Opening and closing device diagnosis system

Abstract

Provided is an opening and closing device diagnosis system capable of easily and precisely identifying an abnormal site in addition to a sign of abnormality in an opening and closing device. In an opening and closing device diagnosis system according to an embodiment, an abnormality occurrence estimation unit obtains abnormality occurrence estimation data by estimating the presence or absence of abnormality occurrence in an opening and closing device on the basis of measurement data stored in a storage unit. On the basis of the measurement data stored in the storage unit, an abnormal site estimation unit obtains abnormal site estimation data by estimating an abnormal site where an abnormality is occurring in a plurality of opening and closing device components. An estimation result integration unit obtains integrated data by integrating the abnormality occurrence estimation data obtained by the abnormality occurrence estimation unit and the abnormal site estimation data obtained by the abnormal site estimation unit. An output unit outputs the integrated data obtained by the estimation result integration unit.

Description

Switchgear Diagnostic System

[0001] The embodiment relates to a switchgear diagnostic system.

[0002] Switchgear, such as circuit breakers and disconnectors, are installed in substations and other facilities to open and close electrical circuits. There is a need for a shift in switchgear maintenance from time-based maintenance to condition-based maintenance. In time-based maintenance, maintenance of switchgear is performed according to the elapsed time since the start of operation of the switchgear. In contrast, in condition-based maintenance, maintenance of switchgear is performed according to the condition of the switchgear.

[0003] In order to perform condition-based maintenance, techniques related to diagnostic systems for diagnosing the operating state of switchgears have been proposed.

[0004] Japanese Patent Application Laid-Open No. 2021-149115

[0005] Conventional diagnostic systems detect abnormalities in switching gears and signs of such abnormalities. However, it has been difficult to easily and accurately determine the location of the abnormality in the switching gear.

[0006] Therefore, an object of the present invention is to provide a switchgear diagnostic system that can easily and accurately grasp the estimated result of the abnormality location in addition to the sign of abnormality in the switchgear.

[0007] A switchinggear diagnostic system according to an embodiment is a system for diagnosing an operating state of a switchinggear configured with a plurality of switchinggear components so as to perform switching operations, and includes a storage unit, an abnormality occurrence estimation unit, an abnormality location estimation unit, an estimation result integration unit, and an output unit. The storage unit stores measurement data measured regarding the switching operations of the switchinggear when the switchinggear is performing the switching operations, in association with a time axis. The abnormality occurrence estimation unit obtains abnormality occurrence estimation data by estimating the presence or absence of an abnormality in the switchinggear based on the measurement data stored in the storage unit. The abnormality location estimation unit obtains abnormality location estimation data by estimating an abnormal location where an abnormality has occurred in the plurality of switchinggear components based on the measurement data stored in the storage unit. The estimation result integration unit obtains integrated data by integrating the abnormality occurrence estimation data obtained by the abnormality occurrence estimation unit and the abnormality location estimation data obtained by the abnormality location estimation unit. The output unit outputs the integrated data obtained by the estimation result integration unit. Here, the abnormality occurrence estimation unit estimates the presence or absence of an abnormality through machine learning using normal state data obtained in advance regarding the switching operations associated with a time axis when the switchinggear is in a normal state as learning data. The abnormality location estimation unit estimates the abnormality location by machine learning using abnormality state data obtained in advance in association with a time axis regarding the opening and closing operation when the switching device is in an abnormal state as learning data.

[0008] FIG. 1A is a diagram schematically showing an example of a switchinggear 1 to be diagnosed by the switchinggear diagnostic system according to the first embodiment. FIG. 1B is a diagram schematically showing an example of a switchinggear 1 to be diagnosed by the switchinggear diagnostic system according to the first embodiment. FIG. 2 is a block diagram schematically showing a configuration of a switchinggear diagnostic system 4 according to the first embodiment. FIG. 3 is a functional block diagram showing main functions of the switchinggear diagnostic system 4 according to the first embodiment. FIG. 4 is a flow diagram showing the operation of the switchinggear diagnostic system 4 according to the first embodiment. FIG. 5A is a diagram showing an example of measurement data D3 input to an input unit 41 in the switchinggear diagnostic system 4 according to the first embodiment. FIG. 5B is a diagram showing an example of measurement data D42 stored in a storage unit 42 in the switchinggear diagnostic system 4 according to the first embodiment. FIG. 6A is a diagram schematically showing an example of normal state data D61 in the switchinggear diagnostic system 4 according to the first embodiment. FIG. 6B is a diagram schematically showing an example of a normal waveform pattern PS in the switchinggear diagnostic system 4 according to the first embodiment. FIG. 6C is a diagram schematically showing an example of how the degree of deviation between the normal waveform pattern PS and the normal state data D61 is calculated in the switchinggear diagnostic system 4 according to the first embodiment. FIG. 6D is a diagram schematically showing an example of how the normal feature range TS is calculated in the switchinggear diagnostic system 4 according to the first embodiment. FIG. 7A is a diagram schematically showing an example of how the abnormality occurrence estimation unit 43 estimates the presence or absence of an abnormality based on the normal feature range TS in the switchinggear diagnostic system 4 according to the first embodiment. FIG. 7B is a diagram schematically showing an example of how the abnormality occurrence estimation unit 43 estimates the presence or absence of an abnormality based on the normal feature range TS in the switchinggear diagnostic system 4 according to the first embodiment. FIG. 8A is a diagram schematically showing an example of abnormal state data D62 in the switchinggear diagnostic system 4 according to the first embodiment. FIG. 8B is a diagram schematically showing an example of an abnormal waveform pattern PD in the switchinggear diagnostic system 4 according to the first embodiment. FIG. 8C is a diagram schematically showing an example of how the degree of deviation between the abnormal waveform pattern PD and the abnormal state data D62 is calculated in the switching device diagnostic system 4 according to the first embodiment.FIG. 8D is a diagram schematically illustrating an example of how the feature quantity abnormality range TD is calculated in the switchinggear diagnosis system 4 according to the first embodiment. FIG. 9A is a diagram schematically illustrating an example of how the abnormality location estimation unit 44 estimates an abnormality location based on the feature quantity abnormality range TD in the switchinggear diagnosis system 4 according to the first embodiment. FIG. 9B is a diagram schematically illustrating an example of how the abnormality location estimation unit 44 estimates an abnormality location based on the feature quantity abnormality range TD in the switchinggear diagnosis system 4 according to the first embodiment. FIG. 10 is a diagram schematically illustrating an example of an image related to integrated data D45 output by the output unit 46 in the switchinggear diagnosis system 4 according to the first embodiment. FIG. 11 is a functional block diagram illustrating main functions of the switchinggear diagnosis system 4 according to the second embodiment. FIG. 12 is a flow chart illustrating the operation of the switchinggear diagnosis system 4 according to the second embodiment. FIG. 13A is a diagram schematically illustrating an example of abnormal state data D62 and normal state data D61 in the switchinggear diagnosis system 4 according to the second embodiment. Fig. 13B is a diagram schematically showing an example of a feature quantity abnormality range TD, etc., in the switching gear diagnostic system 4 according to the second embodiment. Fig. 13C is a diagram schematically showing an example of a state when an abnormality occurs and an abnormal location is estimated in the switching gear diagnostic system 4 according to the second embodiment. Fig. 14 is a diagram schematically showing an example of an image related to integrated data D45 output by the output unit 46 in the switching gear diagnostic system 4 according to the first modification.

[0009] First Embodiment [A] Switchgear 1 Before describing the switchgear diagnostic system of the first embodiment, a switchgear 1 to be diagnosed by the switchgear diagnostic system will be described.

[0010] 1A and 1B are diagrams schematically illustrating an example of a switching device 1 to be diagnosed by the switching device diagnostic system according to the first embodiment. Fig. 1A illustrates the case where the switching device is in an open state (shut-off state), and Fig. 1B illustrates the case where the switching device is in a closed state (closed state).

[0011] The switchgear 1 is, for example, a gas circuit breaker, and is configured with a plurality of switchgear components to perform opening and closing operations as shown in Figures 1A and 1B. Here, the switchgear 1 includes a switchgear main body 11 and an operation mechanism 12, and is configured to open or close an electric circuit by performing opening and closing operations.

[0012] In the switchgear 1, the switchgear body 11 includes a pressure vessel 20, a movable contact 21, and a fixed contact 22. The pressure vessel 20 is filled with, for example, insulating gas. The insulating gas is, for example, SF ６ The movable contact portion 21 and the fixed contact portion 22 are arranged coaxially. The movable contact portion 21 is supported by the pressure vessel 20 so as to be movable along the axial direction. The fixed contact portion 22 is fixed to the pressure vessel 20.

[0013] In the switching device 1, the operating mechanism 12 is configured to move the movable contact 21 along the axial direction. When switching to an open state, the operating mechanism 12 separates the movable contact 21 from the fixed contact 22 (see FIG. 1A). When switching to a closed state, the operating mechanism 12 brings the movable contact 21 into contact with the fixed contact 22 (see FIG. 1B). The operating mechanism 12 is, for example, a spring operating mechanism and is configured with a plurality of switching device components (springs, links, etc.).

[0014] [B] Overview of Switchgear Diagnostic System 4 The configuration of the switchgear diagnostic system 4 used to diagnose the operating state of the above-described switchgear 1 (see FIGS. 1A and 1B) will be described.

[0015] FIG. 2 is a block diagram that schematically shows the configuration of the switching device diagnostic system 4 according to the first embodiment.

[0016] The switching device diagnostic system 4 receives as input the measurement data D3 obtained by the sensor 3 detecting the switching device 1 while the switching device 1 is performing an opening and closing operation.

[0017] The sensor 3 is, for example, a displacement sensor attached to the operation mechanism unit 12, and detects information regarding the displacement X (see FIG. 1B ) of the movable contact unit 21 relative to the fixed contact unit 22 as measurement data D3, and the measurement data D3 is input to the switching device diagnostic system 4.

[0018] The switchgear diagnostic system 4 is configured to diagnose the operating state of the switchgear 1 based on the input measurement data D3. Here, the switchgear diagnostic system 4 diagnoses, for example, the operating state of the operation mechanism unit 12 that constitutes the switchgear 1. It is preferable that the switchgear diagnostic system 4 is provided with a noise shield (not shown) inside to prevent it from receiving electromagnetic noise from noise sources including the switchgear 1.

[0019] [C] Configuration of Switchgear Diagnostic System 4 FIG. 3 is a functional block diagram showing the main functions of the switching gear diagnostic system 4 according to the first embodiment.

[0020] 3, the switching gear diagnostic system 4 includes an input unit 41, a storage unit 42, an abnormality occurrence estimation unit 43, an abnormality location estimation unit 44, an estimation result integration unit 45, and an output unit 46. The switching gear diagnostic system 4 is configured to be, for example, a computer, and is configured to function as each unit by the computer executing a program.

[0021] Each component of the switching device diagnostic system 4 will be described in turn.

[0022] [C-1] Input Unit 41 The input unit 41 includes, for example, an input interface, and receives measurement data D3 (time-series data) measured regarding the opening and closing operation of the opening and closing device 1 from the sensor 3. The input unit 41 constantly receives the measurement data D3.

[0023] [C-2] Storage Unit 42 The storage unit 42 includes, for example, a storage device such as a hard disk drive and memory, and stores the measurement data D3 output from the input unit 41 in association with a time axis.

[0024] [C-3] Abnormality Occurrence Estimation Unit 43 The abnormality occurrence estimation unit 43 receives the measurement data D42 stored in the memory unit 42 in association with a time axis. The abnormality occurrence estimation unit 43 then obtains abnormality occurrence estimation data D43 by estimating the presence or absence of an abnormality in the switching device 1 based on the measurement data D42. The abnormality occurrence estimation unit 43 is configured, for example, by a calculation device including a central processing unit (CPU), and executes estimation of abnormality occurrence using a program. Here, the abnormality occurrence estimation unit 43 is, for example, a one-class classifier.

[0025] In this embodiment, after the opening and closing device 1 completes the opening and closing operation, the abnormality occurrence estimation unit 43 estimates whether or not an abnormality has occurred.

[0026] Although details will be described later, the abnormality occurrence estimation unit 43 estimates the presence or absence of an abnormality by machine learning using normal state data D61 obtained in advance in association with a time axis regarding the switching operation when the switching device 1 is in a normal state as learning data. In the present embodiment, the normal state data D61 is information obtained using a physical simulator 5 of the switching device 1. The physical simulator 5 is a mechanical analysis model that can reproduce the operation of the switching device 1 and is capable of simulating the operation of the switching device 1. In the present embodiment, information corresponding to a displacement X (see FIG. 1B ) by which the movable contact unit 21 moves relative to the fixed contact unit 22 when the switching device 1 is in a normal state is obtained as the normal state data D61 by simulation using the physical simulator 5.

[0027] [C-4] Abnormality Location Estimation Unit 44 The abnormality location estimation unit 44 receives the measurement data D42 stored in the memory unit 42 in association with a time axis. Based on the measurement data D42, the abnormality location estimation unit 44 estimates an abnormality location where an abnormality has occurred in a plurality of switching device components that make up the switching device 1, thereby obtaining abnormality location estimation data D44. Like the abnormality occurrence estimation unit 43, the abnormality location estimation unit 44 is configured, for example, by a computing device, and estimates the abnormality location using a program. Here, the abnormality location estimation unit 44 is, for example, a multi-class classifier.

[0028] In this embodiment, the abnormality location estimation unit 44 estimates the abnormality location after the opening and closing operation of the switching device 1 is completed. Furthermore, in this embodiment, the abnormality location estimation unit 44 estimates the abnormality location when the abnormality occurrence estimation unit 43 estimates that an abnormality has occurred in the switching device 1.

[0029] Although details will be described later, the abnormality location estimation unit 44 estimates the abnormality location by machine learning using abnormal state data D62 obtained in advance in association with a time axis regarding the switching operation when the switching device 1 is in an abnormal state as learning data. In this embodiment, the abnormal state data D62 is data obtained using a physical simulator 5 of the switching device 1. Here, information corresponding to a displacement X (see FIG. 1B ) by which the movable contact portion 21 moves relative to the fixed contact portion 22 when the switching device 1 is in an abnormal state is obtained by simulation as the abnormal state data D62.

[0030] [C-5] Estimation Result Integration Unit 45 The estimation result integration unit 45 obtains integrated data D45 by integrating the abnormality occurrence estimation data D43 obtained by the abnormality occurrence estimation unit 43 and the abnormality location estimation data D44 obtained by the abnormality location estimation unit 44. The estimation result integration unit 45, like the abnormality occurrence estimation unit 43 and the abnormality location estimation unit 44, is configured by, for example, a computing device, and performs the above integration using a program.

[0031] [C-6] Output Unit 46 The output unit 46 is configured to output the integrated data D45 obtained by the estimation result integration unit 45. The output unit 46 is, for example, an information output device including a display. In addition to the display, the output unit 46 may further include an information output device such as a printer for printing and displaying output.

[0032] [D] Operation of Switchgear Diagnostic System 4 FIG. 4 is a flow diagram showing the operation of the switching gear diagnostic system 4 according to the first embodiment.

[0033] As shown in FIG. 4 , the switching device diagnostic system 4 of this embodiment sequentially executes the following steps: input of measurement data D3 (ST41), storage of the measurement data D3 (ST42), estimation of abnormality occurrence (ST43), determination of whether or not an abnormality has occurred (ST431), estimation of the location of the abnormality (ST44), integration of the estimation results of the abnormality occurrence and the abnormality location (ST45), and output of the integration result (ST46).

[0034] Each operation of the switching device diagnostic system 4 of this embodiment will be described in order with reference to FIG. 4 as well as the above-mentioned FIG. 3 and the like.

[0035] [D-1] Input of Measurement Data D3 (ST41) As shown in FIG. 4, in the switching device diagnostic system 4 of this embodiment, first, input of measurement data D3 (ST41) is executed.

[0036] The input of the measurement data D3 (ST41) is executed by the input unit 41 (see FIG. 3). The measurement data D3 measured in relation to the opening and closing operation of the opening and closing device 1 is constantly input to the input unit 41 from the sensor 3.

[0037] FIG. 5A is a diagram showing an example of measurement data D3 input to the input unit 41 in the switching device diagnostic system 4 according to the first embodiment.

[0038] 5A, the measurement data D3 is, for example, information relating to the displacement X (see FIG. 1B) of the movable contact 21 relative to the fixed contact 22 over time t. Here, the measurement data D3 is an example of data measured when the displacement X1 of the "fully open state (shut-off state)" changes to the displacement X0 of the "closed state (turned-on state)."

[0039] The input unit 41 converts the measurement data D3 into a digital signal having a preset sampling frequency. Specifically, if the sampling frequency of the measurement data D3 input from the sensor 3 is higher than the preset sampling frequency, the measurement data D3 is subjected to, for example, thinning processing. On the other hand, if the sampling frequency of the measurement data D3 input from the sensor 3 is lower than the preset sampling frequency, the measurement data D3 is subjected to, for example, interpolation processing.

[0040] [D-2] Storing the Measurement Data D3 (ST42) Next, as shown in FIG. 4, the measurement data D3 is stored (ST42).

[0041] The measurement data D3 is stored (ST42) in the storage unit 42 (see FIG. 3).

[0042] FIG. 5B is a diagram showing an example of measurement data D42 stored in the storage unit 42 in the switching device diagnostic system 4 according to the first embodiment.

[0043] As shown in Figure 5B, the measurement data D42 stored in the memory unit 42 (solid line portion in Figure 5B) corresponds to a portion of the measurement data D3 (see Figure 5A; dashed line portion in Figure 5B) output from the input unit 41 to the memory unit 42.

[0044] Specifically, the storage unit 42 defines the time t in the measurement data D3 when the displacement X1 decreases by a predetermined threshold to become a displacement X11 as "0." The storage unit 42 then calculates the period from time t1, a predetermined time before the time t of "0," to time t1 after the predetermined time has elapsed. The period from time t1 to time t1 is set to, for example, a period longer than the period when the operating state of the opening and closing device 1 is normal. The storage unit 42 then extracts a portion of the measurement data D3 corresponding to that period (t0 to t1) and stores the extracted measurement data D42 in association with a time axis. Data in the measurement data D3 that was not extracted is, for example, deleted.

[0045] Although not shown, when an operation is performed to change from a displacement X0 of the "closed state (on state)" to a displacement X1 of the "fully open state (off state)," the storage unit 42 similarly stores measurement data D42 obtained by cutting out a portion of the measurement data D3. In other words, the portion of the measurement data D3 obtained when the on operation or the off operation is performed is cut out and stored as measurement data D42.

[0046] [D-3] Estimation of Abnormality Occurrence (ST43) Next, as shown in FIG. 4, estimation of abnormality occurrence (ST43) is performed.

[0047] The estimation of the occurrence of an abnormality (ST43) is performed by the abnormality occurrence estimation unit 43 (see FIG. 3).

[0048] As described above, the abnormality occurrence estimation unit 43 estimates the presence or absence of an abnormality by machine learning using, as learning data, the normal state data D61 output from the physical simulator 5 of the switching device 1. Here, the normal state data D61 is information corresponding to the displacement X (see FIG. 1B ) by which the movable contact unit 21 moves relative to the fixed contact unit 22 when the switching device 1 is in a normal state.

[0049] In this embodiment, the abnormality occurrence estimation unit 43 obtains the range of features in the feature space when the opening and closing device 1 is in a normal state as a feature normal range TS from the normal state data D61, and estimates whether or not an abnormality has occurred based on the feature normal range TS to obtain abnormality occurrence estimation data D43.

[0050] 6A to 6D are diagrams schematically showing an example of how the abnormality occurrence estimation unit 43 determines the normal feature range TS from the normal state data D61 in the switching device diagnostic system 4 according to the first embodiment.

[0051] 6A , when determining the feature quantity normal range TS, the abnormality occurrence estimation unit 43 uses a plurality of normal state data D61 output from the physical simulator 5 of the switching device 1. Here, the physical simulator 5 determines the normal state data D61 for each of a plurality of parameters (such as dimensions of components constituting the switching device 1 and operating temperature conditions), and outputs the normal state data D61 determined corresponding to each parameter to the abnormality occurrence estimation unit 43. The normal state data D61 is generated by changing the parameters in the physical simulator 5 within the allowable ranges.

[0052] 6B , the abnormality occurrence estimation unit 43 learns the normal waveform pattern PS when the switching device 1 is in a normal state from the plurality of normal state data D61. Here, an example is shown in which a normal waveform pattern PS1 from time t0 to t11 and a normal waveform pattern PS2 from time t12 to t1 are learned as the normal waveform pattern PS.

[0053] 6C , the abnormality occurrence estimation unit 43 calculates the degree of deviation between the normal waveform pattern PS and the normal state data D61. The degree of deviation is, for example, Euclidean distance, and the degree of deviation from the normal waveform pattern PS is calculated as a feature value for each of the multiple normal state data D61. Here, for example, the degree of deviation DP1 between the normal waveform pattern PS1 and the normal state data D61 and the degree of deviation DP2 between the normal waveform pattern PS2 and the normal state data D61 are calculated as the feature values ​​of the normal state data D61.

[0054] 6D , the abnormality occurrence estimation unit 43 calculates a feature normal range TS based on the degrees of deviation DP1, DP2 calculated as the feature amounts of the plurality of normal state data D61. The feature normal range TS is a range in the feature space classified by the feature amounts of the plurality of normal state data D61 (corresponding to the black dots in FIG. 6D ), and includes the feature amounts of the plurality of normal state data D61.

[0055] 7A and 7B are diagrams schematically illustrating an example of how the abnormality occurrence estimation unit 43 in the opening / closing device diagnostic system 4 according to the first embodiment estimates whether an abnormality has occurred based on the normal feature range TS.

[0056] When estimating whether or not an abnormality has occurred, as shown in Figure 7A, the abnormality occurrence estimation unit 43 calculates the feature quantities of the measurement data D42 by calculating the degree of deviation DP1, DP2 between the measurement data D42 stored in the memory unit 42 and the normal waveform pattern PS.

[0057] Here, as shown in Figure 7B, for example, a value T421 calculated as the degree of deviation DP1 between the normal waveform pattern PS1 and the measurement data D42, and a value T421 calculated as the degree of deviation DP2 between the normal waveform pattern PS2 and the measurement data D42 are calculated as the feature T42 of the measurement data D42 in the feature space.

[0058] 7B , the abnormality occurrence estimation unit 43 estimates whether the position of the feature quantity T42 of the measurement data D42 is within the feature quantity normal range TS in the feature quantity space. When the feature quantity T42 of the measurement data D42 is within the feature quantity normal range TS in the feature quantity space (see FIG. 7B ), the abnormality occurrence estimation unit 43 estimates that no abnormality has occurred in the switching device 1. On the other hand, when the feature quantity T42 of the measurement data D42 is outside the feature quantity normal range TS in the feature quantity space (not shown), the abnormality occurrence estimation unit 43 estimates that an abnormality has occurred in the switching device 1.

[0059] 4, if it is determined that an abnormality has occurred (Yes in ST431), the abnormality location is estimated (ST44). On the other hand, if it is determined that no abnormality has occurred (No in ST431), the abnormality location estimation (ST44) is not performed, and the estimation results of the abnormality occurrence and the abnormality location are integrated (ST45).

[0060] The estimation of the abnormal part (ST44) is performed by the abnormal part estimation unit 44 (see FIG. 3).

[0061] As described above, the abnormality location estimation unit 44 estimates the abnormality location by machine learning using, as learning data, the abnormal state data D62 output from the physical simulator 5 of the switching device 1. Here, the abnormal state data D62 is information corresponding to the displacement X (see FIG. 1B ) by which the movable contact portion 21 moves relative to the fixed contact portion 22 when the switching device 1 is in an abnormal state.

[0062] In this embodiment, the abnormality location estimation unit 44 obtains, in the feature space, a range of feature values ​​when each of the plurality of switching device components is in an abnormal state as a feature value abnormality range TD in association with the plurality of switching device components. Then, the abnormality location estimation unit 44 estimates an abnormality location based on the plurality of feature value abnormality ranges TD obtained for each of the plurality of switching device components, and obtains abnormality location estimation data D44.

[0063] 8A to 8D are diagrams that schematically show an example of how the abnormality location estimation unit 44 determines the feature abnormality range TD from the abnormal state data D62 in the switching device diagnostic system 4 according to the first embodiment.

[0064] When calculating the feature quantity abnormality range TD, the abnormality location estimation unit 44 uses a plurality of abnormal state data D62 output from the physical simulator 5 of the switching device 1, as shown in Fig. 8A . Here, the physical simulator 5 calculates the abnormal state data D62 for each of a plurality of parameters (e.g., dimensions of components constituting the switching device 1, operating temperature conditions, etc.), and outputs the abnormal state data D62 calculated corresponding to each parameter to the abnormality location estimation unit 44. The abnormal state data D62 is generated by changing the parameters outside the allowable range in the physical simulator 5. The abnormal state data D62 is generated for each of a plurality of switching device components, and Fig. 8A illustrates, as an example, a plurality of abnormal state data D62a for a first switching device component and a plurality of abnormal state data D62b for a second switching device component.

[0065] 8B , the abnormal part estimation unit 44 learns, from the plurality of abnormal state data D62, an abnormal waveform pattern PD when each of the plurality of switching device components is in an abnormal state. Here, an example is shown in which an abnormal waveform pattern PD1 between time t0 and time t11d and an abnormal waveform pattern PD2 between time t12d and time t1d are learned as the abnormal waveform patterns PD.

[0066] 8C , the abnormal portion estimation unit 44 calculates the degree of deviation between the abnormal waveform pattern PD and the abnormal state data D62. The degree of deviation is, for example, Euclidean distance, and the degree of deviation from the abnormal waveform pattern PD is calculated as a feature value for each of the multiple abnormal state data D62. Here, for example, the degree of deviation DP1b between the abnormal waveform pattern PD1 and the abnormal state data D62a, and the degree of deviation DP2b between the abnormal waveform pattern PD2 and the abnormal state data D62a are calculated as feature values ​​for the abnormal state data D62a. Although not shown in the figure, the feature values ​​for the abnormal state data D62b are also calculated in the same way as the feature values ​​for the abnormal state data D62a.

[0067] Then, as shown in FIG. 8D , the abnormality location estimation unit 44 calculates a feature abnormality range TD based on the degrees of deviation DP1b, DP2b calculated as the feature quantities of the plurality of abnormal state data D62. The feature abnormality range TD is a range in the feature space classified by the feature quantities of the plurality of abnormal state data D62 (corresponding to the black dots in FIG. 8D ), and includes the feature quantities of the plurality of abnormal state data D62. The feature abnormality range TD is calculated for each of the plurality of switching device components. FIG. 8D illustrates an example in which a feature abnormality range TD1 for a first switching device component and a feature abnormality range TD2 for a second switching device component are calculated as the feature abnormality range TD.

[0068] 9A and 9B are diagrams schematically illustrating an example of how the abnormality location estimation unit 44 estimates the abnormality location based on the feature abnormality range TD in the opening and closing device diagnostic system 4 according to the first embodiment.

[0069] When estimating the abnormal location, as shown in FIG. 9A, the abnormal location estimation unit 44 obtains the feature quantity of the measurement data D42 by determining the degree of deviation DP1b, DP2b between the measurement data D42 stored in the memory unit 42 and the abnormal waveform pattern PD.

[0070] Here, as shown in FIG. 9B, for example, a value TD421 calculated as the degree of deviation DP1b between the abnormal waveform pattern PD1 and the measurement data D42, and a value TD422 calculated as the degree of deviation DP2b between the abnormal waveform pattern PD2 and the measurement data D42 are calculated as the feature TD42 of the measurement data D42 in the feature space.

[0071] 9B , the abnormality location estimation unit 44 estimates whether the position of the feature TD42 of the measurement data D42 is within a feature abnormality range TD in the feature space. When the feature TD42 of the measurement data D42 is within a feature abnormality range TD1 among the multiple feature abnormality ranges TD in the feature space, the abnormality location estimation unit 44 estimates that the first switching device component is the abnormal location. When the feature TD42 of the measurement data D42 is within a feature abnormality range TD2 among the multiple feature abnormality ranges TD in the feature space, the abnormality location estimation unit 44 estimates that the second switching device component is the abnormal location (not shown).

[0072] Although not shown in the figure, if the feature of the measurement data D42 is not within the multiple feature abnormality ranges TD for each of the multiple switching device components, the abnormality location estimation unit 44 determines whether the degree of deviation between the feature of the measurement data D42 and the feature of the measurement data D42 for the multiple switching device components is equal to or less than a predetermined threshold value.

[0073] Then, among the feature amount abnormality ranges TD related to the plurality of switching gear components, if there is one feature amount abnormality range TD in which the degree of deviation from the feature amount of the measurement data D42 is equal to or less than a predetermined threshold, the switching gear component related to that one feature amount abnormality range TD is estimated to be the abnormal location (not shown). For example, among the feature amount abnormality ranges TD1 and TD2, if the degree of deviation from the feature amount abnormality range TD1 is equal to or less than a threshold, but the degree of deviation from the feature amount abnormality range TD2 exceeds the threshold, the first switching gear component is estimated to be the abnormal location.

[0074] Furthermore, when the feature quantities of the measurement data D42 are not within the plurality of feature quantity abnormality ranges TD for each of the plurality of switching device components, and all of the degrees of deviation between each of the feature quantity abnormality ranges for the plurality of switching device components and the feature quantities of the measurement data D42 exceed a threshold value (not shown), the abnormality location estimation unit 44 estimates that the abnormality location is unknown (not shown). Note that the degrees of deviation are, for example, Euclidean distances.

[0075] [D-5] Integration of the estimation results of abnormality occurrence and abnormal location (ST45) After the estimation of the abnormal location (ST44) etc. is performed, the estimation results of abnormality occurrence and abnormal location are integrated (ST45) as shown in FIG.

[0076] The integration of the estimation results of the abnormality occurrence and the abnormality location (ST45) is executed by the estimation result integration unit 45 (see FIG. 3 ). The estimation result integration unit 45 integrates the abnormality occurrence estimation data D43 obtained by the abnormality occurrence estimation unit 43 and the abnormality location estimation data D44 obtained by the abnormality location estimation unit 44 to obtain integrated data D45.

[0077] Specifically, when the abnormality occurrence estimation unit 43 estimates that no abnormality has occurred, the estimation result integrating unit 45 integrates the estimation result of no abnormality with the estimation result of no abnormal location. When the abnormality occurrence estimation unit 43 estimates that no abnormality has occurred and the abnormal location estimation unit 44 estimates that one of the multiple switching device components is the abnormal location, the estimation result integrating unit 45 integrates the estimation result of an abnormality with the estimation result in which the abnormal location is one switching device component. When the abnormality occurrence estimation unit 43 estimates that no abnormality has occurred and the abnormal location estimation unit 44 estimates that an unknown switching device component other than the multiple switching device components is the abnormal location, the estimation result integrating unit 45 integrates the estimation result of an abnormality with the estimation result in which the abnormal location is an unknown switching device component.

[0078] [D-6] Output of Estimation Results (ST46) Next, as shown in FIG. 4, output of the integration results (ST46) is executed.

[0079] The output of the estimation result (ST46) is executed by the output unit 46 (see FIG. 3).

[0080] The output unit 46 displays, for example, an image relating to the integrated data D45 obtained by the estimation result integration unit 45 on the screen of a display.

[0081] FIG. 10 is a diagram schematically showing an example of an image related to the integrated data D45 output by the output unit 46 in the switching device diagnostic system 4 according to the first embodiment.

[0082] As shown in FIG. 10, an image M45 relating to the integrated data D45 includes an image M451 relating to the estimated result of the occurrence of an abnormality and an image M452 relating to the estimated result of the abnormal location.

[0083] Specifically, when the estimation result of the abnormality occurrence indicates that an abnormality is present, the image M451 relating to the estimation result of the abnormality occurrence is, for example, text information of "Operating status: Abnormal". Although not shown in the figure, when the estimation result of the abnormality occurrence indicates that an abnormality is not present, text information of "Operating status: Normal" is displayed as the image M451 relating to the estimation result of the abnormality occurrence.

[0084] The image M452 regarding the estimation result of the abnormal location includes, for example, if the estimation result indicates that part Y (first opening / closing device component) is the abnormal location, text information M452a stating "Abnormality in part Y" and a schematic diagram M452b emphasizing the position of part Y in the opening / closing device 1 (the area indicated by the dashed dotted line in Figure 10).

[0085] In addition, the image M452 relating to the estimation result of the abnormality location may include text information M432c such as "inspection and replacement of part Y" as maintenance and inspection information. The maintenance and inspection information may include, for example, inspection items corresponding to the estimation result of the abnormality location. In addition, the maintenance and inspection information may include arrangement information for parts to be arranged for the maintenance and inspection work. Although not shown in the figure, if the estimation result of the abnormality occurrence is that no abnormality exists, text information M432c such as "none" is displayed as the maintenance and inspection information.

[0086] [E] Summary As described above, the switchinggear diagnostic system 4 of this embodiment diagnoses the operating state of the switchinggear 1, which is configured to perform a switching operation by a plurality of switchinggear components, each time the switching operation is performed. In the switchinggear diagnostic system 4 of this embodiment, in diagnosing the operating state of the switchinggear 1, the abnormality occurrence estimation unit 43 estimates whether or not an abnormality has occurred in the switchinggear 1, thereby obtaining abnormality occurrence estimation data D43, and the abnormality location estimation unit 44 estimates an abnormality location where an abnormality has occurred in the plurality of switchinggear components, thereby obtaining abnormality location estimation data D44. Then, the estimation result integration unit 45 integrates the abnormality occurrence estimation data D43 and the abnormality location estimation data D44 to obtain integrated data D45, and the output unit 46 outputs the integrated data D45.

[0087] As a result, in this embodiment, it is possible to easily and accurately grasp the estimated result of the abnormality location in addition to the signs of abnormality in the switching device. As a result, according to this embodiment, it is possible to replace parts, etc. before the current interruption capability of the switching device 1 is adversely affected. Furthermore, since the work of identifying the abnormality location can be simplified, it is possible to shorten the time required for the part replacement work and also the downtime required to stop the switching device 1 for the replacement work, etc.

[0088] In this embodiment, the abnormality occurrence estimation unit 43 estimates the presence or absence of an abnormality by machine learning using, as learning data, normal state data D61 obtained in advance in association with a time axis regarding the opening and closing operation when the switching device 1 is in a normal state. Furthermore, the abnormality location estimation unit 44 estimates the abnormal location by machine learning using, as learning data, abnormal state data D62 obtained in advance in association with a time axis regarding the opening and closing operation when the switching device 1 is in an abnormal state. Therefore, in this embodiment, it is possible to estimate the occurrence of an abnormality and the abnormal location with high sensitivity.

[0089] In the present embodiment, the normal state data D61 and the abnormal state data D62 are data obtained using the physical simulator 5 of the switching device 1. Therefore, in the present embodiment, the normal state data D61 and the abnormal state data D62 required for machine learning can be sufficiently obtained without operating the actual switching device 1 to perform machine learning.

[0090] Second Embodiment [A] Configuration of Switchgear Diagnostic System 4 FIG. 11 is a functional block diagram showing the main functions of a switching gear diagnostic system 4 according to a second embodiment.

[0091] As shown in Fig. 11 , in the switching gear diagnostic system 4 of the present embodiment, the data used by the abnormality location estimation unit 44 when estimating the abnormality location is different from that in the first embodiment (see Fig. 3 ). Except for this point and related points, the present embodiment is similar to the first embodiment. Therefore, in the description of the present embodiment, descriptions of matters that overlap with the description of the above embodiment will be omitted as appropriate.

[0092] In this embodiment, as shown in FIG. 11, the abnormality location estimation unit 44 is configured to estimate the abnormality location by machine learning using the normal state data D61 as learning data in addition to the abnormal state data D62.

[0093] [B] Operation of Switchgear Diagnostic System 4 FIG. 12 is a flow diagram showing the operation of the switching gear diagnostic system 4 according to the second embodiment.

[0094] 12 , the switching gear diagnostic system 4 of the present embodiment sequentially executes the following steps: input of measurement data D3 (ST41), storage of the measurement data D3 (ST42), estimation of an abnormality occurrence (ST43), estimation of the abnormality location (ST44), integration of the estimation results of the abnormality occurrence and the abnormality location (ST45), and output of the integration result (ST46). Unlike the first embodiment (see FIG. 3 ), the present embodiment does not execute the determination of whether an abnormality has occurred (ST431). That is, in the present embodiment, estimation of the abnormality location (ST45) is executed regardless of whether an abnormality has been estimated (ST43).

[0095] [B-1] Estimation of abnormality occurrence from input of measurement data D3 (ST41) (ST43) In this embodiment, input of measurement data D3 (ST41), storage of measurement data D3 (ST42), and estimation of abnormality occurrence (ST43) are performed in the same manner as in the first embodiment.

[0096] [B-2] Estimation of Abnormal Location (ST44) In this embodiment, the content of estimation of abnormal location (ST44) differs from that in the first embodiment.

[0097] 13A to 13C are diagrams schematically illustrating an example of the state in which the abnormality location estimation unit 44 in the opening / closing device diagnostic system 4 according to the second embodiment obtains the characteristic quantity abnormality range TD and estimates the abnormality location.

[0098] When determining the feature quantity abnormal range TD, the abnormal point estimation unit 44 uses normal state data D61 in addition to a plurality of abnormal state data D62 output from the physical simulator 5 of the switching device 1, as shown in FIG. 8A.

[0099] The abnormal part estimation unit 44 learns an abnormal waveform pattern PD from the plurality of abnormal state data D62 and the plurality of normal state data D61 (see FIG. 8B ). Here, as in the first embodiment (see FIG. 8B ), for example, an abnormal waveform pattern PD1 between time t0 and time t11d and an abnormal waveform pattern PD2 between time t12d and time t1d are learned as the abnormal waveform pattern PD.

[0100] The abnormal portion estimation unit 44 then determines the degree of deviation between the abnormal waveform pattern PD and the normal state data D61, in addition to the degree of deviation between the abnormal waveform pattern PD and the abnormal state data D62 (D62a, D62b). That is, similar to the first embodiment (see FIG. 8C ), the degree of deviation DP1b between the abnormal waveform pattern PD1 and the abnormal state data D62 (D62a, D62b) and the degree of deviation DP2b between the abnormal waveform pattern PD2 and the abnormal state data D62 (D62a, D62b) are determined as feature quantities of the abnormal state data D62 (D62a, D62b). Furthermore, the degree of deviation DP1b between the abnormal waveform pattern PD1 and the normal state data D61 and the degree of deviation DP2b between the abnormal waveform pattern PD2 and the normal state data D61 are determined as feature quantities of the normal state data D61.

[0101] 13B , the abnormality location estimation unit 44 calculates a feature quantity abnormality range TD based on the degree of deviation calculated as the feature quantity of each of the plurality of abnormal state data D62. Here, a case is illustrated in which a feature quantity abnormality range TD1 related to the first switching device component and a feature quantity abnormality range TD2 related to the second switching device component are calculated as the feature quantity abnormality range TD. In addition, in this embodiment, the abnormality location estimation unit 44 calculates a feature quantity normal range TS based on the degree of deviation calculated as the feature quantity of each of the plurality of normal state data D61.

[0102] Thereafter, when estimating the abnormal location, the abnormal location estimation unit 44, as in the first embodiment, calculates the degree of deviation between the measurement data D42 stored in the memory unit 42 and the abnormal waveform pattern PD, thereby calculating the feature value of the measurement data D42 (see FIG. 9A).

[0103] Here, as shown in FIG. 13C, for example, a value TD421 obtained as the degree of deviation DP1b between the abnormal waveform pattern PD1 and the measurement data D42, and a value TD422 obtained as the degree of deviation DP2b between the abnormal waveform pattern PD2 and the measurement data D42 are obtained as the feature quantity TD42 of the measurement data D42.

[0104] Then, similarly to the first embodiment (see FIG. 9B), the abnormality part estimation unit 44 estimates the abnormality part based on the positional relationship between the feature amount TD42 of the measurement data D42 and the feature amount abnormal range TD.

[0105] Although not shown in the figure, if the feature amount TD42 of the measurement data D42 is located within the feature amount normal range TS, for example, the abnormality portion estimation unit 44 estimates that no abnormality has occurred (i.e., normal). Even if the feature amount of the measurement data D42 is not within the feature amount normal range TS, the abnormality portion estimation unit 44 may estimate that no abnormality has occurred if the degree of deviation between the feature amount normal range TS and the feature amount of the measurement data D42 is equal to or less than a predetermined threshold.

[0106] [B-3] Integration of the estimation results of abnormality occurrence and abnormal location (ST45) After the estimation of the abnormal location (ST44) is performed, the estimation results of the abnormality occurrence and abnormal location are integrated (ST45) as shown in FIG. 11 .

[0107] The integration of the estimation results of the abnormality occurrence and the abnormality location (ST45) is executed by the estimation result integration unit 45 (see FIG. 3 ). The estimation result integration unit 45 integrates the abnormality occurrence estimation data D43 obtained by the abnormality occurrence estimation unit 43 and the abnormality location estimation data D44 obtained by the abnormality location estimation unit 44 to obtain integrated data D45.

[0108] In the present embodiment, even when the estimation result of the abnormality occurrence estimation unit 43 indicates that no abnormality has occurred, the abnormality location estimation unit 44 may estimate an abnormal location. Also, even when the estimation result of the abnormality occurrence estimation unit 43 indicates that an abnormality has occurred, the estimation result of the abnormality location estimation unit 44 may indicate that no abnormality has occurred. In this way, when the estimation result of the abnormality occurrence estimation unit 43 and the estimation result of the abnormality location estimation unit 44 differ from each other, the estimation result integration unit 45 preferentially adopts, for example, the estimation result of the abnormality occurrence estimation unit 43.

[0109] [B-4] Output of Estimation Result (ST46) Next, the integration result is output (ST46) as shown in Fig. 11. The estimation result is output (ST46) by the output unit 46, as in the first embodiment.

[0110] [C] Summary As described above, in the switching gear diagnostic system 4 of the present embodiment, the abnormality location estimation unit 44 estimates an abnormality location by machine learning using the normal state data D61 as learning data in addition to the abnormal state data D62. Therefore, in the present embodiment, estimation of an abnormality occurrence is essentially performed not only by the abnormality occurrence estimation unit 43 but also by the abnormality location estimation unit 44. Therefore, in the present embodiment, it is possible to improve the accuracy of estimation.

[0111] <Modifications> Modifications of the above embodiment will now be described.

[0112] [A] Modification 1 In the above embodiment, the case where the diagnosis target of the switching gear diagnostic system 4 is one switching gear 1 has been exemplified, but this is not limiting. The diagnosis target of the switching gear diagnostic system 4 may be a plurality of switching gears 1. In this case, the output unit 46 of the switching gear diagnostic system 4 outputs the integrated data D45 acquired for each of the plurality of switching gears 1. At this time, the output unit 46 displays, for example, an image in which each of the integrated data D45 of the plurality of switching gears 1 is lined up on the screen of the display.

[0113] FIG. 14 is a diagram schematically showing an example of an image related to the integrated data D45 output by the output unit 46 in the switching device diagnostic system 4 according to the first modification.

[0114] 14 illustrates an example in which the switching gears 1 to be diagnosed by the switching gear diagnostic system 4 are the first switching gear 1a and the second switching gear 1b. In this case, as shown in Fig. 14, the image M45 related to the integrated data D45 includes an image M451 related to the estimated result of abnormality occurrence and an image M452 related to the estimated result of abnormality location for the first switching gear 1a and the second switching gear 1b, as in the case of the above embodiment (see Fig. 10). In addition, as shown in Fig. 14, the image M45 related to the integrated data D45 includes an image M453 that shows all the estimated results in a table format.

[0115] Therefore, in this modification, even if the diagnosis target is a plurality of switching devices 1, it is possible to quickly and accurately grasp the estimation results for the plurality of switching devices 1.

[0116] [B] Modification 2 In the above embodiment, the switchinggear diagnostic system 4 may be configured using a cloud system. This eliminates the need to install the switchinggear diagnostic system 4, which is configured from a physical computing device or the like, near the switchinggear 1, making it possible to easily diagnose the operating state of the switchinggear 1.

[0117] [C] Modification 3 In the above embodiment, the abnormality occurrence estimation unit 43 may be configured to perform an operation using a convolution kernel on the measurement data D42 stored in the storage unit 42, and then estimate the presence or absence of an abnormality. The operation using the convolution kernel may be performed multiple times. In other words, the result of an aggregation operation performed by applying multiple convolution kernels to the measurement data D42 may be obtained as the feature quantity of the measurement data D42. In this case, the accuracy of the feature quantity of the measurement data D42 is improved, making it possible to accurately estimate the presence or absence of an abnormality.

[0118] [D] Modification 4 In the above embodiment, the switchgear 1 to be diagnosed by the switchgear diagnostic system 4 is a circuit breaker, but this is not limiting. The switchgear 1 may be a device other than a circuit breaker, such as a disconnecting switch or a grounding switch.

[0119] In the above embodiment, the operating mechanism 12 of the opening and closing device 1 is a spring operating mechanism, but this is not limiting. The operating mechanism 12 may be another mechanism such as a hydraulic operating mechanism.

[0120] [E] Modification 5 In the above embodiment, the measurement data D3 is information relating to the displacement X (see FIG. 1B ) of the movable contact portion 21 relative to the fixed contact portion 22 over time t. However, this is not limiting. The measurement data D3 (time-series data) may be other data as long as it is information measured regarding the opening and closing operation of the opening and closing device 1.

[0121] For example, in the case where the operation mechanism unit 12 performs the opening and closing operation of the switching device 1 using electrical equipment such as an electromagnetic actuator (not shown) or an electric motor (not shown), the current measured in the electrical equipment such as the electromagnetic actuator (not shown) or the electric motor (not shown) may be used as the measurement data D3. Furthermore, the acceleration measured using an acceleration sensor (not shown) during the opening and closing operation of the switching device 1 may be used as the measurement data D3. Furthermore, the sound intensity measured using a sound sensor (not shown) during the opening and closing operation of the switching device 1 may be used as the measurement data D3. Other operating environment values ​​(control voltage, operating oil pressure, temperature, time interval from the previous operation) may also be used as the measurement data D3.

[0122] [F] Modification 6 In the above embodiment, the abnormality location estimation unit 44 calculates the characteristic quantity abnormality range TD for two switching device components, but this is not limiting. The characteristic quantity abnormality range TD may be calculated for three or more switching device components, or the characteristic quantity abnormality range TD may be calculated for a single switching device component.

[0123] [G] Modification 7 In the above embodiment, the deviation degree is the Euclidean distance, but this is not limiting. The deviation degree may be a DTW (Dynamic Time Warping) distance or the like.

[0124] [H] Modification 8 In the above embodiment, the image M452 relating to the estimation result of the abnormal location shows information about the switching device component (component Y) estimated to be the abnormal location, but this is not limiting. Information about switching device components other than the switching device component (component Y) estimated to be the abnormal location may also be output as the image M452 relating to the estimation result of the abnormal location.

[0125] [I] Modification 9 In the above embodiment, the abnormality occurrence estimation data D43 may include information that quantifies the estimation result of the abnormality occurrence, and similarly, the abnormality location estimation data D44 may include information that quantifies the estimation result of the abnormal location. Then, the image M45 related to the integrated data D45 output by the output unit 46 may include this quantified information. The quantified information makes it possible to grasp the extent of the abnormality occurrence and the abnormal location.

[0126] The quantified information is the probability of normality, the probability of abnormality, etc., and is calculated based on the positional relationship between the feature amount T42 of the measurement data D42 calculated by the abnormality occurrence estimation unit 43 and the normal feature amount range TS (see FIG. 7B ). For example, the probability of normality increases as the feature amount T42 of the measurement data D42 approaches the center within the normal feature amount range TS. For example, the probability of abnormality increases as the feature amount T42 of the measurement data D42 moves away from the normal feature amount range TS. Images showing statuses such as "healthy," "caution," and "investigation required" according to the probability may be included in the image M45 related to the integrated data D45.

[0127] Other quantified information includes the probability of abnormality occurrence at the abnormal location, which is calculated based on the positional relationship between the feature value TD42 of the measurement data D42 determined by the abnormal location estimation unit 44 and the feature value abnormality range TD (see FIG. 9B ). For example, the probability of abnormality occurrence at the abnormal location increases as the feature value T42 of the measurement data D42 approaches the center within the feature value abnormality range TD.

[0128] Note that, as in the first modification, when the diagnosis targets of the switching gear diagnostic system 4 are multiple switching gears 1, the image M45 related to the integrated data D45 may include quantified information regarding the estimation results of each of the multiple switching gears 1. In this case, for example, when the numerical value related to one switching gear 1 is larger than the numerical values ​​related to other switching gears 1, the output unit 46 may output the image M45 so as to preferentially display the estimation result related to one switching gear 1 over the estimation results related to the other switching gears 1.

[0129] <Others> Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their modifications are intended to be included in the scope of the invention and its equivalents as defined in the claims, as well as in the scope and spirit of the invention.

[0130] 1: Switchgear, 1a: First Switchgear, 1b: Second Switchgear, 3: Sensor, 4: Switchgear Diagnosis System, 5: Physical Simulator, 11: Switchgear Main Body, 12: Operation Mechanism, 20: Pressure Vessel, 21: Moving Contact, 22: Fixed Contact, 41: Input Unit, 42: Memory Unit, 43: Abnormality Occurrence Estimation Unit, 44: Abnormal Location Estimation Unit, 45: Estimation Result Integration Unit, 46: Output Unit, D3: Measurement Data, D42: Measurement Data, D43: Abnormality Occurrence Estimation Data, D44: Abnormality Normal location estimation data, D45: integrated data, D61: normal state data, D62: abnormal state data, D62a: abnormal state data, D62b: abnormal state data, PD: abnormal waveform pattern, PD1: abnormal waveform pattern, PD2: abnormal waveform pattern, PS: normal waveform pattern, PS1: normal waveform pattern, PS2: normal waveform pattern, TD: feature amount abnormal range, TD1: feature amount abnormal range, TD2: feature amount abnormal range, TS: feature amount normal range, X: displacement

Claims

1. A switchinggear diagnosis system for diagnosing an operating state of a switchinggear constituted by a plurality of switchinggear components so as to perform a switching operation, comprising: a memory unit for storing measurement data measured regarding the switching operation while the switchinggear is performing the switching operation in association with a time axis; an abnormality occurrence estimation unit for obtaining abnormality occurrence estimation data by estimating the presence or absence of an abnormality occurring in the switchinggear based on the measurement data stored in the memory unit; an abnormality location estimation unit for obtaining abnormality location estimation data by estimating an abnormality location where an abnormality is occurring in the plurality of switchinggear components based on the measurement data stored in the memory unit; an estimation result integration unit for obtaining integrated data by integrating the abnormality occurrence estimation data obtained by the abnormality occurrence estimation unit and the abnormality location estimation data obtained by the abnormality location estimation unit; and an output unit for outputting the integrated data obtained by the estimation result integration unit, wherein the abnormality occurrence estimation unit estimates the presence or absence of an abnormality by machine learning using normal state data obtained in advance in association with the time axis regarding the switching operation when the switchinggear is in a normal state as learning data, the abnormality location estimation unit estimates the abnormality location by machine learning using abnormal state data, which is obtained in advance in association with a time axis regarding the switching operation when the switching gear is in an abnormal state, as learning data.

2. The switching gear diagnosis system according to claim 1, wherein the abnormality location estimation unit estimates the abnormality location when the abnormality occurrence estimation unit estimates that an abnormality has occurred in the switching gear.

3. A switching device diagnosis system as described in claim 1 or 2, wherein, after the switching device has completed the switching operation, the abnormality occurrence estimation unit estimates whether or not the abnormality has occurred, and the abnormality location estimation unit estimates the location of the abnormality.

4. A switching gear diagnosis system as claimed in any one of claims 1 to 3, wherein the abnormality occurrence estimation unit determines, in a feature space, a range of features when the switching gear is in a normal state as a feature normal range from the normal state data, and estimates the presence or absence of the abnormality based on the feature normal range.

5. The switching gear diagnosis system according to claim 4, wherein the abnormality occurrence estimation unit learns a normal waveform pattern when the switching gear is in a normal state from the normal state data, and determines the normal range of the feature based on the degree of deviation between the normal waveform pattern and the normal state data.

6. The switching gear diagnosis system according to claim 5, wherein the abnormality occurrence estimation unit determines a feature of the measurement data based on the degree of deviation between the measurement data stored in the memory unit and the normal waveform pattern, and estimates that no abnormality has occurred in the switching gear when the feature of the measurement data is within the normal range of the feature.

7. A switching gear diagnosis system as described in any one of claims 1 to 6, wherein the abnormality location estimation unit determines, in a feature space, a range of features when each of the multiple switching gear components is in an abnormal state as a feature abnormality range associated with the multiple switching gear components, and estimates the abnormality location based on the multiple feature abnormality ranges determined for each of the multiple switching gear components.

8. The switching gear diagnosis system according to claim 7, wherein the abnormality location estimation unit learns an abnormal waveform pattern when each of the multiple switching gear components is in an abnormal state from the abnormal state data, and determines the multiple feature abnormality ranges for each of the multiple switching gear components based on the degree of deviation between the abnormal waveform pattern and the abnormal state data.

9. The switching gear diagnosis system according to claim 8, wherein the abnormality location estimation unit determines a feature of the measurement data based on the degree of deviation between the measurement data stored in the memory unit and the abnormal waveform pattern, and estimates the abnormal location according to the degree of deviation between the feature of the measurement data and the feature abnormality range for each of the multiple switching gear components.

10. The abnormality location estimation unit, when the feature of the measurement data is within the feature abnormality range for one of the multiple switching gear components, estimates that one switching gear component is the abnormal location, and when the feature of the measurement data is not within the multiple feature abnormality ranges for each of the multiple switching gear components, estimates that one switching gear component related to a feature abnormality range among the multiple switching gear components in which the degree of deviation between the feature of the measurement data is below a predetermined threshold, is the abnormal location. The switching gear diagnosis system described in claim 9.

11. The switching gear diagnosis system of claim 10, wherein the abnormality location estimation unit estimates that the abnormality location is unknown when the feature of the measurement data is not within the multiple feature abnormality ranges for each of the multiple switching gear components and the degree of deviation between each of the feature abnormality ranges for the multiple switching gear components and the feature of the measurement data exceeds the threshold value.

12. A switching device diagnosis system according to any one of claims 1 to 11, wherein the abnormality location estimation unit estimates the abnormality location by machine learning using the normal state data as learning data in addition to the abnormal state data.

13. A switching device diagnosis system according to any one of claims 1 to 12, wherein the abnormality occurrence estimation unit performs a calculation using a convolution kernel on the measurement data stored in the memory unit, and then estimates the presence or absence of the abnormality.

14. A switchgear diagnosis system according to any one of claims 1 to 13, wherein the normal state data and the abnormal state data are data obtained using a physical simulator of the switchgear.

15. The switching device diagnosis system according to any one of claims 1 to 14, wherein the output unit outputs the integrated data as a numerical value.

Images