JP2020012726A - Partial discharge detection system, learning system, method for detecting partial discharge, computer program, and electric apparatus - Google Patents

Abstract

To provide a partial discharge detection system, a learning system, a method for detecting a partial discharge, a computer program, and an electric apparatus which can precisely detect a partial discharge of an electric apparatus.SOLUTION: The partial discharge detection system according to an embodiment includes a sensor and a partial discharge determination unit. The sensor generates a signal according to a phenomenon caused by the discharge of the electric apparatus. The partial discharge determination unit determines whether a partial discharge has occurred in the electric apparatus on the basis of a learning model that has learned a signal including noise near the electric apparatus and a signal generated after a predetermined period has passed, by conducting a mechanical learning on the signals generated within a predetermined length of time from the start of operation of the electric apparatus.SELECTED DRAWING: Figure 1

Description

  Embodiments of the present invention relate to a partial discharge detection system, a learning system, a partial discharge detection method, a computer program, and an electric device.

  In the electrical equipment, the insulation performance of the insulator on the surface of the electrical equipment or inside the electrical equipment deteriorates with aging. When the insulation performance deteriorates, partial discharge occurs at the deteriorated portion. If the insulation performance further deteriorates, dielectric breakdown occurs. Serious accidents such as ground faults occur due to insulation breakdown. For this reason, in the maintenance and maintenance of electrical equipment, detection of partial discharge is performed. The partial discharge detection device detects a partial discharge by catching a signal of the partial discharge and estimating a state of deterioration of insulation performance from a magnitude or a frequency of occurrence of the signal.

  2. Description of the Related Art A conventional partial discharge detection device detects a partial discharge by performing machine learning on frequency characteristic data and phase characteristic data measured in a laboratory or the like using a neural network or the like. The partial discharge device detects a partial discharge by acquiring an electric signal at a site where an electric device is provided. The electrical signal measured on site includes noise specific to the site. For this reason, there is a possibility that a partial discharge generated in a laboratory cannot be accurately detected from an electric signal measured on site by a neural network learning the partial discharge.

JP-A-7-181218

  The problem to be solved by the present invention is to provide a partial discharge detection system, a learning system, a partial discharge detection method, a computer program, and an electric device that can more accurately detect a partial discharge of an electric device.

  The partial discharge detection system according to the embodiment has a sensor and a partial discharge determination unit. The sensor generates a signal in response to a phenomenon caused by a discharge of the electric device. The partial discharge determination unit performs machine learning on a signal generated during a predetermined period from the start of operation of the electrical device, thereby learning a signal including noise near the electrical device, and a learning model. Based on the signal generated after the lapse of the period, the presence or absence of the partial discharge of the electric device is determined.

FIG. 2 is a perspective view of a switchboard provided with the partial discharge detection system according to the first embodiment. FIG. 2 is a functional block diagram illustrating a functional configuration of the partial discharge detection system according to the first embodiment. 5 is a flowchart illustrating a flow of a process of detecting partial discharge according to the first embodiment. 5 is a flowchart illustrating a flow of generating a learning model according to the first embodiment. 5 is a flowchart illustrating a flow of determining partial discharge according to the first embodiment. The perspective view of the switchboard provided with the partial discharge detection system of 2nd Embodiment. FIG. 5 is a functional block diagram illustrating a functional configuration of a partial discharge determination device according to a second embodiment. FIG. 9 is a functional block diagram illustrating a functional configuration of the information processing apparatus according to the second embodiment.

  Hereinafter, a partial discharge detection system, a learning system, a partial discharge detection method, a computer program, and an electric device according to an embodiment will be described with reference to the drawings.

(First embodiment)
FIG. 1 is a perspective view of a switchboard 10 provided with a partial discharge detection system 100 according to the first embodiment. The partial discharge detection system 100 includes one or more sensors 101, a signal line 102, and a partial discharge determination device 103. The partial discharge detection device 100 according to the first embodiment is configured by a single device. The partial discharge detection system 100 is provided in the vicinity of a switchboard 10 in which electric devices are housed. The partial discharge detection system 100 detects a partial discharge generated from an electric device based on an electromagnetic wave, a sound wave, a transient ground current, a current on a wall derived from an electromagnetic wave, and the like received by the sensor 101. The switchboard 10 is provided with a sensor 101. Electric equipment is housed inside the switchboard 10. The electric device is configured by devices (all not shown) such as a circuit breaker, a disconnector, a current transformer, or a transformer. Electrical equipment supplies a high voltage and a large current from the outside via a power cable. The electric device has a function of cutting off the current when an abnormality occurs. In addition, a ground electrode is connected to a lower portion of the switchboard 10. The electric device may be any device, such as a power transformer, a gas insulated switch, a generator, an electric motor, a reactor, or the like, as long as it has a possibility of generating partial discharge.

  FIG. 2 is a functional block diagram illustrating a functional configuration of the partial discharge detection system 100 according to the first embodiment. The partial discharge detection system 100 includes a plurality of sensors 101, a signal line 102, and a partial discharge determination device 103 as described with reference to FIG.

  The sensor 101 is configured to include a metal electrode. The sensor 101 generates an electric signal according to a phenomenon caused by the operation of the electric device. The phenomenon caused by the operation of the electric device is, for example, electric current, electromagnetic wave, vibration, sound wave, vibration, or the like. The electric signal represents a physical quantity obtained from a phenomenon caused by driving. The sensor 101 is provided on an outer wall surface of the electric device. The sensor 101 is connected to a partial discharge determination device 103 via a signal line 102. The sensor 101 outputs the generated electric signal to the partial discharge determination device 103. The sensor 101 is a sensor that can measure a current, an electromagnetic wave, a sound wave, or a vibration derived from operation of an electric device such as a CT (Current Transformer) sensor, a TEV (Transient Earth Voltage) sensor, an AE (Acoustic Emission) sensor, and an antenna. Any sensor may be used. As the sensor 101, one of the above-described sensors may be used, or a plurality of types may be used in combination.

  The signal line 102 inputs the electric signal generated by the sensor 101 to the partial discharge determination device 103. The signal line 102 in FIG. 2 receives electric signals output from the plurality of sensors 101 through one signal line 102, but is not limited thereto. For example, the signal line 102 may be provided for each sensor 101.

  The partial discharge determination device 103 is an information processing device such as a personal computer, a smartphone, or a tablet computer. The partial discharge determination device 103 determines whether or not a partial discharge has occurred based on the electric signal output from the sensor 101. The partial discharge determination device 103 executes the partial discharge determination program to execute the communication unit 104, the input unit 105, the display unit 106, the signal storage unit 107, the learning model storage unit 108, the discharge model storage unit 109, and the deterioration model storage unit 110. And a device including the control unit 111.

  The communication unit 104 is a network interface. The communication unit 104 communicates with an external communication device via a network (not shown). The communication unit 104 may communicate using a communication method such as a wireless LAN (Local Area Network), a wired LAN, Bluetooth (registered trademark), or LTE (Long Term Evolution) (registered trademark). The external communication device may be an information processing device such as a personal computer or a server, or may be a cloud computing system.

  The input unit 105 is configured using input devices such as a touch panel, a mouse, and a keyboard. The input unit 105 may be an interface for connecting an input device to the partial discharge detection system 100. In this case, the input unit 105 generates input data (for example, instruction information indicating an instruction to the partial discharge detection system 100) from an input signal input in the input device, and inputs the generated data to the partial discharge detection system 100.

  The display unit 106 is an output device such as a CRT (Cathode Ray Tube) display, a liquid crystal display, and an organic EL (Electro Luminescence) display. The display unit 106 may be an interface for connecting the output device to the partial discharge detection system 100. In this case, the display unit 106 generates a video signal from the video data and outputs the video signal to a video output device connected thereto.

  The signal storage unit 107 is configured using a storage device such as a magnetic hard disk device or a semiconductor storage device. The signal storage unit 107 stores signal data. The signal data includes at least the date and time when the electric signal was acquired and the physical quantity represented by the electric signal. When the partial discharge detection system 100 includes a plurality of sensors 101, the signal data may hold identification information of the sensors 101. The identification information may be any information as long as it can identify the sensor 101.

  The learning model storage unit 108 stores a learning model. The learning model is generated by a learning unit 115 described later. The learning model is generated by performing machine learning on the electric signal generated during a predetermined period from the start of the operation of the electric device. The learning model is generated by performing deep learning or machine learning on learning data obtained from an electric signal. The learning model in the present embodiment is a model that has learned an electric signal including noise near an electric device. The noise is an electric signal other than the partial discharge generated by the operation of the electric device included in the electric signal. The learning data will be described later.

  The discharge model storage unit 109 stores a discharge model. The discharge model is recorded in the discharge model storage unit 109 in advance. The discharge model is a model in which the waveform of the partial discharge generated from the electric device has been learned. The discharge model is a model learned for each partial discharge waveform such as an insulator internal discharge, a void discharge, a creeping discharge, or a corona discharge. The discharge model is generated by machine learning such as a neural network or deep learning. Note that the discharge model may be generated by a machine learning method other than neural network or deep learning.

  The deterioration model storage unit 110 stores a deterioration model. The deterioration model is recorded in the deterioration model storage unit 110 in advance. The deterioration model is a model that can specify the deterioration state of the electric device when the internal discharge of the insulator occurs. The deterioration model is a model learned for each characteristic of an electric signal (for example, a waveform, a voltage, a current, a vibration, or a sound generated according to a partial discharge) according to the degree of deterioration of the electric device. The deterioration model is recorded in advance according to the type of defect of the electric device. Defects are, for example, voids, aging of parts or breakage of parts. The degraded state indicates a state in which performance is reduced due to use of the electric device. The deterioration state may represent the remaining life of the electric device by years or the failure rate of the electric device every N years (N is a natural number). The deterioration state may indicate a remaining life or a failure rate for each specific part of the electric device. Alternatively, the determination may be made in n stages such as the degree of deterioration. The degradation model is generated by machine learning such as a neural network or deep learning. The deterioration model may be generated by a machine learning method other than a neural network or deep learning.

  The control unit 111 controls the operation of each unit of the partial discharge determination device 103. The control unit 111 is executed by an apparatus including a processor such as a CPU (Central Processing Unit) and a RAM (Random Access Memory). The control unit 111 executes the partial discharge determination program to execute the electrical signal acquisition unit 112, the signal data generation unit 113, the learning data generation unit 114, the learning unit 115, the partial discharge determination unit 116, the noise removal unit 117, the discharge type It functions as the specifying unit 118, the deterioration specifying unit 119, and the alarm unit 120.

  The electric signal acquisition unit 112 acquires the electric signal input by the sensor 101. The electric signal acquisition unit 112 associates the date and time when the electric signal was acquired with the reception intensity represented by the electric signal. When the partial discharge detection system 100 includes a plurality of sensors 101, the electric signal acquisition unit 112 associates the date and time when the electric signal was obtained, the reception intensity indicated by the electric signal, and the identification information of the sensor 101.

The signal data generator 113 generates signal data based on the obtained electric signal. The signal data generation unit 113 records the generated signal data in the signal storage unit 107.
The learning data generation unit 114 generates learning data based on the generated signal data. The learning data is data suitable for generating a learning model. The learning data may be, for example, data representing a waveform of an electric signal, or data representing an electric signal in a ΦQN pattern. Note that the learning data generation unit 114 may generate learning data based on signal data generated during the learning period. The learning period is a period during which signal data used for generating a learning model is acquired. The learning period may be specified in advance, for example. The learning period may be, for example, from the start of operation of the electrical device after the shipping test to a predetermined period. The learning period may be, for example, a period from the start of the operation of the electrical device after the shipping test to the period until the first partial discharge is detected. The learning period may be any period as long as the signal data does not include the electric signal indicating the partial discharge.

  The learning unit 115 generates a learning model based on the learning data. The learning unit 115 generates a learning model by using machine learning such as a neural network or deep learning on the learning data. The learning unit 115 generates a learning model using learning data generated based on signal data acquired during a learning period. The learning data represents noise data in the vicinity of the electric device. The learning data is generated based on signal data that does not include an electric signal indicating a partial discharge. Therefore, the learning unit 115 generates a learning model that has learned unique noise near the electric device based on the learning data. The learning unit 115 records the generated learning model in the learning model storage unit 108. Note that the learning unit 115 may generate a learning model using a method other than neural network or deep learning as a method of machine learning. The learning unit 115 may regenerate a learning model at a predetermined timing. The predetermined timing may be a timing designated in advance, or may be a regular timing such as every year. With this configuration, the partial discharge detection system 100 can determine the partial discharge without continuing to use the old learning model even when the environment around the electric device is changed. Becomes possible.

  The partial discharge determination unit 116 determines whether or not the electric device has a partial discharge based on the acquired electric signal. More specifically, based on the waveform obtained from the electric signal and the learning model recorded in the learning model storage unit 108, the partial discharge determination unit 116 determines the shape of the waveform obtained from the electric signal that has not been learned by the learning model. Is determined to be "present" when the pulse signal or waveform is included. Based on the waveform obtained from the electric signal and the learning model recorded in the learning model storage unit 108, the partial discharge determination unit 116 generates a pulse signal or waveform having a shape not learned by the learning model from the waveform obtained from the electric signal. Is not included, it is determined that the partial discharge is “absent”.

  When the partial discharge determination unit 116 determines that partial discharge is present, the noise removal unit 117 removes noise included in the electric signal. Specifically, the noise removal unit 117 specifies, as noise, a waveform learned by the learning model among the waveforms included in the electric signal. The noise removing unit 117 sets the signal strength representing the waveform of the specified electric signal to a predetermined value. For example, when the electric signal is a signal in which a PD signal and noise are superimposed on each other, the noise removing unit 117 may set the noise component to 0 and leave the PD signal as it is. .

  When the partial discharge determination unit 116 determines that the partial discharge is “present”, the discharge type specifying unit 118 specifies the type of the partial discharge. Specifically, the discharge type specifying unit 118 specifies the type of the partial discharge based on the discharge model recorded in the discharge model storage unit 109 and the waveform of the electric signal. The discharge type identification unit 118 compares the waveform learned by the discharge model with the waveform of the electric signal. As a result of the comparison, the discharge type specifying unit 118 determines the type of partial discharge associated with the waveform closest to the waveform of the electric signal among the waveforms learned by the discharge model, May be specified. When the type of the partial discharge is specified, the user can determine the type of the defect of the electric device. Note that the type of the defect of the electric device may be displayed on the display unit 106. The user may be a service person or the like who determines whether or not the electric device has a partial discharge using the partial discharge detection system 100.

  The deterioration specifying unit 119 diagnoses the deterioration state of the electric device when the discharge type specifying unit 118 specifies “partial discharge in the insulator” as the type of the partial discharge. Specifically, the deterioration specifying unit 119 specifies the deterioration state of the electric device based on the deterioration model recorded in the deterioration model storage unit 110 and the characteristics of the electric signal. The deterioration specifying unit 119 compares the characteristic learned by the deterioration model with the characteristic of the electric signal. As a result of the comparison, the discharge type specifying unit 118 may specify, as the deterioration state of the electric device, the deterioration state associated with the characteristic closest to the characteristic of the electric signal among the characteristics learned by the deterioration model. . Note that the deterioration state of the electric device may be displayed on the display unit 106. The deterioration specifying unit 119 is an example of a state specifying unit.

  When the partial discharge determination unit 116 determines that partial discharge is present, the alarm unit 120 performs an alarm process for notifying that partial discharge has occurred. In the warning process, a warning sound may be issued, or a message such as an e-mail or SMS (Short Message Service) may be transmitted to a destination designated in advance. The alarm process may be any process as long as it is a process specified in advance. The warning unit 120 may perform different warning processing according to the type of the specified partial discharge or the deterioration state. For example, the warning unit 120 may include the type of the specified partial discharge in the transmitted message. For example, the warning unit 120 may determine the transmission destination of the message according to the specified deterioration state. By performing the alarm process by the alarm unit 120 in this manner, it is possible to notify the user or the maker of the electric device of the occurrence of the partial discharge.

  FIG. 3 is a flowchart illustrating a flow of processing of the partial discharge detection system 100 according to the first embodiment. For example, when the operation of the electric device is started, the partial discharge detection system 100 detects the partial discharge along the flow in FIG. The partial discharge detection system 100 generates a learning model (Step S101). For example, the partial discharge detection system 100 generates a learning model based on an electric signal acquired from an electric device during a learning period. The partial discharge detection system 100 monitors whether a partial discharge has occurred in the electric device after the elapse of the learning institution (step S102). For example, the partial discharge detection system 100 monitors whether or not a partial discharge has occurred in the electric device by determining whether or not a signal representing the partial discharge is included in the electric signal acquired from the electric device. I do. Step S102 is periodically performed at a predetermined timing.

  FIG. 4 is a flowchart illustrating a flow of generating a learning model according to the first embodiment. The learning model is generated based on learning data generated within a previously specified learning period. The electric signal acquisition unit 112 acquires the electric signal received from the sensor 101 (Step S201). The signal data generator 113 generates signal data based on the obtained electric signal. The signal data generation unit 113 records the generated signal data in the signal storage unit 107 (Step S202). The learning data generator 114 generates learning data based on the signal data generated during the learning period (Step S203). The learning unit 115 generates a learning model based on the learning data (Step S204). The learning unit 115 records the generated learning model in the learning model storage unit 108.

  FIG. 5 is a flowchart illustrating the flow of the determination of the partial discharge according to the first embodiment. The determination of the partial discharge is performed at a predetermined timing after the learning period. In this flowchart, the determination of the partial discharge is performed by comparing the waveform obtained from the electric signal with the learning model.

  The electric signal acquisition unit 112 acquires the electric signal received from the sensor 101 (Step S301). The signal data generator 113 generates signal data based on the obtained electric signal. The signal data generation unit 113 records the generated signal data in the signal storage unit 107 (Step S302).

  The partial discharge determination unit 116 compares the waveform obtained from the electric signal with the waveform learned by the learning model to determine whether the waveform obtained from the electric signal is included in the learning model (step S303). When the waveform obtained from the electric signal is included in the learning model (step S303: YES), the partial discharge determination unit 116 determines that there is no partial discharge. If it is determined that the partial discharge is “none”, the process ends. When the waveform obtained from the electric signal is not included in the learning model (step S303: YES), the partial discharge determination unit 116 determines that there is the partial discharge.

  The noise removing unit 117 specifies, as noise, a waveform learned by the learning model among the waveforms included in the electric signal. The noise removing unit 117 sets the value of the signal strength representing the waveform of the specified electric signal to 0 (Step S304). The discharge type specifying unit 118 specifies the type of the partial discharge based on the discharge model and the waveform of the electric signal (step S305).

  The deterioration specifying unit 119 determines whether or not the type of the specified partial discharge is “insulator internal discharge” (Step S306). If the type of the specified partial discharge is not “insulator internal discharge” (step S306: NO), the process proceeds to step S308. If the type of the specified partial discharge is “insulator internal discharge” (step S306: YES), the deterioration specifying unit 119 specifies the deterioration state of the electric device based on the deterioration model and the characteristics of the electric signal. (Step S307). The alarm unit 120 performs an alarm process for notifying that partial discharge has occurred (step S308). Note that the alarm unit 120 may perform different alarm processing according to the type of the specified partial discharge or the deterioration state.

  In the partial discharge detection system 100 configured as described above, the sensor 101 generates an electric signal when the operation of the electric device is started. The electric signal does not include the signal indicating the partial discharge because the electric device immediately after the start of the operation has not generated the partial discharge. For this reason, the learning unit 115 generates a learning model using the learning data generated based on the signal data not including the electric signal indicating the partial discharge. For this reason, the generated learning model is a learning model that has learned noise unique to the location where the electric device is provided. Therefore, when the waveform obtained from the electric signal includes a pulse signal or waveform having a shape that is not learned by the learning model, the partial discharge determination unit 116 determines that partial discharge is “present”, and thereby more accurately performs partial discharge. Can be detected.

(Second embodiment)
Next, a partial discharge detection device according to a second embodiment will be described. FIG. 6 is a perspective view of the switchboard 10 provided with the partial discharge detection system 100a of the second embodiment. The partial discharge detection system 100a includes one or more sensors 101, a signal line 102, a partial discharge determination device 103a, and an information processing device 200. The partial discharge determination device 103a is provided adjacent to the switchboard 10 in which electric equipment is housed. The partial discharge determination device 103a is communicably connected to the information processing device 200 via the network 300. The network 300 may be constructed with any network. For example, the network 300 may be configured by the Internet.

FIG. 7 is a functional block diagram illustrating a functional configuration of the partial discharge determination device 103a according to the second embodiment.
The partial discharge determination device 103a is an information processing device such as a personal computer, a smartphone, or a tablet computer. The partial discharge determination device 103a determines whether or not a partial discharge has occurred based on the electric signal output from the sensor 101. The partial discharge determination device 103 functions as a device including the communication unit 104, the input unit 105, the display unit 106, and the control unit 111a by executing the partial discharge determination program. Hereinafter, points different from the first embodiment will be described.

  The control unit 111a controls the operation of each unit of the partial discharge determination device 103a. The control unit 111a is executed by a device including a processor such as a CPU and a RAM, for example. The control unit 111a functions as the electric signal acquisition unit 112 and the signal data generation unit 113a by executing the partial discharge determination program.

  The signal data generation unit 113a generates signal data based on the obtained electric signal. The signal data generation unit 113a transmits the generated signal data to the information processing device 200 via the network 300.

  FIG. 8 is a functional block diagram illustrating a functional configuration of the information processing device 200 according to the second embodiment. The information processing device 200 is an information processing device such as a personal computer or a server. The information processing device 200 executes the partial discharge determination program to execute the communication unit 201, the input unit 202, the display unit 203, the signal storage unit 204, the learning model storage unit 205, the discharge model storage unit 206, the deterioration model storage unit 207, It functions as a device including the control unit 208.

  The communication unit 201 is a network interface. The communication unit 201 communicates with the partial discharge determination device 103a via the network 300. The communication unit 201 may communicate using a communication method such as a wireless LAN, a wired LAN, Bluetooth, or LTE.

  The input unit 202 is configured using input devices such as a touch panel, a mouse, and a keyboard. The input unit 202 may be an interface for connecting the input device to the information processing device 200. In this case, the input unit 202 generates input data (for example, instruction information indicating an instruction to the information processing device 200) from an input signal input by the input device, and inputs the generated data to the partial discharge detection system 100a.

  The display unit 203 is an output device such as a CRT display, a liquid crystal display, and an organic EL display. The display unit 203 may be an interface for connecting the output device to the information processing device 200. In this case, the display unit 203 generates a video signal from the video data and outputs the video signal to a video output device connected thereto.

  The signal storage unit 204, the learning model storage unit 205, the discharge model storage unit 206, and the deterioration model storage unit 207 are the signal storage unit 107, the learning model storage unit 108, the discharge model storage unit 109, and the deterioration model storage in the first embodiment. The description is omitted because it is the same as the unit 110.

  The control unit 208 controls the operation of each unit of the information processing device 200. The control unit 208 is executed by a device including a processor such as a CPU and a RAM, for example. The control unit 208 executes the partial discharge determination program to execute the signal data acquisition unit 209, the learning data generation unit 210, the learning unit 211, the partial discharge determination unit 212, the noise removal unit 213, the discharge type identification unit 214, the deterioration identification It functions as the unit 215 and the alarm unit 216.

  The signal data acquisition unit 209 acquires the signal data transmitted by the partial discharge determination device 103a. The signal data acquisition unit 209 records the acquired signal data in the signal storage unit 204.

  The learning data generation unit 210, the learning unit 211, the partial discharge determination unit 212, the noise removal unit 213, the discharge type identification unit 214, the deterioration identification unit 215, and the alarm unit 216 are the learning data generation unit 114 in the first embodiment. It is the same as the unit 115, the partial discharge determination unit 116, the noise removal unit 117, the discharge type identification unit 118, the deterioration identification unit 119, and the alarm unit 120, and thus the description is omitted.

  In the partial discharge detection system 100a thus configured, the partial discharge determination device 103a acquires an electric signal. The signal data generation unit 113 of the partial discharge determination device 103a generates signal data based on the electric signal and transmits the signal data to the information processing device 200. The information processing device 200 performs generation of a learning model, determination of occurrence of partial discharge, and the like based on the acquired signal data. Therefore, on the site, it is possible to acquire information on the electric signal acquired from the electric device and manage the acquired information in different places.

  In the above embodiment, when the partial discharge determination unit determines that the partial discharge is “present”, the noise removal unit removes noise included in the electric signal. However, the noise removing unit may be configured to remove noise from all electric signals regardless of the determination result of the partial discharge determining unit. In this case, since an unknown signal remains in the electric signal, the user can determine whether it is necessary to continue learning by checking the waveform from which noise has been removed. For example, when the user determines that the noise of the electric signal has not been sufficiently removed even though the partial discharge has not occurred, the learning unit is caused to generate a learning model based on more learning data. Is also good.

  The partial discharge detection system 100 may be configured as a plurality of devices. The partial discharge detection system 100 may be configured by a cloud computing system.

  In the above embodiments, the electric signal acquisition unit 112, the signal data generation unit 113, the learning data generation unit 114, the learning unit 115, the partial discharge determination unit 116, the noise removal unit 117, the discharge type identification unit 118, the deterioration identification unit 119, The alarm unit 120 is a software function unit, but may be a hardware function unit such as an LSI.

  According to at least one embodiment described above, the provision of the learning unit 115 and the partial discharge determination unit 116 makes it possible to more accurately detect the partial discharge of the electric device.

  Although several embodiments of the present invention have been described, these embodiments are provided by way of example and are not intended to limit the scope of the invention. These embodiments can be implemented in other various forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are also included in the invention described in the claims and equivalents thereof.

DESCRIPTION OF SYMBOLS 10 ... switchboard, 100 ... partial discharge detection system, 101 ... sensor, 102 ... signal line, 103 ... partial discharge determination apparatus, 104 ... communication part, 105 ... input part, 106 ... display part, 107 ... signal storage part, 108 ... Learning model storage unit, 109: discharge model storage unit, 110: deterioration model storage unit, 111: control unit, 112: electric signal acquisition unit, 113: signal data generation unit, 114: learning data generation unit, 115: learning unit 116, a partial discharge determination unit, 117, a noise removal unit, 118, a discharge type identification unit, 119, a deterioration identification unit, 120, an alarm unit, 100a, a partial discharge detection system, 103a, a partial discharge determination unit, 113a, signal data Generation unit, 200 information processing device, 201 communication unit, 202 input unit, 203 display unit, 204 signal storage unit, 205 learning model storage unit, 206: discharge model storage unit, 207: deterioration model storage unit, 208: control unit, 209: signal data acquisition unit, 210: learning data generation unit, 211: learning unit, 212: partial discharge determination unit, 213: noise removal Unit, 214: discharge type specifying unit, 215: deterioration specifying unit, 216: alarm unit, 300: network

Claims (8)

A sensor that generates a signal in response to a phenomenon caused by electric device discharge;
By performing machine learning on a signal generated during a predetermined period from the start of operation of the electric device, a learning model that has learned a signal including noise near the electric device, and a learning model generated after the predetermined period has elapsed. A partial discharge determination unit that determines whether or not a partial discharge has occurred in the electric device based on the signal.
A partial discharge detection system comprising: As a result of the determination, when it is determined that partial discharge has occurred in the electrical device,
A noise removing unit that removes the noise from the signal based on the learning model from the signal;
Based on the signal from which noise has been removed, a discharge type identification unit that identifies the type of the partial discharge,
The partial discharge detection system according to claim 1, further comprising: When the type of the specified partial discharge is a predetermined partial discharge, the apparatus further includes a state specifying unit that specifies a state of performance degradation of the electric device based on a waveform of the signal from which noise has been removed. ,
The partial discharge detection system according to claim 2. As a result of the determination, when it is determined that a partial discharge has occurred in the electric device, an alarm unit that performs an alarm process for notifying that a partial discharge has occurred is further provided. The partial discharge detection system according to claim 1. A signal acquisition unit that acquires a signal generated in accordance with a phenomenon caused by electric device discharge,
A learning unit that generates a learning model that learns a signal including noise near the electric device by performing machine learning on a signal generated during a predetermined period from the start of operation of the electric device.
A learning system comprising: A generating step of generating a signal according to a phenomenon caused by a discharge of the electric device,
By performing machine learning on a signal generated during a predetermined period from the start of operation of the electric device, a learning model that has learned a signal including noise near the electric device, and a learning model generated after the predetermined period has elapsed. A partial discharge determination step of determining whether or not a partial discharge has occurred in the electrical device based on the received signal,
A partial discharge detection method comprising: A generating step of generating a signal in response to a phenomenon caused by a discharge of the electric device,
By performing machine learning on a signal generated during a predetermined period from the start of operation of the electric device, a learning model that has learned a signal including noise near the electric device, and a learning model generated after the predetermined period has elapsed. A partial discharge determination step of determining whether or not a partial discharge has occurred in the electrical device based on the received signal,
A computer program for causing a computer to execute. A sensor that generates a signal in response to a phenomenon caused by electric device discharge;
By performing machine learning on a signal generated during a predetermined period from the start of operation of the electric device, a learning model that has learned a signal including noise near the electric device, and a learning model generated after the predetermined period has elapsed. A partial discharge determination unit that determines whether or not a partial discharge has occurred in the electric device based on the signal.
An electrical device comprising:

Images