of the insulator; and analyze these danger indicators through hysteresis cycles that allow early detection of surface discharges in the insulators. The method of claim 8, CHARACTERIZED because the artificial intelligence algorithm is through neural networks.
0. El método de la reivindicación 9, CARACTERIZADO porque mediante las redes neuronales se analizan por separado las señales medidas por la antena y el sensor acústico ultrasónico. 1. El método de la reivindicación 8, CARACTERIZADO porque el procesamiento de los datos provenientes de la antena y del sensor ultrasónico antes que ingrese a la red neuronal, utiliza las transformadas de Hilbert y la transformada de Fourier. 2. El método de la reivindicación 8, CARACTERIZADO porque la adquisición de señales de la antena se realiza en grupos de 5 y las señales adquiridas por el sensor acústico se realiza en grupos de a 10. 3. El método de la reivindicación 8, CARACTERIZADO porque los algoritmos de inteligencia artificial son entrenados en base a 5 series de mediciones experimentales. 4. El método de la reivindicación 8, CARACTERIZADO porque los algoritmos de inteligencia artificial entregan como resultado un valor numérico entre 0 y 1 , que se interpreta como un índice de peligrosidad. 5. El método de la reivindicación 14, CARACTERIZADO porque dichos valores que se interpretan como índice de peligrosidad incluyen un rango entre 0 y 0,35 cuyo valor se considera como actividad normal, un rango entre 0,35 y 0,8 cuyo valor se considera como una actividad a la que se debe prestar atención, y un valor superior a 0,8 que se considera como una situación de alarma o peligro. 6. El método de las reivindicaciones 8 y 15, CARACTERIZADO porque la histéresis consiste en esperar 20 decisiones sucesivas del tipo de actividad normal para bajar a este estado desde un estado de una actividad a la que se le debe prestar atención; 20 decisiones sucesivas del tipo de una actividad a la que se debe prestar atención para bajar a este estado desde una situación de alarma o peligro; y 40 decisiones sucesivas del tipo de actividad normal para bajar a este estado desde una situación de alarma o peligro. 0. The method of claim 9, CHARACTERIZED because the signals measured by the antenna and the ultrasonic acoustic sensor are analyzed separately by means of neural networks. 1. The method of claim 8, CHARACTERIZED because the processing of the data coming from the antenna and the ultrasonic sensor before it enters the neural network, uses the Hilbert transforms and the Fourier transform. 2. The method of claim 8, CHARACTERIZED because the acquisition of antenna signals is carried out in groups of 5 and the signals acquired by the acoustic sensor are carried out in groups of 10. 3. The method of claim 8, CHARACTERIZED because the artificial intelligence algorithms are trained based on 5 series of experimental measurements. 4. The method of claim 8, CHARACTERIZED because the artificial intelligence algorithms deliver a numerical value between 0 and 1 as a result, which is interpreted as a danger index. 5. The method of claim 14, CHARACTERIZED because said values that are interpreted as a hazard index include a range between 0 and 0.35, whose value is considered normal activity, a range between 0.35 and 0.8, whose value is It is considered an activity that should be paid attention to, and a value greater than 0.8 is considered an alarm or danger situation. 6. The method of claims 8 and 15, CHARACTERIZED because the hysteresis consists of waiting for 20 successive decisions of the type of normal activity to descend to this state from a state of an activity to which attention must be paid; 20 successive decisions of the type of an activity that must be paid attention to in order to descend to this state from an alarm or danger situation; and 40 successive decisions of the type of normal activity to descend to this state from an alarm or danger situation.
PCT/CL2023/050060 2022-07-25 2023-07-24 Multisensory method and system for detecting surface discharges in insulators WO2024020706A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CL2007-2022 2022-07-25
CL2022002007 2022-07-25

Publications (1)

Publication Number Publication Date
WO2024020706A1 true WO2024020706A1 (en) 2024-02-01

Family

ID=89704815

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CL2023/050060 WO2024020706A1 (en) 2022-07-25 2023-07-24 Multisensory method and system for detecting surface discharges in insulators

Country Status (1)

Country Link
WO (1) WO2024020706A1 (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101235777B1 (en) * 2011-09-26 2013-02-21 유성훈 Artificial intelligent utilization on judgement diagnostic system for electrical power ficilities using comples diagnosis eqipment
CN104569764A (en) * 2015-01-06 2015-04-29 国家电网公司 Live detection system for creeping discharge of composite apparatus and method thereof
US20150120218A1 (en) * 2010-05-31 2015-04-30 Universidad Politecnica De Madrid Novel method for real time tests and diagnosis of partial discharge sources in high voltage equipment and installations, which are in service or out of service, and physical system for the practical use of the method
CN105911438A (en) * 2016-04-13 2016-08-31 国网湖南省电力公司 GIS risk evaluation method and GIS risk evaluation system based on partial discharge live detection

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150120218A1 (en) * 2010-05-31 2015-04-30 Universidad Politecnica De Madrid Novel method for real time tests and diagnosis of partial discharge sources in high voltage equipment and installations, which are in service or out of service, and physical system for the practical use of the method
KR101235777B1 (en) * 2011-09-26 2013-02-21 유성훈 Artificial intelligent utilization on judgement diagnostic system for electrical power ficilities using comples diagnosis eqipment
CN104569764A (en) * 2015-01-06 2015-04-29 国家电网公司 Live detection system for creeping discharge of composite apparatus and method thereof
CN105911438A (en) * 2016-04-13 2016-08-31 国网湖南省电力公司 GIS risk evaluation method and GIS risk evaluation system based on partial discharge live detection

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
POLISETTY SATISH KUMAR: "Partial Discharge Classification Using Acoustic Signals and Artificial Neural Networks and its Application in detection of Defects in Ceramic Insulators", MASTER'S THESIS, UNIVERSITY OF WATERLOO, UNIVERSITY OF WATERLOO, 24 January 2019 (2019-01-24), XP093135719, Retrieved from the Internet <URL:https://uwspace.uwaterloo.ca/bitstream/handle/10012/14415/Polisetty_SatishKumar.pdf?sequence=5&isAllowed=y> [retrieved on 20240228] *

Similar Documents

Publication Publication Date Title
KR100997223B1 (en) Method and system for detecting partial electric discharge of gas-insulated device
Griffiths et al. The invisible fray: A critical analysis of the use of reflectometry for fray location
JP2885223B2 (en) Fault diagnosis system for high voltage power equipment
CN103140766B (en) For being monitored the apparatus and method of electrical power transmission system by part Discharge analysis
US5463317A (en) Shield integrity monitor
KR102491071B1 (en) Switchboard With A Built-in Hybrid Monitoring Device Equipped With The Function Of Diagnosing The Deterioration State Of Power Equipment And Preventing Electric Shock Accident In The Energized Parts
KR101606925B1 (en) Precision measurement for Electric field Intensity and lightning warning system
CN116184102B (en) Inspection system and method for transformer protection system
RU2620021C1 (en) Device (versions) and procedure for determination of condition insulator sets
RU2720638C1 (en) Device for monitoring and diagnostics of high-voltage linear polymer insulators
WO2024020706A1 (en) Multisensory method and system for detecting surface discharges in insulators
JP2006208017A (en) Partial discharge detector device in gas insulation equipment
Pinpart et al. Experimental comparison of UHF sensor types for PD location applications
KR102486079B1 (en) Lightning rod system based on internet of things
JPWO2014125563A1 (en) Partial discharge sensor evaluation method
KR100632078B1 (en) Noise discriminating device and method for detect the partial discharge in underground power cable
CN110716113A (en) GIL fault positioning system based on ultrasonic wave and earth electric wave technology
Gemme et al. Partial Discharge alert system in medium voltage switchgear
BR102016016197A2 (en) INSULATOR FAULT SIGNALING SYSTEM IN HIGH-VOLTAGE STRUCTURES USING ULTRASOUND.
CN206863197U (en) A kind of verifying attachment of the extra-high video sensor based on GHz transverse electro-magnetic wave signal
CN205608069U (en) Support type pipe generating line and suspension type pipe generating line examine electric installation on line
Lopez-Roldan et al. Development of a pocket ultrahigh frequency partial-discharge detector to warn switchyard personnel of imminent failure of a nearby high-voltage plant
Kim et al. Optimal sensor placement methodology based on FDTD for partial discharge detection in GIS
Leonida et al. Remote Diagnosis of insulator strings using radiated emission
CN112162181B (en) Monitoring method, device and computer readable storage medium

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: 23844745

Country of ref document: EP

Kind code of ref document: A1