WO2022130626A1 - Dispositif de diagnostic d'anomalies, programme et procédé de diagnostic d'anomalies - Google Patents

Dispositif de diagnostic d'anomalies, programme et procédé de diagnostic d'anomalies Download PDF

Info

Publication number
WO2022130626A1
WO2022130626A1 PCT/JP2020/047483 JP2020047483W WO2022130626A1 WO 2022130626 A1 WO2022130626 A1 WO 2022130626A1 JP 2020047483 W JP2020047483 W JP 2020047483W WO 2022130626 A1 WO2022130626 A1 WO 2022130626A1
Authority
WO
WIPO (PCT)
Prior art keywords
operating state
feature amount
abnormality diagnosis
setting information
electric device
Prior art date
Application number
PCT/JP2020/047483
Other languages
English (en)
Japanese (ja)
Inventor
新治 村田
久幸 折田
広考 高橋
Original Assignee
株式会社日立製作所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社日立製作所 filed Critical 株式会社日立製作所
Priority to PCT/JP2020/047483 priority Critical patent/WO2022130626A1/fr
Publication of WO2022130626A1 publication Critical patent/WO2022130626A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R21/00—Arrangements for measuring electric power or power factor

Definitions

  • the present invention relates to an abnormality diagnosis device, a program, and an abnormality diagnosis method for electrical equipment.
  • NILM Non-Intrusive Load Monitoring
  • Patent Document 1 a measurement sensor installed near the power supply line inlet of a power consumer and data extraction for extracting data on the currents of fundamental waves and harmonics and their phases with respect to their voltages from the measurement data detected by the measurement sensors.
  • a means and a pattern recognition means for estimating the operating state of the electric equipment used by the electric power consumer based on the data on the currents of the fundamental waves and harmonics from the data extraction means and the phase with respect to their voltages.
  • An electrical equipment monitoring system characterized by being equipped is described.
  • Patent Document 2 describes a device state detecting device for detecting the state of one or a plurality of devices, which is a measuring means for measuring a physical quantity of an environment in which the device is installed, and a measured value measured by the measuring means.
  • a feature amount calculation means for calculating the feature amount of the above, a storage means for storing the feature amount for each device and the corresponding device state as dictionary data in advance, and a feature amount calculated by the feature amount calculation means are searched.
  • the device state is characterized by having a device state detecting means for searching the feature amount stored in the dictionary data as a key and detecting the device state based on the device state corresponding to the feature amount specified as the search result. The detector is described.
  • Patent Document 1 and Patent Document 2 in order to estimate (monitor) the operating state, it is necessary to give and learn the operating state of the electric device as teacher data in advance.
  • data on the operating state of the electric device in the abnormal state is required, but it is often difficult to acquire data on the abnormal state of the electric device.
  • the present invention has been made in view of such a background, and provides an abnormality diagnosis device, a program, and an abnormality diagnosis method capable of diagnosing an abnormality even when there is no data at the time of abnormality of an electric device. Is the subject.
  • the abnormality diagnosis device includes a power data collection unit that collects power data of an electric device, a feature amount extraction unit that acquires a feature amount from the collected power data, and the above-mentioned.
  • the operating state feature amount data including the relationship between the operating state of the electric device and the feature amount of the power data in the operating state, the operating state estimation unit that estimates the operating state of the electric device from the acquired feature amount, and the electricity.
  • a setting information collection unit that collects setting information that sets the operating state of the device, an abnormality diagnosis unit that determines that the electrical device is abnormal if the estimated operating state does not correspond to the collected setting information, To prepare for.
  • an abnormality diagnosis device a program, and an abnormality diagnosis method capable of diagnosing an abnormality even when there is no data at the time of abnormality of an electric device. Issues, configurations and effects other than those described above will be clarified by the description of the following embodiments.
  • the abnormality diagnosis device acquires the power data used by the electric device and the setting information of the electric device.
  • the setting information is information for setting various operating states for an electric device, and for example, in an air conditioner, there are stop, cooling, heating, dehumidification, temperature setting, wind direction, and the like.
  • the abnormality diagnosis device estimates the operating state of the electric device from the power data, and if it is different from the acquired setting information, determines that it is an abnormality. In order to estimate the abnormal state by the conventional technique, it is necessary to collect the power data of the electric device in the abnormal state.
  • the number of abnormal states is not limited to one, and it is difficult to collect power data of all electric devices in abnormal states including unexpected ones.
  • the operating state estimated from the power data in the embodiment of the present invention is the normal operating state of the electrical equipment. Therefore, it is easy to collect the data necessary for estimation (features that characterize the operating state), and it is possible to diagnose abnormalities in electrical equipment at low cost.
  • FIG. 1 is an overall configuration diagram of the abnormality diagnosis system 10 according to the first embodiment.
  • the abnormality diagnosis system 10 includes a measuring device 300 and an abnormality diagnosis device 100 (see FIG. 2 described later).
  • the measuring device 300 measures the electric power supplied from the power source 400 to the electric device 200, and transmits the measured electric power data to the abnormality diagnosis device 100 via the network 500.
  • the measuring device 300 may be a device or system having measurement data, for example, an energy management system.
  • the abnormality diagnosis device 100 acquires setting information from the electric device 200 via the network 500.
  • the electric device 200 is a device that operates by using electric power, and is, for example, an air conditioner, a refrigerator, a lighting fixture, an electric water heater, and the like.
  • the power source 400 is a power supply source, and supplies power to the electric device 200 via a power line such as a power system or a premises distribution network.
  • the electric power data is data related to electric power, and is an instantaneous value such as electric energy (WH), electric power consumption (W), voltage (V), current (A), power factor, and time series value.
  • the network 500 is used for transmitting power data and setting information.
  • the network 500 may use a communication network such as a public communication network or a private communication network regardless of whether it is wired or wireless.
  • FIG. 2 is a functional block diagram of the abnormality diagnosis device 100 according to the first embodiment.
  • the abnormality diagnosis device 100 includes a control unit 110, a storage unit 130, and a communication unit 180.
  • the communication unit 180 is composed of a communication device such as a NIC (Network Interface Card), and transmits / receives communication data to / from other devices including the electric device 200 and the measuring device 300.
  • the storage unit 130 is composed of a storage device such as a ROM (Read Only Memory), a RAM (Random Access Memory), and a flash memory.
  • the storage unit 130 stores the operating state feature amount database 140 (described as the operating state feature amount DB (database) in FIG. 2, see FIG. 3 described later) and the program 131.
  • the program 131 describes the procedure for the abnormality diagnosis process (see FIG. 4 described later).
  • FIG. 3 is a data structure diagram of the operating state feature amount database 140 according to the first embodiment.
  • the operating state feature amount database 140 stores the operating state of the electric device 200 and the feature amount of the electric power data in the operating state in association with each other.
  • the operating state feature amount database 140 is tabular data, and one row (record) includes an operating state 141, an average value 142, a time series value 143, and a column (attribute) of a frequency distribution 144.
  • the operating state 141 is an operating state of the electric device 200, and there are states (set state, operating state) such as "OFF" (stop), "dehumidification", and "cooling".
  • the average value 142 and the time series value 143 are the average and time series value (time series data) of the electric power data when the electric device 200 is in the operating state 141.
  • the frequency distribution 144 is, for example, a frequency distribution obtained by wavelet transforming or discrete Fourier transforming power data having one or two cycles (20 to 40 ms for a commercial power supply of 50 Hz).
  • the power data is, as described above, power consumption, power consumption, voltage, current, power factor, and the like.
  • the feature amount may be acquired from the electric device 200 in advance, or data of the same type of electric device may be used.
  • the electric device of the same type may be, for example, a model such as an air conditioner or a refrigerator, an air conditioner having the same power consumption, or an air conditioner having the same model number.
  • it may be acquired from a server (not shown) connected to the network 500 (see FIG. 1) and storing the feature amount.
  • the feature amount on the recording medium connected to the input / output unit (not shown) of the abnormality diagnosis device 100 may be copied and acquired.
  • the control unit 110 includes a power data collection unit 111, a feature amount extraction unit 112, an operating state estimation unit 113, a setting information collection unit 114, an abnormality diagnosis unit 115, and a notification unit 116.
  • the power data collection unit 111 acquires power data from the measuring device 300 (see FIG. 1) and outputs the power data to the feature amount extraction unit 112.
  • the feature amount extraction unit 112 extracts, for example, an average value, a time series value, and a frequency distribution as feature amounts from the power data.
  • the operating state estimation unit 113 searches for a record having a feature amount most similar to the feature amount extracted by the feature amount extracting unit 112 among the records of the operating state feature amount database 140 (see FIG. 3).
  • the operating state estimation unit 113 outputs the operating state 141 of the search result record to the abnormality diagnosis unit 115 as the estimation result of the operating state of the electric device 200.
  • the operating state estimation unit 113 outputs, for example, the operating state 141 of the record having the highest correlation between the time series value and the frequency distribution as the estimation result.
  • the setting information collecting unit 114 acquires the setting information for setting the operating state from the electric device 200 and outputs it to the abnormality diagnosis unit 115.
  • the abnormality diagnosis unit 115 compares the operating state estimated by the operating state estimation unit 113 with the setting information acquired by the setting information collecting unit 114, and if there is a difference, determines that the electric device 200 is abnormal. If there is no difference, the abnormality diagnosis unit 115 determines that there is no abnormality.
  • the abnormality diagnosis unit 115 outputs the determination result to the notification unit 116.
  • the power data is abnormal enough to be erroneously estimated by the operating state estimation unit 113, and there is a difference from the power data at the time when the feature amount is collected. Is. In such a case, it is considered that the electric device 200 has an abnormality.
  • the notification unit 116 notifies the determination result of the abnormality diagnosis unit 115.
  • the notification unit 116 may notify only when there is an abnormality.
  • the notification may be displayed on a display connected to an input / output unit (not shown) of the abnormality diagnosis device 100.
  • the notification may be a notification to the administrator of the electric device 200 via e-mail or a messaging service, or may be a notification to the existing equipment management system.
  • FIG. 4 is a flowchart of the abnormality diagnosis process according to the first embodiment.
  • the abnormality diagnosis process is executed at a predetermined timing, for example, periodically.
  • the power data collection unit 111 acquires power data from the measuring device 300.
  • the feature amount extraction unit 112 extracts the feature amount from the power data.
  • step S13 the operating state estimation unit 113 estimates the operating state of the electric device 200 from the feature amount with reference to the operating state feature amount database 140.
  • step S14 the setting information collecting unit 114 acquires setting information from the electric device 200.
  • step S15 the abnormality diagnosis unit 115 compares the operating state estimated by the operating state estimation unit 113 in step S13 with the setting information collected by the setting information collecting unit 114 in step S14. In other words, the abnormality diagnosis unit 115 determines whether or not the estimated operating state corresponds to the collected setting information.
  • step S16 the abnormality diagnosis unit 115 ends the abnormality diagnosis process if the comparison results in step S15 match (step S16 ⁇ YES), and proceeds to step S17 if they do not match (step S16 ⁇ NO).
  • step S17 the notification unit 116 notifies the abnormality of the electric device 200.
  • the abnormality diagnosis device 100 detects and notifies an abnormality of the electric device with reference to the feature amount.
  • the feature amount is data collected when the electric device 200 is operating normally (see the operating state feature amount database 140 shown in FIG. 3). Specifically, the abnormality diagnosis device 100 determines that the electric device is abnormal if the estimated operating state and the collected setting information do not correspond to each other.
  • the abnormality diagnosis device 100 can perform abnormality diagnosis without a feature amount at the time of abnormality, which is difficult to acquire and collect, and can be realized at low cost. Further, the abnormality diagnosis device 100 can detect the abnormal state even if there is no feature amount that cannot be collected or a feature amount when the abnormal state is unexpected, and the detection accuracy is high.
  • the feature amount (see the operating state feature amount database 140 shown in FIG. 3) is collected in advance from the electric device 200 or the same type of electric device. Instead of this, it may be acquired during abnormality diagnosis (during monitoring of electrical equipment). By eliminating the need to collect features in advance, it is possible to reduce the labor and cost of collecting them. In addition, the introduction period of the abnormality diagnosis device can be shortened.
  • FIG. 5 is a functional block diagram of the abnormality diagnosis device 100A according to the second embodiment.
  • the feature amount setting unit 117 is added to the control unit 110.
  • the program 131A stored in the storage unit 130 describes the procedure of the feature amount setting process (see FIG. 6 described later).
  • the feature amount setting unit 117 acquires the setting information of the electric device 200 from the setting information collecting unit 114, and whether or not the record of the operating state 141 corresponding to the setting information is stored in the operating state feature amount database 140. Is determined. When not stored, the feature amount setting unit 117 stores a record having the feature amount of the power data extracted by the feature amount extraction unit 112 and the operating state 141 corresponding to the setting information in the operating state feature amount database 140. to add.
  • FIG. 6 is a flowchart of the feature amount setting process according to the second embodiment.
  • the feature amount setting process is executed at a predetermined timing, for example, periodically.
  • the feature amount setting process may be executed at the timing when the setting information collecting unit 114 acquires the setting information. Further, the feature amount setting process may be executed in a short cycle at the beginning of the introduction of the abnormality diagnosis device 100A, and may be executed in a long cycle or at the timing of acquiring the setting information after a predetermined period after the introduction. good.
  • step S21 the feature amount setting unit 117 acquires the setting information of the electric device 200 from the setting information collecting unit 114.
  • the feature amount setting unit 117 determines whether or not the record of the operating state 141 corresponding to the setting information acquired in step S21 is stored in the operating state feature amount database 140 (see FIG. 3). The feature amount setting unit 117 ends the feature amount setting process if it is stored (step S22 ⁇ YES), and proceeds to step S23 if it is not stored (step S22 ⁇ NO).
  • step S23 the power data collection unit 111 acquires power data from the measuring device 300.
  • step S24 the feature amount extraction unit 112 extracts the feature amount from the power data.
  • step S25 the feature amount setting unit 117 adds a new record to the operating state feature amount database 140.
  • the feature amount setting unit 117 sets the operating state 141 of the record to the operating state corresponding to the setting information acquired in step S21. Further, the feature amount setting unit 117 sets the average value 142, the time series value 143, and the frequency distribution 144 of the record as the average value, the time series value, and the frequency distribution extracted in step S24.
  • Second Embodiment Features of the abnormality diagnosis device >> If the operating state corresponding to the acquired setting information is not included in the feature amount, the abnormality diagnosis device 100A determines that it is a new operating state, and transfers the feature amount of the power data to the operating state feature amount database 140. Add. By doing so, the abnormality diagnosis device 100A does not need to collect the feature amount in advance, and the labor and cost for collecting the feature amount can be reduced. In addition, the introduction period of the abnormality diagnosis device 100A can be shortened.
  • FIG. 7 is an overall configuration diagram of the abnormality diagnosis system 10B according to the third embodiment.
  • the abnormality diagnosis system 10B includes a measuring device 300 and an abnormality diagnosis device 100B.
  • the abnormality diagnosis device 100B detects an abnormality in a plurality of electric devices 211 to 213.
  • a plurality of electric devices 211 to 213 are collectively referred to as electric devices 210.
  • FIG. 8 is a functional block diagram of the abnormality diagnosis device 100B according to the third embodiment. Compared with the abnormality diagnosis device 100 (see FIG. 2) in the first embodiment, the operation state estimation unit 113B, the setting information collection unit 114B, the abnormality diagnosis unit 115B, and the operation state feature amount database 140B are different. The differences will be described below.
  • FIG. 9 is a data configuration diagram of the operating state feature amount database 140B according to the third embodiment.
  • the operating state 141B is a combination of the operating states of the electric devices 211 to 213.
  • the third record "cooling, OFF, " Indicates that the operating state (setting state) of the electric device 211, which is an air conditioner, is "cooling" and the electric device 212 is "OFF”. ing.
  • the operating state estimation unit 113B estimates the operating state of each electric device 210 with reference to the operating state feature amount database 140B. Specifically, among the records of the operating state feature amount database 140B, the record having the feature amount most similar to the feature amount extracted by the feature amount extraction unit 112 is searched. The operating state estimation unit 113B outputs the operating state 141B of the search result record to the abnormality diagnosis unit 115B as the estimation result of the operating state of each of the electric devices 210. The setting information collecting unit 114B acquires the operating state of each of the electric devices 210.
  • the abnormality diagnosis unit 115B individually compares the estimated operating state of each electric device 210 with the setting information of each electric device 210 acquired by the setting information collecting unit 114, and if there is a difference, the electric device 210 is concerned. Judge that the electrical equipment is abnormal.
  • FIG. 10 is a flowchart of the abnormality diagnosis process according to the third embodiment.
  • the abnormality diagnosis process is executed at a predetermined timing, for example, periodically.
  • the power data collection unit 111 acquires power data from the measuring device 300.
  • the feature amount extraction unit 112 extracts the feature amount from the power data.
  • step S33 the operating state estimation unit 113B estimates the operating state of each of the electric devices 210 from the feature amount with reference to the operating state feature amount database 140B.
  • step S34 the setting information collecting unit 114B acquires setting information from each of the electric devices 210.
  • step S35 the abnormality diagnosis unit 115B compares the estimation result estimated by the operating state estimation unit 113B in step S33 with the setting information collected by the setting information collecting unit 114B in step S34 for each electric device 210.
  • step S36 the abnormality diagnosis unit 115B ends the abnormality diagnosis process if the comparison results match for all the electrical devices 210 in step S35 (step S36 ⁇ YES), and if there is a mismatched electrical device 210 (step S36 ⁇ NO). ) Proceed to step S37.
  • step S37 the notification unit 116 notifies the abnormality of the electric device 210 in which the estimated operating state and the acquired setting information do not match.
  • the abnormality diagnosis device 100B detects and notifies an abnormality of a plurality of electric devices by referring to the feature amount.
  • the feature amount is data collected when the electric device 210 is operating normally (see the operating state feature amount database 140B shown in FIG. 9).
  • the abnormality diagnosis device 100B can perform abnormality diagnosis without a feature amount at the time of abnormality, which is difficult to acquire and collect, and can be realized at low cost. Further, the abnormality diagnosis device 100B can detect the abnormal state even if there is no feature amount that cannot be collected or a feature amount when the abnormal state is unexpected, and the detection accuracy is high.
  • the feature amount (see the operating state feature amount database 140B shown in FIG. 9) is collected in advance from the electric device 210 or the same type of electric device. Instead of this, it may be acquired during abnormality diagnosis (during monitoring of electrical equipment). By eliminating the need to collect features in advance, it is possible to reduce the labor and cost of collecting them. In addition, the introduction period of the abnormality diagnosis device can be shortened.
  • FIG. 11 is a functional block diagram of the abnormality diagnosis device 100C according to the fourth embodiment.
  • the feature amount setting unit 117C is added to the control unit 110.
  • the program 131C stored in the storage unit 130 describes the procedure of the feature amount setting process (see FIG. 12 described later).
  • the feature amount setting unit 117C acquires the setting information of the electric device 210 from the setting information collecting unit 114B, and determines whether or not the record of the operating state 141B corresponding to the setting information is stored in the operating state feature amount database 140B. judge. If it is not stored, a record having the feature amount (average value, time series value, frequency distribution) of the power data extracted by the feature amount extraction unit 112 and the operating state 141B corresponding to the setting information is stored as the operating state feature amount. Add to database 140B.
  • FIG. 12 is a flowchart of the feature amount setting process according to the fourth embodiment.
  • the feature amount setting process is processed at a predetermined timing, for example, periodically.
  • the feature amount setting process may be executed at the timing when the setting information collecting unit 114B acquires the setting information.
  • the feature amount setting unit 117C acquires the setting information of each electric device 210 from the setting information collecting unit 114B.
  • step S42 the feature amount setting unit 117C determines whether or not the record of the operating state 141B corresponding to the setting information acquired in step S41 is stored in the operating state feature amount database 140B (see FIG. 9).
  • the feature amount setting unit 117C ends the feature amount setting process if it is stored (step S42 ⁇ YES), and proceeds to step S43 if it is not stored (step S42 ⁇ NO).
  • step S43 the power data collection unit 111 acquires power data from the measuring device 300.
  • step S44 the feature amount extraction unit 112 extracts the feature amount from the power data.
  • step S45 the feature amount setting unit 117C adds a new record to the operating state feature amount database 140B.
  • the feature amount setting unit 117C sets the operating state 141B of the record to the operating state corresponding to the setting information of each electric device 210 acquired in step S41. Further, the feature amount setting unit 117C sets the average value 142, the time series value 143, and the frequency distribution 144 of the record as the average value, the time series value, and the frequency distribution extracted in step S44.
  • the abnormality diagnosis device 100C determines that it is a new operating state (combination of operating states). , The feature amount of the power data is added to the operating state feature amount database 140B. By doing so, the abnormality diagnosis device 100C does not need to collect the feature amount in advance, and the labor and cost for collecting the feature amount can be reduced. In addition, the introduction period of the abnormality diagnosis device 100C can be shortened.
  • the abnormality diagnosis device 100 notifies that the electric device 200 is in an abnormal state.
  • the notification includes the abnormality type (cause of the abnormality).
  • FIG. 13 is a functional block diagram of the abnormality diagnosis device 100D according to the fifth embodiment.
  • the storage unit 130 stores the abnormality state feature amount database 150D (see FIG. 14 described later), and the notification unit 116D of the control unit 110. Is different.
  • the program 131D stored in the storage unit 130 describes the procedure for the abnormality diagnosis process (see FIG. 15 described later).
  • FIG. 14 is a data structure diagram of the abnormal state feature amount database 150D according to the fifth embodiment.
  • the abnormal state feature amount database 150D has a data structure in which the operating state 141 of the operating state feature amount database 140 (see FIG. 3) is replaced with the abnormal state 151.
  • the abnormal state 151 indicates the type of the abnormal state of the electric device 200 or the cause of the abnormality.
  • the notification unit 116D searches for a record having a feature amount most similar to the feature amount extracted by the feature amount extraction unit 112 among the records of the abnormal state feature amount database 150D.
  • the notification unit 116D notifies the abnormal state 151 of the search result record as an estimation result of the cause of the abnormal state in the electric device 200.
  • the notification unit 116D outputs, for example, the abnormal state 151 of the record having the highest correlation between the time series value and the frequency distribution as the estimation result.
  • FIG. 15 is a flowchart of the abnormality diagnosis process according to the fifth embodiment.
  • the abnormality diagnosis process is executed at a predetermined timing, for example, periodically.
  • Steps S51 to S56 are the same as steps S11 to S16 (see FIG. 4), respectively.
  • step S57 the notification unit 116D estimates the cause of the abnormality of the electric device 200.
  • step S58 the notification unit 116D notifies the abnormality of the electric device 200 including the cause of the abnormality estimated in step S57.
  • the abnormality diagnosis device 100D notifies the electric device 200 determined to be abnormal including the cause of the abnormality. By knowing the cause of the abnormality, the manager of the electric device 200 can efficiently deal with the abnormality of the electric device 200.
  • the notification unit 116D estimates the cause of the abnormality, but the abnormality diagnosis unit 115 may estimate the cause.
  • the average value, time series value, and frequency distribution of power data are used as the feature quantities of the abnormal state. Among them, it is desirable to estimate by looking at the frequency distribution, especially the high frequency distribution (correlation).
  • the operating state estimation units 113 and 113B search for the record of the feature amount closest to the feature amount of the power data in the operating state feature amount databases 140 and 140B (see FIGS. 3 and 9). ..
  • the operating state estimation units 113 and 113B use the operating states 141 and 141B of the record as the estimation result. Instead of such processing, machine learning techniques may be used.
  • FIG. 16 is a functional block diagram of the abnormality diagnosis device 100E according to the modified example of the first embodiment.
  • the learning unit 117E is added to the control unit 110, and the operating state estimation unit 113E is different. Further, the operating state learning model 140E is stored in the storage unit 130 instead of the operating state feature amount database 140.
  • the operating state learning model 140E is a machine learning model trained using teacher data in which the operating states 141 and 141B of the electric devices 200 and 210 are added as correct labels to the feature quantities of the electric power data.
  • a machine learning technique a neural network, SVM (Support Vector Machine), clustering, or the like can be used.
  • the machine learning technique may be used for the estimation of the cause of the abnormality performed by the notification unit 116D (see FIG. 13).
  • the operating state estimation unit 113E estimates the operating states of the electric devices 200 and 210 from the feature amount of the electric power data with reference to the operating state learning model 140E.
  • the setting information collecting unit 114 acquires the setting information (operating state) that is not in the estimation result of the operating state learning model 140E
  • the learning unit 117E determines that the setting information is correct for the feature amount of the power data at the time of acquisition.
  • the operating state learning model 140E is updated by additionally learning (training) the teacher data to be used.
  • the number of records in the operating state feature amount database 140B is the product of the number of operating states in each electric device 210. It increases exponentially as the number of devices 210 increases.
  • the shift number and the value of the expansion coefficient having the maximum value after the wavelet transform of the electric power data may be registered (stored) as a feature amount.
  • the operating state of each electric device 210 can be estimated by obtaining the expansion coefficient of the power data after wavelet transform and taking the correspondence between the number of shifts and the value in the feature data.
  • the number of feature amounts (the number of records in the operating state feature amount database 140B) is the sum of the number of operating states in each electric device 210, and is proportional to the number of electric devices 210.
  • the number of registered features can be reduced, which in turn improves the speed of estimation processing.
  • the cost of collecting features can be reduced.
  • the notification unit 116D (see FIG. 13) may estimate the abnormal state (cause of the abnormality) of the electric device.
  • the operating state feature amount database 140 may include only a part of the average value 142, the time series value 143, and the frequency distribution 144 of the power data.
  • the operating state estimation units 113 and 113B estimate the operating state by comparing some of the features.
  • the programs 131, 131A, 131B, 131C, 131D, 131E are stored in the abnormality diagnosis devices 100, 100A, 100B, 100C, 100D, 100E, which are computers.
  • the programs 131, 131A, 131B, 131C, 131D, 131E in the recording medium 910 may be read and loaded into the storage unit 130 and executed, or may be installed and executed from the recording medium 910.
  • FIG. 17 is a diagram showing a recording medium 910 according to the above-described embodiment. By installing the recording medium 910 on the computer 900, the computer can function as the abnormality diagnostic devices 100, 100A, 100B, 100C, 100D, 100E.
  • the average value, the time series value, the frequency distribution (including the expansion coefficient after the wavelet transform) and the like are given as the feature amount of the power data, but other feature amounts such as the instantaneous maximum value are used. You may.
  • the feature amount (operating state feature amount database) is stored in the storage unit of the abnormality diagnosis device, but the present invention is not limited to this.
  • the operating state estimation unit may estimate the operating state of the electric device by referring to the feature amount stored in the outside.
  • the present invention can take various other embodiments, and further, various changes such as omission and replacement can be made without departing from the gist of the present invention.
  • These embodiments and variations thereof are included in the scope and gist of the invention described in the present specification and the like, and are also included in the scope of the invention described in the claims and the equivalent scope thereof.
  • Abnormality diagnosis system 100, 100A, 100B, 100C, 100D, 100E Abnormality diagnosis device 111 Power data collection unit 112 Feature quantity extraction unit 113, 113B, 113E Operation status estimation unit 114, 114B Setting information collection unit 115, 115B Abnormality Diagnosis unit 116, 116D Notification unit 117, 117C Feature amount setting unit 117E Learning unit 140, 140B Operating state feature amount database (operating state feature amount data) 140E Operating state learning model (machine learning model) 150D Abnormal state feature database (fault feature data) 200, 210, 211,212,213 Electrical equipment

Abstract

La présente invention a pour objectif de résoudre le problème consistant à diagnostiquer une anomalie même lorsque les données correspondant au moment où une anomalie de dispositif électrique se produit ne sont pas disponibles. Un dispositif de diagnostic d'anomalie (100) est équipé : d'une unité de collecte de données d'alimentation (111) permettant de collecter des données d'alimentation concernant un dispositif électrique ; d'une unité d'extraction de caractéristique (112) permettant d'acquérir une caractéristique à partir des données d'alimentation collectées ; d'une unité d'estimation d'état de fonctionnement (113) permettant d'estimer l'état de fonctionnement du dispositif électrique à partir des données d'état de fonctionnement/de caractéristique (base de données d'états de fonctionnement/caractéristiques (140)), cet état incluant la relation entre des caractéristiques d'état de fonctionnement du dispositif électrique et des caractéristiques de données d'alimentation associées audit état de fonctionnement et également à partir de ladite caractéristique acquise ; d'une unité de collecte d'informations de réglage (114) permettant de collecter les informations de réglage pour régler l'état de fonctionnement du dispositif électrique ; et d'une unité de diagnostic d'anomalie (115) permettant de déterminer que le dispositif électrique présente une anomalie lorsque l'état de fonctionnement estimé et les informations de réglage collectées ne correspondent pas les uns aux autres.
PCT/JP2020/047483 2020-12-18 2020-12-18 Dispositif de diagnostic d'anomalies, programme et procédé de diagnostic d'anomalies WO2022130626A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/JP2020/047483 WO2022130626A1 (fr) 2020-12-18 2020-12-18 Dispositif de diagnostic d'anomalies, programme et procédé de diagnostic d'anomalies

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2020/047483 WO2022130626A1 (fr) 2020-12-18 2020-12-18 Dispositif de diagnostic d'anomalies, programme et procédé de diagnostic d'anomalies

Publications (1)

Publication Number Publication Date
WO2022130626A1 true WO2022130626A1 (fr) 2022-06-23

Family

ID=82059368

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2020/047483 WO2022130626A1 (fr) 2020-12-18 2020-12-18 Dispositif de diagnostic d'anomalies, programme et procédé de diagnostic d'anomalies

Country Status (1)

Country Link
WO (1) WO2022130626A1 (fr)

Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH04110779A (ja) * 1990-08-31 1992-04-13 Toshiba Lighting & Technol Corp 負荷異常監視装置
JP2000292465A (ja) * 1999-02-01 2000-10-20 Central Res Inst Of Electric Power Ind 電気機器モニタリングシステム及び動作異常警報システム
JP2003259569A (ja) * 2001-12-28 2003-09-12 Atsushi Matsushita 電気機器の遠隔検出方法及び遠隔制御方法並びに遠隔制御システムと、これに用いるコンセント
JP2004222374A (ja) * 2003-01-10 2004-08-05 Matsushita Electric Works Ltd 電気機器監視システム
WO2009125659A1 (fr) * 2008-04-11 2009-10-15 三菱電機株式会社 Dispositif de détection d’état de dispositif, procédé de détection d’état de dispositif, serveur de détection d’état de dispositif, système de détection d’état de dispositif, dispositif de détection d’anomalie de dispositif sous tension, système de détection d’anomalie de dispositif sous tension, procédé de détection d’anomalie de dispositif sous tension, et serveur de maintien de base de données d’états de dispositifs
JP2014072561A (ja) * 2012-09-27 2014-04-21 Nitto Denko Corp 家電機器遠隔監視システム
KR20160097680A (ko) * 2015-02-09 2016-08-18 (주)제이비엘 파워 모니터링 시스템 및 장치
JP2016156643A (ja) * 2015-02-23 2016-09-01 株式会社日立製作所 予兆診断システム、予兆診断方法及び予兆診断装置
JP2017017776A (ja) * 2015-06-26 2017-01-19 東芝ライテック株式会社 電力管理装置、電力管理システム、電力管理方法、及び、プログラム
US20180231603A1 (en) * 2017-02-15 2018-08-16 Abhay Gupta Systems and Methods for Detecting Occurence of an Event in a Household Environment
JP2019170100A (ja) * 2018-03-23 2019-10-03 積水化学工業株式会社 機器管理システム
JP2020056744A (ja) * 2018-10-04 2020-04-09 シャープ株式会社 スマートタップ
JP3227337U (ja) * 2020-03-17 2020-08-20 明宗 陳 電子装置監視システム

Patent Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH04110779A (ja) * 1990-08-31 1992-04-13 Toshiba Lighting & Technol Corp 負荷異常監視装置
JP2000292465A (ja) * 1999-02-01 2000-10-20 Central Res Inst Of Electric Power Ind 電気機器モニタリングシステム及び動作異常警報システム
JP2003259569A (ja) * 2001-12-28 2003-09-12 Atsushi Matsushita 電気機器の遠隔検出方法及び遠隔制御方法並びに遠隔制御システムと、これに用いるコンセント
JP2004222374A (ja) * 2003-01-10 2004-08-05 Matsushita Electric Works Ltd 電気機器監視システム
WO2009125659A1 (fr) * 2008-04-11 2009-10-15 三菱電機株式会社 Dispositif de détection d’état de dispositif, procédé de détection d’état de dispositif, serveur de détection d’état de dispositif, système de détection d’état de dispositif, dispositif de détection d’anomalie de dispositif sous tension, système de détection d’anomalie de dispositif sous tension, procédé de détection d’anomalie de dispositif sous tension, et serveur de maintien de base de données d’états de dispositifs
JP2014072561A (ja) * 2012-09-27 2014-04-21 Nitto Denko Corp 家電機器遠隔監視システム
KR20160097680A (ko) * 2015-02-09 2016-08-18 (주)제이비엘 파워 모니터링 시스템 및 장치
JP2016156643A (ja) * 2015-02-23 2016-09-01 株式会社日立製作所 予兆診断システム、予兆診断方法及び予兆診断装置
JP2017017776A (ja) * 2015-06-26 2017-01-19 東芝ライテック株式会社 電力管理装置、電力管理システム、電力管理方法、及び、プログラム
US20180231603A1 (en) * 2017-02-15 2018-08-16 Abhay Gupta Systems and Methods for Detecting Occurence of an Event in a Household Environment
JP2019170100A (ja) * 2018-03-23 2019-10-03 積水化学工業株式会社 機器管理システム
JP2020056744A (ja) * 2018-10-04 2020-04-09 シャープ株式会社 スマートタップ
JP3227337U (ja) * 2020-03-17 2020-08-20 明宗 陳 電子装置監視システム

Similar Documents

Publication Publication Date Title
JP6301791B2 (ja) 配電網の故障予兆診断システムおよびその方法
CN107462863B (zh) 一种智能电能表运行误差运行诊断分析方法及系统
Xue et al. Fault diagnosis and abnormality detection of lithium-ion battery packs based on statistical distribution
US8209062B2 (en) Method for non-intrusive load monitoring using a hybrid systems state estimation approach
Seem Using intelligent data analysis to detect abnormal energy consumption in buildings
JP5306902B2 (ja) 資産システムの高性能条件監視のためのシステムおよび方法
CN103403463B (zh) 故障检测和诊断算法
Catterson et al. Online conditional anomaly detection in multivariate data for transformer monitoring
US20110307200A1 (en) Recognizing multiple appliance operating states using circuit-level electrical information
CN109298697A (zh) 基于动态基线模型的火电厂系统各部件工作状态评估方法
KR101214765B1 (ko) 전력설비 모니터링 시스템 및 방법
JP2018169161A (ja) 電池の劣化診断装置、劣化診断方法、及び劣化診断システム
CN105804981A (zh) 动车组牵引变压器油泵远程故障监测和预警系统及其方法
Mosallam et al. Component based data-driven prognostics for complex systems: Methodology and applications
GB2476246A (en) Diagnosing an operation mode of a machine
KR102417800B1 (ko) 전력 케이블 노후도 판단 장치 및 방법
CN115128513A (zh) 一种基于热量的电容器异常检测方法及相关装置
WO2022130626A1 (fr) Dispositif de diagnostic d'anomalies, programme et procédé de diagnostic d'anomalies
US11339763B2 (en) Method for windmill farm monitoring
CN113591396B (zh) 一种基于朴素贝叶斯网络的电网元器件故障诊断方法
Peng et al. IEPE accelerometer fault diagnosis for maintenance management system information integration in a heavy industry
WO2022165792A1 (fr) Procédé et système de gestion de défaut de capteur
JP7044170B2 (ja) 異常検知装置、方法、およびプログラム
Trstenjak et al. A Decision Support System for the Prediction of Wastewater Pumping Station Failures Based on CBR Continuous Learning Model.
BR112021004932A2 (pt) método, sistema e produto de software para identificar instalações com tendência a apresentar uma não conformidade elétrica

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20966013

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20966013

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP