Partial discharge diagnosis device for power apparatus and partial discharge diagnosis method for power apparatus

Abstract

This partial discharge diagnosis device (1) comprises: a signal acquisition unit (11) that acquires a signal detected by a detector (2) of a power apparatus (100); a signal separation unit (121) that separates the signal acquired by the signal acquisition unit (11) into a signal for each of a plurality of signal sources; a feature amount extraction unit (122) that extracts a feature amount of the signal for each signal source separated by the signal separation unit (121); and a partial discharge factor determination unit (123) that detects the occurrence of partial discharge in the power apparatus (100) on the basis of the feature amount extracted by the feature amount extraction unit (122), and determines a factor of the occurrence of the partial discharge when the partial discharge occurs.

Description

Partial discharge diagnostic device for electric power equipment and partial discharge diagnostic method for electric power equipment

[0001] The present disclosure relates to a partial discharge diagnostic device for electric power equipment and a partial discharge diagnostic method for electric power equipment.

[0002] In electric power equipment, when the insulation performance deteriorates due to the deterioration of solid insulating components constituting the equipment or the generation of metallic foreign matter during operation of the equipment, partial discharges may occur due to localized insulation breakdown. This partial discharge is a precursor to failure, and monitoring this partial discharge can prevent failure of electric power equipment. In recent years, machine learning has been used in partial discharge detection devices that diagnose the causes of partial discharges in electric power equipment (see Patent Document 1 below).

[0003] Japanese Patent Application Laid-Open No. 2023-180397

[0004] However, there is a problem that partial discharge signals are mixed due to the influence of signals from multiple partial discharge sources or external noise in the entire substation or in specific power equipment, making it difficult to determine the cause of partial discharge.

[0005] The present disclosure discloses a technique for solving the above-described problems, and provides a partial discharge diagnostic device for electric power equipment and a partial discharge diagnostic method for electric power equipment that can accurately determine the cause of partial discharge even when multiple partial discharge sources exist within the electric power equipment.

[0006] A partial discharge diagnosis device for electric power equipment according to the present disclosure comprises: a signal acquisition unit that acquires a signal detected by a detector attached to the electric power equipment; a signal separation unit that separates the signal acquired by the signal acquisition unit into a plurality of signals for each signal source; a feature extraction unit that extracts feature quantities of the signals for each signal source separated by the signal separation unit; and a partial discharge cause determination unit that detects the occurrence of a partial discharge in the electric power equipment based on the feature quantities extracted by the feature extraction unit, and determines the cause of the partial discharge if a partial discharge has occurred. Also, a partial discharge diagnosis method for electric power equipment according to the present disclosure comprises: a signal acquisition step that acquires a signal detected by a detector attached to the electric power equipment; a signal separation step that separates the signal acquired in the signal acquisition step into a plurality of signals for each signal source; a feature extraction step that extracts feature quantities of the signals for each signal source separated in the signal separation step; and a partial discharge cause determination step that detects the occurrence of a partial discharge in the electric power equipment based on the feature quantities extracted in the feature extraction step, and determines the cause of the partial discharge if a partial discharge has occurred.

[0007] According to the partial discharge diagnostic device and partial discharge diagnostic method for electric power equipment of the present disclosure, it is possible to determine the cause of partial discharge with high accuracy even when multiple partial discharge sources exist in the electric power equipment.

[0008] FIG. 1 is a block diagram showing a partial discharge diagnostic device for electric power equipment according to embodiment 1. FIG. 2 is a block diagram showing a signal acquisition unit of the partial discharge diagnostic device according to embodiment 1. FIG. 3 is a block diagram showing a signal separation unit of the partial discharge diagnostic device according to embodiment 1. FIG. 4 shows a specific example of a signal waveform of partial discharge according to embodiment 1. FIG. 5 shows a specific example of a signal waveform of partial discharge according to embodiment 1. FIG. 6 shows a specific example of a signal waveform of partial discharge according to embodiment 1. FIG. 7 shows a partial discharge waveform obtained by mixing a plurality of signal sources and input to an encoder unit according to embodiment 1. FIG. 8 is a diagram for explaining the arithmetic processing of the encoder unit according to embodiment 1. FIG. 9 is a diagram for explaining the arithmetic processing of the separator unit according to embodiment 1. FIG. 10 is a diagram for explaining the arithmetic processing of the decoder unit according to embodiment 1. FIG. 11 is an explanatory diagram of a φ-q-n pattern shape of partial discharge according to embodiment 1. FIG. 12 is an explanatory diagram of a φ-q-n pattern shape of partial discharge according to embodiment 1. FIG. 13 is an explanatory diagram of a φ-q-n pattern shape of partial discharge according to embodiment 1. FIG. 14 is a diagram showing an example of a display screen of a determination result display unit according to embodiment 1. Fig. 1 is a flowchart of a partial discharge diagnostic method for electric power equipment according to embodiment 1. Fig. 2 is an explanatory diagram of a dt-dq-n pattern shape of partial discharge according to embodiment 2. Fig. 3 is an explanatory diagram of a dt-dq-n pattern shape of partial discharge according to embodiment 2. Fig. 4 is an explanatory diagram of a dt-dq-n pattern shape of partial discharge according to embodiment 2. Fig. 5 is a diagram showing an example of hardware of a partial discharge diagnostic device according to embodiment 1 and embodiment 2.

[0009] Hereinafter, a partial discharge diagnostic device for electric power equipment according to an embodiment of the present disclosure will be described in detail with reference to the drawings. In addition, in the following embodiment, a gas-insulated switchgear will be described as an example of electric power equipment to be diagnosed for partial discharge, but the present disclosure can be applied to any electric power equipment in which partial discharge occurs.

[0010] Embodiment 1. Fig. 1 is a block diagram showing a partial discharge diagnostic device for electric power equipment according to embodiment 1. A gas-insulated switchgear 100 as electric power equipment according to this embodiment has a metal container 101 in which circuit breakers, disconnecting switches, and the like (not shown) are housed. The interior of the metal container 101 is filled with an insulating gas such as SF6 (sulfur hexafluoride gas). The gas-insulated switchgear 100 is electrically connected to an external transformer and the like (not shown), and the voltage applied to the gas-insulated switchgear 100 is referred to as the system voltage.

[0011] The detector 2 is a detector that detects signals inside the gas-insulated switchgear 100, and detects the signals inside the gas-insulated switchgear 100 by electrical or acoustic means. The detector 2 may be a detector that detects the AC voltage or AC current of the AC power input to the gas-insulated switchgear, or may be a detector that detects electromagnetic waves or acoustic waves. The detector 2 may also be arranged outside the gas-insulated switchgear 100.

[0012] The partial discharge diagnostic device 1 is electrically connected to the detector 2 and includes a signal acquisition unit 11, a signal processing unit 12, a storage unit 13, a communication unit 14, and a determination result display unit 15. The signal acquisition unit 11 acquires a signal detected by the detector 2 and converts it into digital information (signal information). The signal processing unit 12 determines the cause of partial discharge from the signal information acquired by the signal acquisition unit 11. The storage unit 13 stores partial discharge causes or external noise causes that may occur inside the power equipment, their signal information, an operation program for the signal processing unit 12, calculation results of the signal processing unit 12, etc. The communication unit 14 is means for communicating with external devices. The determination result display unit 15 is display means, such as a display, that displays the cause of partial discharge.

[0013] The signal processing unit 12 includes a signal separation unit 121, a feature extraction unit 122, a partial discharge cause determination unit 123, and a display control unit 124. The signal separation unit 121 refers to signal waveforms of partial discharge causes that may occur inside the power equipment or external noise causes that are recorded in the storage unit 13, and separates the signal acquired by the signal acquisition unit 11 into signals of multiple signal sources based on a machine learning algorithm that has been trained to have the function of separating the signals into signals for each signal source. Here, the machine learning algorithm is, for example, a neural network (NN). The feature extraction unit 122 calculates feature amounts for identifying partial discharge causes for the signals separated for each signal source by the signal separation unit 121. The partial discharge cause determination unit 123 determines the cause of partial discharge occurrence based on a comparison result between the feature amount calculated by the feature extraction unit 122 and a predetermined determination value. The display control unit 124 causes the determination result display unit 15 to display various information relating to the partial discharge occurrence cause determined by the partial discharge cause determination unit 123 .

[0014] FIG. 2 is a configuration diagram of the signal acquisition unit 11 according to the first embodiment. The signal acquisition unit 11 in the first embodiment includes an overvoltage protection circuit unit 111, an amplifier unit 112, a filter unit 113, and an AD (Analog-Digital) converter unit 114. The overvoltage protection circuit unit 111 removes overvoltage components contained in the signal input from the detector 2. The amplifier unit 112 amplifies the voltage pulse from which the overvoltage components have been removed. The filter unit 113 removes noise components contained in the voltage pulse and whose frequency is different from that of the system voltage. The AD converter unit 114 converts the voltage pulse from which the noise components have been removed from analog to digital. The order of the amplifier unit 112 and the filter unit 113 may be reversed.

[0015] Fig. 3 is a block diagram showing the configuration of a signal separation unit of the partial discharge diagnostic device according to embodiment 1. In Fig. 3, the signal separation unit 121 plays a role of separating a signal including partial discharge signals of multiple signal sources input from the signal acquisition unit 11 into signals including a partial discharge signal of a single signal source using a neural network, and includes an encoder unit 1211, a separation unit 1212, and a decoder unit 1213. Note that Fig. 3 shows a case where a signal including mixed partial discharge signals of two signal sources is separated into signals each including a partial discharge signal of a single signal source, but it is also possible to separate a signal including mixed partial discharge signals of three or more signal sources into signals each including a partial discharge signal of a single signal source.

[0016] Here, we will explain the causes of partial discharges that occur inside power equipment and their signal waveforms. Causes of partial discharges can be classified into, for example, conductor protrusions, foreign objects creeping on the surface of insulators, floating electrodes, and voids in insulators. A conductor protrusion is a state in which there is a protrusion on the surface of an electrode to which system voltage is applied. A foreign object creeping on the surface of an insulator is a state in which foreign objects are attached to the surface of an insulating material. A floating electrode is a state in which a metal member, such as a bolt used to secure the conductor of a gas switchgear, is not conductive to the conductor and has a floating potential. A void in an insulator is a state in which a gap exists within an insulating material.

[0017] Fig. 4 shows a partial discharge waveform W1 in the case of a conductor protrusion, Fig. 5 shows a partial discharge waveform W2 in the case of a foreign object creeping on the surface of an insulator, Fig. 6 shows a partial discharge waveform W3 in the case of a floating electrode, and Fig. 7 shows a partial discharge waveform W4 in the case of a gap in an insulator, which show specific examples of signal waveforms of partial discharges occurring in one cycle of the AC voltage Vac. These partial discharge waveforms and waveforms of external noise such as inverter noise are stored in advance for each signal source in the memory unit 13 of the partial discharge diagnostic device 1. Note that in this disclosure, partial discharge waveforms of external noise such as inverter noise are not shown.

[0018] Next, the operation of the signal separation unit 121 according to the first embodiment will be described in detail. The encoder unit 1211 of the signal separation unit 121 receives a partial discharge waveform (signal) resulting from a mixture of multiple signal sources output from the signal acquisition unit 11. The signal received by the encoder unit 1211 includes the partial discharge waveform resulting from a mixture of multiple signal sources and time information (sampling rate). As shown in FIG. 8 , the encoder unit 1211 divides the time-series data of length N corresponding to one cycle of the AC voltage Vac into segments of length L (N>L), and inputs data from each segment (A1, A2, A3, ...) of length L into a neural network. In FIG. 8 , Wcom represents the partial discharge waveform resulting from a mixture of multiple signal sources. As shown in FIG. 9 , the neural network generates an S×T-dimensional feature map by performing a convolution operation or the like on the input data via an intermediate layer (e.g., an X×Y-dimensional matrix). The S×T-dimensional feature map generated here is referred to as a first feature map TM1. The number of intermediate layers may be plural.

[0019] 10 , the separation unit 1212 of the signal separation unit 121 adds, for example, a learned matrix M1 for separating signal source 1 or a learned matrix M2 for separating signal source 2 to the S×T-dimensional feature map (first feature map TM1) generated by the encoder unit 1211, or performs a similar matrix operation using multiplication to generate an S×T-dimensional feature map, which is a matrix M3 after separation of signal source 1, or an S×T-dimensional feature map, which is a matrix M4 after separation of signal source 2. In this case, a learned matrix (S×T-dimensional feature map) for separating external noise such as inverter noise is prepared, and the matrix for separating external noise is subtracted. The generated S×T-dimensional feature map for signal source 1 is called a second feature map TM21, and the generated S×T-dimensional feature map for signal source 2 is called a second feature map TM22.

[0020] As shown in FIG. 11 , the decoder unit 1213 of the signal separation unit 121 converts the S×T-dimensional feature maps (second feature maps TM21, TM22) generated by the separation unit 1212 after signal source separation into time-series data of length N by performing the inverse operation of that performed by the encoder unit 1211, and generates a partial discharge waveform of a single signal source.

[0021] The signals separated for each signal source (partial discharge waveform of a single signal source) contain time information and can be compared with the voltage waveform of each phase of the system voltage (AC voltage). As described above, by using a neural network for the calculation process of the signal separation unit 121, calculations such as signal frequency analysis are not required for the determination process, thereby reducing calculation costs. Furthermore, even if the same partial discharge cause occurs at multiple locations, it is possible to separate each signal source, except when the partial discharge waveforms are exactly the same.

[0022] Next, the operation of the feature extraction unit 122 according to the first embodiment will be described. Here, a method using a φ-q-n pattern will be described as a specific example of a method for calculating the feature amounts of signals separated for each signal source in the feature extraction unit 122. The occurrence characteristics of partial discharges are correlated with the cause of their occurrence. The intensity of a partial discharge pulse is q, the occurrence frequency is n, and the phase of the system voltage at the timing when a partial discharge pulse occurs is φ. The cause of partial discharge occurrence is determined from the shape of the φ-q-n pattern.

[0023] Figures 12 to 15 are explanatory diagrams illustrating examples of partial discharge φ-q-n patterns. Here, the partial discharge φ-q-n patterns are depicted using data for approximately 20 AC cycles. In Figures 12 to 15, the sine wave represents the system voltage Vac, and the black dots represent the frequency n of partial discharge occurrence. Note that the darker the black dots, the higher the frequency of occurrence. In the system voltage Vac waveforms shown in Figures 12 to 16, the positions of the maximum and minimum values ​​of the waveform are called peaks, and the positions of the waveform where the voltage becomes zero are called zero crossings. As mentioned above, causes of partial discharges can be classified into, for example, conductor protrusions, foreign objects on the surface of insulators, floating electrodes, and voids in insulators. When conditions such as conductor protrusions, foreign objects on the surface of insulators, floating electrodes, and voids in insulators occur, the electric field concentrates in those areas, becoming the starting point of partial discharges.

[0024] As shown in Figures 12 to 15, the φ-q-n pattern of partial discharges varies depending on the cause of the partial discharges. Figure 12 shows a φ-q-n pattern P1 of partial discharges when the cause of partial discharges is a conductor protrusion. As shown in Figure 12, when a conductor protrusion occurs, the partial discharges are widely distributed around the positive peak of the system voltage Vac. Figure 13 shows a φ-q-n pattern P2 of partial discharges when the cause of partial discharges is a foreign object creeping on the surface of an insulator. As shown in Figure 13, when a foreign object creeps on the surface of an insulator, high-intensity partial discharge pulses occur around the zero crossing of the system voltage Vac and its peak. Figure 14 shows a φ-q-n pattern P3 of partial discharges when the cause of partial discharges is a floating electrode. As shown in Figure 14, when a floating electrode occurs, high-intensity partial discharges occur at the zero crossing of the system voltage Vac. Fig. 15 shows a φ-q-n pattern P4 of partial discharge when the cause of partial discharge is a gap in the insulator. As shown in Fig. 15, when a gap occurs in the insulator, partial discharge occurs in a phase ranging from the zero cross position to the peak of the system voltage Vac. The feature extraction unit 122 converts the signals separated for each signal source by the signal separation unit 121 into a φ-q-n pattern.

[0025] Next, the operation of the partial discharge cause determination unit 123 according to the first embodiment will be described. The partial discharge cause determination unit 123 recognizes the φ-q-n pattern shape of the partial discharge converted by the feature extraction unit 122 as an image shape, and determines the cause of the partial discharge using artificial intelligence technology. Specifically, a determination method using a machine learning algorithm such as a neural network or a decision tree (DT) can be used. By using, for example, a neural network to determine the cause of the partial discharge, the cause of the partial discharge can be estimated with high accuracy. Furthermore, if the partial discharge cause determination unit 123 is trained to learn shape patterns caused by noise, it can separate noise signals. Furthermore, shape patterns that do not match the learned shape patterns are recognized as white noise. The number of partial discharge occurrences can be calculated by comparing the number of signals separated by the signal separation unit 121 with the number of signals determined as noise signals by the partial discharge cause determination unit 123. This configuration can improve the accuracy of determining the cause of partial discharge.

[0026] In the above explanation, the partial discharge cause determination unit 123 recognizes the φ-q-n pattern shape of the partial discharge converted by the feature extraction unit 122 as an image shape and determines the cause of the partial discharge using artificial intelligence technology. However, this is not limiting, and the pulse intensity of the partial discharge waveform for each signal source separated by the signal separation unit 121 may be represented as a histogram, and the cause of the partial discharge may be determined based on the kurtosis, skewness, etc.

[0027] FIG. 16 is a diagram showing an example of a display screen of the judgment result display unit 15 according to the first embodiment. The judgment result display unit 15 includes a final judgment result display unit 151, a measurement result display unit 152, and judgment result display units 153A and 153B for each signal source. The final judgment result display unit 151 displays the partial discharge cause output by the partial discharge cause determination unit 123. The measurement result display unit 152 displays signal information acquired by the signal acquisition unit 11. The judgment result display units 153A and 153B for each signal source include waveform display units 1531A and 1531B for each signal source that display waveforms separated for each signal source, and diagnosis result display units 1532A and 1532B for each signal source that display the diagnosis results. The diagnosis result display units 1532A and 1532B may be configured to display the judgment results on an inspector's terminal instead of a dedicated display. Note that in FIG. 16 , "void" means an air gap in an insulator. Furthermore, trees are a type of deterioration such as scratches or burrs in solid insulation, and this disclosure does not show examples of partial discharge waveforms. The partial discharge diagnostic device configured in this way can determine the cause of partial discharge when partial discharge signals from multiple signal sources are mixed.

[0028] In the above description, an example has been shown in which the electric power equipment includes an AC voltage application unit to which single-phase AC power (AC voltage Vac) is applied, and the partial discharge cause determination unit 123 detects the occurrence of a single-phase partial discharge in the AC voltage application unit, and determines the cause of the partial discharge if a partial discharge occurs. However, the electric power equipment may also include an AC voltage application unit to which multi-phase AC power is applied, and the partial discharge cause determination unit 123 may detect the occurrence of multi-phase partial discharge in the AC voltage application unit, and determine the cause of the partial discharge if a partial discharge occurs. Furthermore, as will be described in detail in a second embodiment below, the electric power equipment may also include a DC voltage application unit to which DC power is applied, and the partial discharge cause determination unit 123 may detect the occurrence of a partial discharge in the DC voltage application unit, and determine the cause of the partial discharge if a partial discharge occurs.

[0029] In this embodiment, an example has been described in which the detector 2 attached to the power equipment and the partial discharge diagnostic device 1 are electrically connected by a cable or the like, but the signal detected by the detector 2 may be AD converted and transmitted via wireless or other communication or a network to the partial discharge diagnostic device 1 which is not electrically connected to the detector 2. Furthermore, the determination result may be transmitted via wireless or other communication to a determination result display unit 15 provided outside the partial discharge diagnostic device 1 and displayed thereon.

[0030] As described above, the partial discharge diagnosis device for electric power equipment of embodiment 1 includes a signal acquisition unit that acquires a signal detected by a detector attached to the electric power equipment, a signal separation unit that separates the signal acquired by the signal acquisition unit into signals for each of a plurality of signal sources, a feature extraction unit that extracts features of the signals for each of the signal sources separated by the signal separation unit, and a partial discharge cause determination unit that detects the occurrence of partial discharge in the electric power equipment based on the features extracted by the feature extraction unit, and determines the cause of the partial discharge if a partial discharge has occurred.Therefore, even if a plurality of partial discharge sources exist within the electric power equipment, the cause of the partial discharge can be determined with high accuracy.

[0031] Furthermore, since the signal separation unit separates a mixed signal of multiple signal sources into signals for each of the multiple signal sources using a neural network, there is no need to use frequency analysis for signal processing, thereby reducing calculation costs. Furthermore, since frequency analysis is not used, even if there are multiple signal sources in the same frequency band, it is possible to separate each signal source.

[0032] Furthermore, the signal separation unit includes an encoder unit that generates a first feature map from a mixed signal of multiple signal sources acquired by the signal acquisition unit using a neural network, a separation unit that separates the first feature map generated by the encoder unit into second feature maps separated for each of the multiple signals, and a decoder unit that generates signals separated for each of the multiple signal sources based on the second feature maps separated by the separation unit, thereby eliminating the need for frequency analysis in signal processing and reducing calculation costs.Furthermore, because frequency analysis is not used, separation for each signal source is possible even when there are multiple signal sources in the same frequency band.

[0033] Furthermore, the signals acquired by the signal acquisition unit include time information, and the signals for each of the multiple signal sources separated by the signal separation unit also include time information. Therefore, the signals can be separated for each partial discharge source and reconstructed as time waveforms, which can be used to analyze the causes of partial discharge.

[0034] Furthermore, the signal separation unit separates external noise from the signal acquired by the signal acquisition unit, so that the cause of partial discharge can be determined with high accuracy by separating the external noise from the signal waveform of partial discharge.

[0035] Furthermore, the electric power equipment includes an AC voltage application unit to which single-phase or multi-phase AC power is applied, and the partial discharge cause determination unit detects the occurrence of single-phase or multi-phase partial discharge in the AC voltage application unit, and determines the cause of the partial discharge if a partial discharge occurs. Therefore, single-phase or multi-phase partial discharge can be diagnosed with a single device, thereby reducing costs.

[0036] Furthermore, the electric power equipment includes a DC voltage application unit to which DC power is applied, and the partial discharge cause determination unit detects the occurrence of partial discharge in the DC voltage application unit, and if partial discharge occurs, determines the cause of the partial discharge, so that it is also possible to diagnose partial discharge in DC electric power equipment.

[0037] Furthermore, since the detector of the electric power equipment and the signal acquisition unit are connected by communication, it is possible to diagnose partial discharge in the electric power equipment even if the electric power equipment and the partial discharge diagnosis device are separated from each other.

[0038] Furthermore, the partial discharge cause determination unit is connected wirelessly or by wire to a determination result display unit that displays the determination result of the partial discharge cause, so that the determination result of the partial discharge cause can be notified to a user or the like.

[0039] FIG. 17 is a flowchart of the partial discharge diagnosis method for electric power equipment according to the first embodiment. In FIG. 17, step S10 is a signal acquisition step for acquiring a signal detected by a detector 2 attached to the electric power equipment. Step S11 is a signal separation step for separating the signal acquired in the signal acquisition step S10 into signals for each of a plurality of signal sources. Step S12 is a feature extraction step for extracting feature quantities of the signals for each of the signal sources separated in the signal separation step S11. Step S13 is a partial discharge cause determination step for detecting the occurrence of partial discharge in the electric power equipment based on the feature quantities extracted in the feature extraction step S12, and determining the cause of the partial discharge if a partial discharge has occurred. The details of steps S10 to S13 are as described above.

[0040] As described above, according to the partial discharge diagnosis method for electric power equipment of the first embodiment, even when a plurality of partial discharge sources exist in the electric power equipment, the cause of the partial discharge can be determined with high accuracy.

[0041] Embodiment 2. A partial discharge diagnostic device according to embodiment 2 will be described, focusing on the differences from embodiment 1. The partial discharge diagnostic device according to embodiment 2 is applied to electric power equipment 100 in which a DC voltage is applied to at least one of the voltage application units. The electric power equipment 100 according to embodiment 2 is assumed to be, for example, an HVDC (High Voltage Direct Current) power converter or a DC-GIS (Gas Insulated Switchgear for Direct Current). The configuration of the electric power equipment partial discharge diagnostic device according to embodiment 2 is the same as the configuration of embodiment 1 shown in FIGS. 1 and 2.

[0042] In the second embodiment, the feature extractor 122 of the partial discharge diagnostic device 1 calculates the feature of the signal separated for each signal source using, for example, a dt-dq-n pattern. Here, the time interval between partial discharge occurrences is dt, the difference in intensity between adjacent partial discharge pulses is dq, and the occurrence frequency is n. The cause of partial discharge occurrence is determined based on the shape of the dt-dq-n pattern.

[0043] 18 to 20 are explanatory diagrams of examples of partial discharge dt-dq-n pattern shapes. Here, for example, 10 seconds of data is used to depict the partial discharge dt-dq-n pattern shapes. In Figs. 18 to 20, the horizontal axis represents the time interval dt at which partial discharges occur, the vertical axis represents the difference in the discharge charge amount of partial discharges, and the black dots represent the frequency of occurrence. Note that the darker the black dots, the higher the frequency of occurrence.

[0044] In the DC voltage application section of the electric power device 100, as in the AC voltage application section, partial discharges occur due to factors such as conductor protrusions, foreign objects on the surface of the insulator, and floating electrodes. Figure 18 shows the dt-dq-n pattern shape when the partial discharge is caused by a conductor protrusion and the voltage polarity is positive. As shown in Figure 18, when a conductor protrusion occurs, the time interval between discharges is long and the difference in discharge charge amount is small, resulting in a horizontally elongated dt-dq-n pattern shape. Figure 19 shows the dt-dq-n pattern shape when the partial discharge is caused by a foreign object on the surface of the insulator and the voltage polarity is positive. As shown in Figure 19, when a foreign object on the surface of the insulator occurs, the dt-dq-n pattern shape is approximately circular. Figure 20 shows the dt-dq-n pattern shape when the partial discharge is caused by a floating electrode and the voltage polarity is positive. As shown in FIG. 20, when a floating electrode occurs, the time interval between discharges is short and the difference in the amount of discharge charge is large, so the pattern shape of dt-dq-n becomes gourd-shaped.

[0045] In the above description, the partial discharge diagnostic device 1 is described as being installed in an electric power device 100 in which a DC voltage is applied to at least one of the voltage application parts, but it can also be applied in cases in which a DC voltage is applied to all of the voltage application parts of the electric power device 100.

[0046] The partial discharge diagnostic device 1 according to the first and second embodiments includes a processor 1000 and a storage device 1010, as shown in FIG. 18 , which illustrates an example of hardware. The storage device 1010 includes a volatile storage device such as a random access memory and a nonvolatile storage device such as a flash memory, both of which are not shown. The storage unit 13 shown in FIG. 1 is included in the nonvolatile storage device of the storage device 1010. Alternatively, a hard disk storage device may be provided instead of the flash memory. The processor 1000 executes a program input from the storage device 1010. In this case, the program is input to the processor 1000 from the nonvolatile storage device via the volatile storage device. The processor 1000 may output data such as calculation results to the volatile storage device of the storage device 1010, or may store the data in the nonvolatile storage device via the volatile storage device.

[0047] Although exemplary embodiments are described in the present disclosure, the various features, aspects, and functions described in the embodiments are not limited to the application of a particular embodiment, but may be applied to the embodiments alone or in various combinations. Therefore, countless variations not illustrated are anticipated within the scope of the technology disclosed in this specification. For example, variations in, addition to, or omission of at least one component are included.

[0048] Various aspects of the present disclosure are summarized below as appendices.

[0049] (Supplementary Note 1) A partial discharge diagnostic device for electric power equipment, comprising: a signal acquisition unit that acquires a signal detected by a detector attached to electric power equipment, a signal separation unit that separates the signal acquired by the signal acquisition unit into signals for each of a plurality of signal sources, a feature extraction unit that extracts feature quantities of the signals for each of the signal sources separated by the signal separation unit, and a partial discharge cause determination unit that detects the occurrence of partial discharge in the electric power equipment based on the feature quantities extracted by the feature extraction unit, and determines a cause of the partial discharge if a partial discharge has occurred. (Supplementary Note 2) The partial discharge diagnostic device for electric power equipment according to Supplementary Note 1, wherein the signal separation unit separates a mixed signal of a plurality of signal sources into signals for each of the plurality of signal sources using a neural network. (Supplementary Note 3) The partial discharge diagnosis device for electric power equipment according to Supplementary Note 2, wherein the signal separation unit includes: an encoder unit that generates a first feature amount map from a mixed signal of a plurality of signal sources acquired by the signal acquisition unit using a neural network; a separation unit that separates the first feature amount map generated by the encoder unit into second feature amount maps separated for each of the plurality of signals; and a decoder unit that generates signals separated for each of the plurality of signal sources based on the second feature amount maps separated by the separation unit. (Supplementary Note 4) The partial discharge diagnosis device for electric power equipment according to any one of Supplementary Note 1 to Supplementary Note 3, wherein the signals acquired by the signal acquisition unit include time information, and the signals for each of the plurality of signal sources separated by the signal separation unit also include the time information. (Supplementary Note 5) The partial discharge diagnosis device for electric power equipment according to any one of Supplementary Note 1 to Supplementary Note 4, wherein the signal separation unit separates external noise from the signals acquired by the signal acquisition unit. (Supplementary Note 6) The partial discharge diagnostic device for electric power equipment according to any one of Supplementary Note 1 to Supplementary Note 5, wherein the electric power equipment includes an AC voltage application unit to which single-phase or multi-phase AC power is applied, and the partial discharge cause determination unit detects occurrence of single-phase or multi-phase partial discharge in the AC voltage application unit, and, if partial discharge has occurred, determines the cause of the occurrence of the partial discharge.(Supplementary Note 7) The partial discharge diagnostic device for electric power equipment according to any one of Supplementary Note 1 to Supplementary Note 5, wherein the electric power equipment includes a DC voltage application unit to which DC power is applied, and the partial discharge cause determination unit detects occurrence of partial discharge in the DC voltage application unit and, if partial discharge has occurred, determines the cause of the partial discharge. (Supplementary Note 8) The partial discharge diagnostic device for electric power equipment according to any one of Supplementary Note 1 to Supplementary Note 7, wherein the detector of the electric power equipment and the signal acquisition unit are connected by communication. (Supplementary Note 9) The partial discharge diagnostic device for electric power equipment according to any one of Supplementary Note 1 to Supplementary Note 8, wherein the partial discharge cause determination unit is connected wirelessly or wired to a determination result display unit that displays the determination result of the cause of the partial discharge. (Supplementary Note 10) A partial discharge diagnosis method for electric power equipment, comprising: a signal acquisition step of acquiring a signal detected by a detector attached to electric power equipment; a signal separation step of separating the signal acquired by the signal acquisition step into signals for each of a plurality of signal sources; a feature extraction step of extracting features of the signals for each of the signal sources separated by the signal separation step; and a partial discharge cause determination step of detecting the occurrence of partial discharge in the electric power equipment based on the feature extracted by the feature extraction step, and determining the cause of the partial discharge if partial discharge has occurred.

[0050] 1 Partial discharge diagnostic device, 2 Detector, 11 Signal acquisition unit, 12 Signal processing unit, 13 Memory unit, 14 Communication unit, 15 Determination result display unit, 100 Power equipment, 101 Metal container, 121 Signal separation unit, 122 Feature extraction unit, 123 Partial discharge cause determination unit, 124 Display control unit, 111 Overvoltage protection circuit unit, 112 Amplification unit, 113 Filter unit, 114 AD conversion unit, 1211 Encoder unit, 1212 Separation unit, 1213 Decoder unit, 1000 Processor, 1010 Storage device.

Claims

1. A partial discharge diagnostic device for electric power equipment comprising: a signal acquisition unit that acquires signals detected by a detector attached to electric power equipment; a signal separation unit that separates the signals acquired by the signal acquisition unit into signals for each of a plurality of signal sources; a feature extraction unit that extracts features of the signals for each of the signal sources separated by the signal separation unit; and a partial discharge cause determination unit that detects the occurrence of partial discharge in the electric power equipment based on the features extracted by the feature extraction unit, and, if partial discharge has occurred, determines the cause of the partial discharge.

2. The partial discharge diagnostic device for electric power equipment according to claim 1, wherein the signal separation unit separates a mixed signal from a plurality of signal sources into signals for each of the plurality of signal sources using a neural network.

3. The partial discharge diagnostic device for electric power equipment according to claim 2, wherein the signal separation unit comprises: an encoder unit that uses a neural network to generate a first feature map from the mixed signals of the multiple signal sources acquired by the signal acquisition unit; a separation unit that separates the first feature map generated by the encoder unit into second feature maps separated for each of the multiple signals; and a decoder unit that generates signals separated for each of the multiple signal sources based on the second feature maps separated by the separation unit.

4. A partial discharge diagnostic device for electric power equipment according to any one of claims 1 to 3, wherein the signal acquired by the signal acquisition unit includes time information, and the signals for each of the multiple signal sources separated by the signal separation unit include time information.

5. The partial discharge diagnostic device for electric power equipment according to any one of claims 1 to 4, wherein the signal separation unit separates external noise from the signal acquired by the signal acquisition unit.

6. A partial discharge diagnostic device for electric power equipment according to any one of claims 1 to 5, wherein the electric power equipment includes an AC voltage application unit to which single-phase or multi-phase AC power is applied, and the partial discharge cause determination unit detects the occurrence of single-phase or multi-phase partial discharge in the AC voltage application unit, and, if partial discharge has occurred, determines the cause of the partial discharge.

7. A partial discharge diagnostic device for electric power equipment according to any one of claims 1 to 5, wherein the electric power equipment includes a DC voltage application unit to which DC power is applied, and the partial discharge cause determination unit detects the occurrence of partial discharge in the DC voltage application unit and, if partial discharge has occurred, determines the cause of the partial discharge.

8. The partial discharge diagnostic device for electric power equipment according to any one of claims 1 to 7, wherein the detector of the electric power equipment and the signal acquisition unit are connected by communication.

9. A partial discharge diagnostic device for electric power equipment according to any one of claims 1 to 8, wherein the partial discharge cause determination unit is connected wirelessly or by wire to a determination result display unit that displays the determination result of the cause of the partial discharge.

10. A partial discharge diagnosis method for electric power equipment, comprising: a signal acquisition step of acquiring a signal detected by a detector attached to electric power equipment; a signal separation step of separating the signal acquired in the signal acquisition step into signals for each of a plurality of signal sources; a feature extraction step of extracting features of the signals for each of the signal sources separated in the signal separation step; and a partial discharge cause determination step of detecting the occurrence of partial discharge in the electric power equipment based on the features extracted in the feature extraction step, and determining the cause of the partial discharge if partial discharge has occurred.

Images