Insulating oil deterioration diagnosis device and insulating oil deterioration diagnosis method

Abstract

To diagnose deterioration of insulating oil of oil-filled electric equipment.SOLUTION: An insulation oil deterioration diagnosis device 1 diagnoses deterioration of the insulating oil of an oil-filled transformer being an example of oil-filled electric equipment. The insulating oil deterioration diagnosis device 1 comprises a data analysis unit 6 that diagnoses the deterioration of the insulating oil on the basis of the feature quantities of an odor component of the insulating oil collected from the oil-filled transformer into a sample container 2. The data analysis unit 6 diagnoses deterioration by providing the feature quantities of the odor component of the collected insulating oil to a support vector machine that has learned the feature quantities of the odor component before and after the deterioration of the insulating oil.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a technique for diagnosing deterioration of insulating oil in oil-filled electrical equipment such as an oil-filled transformer. [Background technology]

[0002] Examples of diagnostic techniques for the deterioration state of oil-filled electrical equipment include the aging deterioration diagnostic method for oil-filled electrical equipment described in Patent Document 1 and the internal condition diagnostic device for oil-filled electrical equipment described in Patent Document 2.

[0003] The diagnostic method of Patent Document 1 involves extracting degradation products dissolved in insulating oil into an aqueous solution and diagnosing the aging of oil-filled electrical equipment using an aqueous solution sensor that is sensitive to ketones, or converting the sampled insulating oil into a gas phase and diagnosing the aging of oil-filled electrical equipment using a synthetic bilayer membrane sensor that is sensitive to ketones.

[0004] The diagnostic device of Patent Document 2 diagnoses the internal state of oil-filled electrical equipment by using a diagnostic device equipped with odor detection means for detecting the odor of gas contained in the insulating oil of the oil-filled electrical equipment. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] Japanese Patent Application Publication No. 6-160379 [Patent Document 2] Japanese Patent Application Publication No. 1-233714 [Non-patent literature]

[0006] [Non-Patent Document 1] B. Wyszynski, R. Yatabe, A. Nakano, M. Nakatani, A. Oki, H. Oka, K. Toko: "Array of Chemosensitive Resistors with Composites of Gas Chromatography (GC) Materials and Carbon Black for Detection and Recognition of VOCs: A Basic Study", Sensors, Issue 17,1606,2017-7 Summary of the Invention [Problem to be solved by the invention]

[0007] The diagnostic methods of Patent Documents 1 and 2 diagnose the aging deterioration and internal state of oil-filled electrical equipment by analyzing the components contained in the insulating oil of the equipment, but do not diagnose deterioration of the insulating oil.

[0008] In view of the above circumstances, an object of the present invention is to provide a method for diagnosing deterioration of insulating oil in oil-filled electrical equipment. [Means for solving the problem]

[0009] Therefore, one aspect of the present invention is an insulating oil deterioration diagnosis method for diagnosing deterioration of insulating oil in an oil-filled transformer, in which the deterioration diagnosis is performed by providing features of odor components of insulating oil collected from the oil-filled transformer to a support vector machine that has learned the features of odor components before and after deterioration of the insulating oil in the oil-filled transformer.

[0010] One aspect of the present invention is an insulating oil degradation diagnosis device that performs degradation diagnosis of insulating oil in an oil-filled transformer, and includes a data analysis unit that performs the degradation diagnosis by providing feature amounts of odor components of insulating oil collected from the oil-filled transformer to a support vector machine that has learned feature amounts of odor components before and after degradation of the insulating oil in the oil-filled transformer.

[0011] In one aspect of the present invention, in the insulating oil deterioration diagnostic device, the result of the deterioration diagnosis is output and displayed.

[0012] In one aspect of the present invention, in the insulating oil deterioration diagnostic device, the result of the deterioration diagnosis is transmitted to a mobile terminal by power-saving wireless communication.

[0013] In one aspect of the present invention, in the insulating oil deterioration diagnostic device, the result of the deterioration diagnosis is transmitted to a remote control device via a network. Effect of the Invention

[0014] According to the present invention as described above, deterioration of insulating oil in oil-filled electrical equipment can be diagnosed. [Brief description of the drawings]

[0015] [Figure 1] 1 is a block diagram showing the configuration of a degradation diagnosis device according to a first embodiment of the present invention. [Diagram 2] FIG. 2 is a block diagram of a control unit according to the first embodiment. [Diagram 3] 5 shows a specific example of a data measurement operation schedule according to the first embodiment. [Figure 4] FIG. 3 is a signal waveform diagram illustrating an output signal from the olfactory sensor of the first embodiment. [Diagram 5] FIG. 2 is a block diagram showing the configuration of a data analysis unit according to the first embodiment. [Figure 6] 4 shows an example of creating feature data obtained by the olfactory sensor. [Figure 7] 13 is a configuration example of the feature data organized into a high-dimensional vector. [Figure 8] FIG. 4 is a flow diagram of degradation diagnosis processing by the diagnosis processing unit of the first embodiment. [Figure 9] FIG. 2 is a block diagram showing the configuration of an insulating oil deterioration diagnosis device according to a second embodiment of the present invention. [Figure 10] 10 is a flowchart of an oil circulation process according to a second embodiment. [Figure 11] FIG. 11 is an explanatory diagram of an inspection operation for an oil-filled transformer to which the insulating oil deterioration diagnosis device of the second embodiment is applied. [Figure 12] FIG. 4 is a block diagram of an insulating oil deterioration diagnosis device according to a third embodiment of the present invention. [Figure 13] FIG. 11 is an explanatory diagram of an inspection operation for an oil-filled transformer to which the insulating oil deterioration diagnosis device of the third embodiment is applied. [Figure 14] FIG. 11 is a block diagram showing the configuration of an insulating oil deterioration diagnosis device according to a fourth embodiment of the present invention. [Figure 15] FIG. 11 is an explanatory diagram of an inspection operation for an oil-filled transformer to which the insulating oil deterioration diagnosis device of the fourth embodiment is applied. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0017] [Embodiment 1] 1 performs deterioration diagnosis of insulating oil collected from an oil-filled electrical device (not shown), such as an oil-filled transformer. The insulating oil deterioration diagnosis device 1 includes a sample container 2, a carrier gas tank 3, a gas mixing container 4, a first solenoid valve V1, a second solenoid valve V2, an olfactory sensor 5, a pump P1, a data analysis unit 6, and a control unit 7.

[0018] The sample container 2 is filled with insulating oil collected from the oil-filled electrical device.

[0019] The carrier gas tank 3 is filled with a carrier gas to be supplied to the sample container 2 and the gas mixing container 4. Nitrogen gas is used as the carrier gas.

[0020] The gas mixing vessel 4 is a vessel for mixing the gas from the sample vessel 2 and the carrier gas from the carrier gas tank 3 .

[0021] The first solenoid valve V1 is provided in a gas supply line between the carrier gas tank 3 and the gas mixing container 4, and the valve opening / closing degree A is arbitrarily adjusted by a valve adjustment command from the control unit .

[0022] The second solenoid valve V2 is provided in a gas supply line between the sample container 2 and the gas mixing container 4, and the valve opening / closing degree B is arbitrarily adjusted by a valve adjustment command from the control unit .

[0023] The olfactory sensor 5 detects the odor components of the gas supplied from the gas mixing container 4 and outputs a detection signal. An example of the olfactory sensor 5 is an artificial olfactory system configured with a sensor array equipped with multiple (e.g., 16-channel) sensors (Non-Patent Document 1). The sensor is made of a different GC stationary phase material for each channel, and is capable of identifying odors learned from multi-channel data. The artificial olfactory system has a receptor membrane in the sensor receptor section made by mixing the column packing material (GC material) of a gas chromatography mass spectrometer (GCMS) and carbon black. When an odor substance is absorbed into the membrane, the conductivity of the membrane changes, and the accompanying change in resistance value is output as a sensor response.

[0024] The pump P1 sucks the gas in the gas mixing container 4 through the olfactory sensor 5. A flow meter is attached to the pump P1 to control the flow rate of the gas.

[0025] The data analysis unit 6 pre-stores the analysis parameters described below that are used to diagnose the deterioration of the insulating oil, and based on a control signal from the control unit 7, performs data analysis of the sensor signal data in which the sensor signals of odor components from the olfactory sensor 5 are organized in chronological order, thereby performing the deterioration diagnosis.

[0026] The control unit 7 controls the first solenoid valve V1, the second solenoid valve V2, the pump P1, and the data analysis unit 6.

[0027] An example of the control unit 7 will be described in detail with reference to FIG.

[0028] The control unit 7 implements a parameter input unit 71, a diagnostic schedule management unit 72, and a storage unit 73.

[0029] A parameter input unit 71 stores control parameters input from the outside in a storage unit 73 via a diagnosis schedule management unit 72. The control parameters include a gas flow rate set value, a sensor cleaning time set value, a sample ON time, a sample OFF time, a sample ON / OFF repetition count, a sample ON opening / closing degree set value, and a sample OFF opening / closing degree set value.

[0030] The diagnosis schedule management unit 72 outputs the gas flow rate, the valve opening / closing degree A, the valve opening / closing degree B, a sensor signal recording command, and a data analysis command based on a preset data measurement operation schedule. The gas flow rate is output to the pump P1. The valve opening / closing degree A is output to the first solenoid valve V1. The valve opening / closing degree B is output to the second solenoid valve V2. The sensor signal recording command and the data analysis command are output to the data analysis unit 6.

[0031] A specific example of a data measurement operation schedule will be described with reference to FIG.

[0032] In the data measurement operation schedule, the values ​​of the gas flow rate, the valve opening / closing degree A, the valve opening / closing degree B, and the sensor signal recording command are set arbitrarily in accordance with each time period.

[0033] In the data measurement operation schedule, the time other than the sensor cleaning time and the data measurement time is the standby time. The data measurement time is defined as one measurement cycle consisting of a set of sample ON time and sample OFF time. The data measurement time is the time required for this measurement cycle to be repeated the number of times that the sample ON / OFF is repeated.

[0034] The gas flow rate is set to 0 during the standby time, and the gas flow rate setting value is used as the gas flow rate value during the sensor cleaning time and data measurement time.

[0035] During standby time, the valve opening / closing degree A value is set to 0. During sensor cleaning time, the sample OFF opening / closing degree set value is set as the valve opening / closing degree A value. During data measurement time, the valve opening / closing degree A value is set by subtracting the sample ON opening / closing degree set value from the sample OFF opening / closing degree set value for the sample ON time, and the sample OFF opening / closing degree A value is set for the sample OFF time.

[0036] During the standby time and the sensor cleaning time, the valve opening / closing degree B value is set to 0. During the data measurement time, the sample ON opening / closing degree setting value is set as the valve opening / closing degree B value for the sample ON time, and the valve opening / closing degree B value is set as 0 for the sample OFF time.

[0037] The sensor signal recording command is set such that the sensor signal recording command value is set to 0 during the standby time and the sensor cleaning time, and is set to 1 during the data measurement time.

[0038] By controlling the pump P1, the first solenoid valve V1, the second solenoid valve V2, and the data analysis unit 6 according to the data measurement operation schedule, it is possible to obtain sawtooth-shaped sensor signal data in which the sensor signals from each sensor of the olfactory sensor 5 are recorded in chronological order, as shown in Fig. 4. During the sample ON time, an output signal from the olfactory sensor 5 is generated in response to the components of the insulating oil. During the sample OFF time, the olfactory sensor 5 is cleaned by the carrier gas, and the output signal decreases.

[0039] The sensor signal data is recorded during the data measurement time in the data analysis unit 6. The data analysis command is set to a data analysis command value of 1 when the data analysis unit 6 is to execute data analysis, and set to 0 when the data analysis unit 6 is not to execute the analysis.

[0040] An example of the data analysis unit 6 will be described in detail with reference to FIG.

[0041] The data analysis unit 6 implements a parameter input unit 61, a command setting unit 62, a sensor signal data creation unit 63, a storage unit 64, a feature extraction unit 65, a deterioration diagnosis unit 66, a result output unit 67, and a sensor signal data output unit 68.

[0042] The parameter input unit 61 stores the analysis parameters input from the outside and the control parameters received from the control unit 7 in the storage unit 64 .

[0043] The command setting unit 62 receives a sensor signal recording command and a data analysis command from the control unit 7, and sets and outputs a value of the analysis control state so that the sensor signal recording command has priority.

[0044] The analysis control state is indicated by either a wait state value, a record sensor signal state value, or a data analysis state value.

[0045] When the sensor signal recording command value and the data analysis command value are 0, the analysis control state is set to a standby state value.

[0046] When the sensor signal recording command value is 1, the analysis control state becomes the sensor signal recording state value.

[0047] If the record sensor signal command value is 0 while the analyze data command value is 1, the analysis control state is the data analysis state value.

[0048] When the analysis control state provided by the command setting unit 62 is a sensor signal recording state value, the sensor signal data creation unit 63 stores the sensor signal data, which is the sensor signal input from the olfactory sensor 5 organized in chronological order, in the memory unit 64.

[0049] When the analysis control state provided from the command setting unit 62 is a data analysis state value, the feature extraction unit 65 stores the feature of the sensor signal data extracted based on the control parameters and the sensor signal data retrieved from the memory unit 64 in the memory unit 64 as feature data.

[0050] The creation of feature data will be described below.

[0051] First, the average value of the sensor signal during the sample OFF time is calculated, and then the difference value of the sensor signal from this average value is calculated discretely and arranged to create feature amount data for each sensor. An example of creating feature amount data for each sensor is shown in Figure 6.

[0052] Next, the feature data calculated for each sensor is organized into high-dimensional vector data. This high-dimensional vector data becomes the feature data. Figure 7 shows an example of organizing the data into high-dimensional vector data to become feature data.

[0053] When the analysis control state is a data analysis state value, the deterioration diagnosis unit 66 performs the deterioration diagnosis by providing the analysis parameters and feature amount data extracted from the storage unit 64 to a support vector machine. The diagnosis result of "deteriorated" or "not deteriorated" obtained by this diagnosis is stored in the storage unit 64 as the deterioration diagnosis result of the insulating oil. The support vector machine is a two-class classifier to which the analysis parameters are applied. The analysis parameters include parameters of linear input elements of the two-class classifier obtained by offline learning using feature amounts based on sensor signal data of odor components of the insulating oil before and after deterioration collected in advance as teacher data.

[0054] The degradation diagnostic process performed by degradation diagnostic unit 66 will be described with reference to FIG.

[0055] S101: The analysis parameters are read from the storage unit 64.

[0056] S102: The support vector machine is started.

[0057] S103: The characteristic amount data of the insulating oil is read out from the storage unit 64.

[0058] S104: The read analysis parameters and feature amount data are provided to the support vector machine to diagnose the presence or absence of deterioration of the insulating oil.

[0059] S105: The presence or absence of deterioration obtained in S104 is stored in the memory unit 64 as a deterioration diagnosis result.

[0060] The result output unit 67 reads out the deterioration diagnosis result from the storage unit 64 and outputs it to the outside.

[0061] The sensor signal data output unit 68 extracts the sensor signal data corresponding to the deterioration diagnosis result from the storage unit 64 and outputs it to the outside.

[0062] According to the insulating oil deterioration diagnostic device 1 of the above-described first embodiment, deterioration diagnosis of insulating oil can be performed based on the characteristic quantities of odor components of the insulating oil sampled from an oil-filled electrical device.

[0063] [Embodiment 2] The insulating oil deterioration diagnostic device 1 of the second embodiment shown in FIG.

[0064] The insulating oil in the oil-filled transformer 10 is automatically sampled periodically by the insulating oil deterioration diagnostic device 1 and diagnosed for deterioration. The result of this deterioration diagnosis is output and displayed from the insulating oil deterioration diagnostic device 1.

[0065] The insulating oil deterioration diagnostic device 1 of this embodiment further includes an oil level gauge 11, a third solenoid valve V3, an oil pump P2, an oil circulation control unit 12, and an indicator 13 in addition to the configuration of the first embodiment.

[0066] The oil level meter 11 measures the amount of insulating oil in the sample container 2 and outputs the amount to the oil circulation control unit 12.

[0067] The third solenoid valve V3 opens and closes based on the valve opening / closing degree C output from the oil circulation control unit 12 to supply insulating oil from the oil-filled transformer 10 to the sample container 2.

[0068] The oil pump P2 operates based on the oil flow rate output from the oil circulation control unit 12 to return the insulating oil from the sample container 2 to the oil-filled transformer 10.

[0069] The oil circulation control unit 12 operates at preset circulation time intervals, and circulates the insulating oil in the sample container 2 by controlling the third solenoid valve V3 and the oil pump P2.

[0070] In the oil circulation control unit 12, a circulation time interval, a sample oil amount, a sample oil replacement amount, a set value of the valve opening / closing degree C, and an oil flow rate set value are preset.

[0071] The oil circulation control unit 12 outputs a control signal to the third solenoid valve V3 and the oil pump P2 at each circulation time interval. The oil circulation control unit 12 sets the value of the oil flow rate to an oil flow rate setting value and outputs it to the oil pump P2 until the value of the oil gauge 11 in the sample container 2 becomes the value obtained by subtracting the sample oil replacement amount from the sample oil amount. Thereafter, the oil flow rate value is set to 0 and output to the oil pump P2. Next, the value of the valve opening / closing degree C is set to the valve opening / closing degree C setting value and output to the third solenoid valve V3 until the value of the oil gauge 11 in the sample container 2 becomes the sample oil amount. Thereafter, the value of the valve opening / closing degree C is set to 0 and output to the third solenoid valve V3. As described above, the insulating oil in the sample container 2 is automatically circulated.

[0072] The process of oil circulation processing by the oil circulation control unit 12 will be described with reference to FIG.

[0073] S201: The clock (timer) of the insulating oil deterioration diagnostic device 1 starts.

[0074] S202: If the cyclic time interval has elapsed (Yes), the process proceeds to S203. On the other hand, if the cyclic time interval has not elapsed (No), the process proceeds to S209 and enters a standby state.

[0075] S203: The value of the oil flow rate is set as an oil flow rate setting value and output to the oil pump P2.

[0076] S204: The oil flow rate is set to an oil flow rate setting value and output to the oil pump P2 until the value of the oil gauge 11 in the sample container 2 becomes the value obtained by subtracting the sample oil replacement amount from the sample oil amount (Yes).

[0077] S205: The oil flow rate value is set to 0 and output to the oil pump P2.

[0078] S206: The value of the valve opening / closing degree C is set to the valve opening / closing degree C setting value until the value of the oil gauge 11 of the sample container 2 reaches the sample oil amount, and is output to the third solenoid valve V3.

[0079] S207: When the value of the oil gauge 11 in the sample container 2 reaches the sample oil amount (Yes), the process proceeds to S208.

[0080] S208: The value of the valve opening / closing degree C is set to 0 and output to the third solenoid valve V3.

[0081] S209: The third solenoid valve V3 and the oil pump P2 are stopped.

[0082] The operations of the first solenoid valve V1, the second solenoid valve V2, the olfactory sensor 5, the pump P1, the data analysis unit 6, and the control unit 7 in this embodiment are the same as those in the first embodiment.

[0083] The indicator 13 outputs and displays the presence or absence of deterioration of the insulating oil, which is the result of the deterioration diagnosis obtained by the data analysis unit 6, by turning on a warning light or the like.

[0084] It is clear that the insulating oil degradation diagnosis device 1 of the above-mentioned embodiment 2 can provide the same effects as those of embodiment 1. In particular, as shown in Fig. 11, in an on-site electrical equipment in which an oil-filled transformer 10 is installed, the insulating oil of the oil-filled transformer 10 is periodically diagnosed for deterioration. The result of this diagnosis is output and displayed on the indicator 13 of the insulating oil degradation diagnosis device 1, so that a maintenance worker 20 can clearly grasp the result of the deterioration diagnosis of the insulating oil when inspecting the electrical equipment.

[0085] [Embodiment 3] The insulating oil deterioration diagnostic device 1 of the third embodiment shown in FIG. 12 is capable of transmitting the deterioration diagnostic result of the insulating oil of the oil-filled transformer 10 to a mobile terminal 21 of a maintenance worker 20 in FIG.

[0086] The insulating oil deterioration diagnostic device 1 of this embodiment, in the aspect of embodiment 2, further includes a deterioration diagnosis result wireless transmission unit 14. The deterioration diagnosis result wireless transmission unit 14 is capable of communicating with a mobile terminal 21 by low-power wireless communication such as Bluetooth (registered trademark).

[0087] The degradation diagnosis result wireless transmission unit 14 temporarily stores the degradation diagnosis result of the insulating oil provided from the data analysis unit 6 in a built-in memory of the degradation diagnosis result wireless transmission unit 14. Then, upon receiving a request to transmit the degradation diagnosis result from the mobile terminal 21, the degradation diagnosis result wireless transmission unit 14 transmits the current degradation diagnosis result stored in the built-in memory to the mobile terminal 21 by the power-saving wireless communication.

[0088] It is clear that the insulating oil deterioration diagnosis device 1 of the above-mentioned embodiment 3 can provide the same effects as those of embodiment 2. In particular, in the on-site electrical equipment where the oil-filled transformer 10 is installed, the mobile terminal 21 of the maintenance worker 20 can receive, record, and confirm the deterioration diagnosis result of the insulating oil of the oil-filled transformer 10 at low cost when inspecting the oil-filled transformer 10.

[0089] [Embodiment 4] The insulating oil deterioration diagnostic device 1 of the fourth embodiment shown in FIG. 14 is capable of transmitting the result of the deterioration diagnosis of the insulating oil of the oil-filled transformer 10 to the remote control device 22 in FIG.

[0090] The insulating oil deterioration diagnostic device 1 of this embodiment is similar to that of the third embodiment, except that it includes a deterioration diagnosis result communication unit 15 instead of the deterioration diagnosis result wireless transmission unit 14. The deterioration diagnosis result communication unit 15 is capable of communicating with a remote control device 22 via a network such as a LAN (local area network).

[0091] When the degradation diagnosis result communication unit 15 receives the deterioration diagnosis result of the insulating oil from the data analysis unit 6, it temporarily stores the result in the built-in memory of the degradation diagnosis result communication unit 15, similarly to the first embodiment. Then, in response to a transmission request of the diagnosis result from the remote control device 22, the current deterioration diagnosis result stored in the built-in memory is transmitted from the degradation diagnosis result communication unit 15 to the remote control device 22 via the network.

[0092] It is clear that the insulating oil deterioration diagnostic device 1 of the above-mentioned embodiment 4 can provide the same effects as those of embodiment 2. In particular, as shown in Fig. 15, the maintenance man 20 can obtain, record and confirm the deterioration diagnostic results of the insulating oil of the oil-filled transformer 10 installed in the on-site electrical equipment away from the remote control device 22 by using the remote control device 22. [Explanation of symbols]

[0093] 1...insulating oil deterioration diagnostic device, 2...sample container, 3...carrier gas tank, 4...gas mixing container, V1...first solenoid valve, V2...second solenoid valve, 5...olfactory sensor, P1...pump 10... oil-filled transformer, 20... maintenance worker, 21... mobile terminal, 22... remote control device 11...oil level gauge, V3...third solenoid valve, P2...oil pump, 12...oil circulation control unit, 13...indicator, 14...deterioration diagnosis result wireless transmission unit, 15...deterioration diagnosis result communication unit 6... data analysis unit, 61... parameter input unit, 62... command setting unit, 63... sensor signal data creation unit, 64... storage unit, 65... feature extraction unit, 66... ​​deterioration diagnosis unit, 67... result output unit, 68... sensor signal data output unit 7... control unit, 71... parameter input unit, 72... diagnostic schedule management unit, 73... storage unit

Claims

1. An insulating oil deterioration diagnosis device for diagnosing deterioration of insulating oil in an oil-filled transformer, an olfactory sensor equipped with a plurality of sensors for detecting odor components of gas from insulating oil collected from the oil-filled transformer; a data analysis unit that performs the deterioration diagnosis by providing the characteristic amounts of the odor components of the insulating oil collected from the oil-filled transformer to a support vector machine that has learned the characteristic amounts of the odor components before and after deterioration of the insulating oil of the oil-filled transformer detected by the olfactory sensor; Equipped with The data analysis unit a feature extraction unit that creates feature data for each sensor based on sensor signal data for each sensor of the olfactory sensor and organizes the feature data for each sensor into high-dimensional vector data; a degradation diagnosis unit that performs the degradation diagnosis by providing the feature amount data consisting of the high-dimensional vector data to a support vector machine that has learned feature amounts of odor components before and after degradation of the insulating oil based on the sensor signal data of each sensor of the olfactory sensor; Equipped with the sensor signal data is acquired during a sample ON time, which is a time period during which an output signal of the olfactory sensor is generated in response to a component of the insulating oil, and during a sample OFF time, which is a time period during which the olfactory sensor is cleaned by a carrier gas; The feature amount data is obtained by discretely calculating difference values ​​of the sensor signal data during the sample OFF time and the sample ON time from an average value of the sensor signal data during the sample OFF time and arranging the difference values ​​in a time series. An insulating oil deterioration diagnosis device characterized by the above.

2. 2. The insulating oil deterioration diagnostic device according to claim 1, wherein the result of the deterioration diagnosis is output and displayed.

3. 3. The insulating oil deterioration diagnostic device according to claim 1, wherein the result of the deterioration diagnosis is transmitted to a mobile terminal by low-power wireless communication.

4. 4. The insulating oil deterioration diagnostic device according to claim 1, wherein the result of the deterioration diagnosis is transmitted to a remote control device via a network.

5. An insulating oil deterioration diagnosis method implemented by an insulating oil deterioration diagnosis device for diagnosing deterioration of insulating oil in an oil-filled transformer, comprising: A data analysis process is provided to a support vector machine that has learned feature amounts of odor components of the insulating oil of the oil-filled transformer before and after deterioration, detected by an olfactory sensor equipped with a plurality of sensors that detects odor components of gas from the insulating oil collected from the oil-filled transformer, to perform the deterioration diagnosis, The data analysis process includes: a feature extraction process of creating feature data for each sensor based on sensor signal data for each sensor of the olfactory sensor, and organizing the feature data for each sensor into high-dimensional vector data; a degradation diagnosis process in which the degradation diagnosis is performed by providing the feature amount data consisting of the high-dimensional vector data to a support vector machine that has learned feature amounts of odor components before and after degradation of the insulating oil based on the sensor signal data of each sensor of the olfactory sensor; having the sensor signal data is acquired during a sample ON time, which is a time period during which an output signal of the olfactory sensor is generated in response to a component of the insulating oil, and during a sample OFF time, which is a time period during which the olfactory sensor is cleaned by a carrier gas; The feature amount data is obtained by discretely calculating difference values ​​of the sensor signal data during the sample OFF time and the sample ON time from an average value of the sensor signal data during the sample OFF time and arranging the difference values ​​in a time series. An insulating oil deterioration diagnosis method comprising the steps of:

Images