Discharge detection device, discharge detection method, and program
The discharge determination device and method enhance discharge type identification by using a combination of past signals and a discharge type determination model to reduce errors and improve accuracy in identifying partial discharge types, especially for unlearned discharges.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- KK TOSHIBA
- Filing Date
- 2021-12-27
- Publication Date
- 2026-06-29
AI Technical Summary
Existing discharge determination methods using machine learning, such as neural networks, suffer from low accuracy in determining the type of partial discharge for unlearned discharges and have a high likelihood of misclassification.
A discharge determination device and method that includes a partial discharge type determination unit and a partial discharge signal determination unit, which utilize a combination of past partial discharge signals and a discharge type determination model, such as a support vector machine, to accurately identify the type of discharge and reduce errors by identifying difficult-to-determine signals.
The proposed solution significantly reduces the probability of errors in discharge type determination by accurately identifying and flagging difficult-to-determine partial discharge signals, thereby improving overall accuracy.
Smart Images
Figure 0007881308000002 
Figure 0007881308000003 
Figure 0007881308000004
Abstract
Description
Technical Field
[0001] Embodiments of the present invention relate to a discharge determination device, a discharge determination method, and a program.
Background Art
[0002] When partial discharge occurs in electrical equipment and power equipment, the solid insulator may be eroded and cause a failure. Therefore, there is a need for a technique for detecting partial discharge and a technique for determining the type of partial discharge (void discharge, peeling discharge, surface discharge, etc.).
[0003] For example, a technique for determining the type of discharge by causing neural networks to learn frequency characteristic data and phase characteristic data has been proposed. However, when determining the type of partial discharge by machine learning such as neural networks, the type of partial discharge can be determined with high accuracy for the partial discharge being learned, but the accuracy of the determination for the partial discharge not being learned may be low. In addition, since a determination result is always output by the determination method using machine learning, there is a high possibility that an incorrect type of partial discharge will be output, and there is a possibility of misrecognizing the type of partial discharge.
Prior Art Documents
Patent Documents
[0004]
Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0005] The problem to be solved by the present invention is to provide a discharge determination device, a discharge determination method, and a program that can reduce the occurrence of determination errors in the type of discharge.
Means for Solving the Problems
[0006] The discharge determination device of the embodiment includes a partial discharge type determination unit and a partial discharge signal determination unit that is difficult to determine. The partial discharge type determination unit determines the type of partial discharge from a target signal, which is the signal of the partial discharge to be determined, based on a combination of past partial discharge signals and the type of partial discharge in question. The determination unit determines whether or not to perform the determination by the partial discharge type determination unit by comparing the target signal with past partial discharge signals related to the combination. [Brief explanation of the drawing]
[0007] [Figure 1] A diagram showing the configuration of the discharge determination system 100 according to the first embodiment. [Figure 2] A functional block diagram showing the functional configuration of the discharge determination system 100 of the first embodiment. [Figure 3] An example of a Φ-qn pattern. [Figure 4] A flowchart illustrating the operation of the discharge detection device 103. [Figure 5] A table showing the results of Experiment 1 using the discharge detection device 103. [Figure 6] A table showing the results of Experiment 2 using the discharge detection device 103. [Modes for carrying out the invention]
[0008] The discharge determination method of the embodiment will be described below with reference to the drawings. (First embodiment) Figure 1 shows the configuration of the discharge determination system 100 according to the first embodiment. The discharge detection system 100 consists of one or more electrodes 101, a signal line 102, and a discharge detection device 103. The discharge detection device 103 in the first embodiment consists of a single device. The discharge detection system 100 is installed near a switchboard 10 that houses electrical equipment. The discharge detection system 100 detects partial discharge generated from the electrical equipment using the electrodes 101. The switchboard 10 is provided with electrodes 101. Electrical equipment is housed inside the switchboard 10. The electrical equipment consists of devices such as circuit breakers, disconnectors, current transformers, or transformers (none of which are shown). High voltage and high current are supplied to the electrical equipment from the outside via power cables. The electrical equipment has a function to cut off the power supply in the event of an abnormality. A grounding electrode is connected to the bottom of the switchboard 10. The electrical equipment can be any equipment that has the potential to generate partial discharge, such as power transformers, gas-insulated switches, generators, motors, or reactors.
[0009] The discharge detection device 103 does not have to be located near the power distribution panel 10. For example, the electrode 101 and the discharge detection device 103 may be connected via a network, and the discharge detection device 103 may not have to be located near the power distribution panel 10. Alternatively, the measurement results from the electrode 101 may be stored in the cloud, and the discharge detection device 103 may acquire this data to obtain the measurement results from the electrode 101.
[0010] Figure 2 is a functional block diagram showing the functional configuration of the discharge determination system 100 of the first embodiment. As described in Figure 1, the discharge determination system 100 consists of an electrode 101, a signal line 102, and a discharge determination device 103.
[0011] Electrode 101 generates an electrical signal in response to phenomena caused by the operation of electrical equipment. These phenomena include, for example, electric current, electromagnetic waves, vibrations, sound waves, or vibrations. The electrical signal represents a physical quantity obtained from these phenomena. Electrode 101 is installed near the electrical equipment, such as on its outer wall. Electrode 101 is connected to the discharge detection device 103 via signal line 102. Electrode 101 outputs the generated electrical signal to the discharge detection device 103. Electrode 101 may also be a sensor capable of measuring electric current, electromagnetic waves, sound waves, or vibrations originating from the operation of electrical equipment, such as a CT (Current Transformer) sensor, a TEV (Transient Earth Voltage) sensor, an AE (Acoustic Emission) sensor, or an antenna. Electrode 101 may use one type of sensor from the above-mentioned sensors, or a combination of multiple types.
[0012] The signal line 102 inputs the electrical signal generated by the electrode 101 to the discharge determination device 103. The device consists of one or more electrodes 101, the signal line 102, and the discharge determination device 103. In the first embodiment, the discharge determination device 103 consists of a single device. The electrical signal generated by the electrode 101 is converted AD and processed by the discharge determination device 103, but the electrical signal may also be converted AD by an AD converter provided on the signal line 102, or the electrical signal input to the discharge determination device 103 may be converted AD by an AD converter provided on the discharge determination device 103.
[0013] The discharge determination device 103 is an information processing device such as a personal computer, smartphone, or tablet computer. The discharge determination device 103 determines the type of partial discharge based on the electrical signal output from the electrode 101. The discharge determination device 103 may randomly extract multiple intervals corresponding to one cycle of the applied voltage from the electrode 101 within a predetermined period, and determine the type of partial discharge based on the electrical signals in those intervals.
[0014] 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, for example, 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, for example, an information processing device such as a personal computer or a server, or may be a cloud computing system.
[0015] 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 discharge determination system 100. In this case, the input unit 105 generates input data (for example, instruction information indicating an instruction to the discharge determination system 100) from an input signal input in the input device and inputs it to the discharge determination system 100.
[0016] The display unit 106 is an output device such as a CRT (Cathode Ray Tube) display, a liquid crystal display, or an organic EL (Electro Luminescence) display. The display unit 106 may be an interface for connecting an output device to the discharge determination system 100. In this case, the display unit 106 generates a video signal from video data and outputs the video signal to a video output device connected to itself.
[0017] 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 electrical signal was acquired and the physical quantity represented by the electrical signal. When the discharge determination system 100 includes a plurality of electrodes 101, the signal data may hold identification information of the electrodes 101. The identification information may be any information as long as it can identify the electrode 101.
[0018] The discharge signal type data storage unit 108 stores data on combinations of partial discharge signals and discharge signal types. These combinations are used for determination by the difficult-to-determine partial discharge signal determination unit 114, which will be described later. The combinations of partial discharge signals and discharge signal types are also used to create discharge type determination models, which will be stored in the discharge type determination model storage unit 109, which will be described later.
[0019] The discharge type determination model storage unit 109 stores the discharge type determination model. The discharge type determination model may be used for determination by the discharge type determination unit 115, which will be described later. The discharge type determination model is a model generated based on data of combinations of partial discharge signals and types of discharge signals. The discharge type determination model is a model generated by inputting data of combinations of partial discharge signals and types of discharge signals into a support vector machine (SVM). Specifically, the discharge type determination model is generated by a multi-class SVM composed of discriminators (e.g., void discriminator, peeling discriminator, creepage discriminator) that determine whether or not a partial discharge signal belongs to a class of partial discharge type (e.g., void discharge, peeling discharge, creepage discharge). Each discriminator determines whether or not the input partial discharge signal belongs to a class based on a hyperplane that separates regions belonging to a class from regions that do not belong to a class obtained through learning. The multi-class SVM determines which class a partial discharge signal belongs to by inputting the partial discharge signal into each discriminator. In the discharge type determination model, the void discriminator is trained by finding the hyperplane with the largest margin that separates partial discharge signals whose label indicating the type of partial discharge is void from other partial discharge signals. Similarly, the peel discriminator and creepage discriminator are trained. The hyperplanes of each discriminator are examples of parameters for the discharge type determination model. In this embodiment, the multiclass SVM is implemented using a one-vs-rest discriminator, but the multiclass SVM in other embodiments may be implemented using a one-vs-one discriminator.
[0020] The discharge type determination model may be represented by a method other than SVM, such as a neural network or a decision tree. When the discharge type determination model is represented by a neural network or a decision tree, it may be configured to discriminate multiple classes by one model. In this case, the discharge type determination model is learned to take a partial discharge signal as an input and output a vector representing the probability for each type of partial discharge.
[0021] The control unit 111 controls the operations of each part of the discharge determination device 103. The control unit 111 is executed by a device including a processor such as a CPU (Central Processing Unit) and a RAM (Random Access Memory). By executing a discharge determination program, the control unit 111 functions as an electrical signal acquisition unit 112, a partial discharge signal extraction unit 113, a determination-difficult partial discharge signal determination unit 114, and a discharge type determination unit 115.
[0022] The electrical signal acquisition unit 112 acquires the electrical signal input by the electrode 101. The electrical signal acquisition unit 112 associates the date and time when the electrical signal is acquired with the reception intensity represented by the electrical signal. When the discharge determination system 100 includes a plurality of electrodes 101, the electrical signal acquisition unit 112 associates the date and time when the electrical signal is acquired, the reception intensity represented by the electrical signal, and the identification information of the electrode 101. The electrical signal acquisition unit 112 may include a filter such as a high-pass filter or a low-pass filter to acquire a signal in a specific frequency band.
[0023] The partial discharge signal extraction unit 113 extracts a partial discharge signal from the acquired electrical signal. When the electrode 101 is a TEV sensor, the electrical signal generated by the occurrence of partial discharge has signal components in the 1 - 20 MHz band and the 100 MHz band. When the electrode 101 is a TEV sensor, the partial discharge signal extraction unit 113 extracts the signal in the 1 - 20 MHz band and the signal in the 100 MHz band from the electrical signal, and determines the signal included in both frequency bands as the partial discharge signal. The partial discharge signal extraction unit 113 may extract the partial discharge signal from the acquired electrical signal by removing noise using a band-pass filter or the like.
[0024] The discharge type determination unit 115 determines the type of discharge of a partial discharge signal based on the combination of the partial discharge signal and the type of partial discharge. The discharge type determination unit 115 determines the type of partial discharge by inputting the partial discharge signal into a discharge type determination model stored in the discharge type determination model storage unit 109, for example, and outputting the type of partial discharge.
[0025] The difficult-to-determine partial discharge signal determination unit 114 determines whether the partial discharge signal extracted by the partial discharge signal extraction unit 113 is a partial discharge signal that is difficult to determine due to low accuracy in determining the type of discharge by the discharge type determination unit 115 (a difficult-to-determine partial discharge signal). The difficult-to-determine partial discharge signal determination unit 114 is an example of a determination unit that determines whether or not to perform a determination by the discharge type determination unit 115 by comparing the target signal with past partial discharge signals that serve as the basis for determination by the discharge type determination unit 115.
[0026] The difficult-to-determine partial discharge signal determination unit 114 determines whether a partial discharge signal is a difficult-to-determine partial discharge signal based on data of combinations of partial discharge signals (partial discharge signal samples) and discharge signal types stored in the discharge signal type data storage unit 108. The difficult-to-determine partial discharge signal determination unit 114 calculates the distance between the partial discharge signal and all partial discharge signal samples, and determines that the partial discharge signal is a difficult-to-determine partial discharge signal if there are no partial discharge signal samples that exceed a predetermined distance.
[0027] The difficult-to-determine partial discharge signal determination unit 114 determines whether a partial discharge signal is a difficult-to-determine partial discharge signal based on, for example, a partial discharge signal pattern which is the Φ-qn pattern of the partial discharge signal and a partial discharge signal sample pattern which is the Φ-qn pattern of a partial discharge signal sample. The Φ-qn pattern is a pattern in which the phase Φ is on the horizontal axis, the charge amount q is on the vertical axis, and the number of occurrences n is shown in the grid. Figure 3 is an example of a Φ-qn pattern. The phase Φ and charge amount q of the Φ-qn pattern may be normalized. For example, in the Φ-qn pattern shown in Figure 3, the phase Φ is divided into 20 parts of 360 degrees, which is one period, and the charge amount q is normalized to the maximum charge amount. The partial discharge signal sample pattern may be stored in the discharge signal type data storage unit 108 for each partial discharge signal.
[0028] The distance between the Φ-qn pattern of a partial discharge signal and the Φ-qn pattern of a partial discharge signal sample is, for example, the Bray-Curtis index. The Bray-Curtis index is expressed by equation (1).
[0029]
number
[0030] Here, m is the number of cells in the Φ-qn pattern, x ai x is the value of n in the i-th cell of the partial discharge signal pattern. bi is the value of n in the i-th cell of the partial discharge signal sample pattern. In other words, the difficult-to-determine partial discharge signal determination unit 114 uses equation (1) to calculate the distance based on all cells of the partial discharge signal pattern and the partial discharge signal sample pattern.
[0031] The difficult-to-determine partial discharge signal determination unit 114 calculates the distance between the partial discharge signal pattern and the partial discharge signal sample pattern for all partial discharge signal sample patterns, and uses the smallest distance to determine the difficult-to-determine partial discharge signal. The difficult-to-determine partial discharge signal determination unit 114 determines that the partial discharge signal sample does not contain a partial discharge signal and that it is a difficult-to-determine partial discharge signal that is difficult to determine by the discharge type determination unit 115 when the smallest distance is greater than or equal to a predetermined value.
[0032] The discharge type determination unit 115 determines the discharge type of a partial discharge signal that was not determined to be a difficult-to-determine partial discharge signal by the difficult-to-determine partial discharge signal determination unit 114. For partial discharge signals that have been determined to be difficult-to-determine partial discharge signals by the difficult-to-determine partial discharge signal determination unit 114, the discharge type determination unit 115 does not determine the discharge type.
[0033] If the partial discharge signal is determined to be a difficult-to-determine partial discharge signal, the determination result data may be output by, for example, the display unit 106, and "Undeterminable" may be displayed on a display or the like. If the partial discharge signal is not determined to be a difficult-to-determine partial discharge signal and the discharge type determination unit 115 determines the type of discharge, the determination result data may be output by, for example, the display unit 106, and the type of discharge may be displayed on a display or the like. The method of outputting the determination result is not limited to this, and for example, the determination result data may be transmitted to an external communication device by the communication unit 104.
[0034] Figure 4 is a flowchart showing the operation of the discharge determination device 103. The electrical signal acquisition unit 112 acquires an electrical signal (step S1). The partial discharge signal extraction unit 113 extracts a partial discharge signal from the acquired electrical signal (step S2). The difficult-to-determine partial discharge signal determination unit 114 determines a difficult-to-determine partial discharge signal for each partial discharge signal (step S3). If the partial discharge signal is determined to be a difficult-to-determine partial discharge signal (step S4: YES), the type of discharge of the partial discharge signal is not determined, and the display unit 106 outputs the result that it was determined to be a difficult-to-determine partial discharge signal (step S5). If the partial discharge signal is not determined to be a difficult-to-determine partial discharge signal (step S4: NO), the discharge type determination unit 115 determines the type of discharge based on the partial discharge signal (step S6), and the display unit 106 outputs the determined type of discharge (step S7).
[0035] (Example of experiment) Figures 5 and 6 are tables showing the results of Experiment 1 and Experiment 2, respectively, using the discharge determination device 103. The electrical signals acquired by the discharge determination device 103 are different in Experiment 1 and Experiment 2. In both Experiment 1 and Experiment 2, the electrical signals acquired by the discharge determination device 103 are denoised signals with mixed noise, but the noise mixed in the electrical signal in Experiment 2 is more difficult to denoise than the noise mixed in the electrical signal in Experiment 1. The electrical signals in the experimental results shown in Figure 6, for example, contain noise with the same frequency as the electrical signal, or the noise is large and the signal-to-noise ratio is small. In Experiment 1 and Experiment 2, the discharge type determination unit 115 determined the type of discharge based on the partial discharge signal, and the accuracy rate was calculated by dividing the number of partial discharge signals for which the determination result was correct by the total number of determined partial discharge signals.
[0036] In the experimental results shown in Figure 5, of the 350 partial discharge signals that were judged, without the difficult-to-judge partial discharge signal judgment unit 114, there were 334 correct answers and 16 incorrect answers, resulting in an accuracy rate of 95.4%. In contrast, with the difficult-to-judge partial discharge signal judgment unit 114, there were 293 correct answers, 3 incorrect answers, and 50 signals that were judged to be difficult-to-judge partial discharge signals, resulting in an accuracy rate of 97.7%.
[0037] In the experimental results shown in Figure 6, of the 350 partial discharge signals that were judged, when the difficult-to-judge partial discharge signal judgment unit 114 was not present, there were 224 correct answers and 126 incorrect answers, resulting in an accuracy rate of 64.0%. In contrast, when the difficult-to-judge partial discharge signal judgment unit 114 was present, there were 129 correct answers, 30 incorrect answers, and 191 signals were judged to be difficult-to-judge partial discharge signals, resulting in an accuracy rate of 81.1%.
[0038] Both Figure 5 and Figure 6 show experimental results where the accuracy rate is higher when the difficult-to-decode partial discharge signal detection unit 114 is present. Furthermore, in the experimental results shown in Figure 6, where noise that is difficult to denoise is mixed into the electrical signal, there are more signals that are judged as difficult-to-decode partial discharge signals, resulting in a larger difference in accuracy rate.
[0039] The control unit 111 may include a second discharge type determination unit that determines the type of partial discharge of a difficult-to-determine partial discharge signal. The second discharge type determination unit determines the type of partial discharge of a difficult-to-determine partial discharge signal by a method different from that of the discharge type determination unit 115. The second discharge type determination unit may determine the type of partial discharge by inputting the partial discharge signal into a discharge type determination model generated based on data of combinations of partial discharge signals and types of discharge signals that are different from the combinations of partial discharge signals and types of discharge signals stored in the discharge signal type data storage unit 108. The second discharge type determination unit may also determine the type of partial discharge by performing a simulation using physical calculations based on the partial discharge signal. The method of performing a simulation using physical calculations does not require consideration of the partial discharge signal used to generate the model, compared to the method of using a discharge type determination model.
[0040] The control unit 111 may include a second difficult-to-determine partial discharge signal determination unit that determines a second difficult-to-determine partial discharge signal, which is a signal that is difficult to determine by the second discharge type determination unit, based on data different from the data stored in the discharge signal type data storage unit. The second difficult-to-determine partial discharge signal determination unit may, for example, determine whether a signal that has been determined to be a difficult-to-determine partial discharge signal by the difficult-to-determine partial discharge signal determination unit 114 is a second difficult-to-determine partial discharge signal, and the second discharge type determination unit may determine the type of discharge based on signals that have not been determined to be a second difficult-to-determine partial discharge signal by the second difficult-to-determine partial discharge signal determination unit.
[0041] The control unit 111 may include several discharge type determination units that determine the type of discharge based on data different from the data stored in the discharge signal type data storage unit 108 and the discharge type determination model storage unit 109, and several difficult-to-determine partial discharge signal determination units that determine whether or not the signal is difficult to determine by the discharge type determination unit.
[0042] In the above embodiment, the discharge type determination unit 115 determines the discharge type of the partial discharge signal based on a discharge type determination model generated by machine learning, but it is not limited to this. For example, the discharge type determination unit 115 may determine the discharge type of the partial discharge signal to be the partial discharge signal corresponding to the partial discharge signal sample with the smallest distance among the distances between the partial discharge signal and the partial discharge signal sample calculated by the difficult-to-determine partial discharge signal determination unit 114.
[0043] According to at least one embodiment described above, the discharge determination device 103 has a difficult-to-determine partial discharge signal determination unit 114, thereby reducing the probability of errors in determining the type of discharge.
[0044] While several embodiments of the present invention have been described, these embodiments are presented as examples only and are not intended to limit the scope of the invention. These embodiments can be carried out in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. [Explanation of Symbols]
[0045] 100...Discharge determination system, 101...Electrode, 102...Signal line, 103...Discharge determination device, 104...Communication unit, 105...Input unit, 106...Display unit, 107...Signal storage unit, 108...Discharge signal type data storage unit, 109...Discharge type determination model storage unit, 111...Control unit, 112...Electrical signal acquisition unit, 113...Partial discharge signal extraction unit, 114...Difficult-to-determine partial discharge signal determination unit, 115...Discharge type determination unit
Claims
1. A partial discharge signal extraction unit extracts the target signal, which is a partial discharge signal, from an electrical signal. A partial discharge type determination unit determines the type of partial discharge from the target signal based on multiple combinations of past partial discharge signals and the type of partial discharge in question. A partial discharge signal determination unit that determines whether or not to perform a determination by the partial discharge type determination unit by comparing the target signal with each of the past partial discharge signals related to the multiple combinations, The difficult-to-determine partial discharge signal determination unit outputs a message indicating that it is difficult to determine the type of partial discharge if it determines that the partial discharge type determination unit will not perform a determination on the target signal, and outputs the result of the partial discharge type determination by the partial discharge type determination unit if it determines that it will perform a determination on the partial discharge type determination unit. Equipped with, The difficult-to-determine partial discharge signal determination unit determines that it will not perform a determination by the partial discharge type determination unit if the distances calculated from the target signal and each of the past partial discharge signals related to the multiple combinations do not exceed a predetermined distance. Discharge determination device.
2. The partial discharge type determination unit determines the type of partial discharge by inputting the target signal to a discharge type determination model, which is a trained model that has been trained by machine learning based on the multiple combinations to take the partial discharge signal as input and output the type of partial discharge. The discharge determination device according to claim 1.
3. The aforementioned distance is the Bray-Curtis index. The discharge determination device according to claim 1.
4. A partial discharge type determination unit determines the type of partial discharge from the target signal, which is the signal of the partial discharge to be determined, based on multiple combinations of past partial discharge signals and the type of partial discharge in question. A partial discharge signal determination unit that determines whether or not to perform a determination by the partial discharge type determination unit by comparing the target signal with each of the past partial discharge signals related to the multiple combinations, If the determination unit for the difficult-to-determine partial discharge signal determines that the determination unit for the partial discharge type should not be performed, a second partial discharge type determination unit determines the type of partial discharge from the target signal using a method different from that of the partial discharge type determination unit. Equipped with, The difficult-to-determine partial discharge signal determination unit determines that it will not perform a determination by the partial discharge type determination unit if the distances calculated from the target signal and each of the past partial discharge signals related to the multiple combinations do not exceed a predetermined distance. Discharge determination device.
5. A partial discharge signal extraction step of extracting a target signal which is a partial discharge signal from an electrical signal; a partial discharge type determination step of determining the type of partial discharge from the target signal based on a plurality of combinations of past partial discharge signals and the type of partial discharge in question; A difficult-to-determine partial discharge signal determination step, which determines whether or not to perform the determination according to the partial discharge type determination step by comparing the target signal with each of the past partial discharge signals related to the multiple combinations, In the difficult-to-determine partial discharge signal determination step, if it is determined that the determination of the type of partial discharge is difficult for the target signal, an output step is provided that indicates that the determination of the type of partial discharge is difficult, and if it is determined that the determination of the type of partial discharge is to be performed, an output step is provided that outputs the determination result of the type of partial discharge performed in the partial discharge type determination step. It has, In the difficult-to-determine partial discharge signal determination step, if the distance calculated from the target signal and each of the past partial discharge signals related to the multiple combinations does not exceed a predetermined distance, it is determined that the determination in the partial discharge type determination step is not performed. Discharge determination method.
6. On the computer, A partial discharge signal extraction step of extracting a target signal which is a partial discharge signal from an electrical signal; a partial discharge type determination step of determining the type of partial discharge from the target signal based on a plurality of combinations of past partial discharge signals and the type of partial discharge in question; A difficult-to-determine partial discharge signal determination step, which determines whether or not to perform the determination according to the partial discharge type determination step by comparing the target signal with each of the past partial discharge signals related to the multiple combinations, In the difficult-to-determine partial discharge signal determination step, if it is determined that the determination of the type of partial discharge is difficult for the target signal, an output step is provided that indicates that the determination of the type of partial discharge is difficult, and if it is determined that the determination of the type of partial discharge is to be performed, an output step is provided that outputs the determination result of the type of partial discharge performed in the partial discharge type determination step. Make it run, In the difficult-to-determine partial discharge signal determination step, if the distance calculated from the target signal and each of the past partial discharge signals related to the multiple combinations does not exceed a predetermined distance, it is determined that the determination in the partial discharge type determination step is not performed. program.