JP2019057056A - Accident cause classification device - Google Patents

Abstract

To provide an accident cause classification device that updates an accident classification model for classifying accident causes according to change of electric power system facilities and ambient environment of electric power system facilities.SOLUTION: An accident cause classification device according to an embodiment has a model storage part, a measurement information acquisition part, an accident information acquisition part, a cause information acquisition part, a result information acquisition part, an accident history information storage part, a tendency determination part, an object information storage part, an object accident information storage part, and an update model generation part. The measurement information acquisition part acquires measurement information representing a result of measurement. The accident information acquisition part acquires accident information representing whether an accident takes place. The cause information acquisition part inputs the measurement information acquired by the measurement information acquisition part in the timing of the acquisition of the accident information to the accident classification model to acquire cause information. The tendency determination part determines whether object facilities change in state based upon the measurement information. The update model generation part generates an after-update accident classification model to be used instead of the accident classification model, and stores it in the model storage part.SELECTED DRAWING: Figure 1

Description

  Embodiments described herein relate generally to an accident cause classification apparatus.

  Conventionally, as a method of determining the cause of an accident that occurs in the power system, an accident classification model that classifies the cause of an accident is created by learning various measurement data collected from facilities at the time of past accidents by machine learning etc. A technique for classifying accident causes by inputting measurement data collected from facilities at the time of an accident into the accident classification model is known.

  In addition, if the power system equipment and the surrounding environment of the power system equipment change, there will be a discrepancy between the learning data used to create the accident classification model and the measurement data used to input the accident classification model, The accuracy of classification may be reduced. In general, such re-learning method in machine learning includes learning data at the time of creating an existing model, data misrecognized in recognition processing using the existing model, and re-learning data on the correction result for the misrecognized data. A re-learning method has been proposed.

  When the power system equipment and the surrounding environment of the power system equipment change, it is required to use the data after the environmental change in the relearning of the accident classification model, and it is not appropriate to use the data before the environmental change. There is a case. In addition, such environmental changes may not occur throughout the entire power system, but may occur only in some power system facilities. For this reason, it is necessary to consider the learning data selection method before and after the environmental change and the selection method of the range of the target equipment that is the measurement target of the measurement data that becomes the learning data. In the conventional technology, such processing is performed. There were cases where it was difficult.

JP-A-11-142466 Japanese Patent Laid-Open No. 2006-143354

  The problem to be solved by the present invention is to provide an accident cause classification device that updates an accident classification model for classifying an accident cause according to changes in the power system equipment and the surrounding environment of the power system equipment.

  The accident cause classification apparatus of the embodiment includes a model storage unit, a measurement information acquisition unit, an accident information acquisition unit, a cause information acquisition unit, a result information acquisition unit, an accident history information storage unit, a tendency determination unit, It has a target information storage unit, a target accident information storage unit, and an update model generation unit. The model storage unit stores an accident classification model that is learned by machine learning and that outputs cause information when measurement information is input. The measurement information acquisition unit acquires measurement information indicating a result of measurement related to electric power in the target facility. The accident information acquisition unit acquires accident information indicating whether an accident has occurred in the target facility. The cause information acquisition unit inputs the measurement information acquired by the measurement information acquisition unit to the accident classification model at the timing when the accident information is acquired by the accident information acquisition unit, and acquires the cause information. The result information acquisition unit is information acquired from a patrolperson of the target facility, and acquires result information indicating the cause of the accident in association with the accident information. The accident history information storage unit stores accident history information in which the result information acquired by the result information acquisition unit is associated with information indicating the timing at which the accident indicated by the result information occurs. The trend determination unit determines whether the state of the target facility has changed based on the measurement information. The target information storage unit stores target information in which the determination result of the tendency determination unit, the measurement information, and information indicating the timing at which the measurement information is acquired are associated with each other. The target accident information storage unit stores target accident information in which the measurement information acquired in response to acquiring the accident information is associated with information indicating the timing at which the accident occurs. The updated model generation unit generates an updated accident classification model used instead of the accident classification model based on the target information acquired or generated at the same time, the target accident information, and the accident history information And stored in the model storage unit.

It is a figure which shows an example of a structure of an accident cause classification device. It is a figure which shows an example of object accident information. It is a figure which shows an example of a tendency determination model. It is a figure which shows an example of object information. It is a figure which shows an example of accident log information. It is a figure which shows an example of an accident classification model. It is a figure which shows an example of the image which a display control part displays on a terminal. It is a figure which shows an example of selection of the range of update range information. It is a flowchart which shows an example of operation | movement of a relearning function. It is a flowchart which shows an example of operation | movement of a model update function. It is a figure which shows an example of the period after the state of a power distribution equipment changes until an accident classification model after an update is produced | generated. It is a figure which shows the other example of the image which a display control part displays on a terminal.

  Hereinafter, an accident cause classification apparatus according to an embodiment will be described with reference to the drawings. The accident cause classification device according to the embodiment is a device that classifies the cause of an accident that occurs in an electric power system facility based on an accident classification model learned by machine learning, and the environment around the electric power system facility and the electric power system facility is It is a device that updates the accident classification model as it changes.

<Embodiment>
FIG. 1 is a diagram illustrating an example of the configuration of the accident cause classification apparatus 1. The accident cause classification device 1 communicates with the measuring instrument M, the accident information output device OD, and the like via the network NW. The network NW includes, for example, the Internet, a WAN (Wide Area Network), a LAN (Local Area Network), a wireless base station, a provider terminal, a dedicated line, and the like. The measuring instrument M is a device that performs measurement related to electric power. Specifically, the measuring instrument M is a device that measures a current flowing through the power distribution facility EQ or a device that measures a voltage applied to the power distribution facility EQ. The distribution facility EQ is a facility that constitutes a part of an electric power system that supplies electric power from a substation to each consumer. The power distribution equipment EQ includes, for example, a switch and a circuit breaker. The measuring instrument M measures the current flowing through the power distribution equipment EQ or the voltage applied to the power distribution equipment EQ, and sends the information indicating the measured value (hereinafter referred to as measurement information MI) to the accident cause classification device 1 at regular or predetermined intervals. Send with. The accident cause classification device 1 classifies (estimates) the cause of an accident occurring in the power system based on the received measurement information MI, and outputs information indicating the cause of the accident (hereinafter, cause information).

  The accident information output device OD is a device that outputs accident information indicating whether or not an accident has occurred in the power system. The accident information output device OD outputs accident information based on the state of a ground fault relay that detects a ground fault, for example. The accident information output device OD transmits, to the accident cause classification device 1, accident information indicating that an accident has occurred in the power system, for example, in response to the detection of a ground fault by a ground fault relay.

  A terminal device TM1 and a terminal device TM2 are connected to the accident cause classification device 1. The terminal device TM1 is a terminal used by a patrol instructor. The terminal device TM2 is a terminal used by a patrolman. The terminal device TM is, for example, a mobile phone such as a smartphone, a tablet terminal, a personal computer, or the like. The patrol instructor instructs the patrol officer on the patrol method of the power distribution equipment EQ and points to be noted in the patrol. The patrolman patrols the power distribution equipment EQ based on the guidance of the patrol instructor. The cause information output from the accident cause classification device 1 is displayed on the terminal device TM1. The patrol instructor refers to the cause information displayed on the terminal device TM1 when an accident occurs in the power system, and causes the accident cause classification apparatus 1 to inspect the patrol person who patrols the power system. The power distribution equipment EQ to be notified is notified. Information (hereinafter, result information) indicating the cause of the accident is input to the terminal device TM2 by a patrolman who patrols the power distribution facility EQ in accordance with the occurrence of the accident in the power system. The accident cause classification device 1 updates a mechanism for classifying (estimating) the cause of an accident occurring in the power system based on the measurement information MI and result information input from the terminal device TM2.

  The accident cause classification apparatus 1 includes, for example, an acquisition function unit 10, a tendency determination function unit 20, a result information acquisition unit 30 by a hardware processor such as a CPU (Central Processing Unit) executing a program (software), The relearning function unit 40, the model update function unit 50, the cause information acquisition function unit 60, and the display control unit 70 are realized. Some or all of these components (excluding the storage unit included) are LSI (Large Scale Integration), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), GPU (Graphics Processing). Unit) or the like (circuit unit; including circuit), or may be realized by cooperation of software and hardware. The accident cause classification apparatus 1 includes a tendency determination model storage unit 22, an update model storage unit 54, a model storage unit 62, a target information storage unit 80, a target accident information storage unit 90, and an accident history information storage unit. 100. Each storage unit is realized by, for example, a ROM (Read Only Memory), a flash memory, an SD card, a RAM (Random Access Memory), a register, and the like.

  The acquisition function unit 10 includes a measurement information acquisition unit 11 and an accident information acquisition unit 12. The measurement information acquisition unit 11 receives measurement information MI transmitted from the measuring instrument M via the network NW. The accident information acquisition unit 12 receives the accident information transmitted by the accident information output device OD via the network NW. The measurement information acquisition unit 11 supplies the received measurement information MI to each unit included in the accident cause classification device 1. The measurement information acquisition unit 11 supplies the measurement information MI to the tendency determination function unit 20 when no accident has occurred in the power system. Moreover, the measurement information acquisition part 11 supplies measurement information MI to the model update function part 50 and the cause information acquisition function part 60, when the accident has generate | occur | produced in the electric power grid | system. Specifically, the measurement information acquisition unit 11 is a function for determining the trend of the measurement information MI when the accident information acquisition unit 12 does not acquire accident information from the accident information output device OD (when no accident has occurred in the power system). To the unit 20. Moreover, the measurement information acquisition part 11 is the case where the accident information acquisition part 12 acquires accident information from the accident information output device OD (when an accident has occurred in the power system), and the measurement information MI and the model update function part 50 To the cause information acquisition function unit 60.

  In addition, the measurement information acquisition unit 11 acquires the measurement information MI when the accident information acquisition unit 12 does not acquire the accident information from the accident information output device OD (when no accident has occurred in the power system), the measurement information MI Information associated with the information MI (hereinafter, target accident information 91) is stored in the target accident information storage unit 90. FIG. 2 is a diagram illustrating an example of the target accident information 91. In the present embodiment, the target accident information 91 includes measurement information MI (hereinafter referred to as accident measurement information AMI) measured at the time of the accident and the date when the accident measurement information AMI was acquired, that is, the date when the accident occurred (accident occurrence). Date AD), information for identifying the distribution facility EQ (hereinafter referred to as distribution facility identification information EQID), the name of the distribution facility EQ, and the region where the distribution facility EQ is installed (hereinafter referred to as installation region AR) Is memorized.

  In the present embodiment, the measuring instrument M is attached to each of the five distribution facilities EQ (distribution facilities EQ1 to 5). The target accident information 91 includes accident measurement information AMI1 to AMI1 of the distribution facilities EQ1 to 5 that are the distribution facility identification information EQID1 to 5 acquired on the accident occurrence date AD “2017/11/15”, and each distribution facility. The EQ installation area AR is stored in association with each other. Further, the target accident information 91 includes the accident measurement information AMI6 to 10 of the distribution facilities EQ1 to 5 that are the distribution facility identification information EQID1 to 5 acquired on the accident occurrence date AD “2017/11/30”, and each distribution facility. The EQ installation area AR is stored in association with each other. In addition, the installation area AR of the distribution facilities EQ1 to EQ3 is “north of prefecture X”. In addition, the installation area AR of the distribution facilities EQ4 to EQ5 is “southern X prefecture”. In the present embodiment, the measuring instrument M transmits power distribution facility identification information EQID and measurement information MI. Moreover, the accident cause classification apparatus 1 has a database (not shown) in which the distribution facility identification information EQID, the name of the distribution facility EQ, and the installation area AR are stored in association with each other. In this case, based on the distribution facility identification information EQID transmitted by the measuring instrument M, the measurement information acquisition unit 11 acquires the name of the distribution facility EQ associated with the distribution facility identification information EQID and the installation area AR from the database. . In addition, the measurement information acquisition unit 11 stores various information acquired from the database, the accident measurement information AMI received from the measuring instrument M, and the accident occurrence date AD in the target accident information storage unit 90 in association with each other.

  Returning to FIG. 1, the trend determination function unit 20 includes a trend determination unit 21 and a trend determination model storage unit 22. The trend determination model storage unit 22 stores a trend determination model 23. The trend determination unit 21 determines whether a change has occurred in the power system and the surrounding environment of the power system facility based on the measurement information MI acquired from the measurement information acquisition unit 11 and the trend determination model 23. In the following description, changes in the environment around the power system and the power system equipment are also collectively referred to as changes in the state of the power distribution equipment EQ. That is, the trend determination unit 21 determines whether or not the state of the distribution facility EQ has changed based on the measurement information MI and the trend determination model 23. The tendency determination model 23 is information indicating a learning model in which the properties of the measurement information MI are learned by machine learning based on the measurement information MI acquired in the past. The tendency determination model 23 is, for example, a learning model in which the nature of the measurement information MI acquired in the past is modeled by a neural network.

  FIG. 3 is a diagram illustrating an example of the tendency determination model 23. As shown in FIG. 3, the measurement information MI acquired from the measurement information acquisition unit 11 is input to the trend determination model 23. The trend determination unit 21 inputs the measurement information MI acquired from the measurement information acquisition unit 11 to the trend determination model 23 and performs a calculation. The trend determination unit 21 determines the trend of the distribution facility EQ based on the measurement information MI input to the trend determination model 23 and the reconstruction error of the value output from the trend determination model 23. For example, when the reconfiguration error is large, the tendency determination unit 21 determines that the state of the power distribution equipment EQ has changed. When the reconfiguration error is small, the tendency determination unit 21 determines that the state of the power distribution equipment EQ has not changed.

  As described above, the measurement information MI is a value indicating a current flowing through the distribution facility EQ or a value indicating a voltage applied to the distribution facility EQ. For example, the trend determination unit 21 may be configured to normalize each measurement information MI and input the normalized value to the trend determination model 23. Further, when the measurement information MI is a value indicating the current flowing through the distribution facility EQ for a predetermined period (for example, several seconds) or a value indicating the voltage applied to the distribution facility EQ, the trend determination model 23 includes the measurement information A value obtained by discretizing MI may be input. Further, the tendency determination unit 21 may be configured to acquire the reconstruction error for all values obtained by discretizing the measurement information MI, and acquire the reconstruction error for a part of the discretized values. It may be a configuration.

  The trend determination unit 21 associates a determination result (hereinafter referred to as a trend determination result) for determining a trend of the distribution equipment EQ, a date when the measurement information MI is acquired (hereinafter, a measurement date MD), and the measurement information MI. Is stored in the target information storage unit 80 as target information 81. FIG. 4 is a diagram illustrating an example of the target information 81. In the present embodiment, the target information 81 includes the distribution facility identification information EQID, the name of the distribution facility EQ, the measurement information MI acquired by the measurement information acquisition unit 11, the measurement date MD, the trend determination result, and the distribution facility. The installation area AR for each EQ is stored in association with each other.

  The target information 81 includes the measurement information MI1-5 of the distribution facilities EQ1-5 that are the distribution facility identification information EQID1-5 acquired on the measurement date MD “2017/11/7”, and the installation area AR of each distribution facility EQ. Are stored in association with each other. Further, in the target information 81, the measurement information MI6 to 10 of the distribution facilities EQ1 to 5 that are the distribution facility identification information EQID1 to 5 acquired on the measurement date MD “2017/12/7”, and the installation of each distribution facility EQ The area AR is associated and stored. Further, in the target information 81, among the measurement information MI1 to the measurement information MI10, the measurement information MI of the distribution facility EQ5 (the measurement information MI5 and the measurement information MI10 shown in the figure) indicates that the state of the distribution facility EQ has changed. The trend determination result indicating that the state of the distribution equipment EQ has not changed corresponds to the other measurement information MI (measurement information MI1 to 4 and measurement information MI6 to 9). Attached.

  Returning to FIG. 1, the result information acquisition part 30 acquires result information from the terminal device TM2. The result information acquisition unit 30 associates the acquired result information with the date when the accident indicated by the result information occurred, that is, the accident occurrence date AD (hereinafter, the accident history information 101). Remember me. FIG. 5 is a diagram illustrating an example of the accident history information 101. In the present embodiment, the accident history information 101 includes the accident occurrence date AD “2017/11/15”, the result information “ground fault” that is the cause of the accident that occurred on the accident occurrence date AD, and the accident occurred. Distribution facility EQ5 and distribution facility identification information EQID5 are stored in association with each other. Further, the accident history information 101 includes the accident occurrence date AD “2017/11/30”, the result information “ground fault” that is the cause of the accident that occurred on the accident occurrence date AD, and the distribution equipment EQ5 where the accident occurred. And distribution facility identification information EQID5 are stored in association with each other.

  Returning to FIG. 1, the cause information acquisition function unit 60 includes a cause information acquisition unit 61 and a model storage unit 62. The model storage unit 62 stores an accident classification model 63 and range information 64. The accident classification model 63 is information indicating a learning model learned by machine learning based on the measurement information MI acquired in the past and the result information acquired in the past. The accident classification model 63 is configured to output cause information when the measurement information MI is input. The accident classification model 63 is a learning model modeled based on, for example, a neural network. The range information 64 is information indicating a range in which cause information can be acquired by the accident classification model 63. In other words, the range information 64 and the accident classification model 63 indicating the accident classification model 63 applicable to the range indicated by the range information 64 are stored in the model storage unit 62 in association with each other. For example, based on the target accident information 91, the cause information acquisition unit 61 acquires the accident measurement information AMI and the installation area AR of the power distribution equipment EQ where the accident measurement information AMI is measured. The cause information acquisition unit 61 determines range information 64 corresponding to the installation area of the power distribution equipment EQ indicated by the installation area AR, and acquires cause information based on the accident classification model 63 associated with the range information 64. The measurement information MI acquired in the past is an example of teacher measurement information. The result information acquired in the past is an example of teacher cause information.

  FIG. 6 is a diagram illustrating an example of the accident classification model 63. As shown in FIG. 6, each accident measurement information AMI acquired from the measurement information acquisition unit 11 is input to the accident classification model 63. The cause information acquisition unit 61 inputs the accident measurement information AMI acquired from the measurement information acquisition unit 11 to the accident classification model 63 and performs calculation. The cause information acquisition unit 61 acquires cause information as a calculation result. The cause information may be information in which the cause of the accident is uniquely indicated, or may be information in which a plurality of accident causes and the probability that each cause of the accident may occur are associated with each other. The cause information acquisition unit 61 of the present embodiment acquires cause information in which a plurality of cause of accidents are associated with the probability that each cause of the accident may occur. In the example illustrated in FIG. 6, the cause information acquisition unit 61 has the probability that the accident cause of “bird contact”, “tree contact”, and “coconut fouling” may occur “0.6”, “0.3”, Cause information that is “0.2” is acquired. The cause information acquired by the accident classification model 63 may include a calculation result in which the cause of the accident is “none”.

  Returning to FIG. 1, the cause information acquisition unit 61 supplies the acquired cause information to the display control unit 70. The display control unit 70 causes the terminal device TM1 to display an image indicating the cause information acquired from the cause information acquisition unit 61 (hereinafter, image IM11). FIG. 7 is a diagram illustrating an example of an image IM1 that the display control unit 70 displays on the terminal device TM1. As described above, in the present embodiment, the cause information includes the probability that an accident cause of “bird contact”, “tree contact” and “coconut fouling” may occur is “0.6”, “0.3” and This is information indicating “0.2”. The display control unit 70 generates the image IM1 based on the cause information acquired from the cause information acquisition function unit 60, and causes the terminal device TM1 to display the image IM1.

  Returning to FIG. 1, the relearning function unit 40 inputs the measurement information MI and the accident measurement information AMI based on the target information 81, the target accident information 91, and the accident history information 101, and outputs the cause information. Thus, the updated accident classification model 55 learned by machine learning is generated. Specifically, the relearning function unit 40 includes an accident frequency determination unit 41, a learning data selection unit 42, and an update model generation unit 43. The number-of-accidents determination unit 41 determines whether two or more accidents have occurred in the same distribution facility EQ based on the target accident information 91. The learning data selection unit 42 selects learning data used when generating the updated accident classification model 55. The update model generation unit 43, when the determination result of the accident frequency determination unit 41 indicates that two or more accidents have occurred in the same distribution facility EQ, the measurement information MI selected by the learning data selection unit 42 and The updated accident classification model 55 is generated based on the accident measurement information AMI. The update model generation unit 43 stores the generated updated accident classification model 55 in the update model storage unit 54.

  The model update function unit 50 updates the accident classification model 63 to the updated accident classification model 55 generated by the re-learning function unit 40. Specifically, the model update function unit 50 includes an adaptive range selection unit 51, an adaptation determination unit 52, an update unit 53, and an update model storage unit 54. The updated model storage unit 54 stores an updated accident classification model 55 and updated range information 56. The update range information 56 is information indicating a range (for example, X prefecture) where the distribution information EQ in which the measurement information MI and the accident measurement information AMI input to the updated accident classification model 55 are measured exists. The update model storage unit 54 stores the update range information 56 and the updated accident classification model 55 that can be applied to the range indicated by the update range information 56 in association with each other.

  The adaptive range selection unit 51 selects (clusters) the update range information 56 based on the accident measurement information AMI acquired from the measurement information acquisition unit 11. In the update range information 56, for example, an initial value (eg, X prefecture) of the range selected by the adaptive range selection unit 51 is stored. The adaptive range selection unit 51 selects, for example, the accident measurement information AMI of the power distribution equipment EQ existing in the range indicated by the update range information 56 from the acquired accident measurement information AMI. The adaptation determination unit 52 causes the cause information based on the accident measurement information AMI selected by the adaptation range selection unit 51 and the updated accident classification model 55 indicated by the updated accident classification model 55 associated with the update range information 56. To get. The adaptation determination unit 52 acquires the cause information acquired by the cause information acquisition unit 61.

  Whether the adaptation determination unit 52 updates the updated accident classification model 55 as the accident classification model 63 based on the difference between the cause information acquired by the adaptation determination unit 52 and the cause information acquired by the cause information acquisition unit 61. Determine whether. When there is no difference between the cause information acquired by the adaptation determination unit 52 and the cause information acquired by the cause information acquisition unit 61, that is, the accuracy of the updated accident classification model 55 is determined according to the accident classification model 63. If it is the same as the accuracy of, it is determined that the updated accident classification model 55 is updated as the accident classification model 63. The adaptation determination unit 52 compares the cause information acquired by the adaptation determination unit 52, the cause information acquired by the cause information acquisition unit 61, and the accident history information 101 stored in the accident history information storage unit 100. When the cause information acquired by the updated accident classification model 55 is more consistent with the result information than the cause information acquired by the accident classification model 63, that is, the accuracy of the updated accident classification model 55 is higher than that of the accident classification model 63. If the accuracy is higher than the accuracy, it is determined that the updated accident classification model 55 is updated as the accident classification model 63. The adaptation determination unit 52 is a case where there is a difference between the cause information acquired by the adaptation determination unit 52 and the cause information acquired by the cause information acquisition unit 61, and the accuracy of the updated accident classification model 55 is When the accuracy is lower than the classification model 63, it is not determined that the updated accident classification model 55 is updated as the accident classification model 63.

  When the determination result of the adaptation determination unit 52 indicates that the updated accident classification model 55 is updated as the accident classification model 63, the update unit 53 uses the content of the updated accident classification model 55 as the accident classification model 63 and the model storage unit 62. To remember. Further, when the determination result of the adaptation determination unit 52 indicates that the updated accident classification model 55 is updated as the accident classification model 63, the update unit 53 stores the content of the update range information 56 in the model storage unit 62 as range information 64. Remember. The updating unit 53 may be configured to update a part of the updated accident classification model 55 indicated by the updated accident classification model 55 as the accident classification model 63. In this case, the updating unit 53 includes an updated accident classification model 55 indicating the updated accident classification model 55 determined to be updated by the adaptation determination unit 52, and a range in which cause information can be acquired by the updated accident classification model 55. The update range information 56 to be shown is associated and stored in the model storage unit 62.

  Here, the adaptation range selection unit 51 can adapt the updated accident classification model 55 when the adaptation determination unit 52 determines that the accuracy of the updated accident classification model 55 is lower than the accuracy of the accident classification model 63. Such a range may be selected again and stored as the update range information 56. FIG. 8 is a diagram illustrating an example of selection of the range of the update range information 56. The adaptive range selection unit 51 reselects the range based on, for example, the distribution facility EQ or the type of the distribution facility EQ that is close to the physical distance. The adaptive range selection unit 51 is, for example, a range indicated by the update range information 56 and a range in which the distribution facility EQ exists (for example, X prefecture), a range in which the distribution facility EQ with a closer physical distance exists ( For example, re-selection (re-clustering) to the north of X prefecture or the like. The adaptive range selection unit 51 stores the reselected range as update range information 56 in the update model storage unit 54. Moreover, the adaptive range selection part 51 selects the measurement information MI and accident measurement information AMI of the distribution equipment EQ which exist in the range of the reselected update range information 56 among measurement information MI and accident measurement information AMI. The adaptation determination unit 52 acquires cause information based on the range reselected by the adaptation range selection unit 51 (in this example, the northern part of prefecture X). Until the adaptive range selection unit 51 determines that the accuracy of the updated accident classification model 55 is similar to the accuracy of the accident classification model 63 or improved than the accuracy of the accident classification model 63, Select a range.

  Of the target accident information 91 stored in the target accident information storage unit 90, the accident measurement information AMI used by the update model generation unit 43 to generate the updated accident classification model 55 is already learned teacher information (hereinafter referred to as “accident measurement information”). As the learned accident information), another storage unit (not shown) or another section in the target accident information storage unit 90 may be stored. In this case, the update model generation unit 43 does not use the already learned target accident information for generation of the updated accident classification model 55. That is, the update model generation unit 43 generates the updated accident classification model 55 based on the target accident information 91 other than the learned target accident information.

  The details of the operation of the relearning function unit 40 will be described below with reference to FIG. FIG. 9 is a flowchart illustrating an example of the operation of the relearning function unit 40. The accident frequency determination unit 41 determines whether an accident has occurred in the power system based on the accident history information 101 (step S110). When the accident has occurred in the power system, the accident frequency determination unit 41 determines whether the accident has occurred twice or more in the same distribution facility EQ (step S120). The relearning function unit 40 is in a period until an accident occurs in the power system (step S110; NO), and until the accident becomes a second or more accident in the same distribution facility EQ (step S120; NO), no processing is performed. The accident number determination unit 41 may be configured to acquire accident information from the accident information acquisition unit 12 and determine whether or not an accident has occurred.

  The learning data selection part 42 selects the learning data used when producing | generating the updated accident classification model 55, when the accident has generate | occur | produced twice or more in the same power distribution equipment EQ (step S130). Specifically, the learning data selection unit 42 selects the measurement information MI stored as the target information 81 and the accident measurement information AMI stored in the target accident information 91 as learning data for the updated accident classification model 55. In addition, when there is measurement information MI associated with a trend determination result indicating that the state has changed, the learning data selection unit 42 performs measurement after the measurement date MD associated with the trend determination result. The information MI is selected as learning data for the updated accident classification model 55. Further, the learning data selection unit 42 selects the accident measurement information AMI associated with the accident occurrence date AD after the measurement date MD as learning data of the updated accident classification model 55. The updated model generation unit 43 inputs the measurement information MI and the accident measurement information AMI based on the learning data selected by the learning data selection unit 42, and outputs the cause information after the updated accident that has been learned by machine learning. A classification model 55 is generated (step S140). The update model generation unit 43 stores the generated updated accident classification model 55 in the update model storage unit 54 (step S150).

  The details of the operation of the model update function unit 50 will be described below with reference to FIG. FIG. 10 is a flowchart showing an example of the operation of the model update function unit 50. The model update function unit 50 determines whether an accident has occurred in the power system based on the accident history information 101 (step S210). The model update function unit 50 does not perform processing until an accident occurs in the power system (step S210; NO). The model update function unit 50 may be configured to acquire accident information from the accident information acquisition unit 12 and determine whether or not an accident has occurred.

  The adaptive range selection unit 51 selects the accident measurement information AMI of the power distribution equipment EQ existing in the range indicated by the update range information 56 from the acquired accident measurement information AMI (step S220). The adaptive range selection unit 51 acquires the installation area AR associated with the accident measurement information AMI based on the target accident information 91, for example. The adaptive range selection unit 51 selects the accident measurement information AMI as input information of the updated accident classification model 55 when the acquired installation area AR is an area (position) existing in the range indicated by the update range information 56. The adaptation determination unit 52 acquires cause information based on the updated accident classification model 55 indicated by the updated accident classification model 55 and the accident measurement information AMI selected by the adaptive range selection unit 51 (step S230). Moreover, the adaptation determination part 52 acquires the cause information which the cause information acquisition part 61 acquired based on accident measurement information AMI and the accident classification model 63 (step S240).

  The adaptive determination unit 52 compares the acquired cause information to determine whether or not the accuracy of the updated accident classification model 55 is the same as that of the accident classification model 63 or higher than the accuracy of the accident classification model 63. (Step S250). When the accuracy of the updated accident classification model 55 is the same as or higher than the accuracy of the accident classification model 63, the adaptive determination unit 52 stores the accident classification model stored in the model storage unit 62. 63 and range information 64 are updated (step S260). Specifically, the adaptation determination unit 52 stores the updated accident classification model 55 in the model storage unit 62 as an accident classification model 63. Further, the adaptation determination unit 52 is a model in which the range that is the range selected by the adaptation range selection unit 51 (the range indicated by the update range information 56) and to which the updated accident classification model 55 can be applied is the range information 64. Store in the storage unit 62.

  When the adaptation determination unit 52 determines that the accuracy of the updated accident classification model 55 is lower than the accuracy of the accident classification model 63 (step S250; NO), the adaptation range selection unit 51 applies the updated accident classification model 55. The range is reselected, and the reselected range is stored in the update model storage unit 54 as the update range information 56 (step S260). The adaptive range selection unit 51 selects the accident measurement information AMI acquired from the distribution facility EQ existing in the reselected range among the acquired accident measurement information AMI as input information of the updated accident classification model 55 (step S270). . The adaptation determination unit 52 acquires cause information based on the accident measurement information AMI selected by the adaptation range selection unit 51 and the updated accident classification model 55 (step S280). The adaptation determination unit 52 compares the cause information acquired based on the accident measurement information AMI in the range reselected by the adaptation range selection unit 51 with the cause information acquired in step S240, and the accuracy of the updated accident classification model 55 is determined. However, it is determined whether or not the accuracy of the accident classification model 63 is the same as or higher than the accuracy of the accident classification model 63 (step S290). The adaptive range selection unit 51 and the adaptation determination unit 52 perform steps S260 to S290 until the accuracy of the updated accident classification model 55 is the same as that of the accident classification model 63 or until the accuracy of the accident classification model 63 is improved. Repeat the process. If the accuracy of the updated accident classification model 55 is the same as or higher than the accuracy of the accident classification model 63, the adaptive determination unit 52 advances the process to step S260.

  Note that the adaptive range selection unit 51 and the adaptation determination unit 52 start from step S260 until the accuracy of the updated accident classification model 55 is the same as that of the accident classification model 63 or until the accuracy of the accident classification model 63 is improved. If the accuracy of the updated accident classification model 55 does not become the same as the accuracy of the accident classification model 63 or the accuracy of the accident classification model 63 decreases even if the process of step S290 is repeated a predetermined number of times, an updated model storage unit The post-update accident classification model 55 may be deleted from 54.

  As described above, the accident cause classification apparatus 1 of the present embodiment includes the measurement information acquisition unit 11, the accident information acquisition unit 12, the tendency determination unit 21, the result information acquisition unit 30, and the update model generation unit 43. The cause information acquisition unit 61, the model storage unit 62, the target information storage unit 80, the target accident information storage unit 90, and the accident history information storage unit 100 are included. The model storage unit 62 stores cause information indicating the cause of an accident that has occurred in the target equipment (in this example, the power distribution equipment EQ), and an accident classification model 63. The accident classification model 63 is the cause information indicating the cause of the accident when the measurement information MI obtained by measuring the voltage applied to the distribution facility EQ or the current flowing through the distribution facility EQ (that is, measuring the power) is input. Is a learning model learned by machine learning. The measurement information acquisition unit 11 acquires measurement information MI indicating a result of measurement related to power in the distribution facility EQ. The accident information acquisition unit 12 acquires accident information indicating whether an accident has occurred in the power distribution equipment EQ. The cause information acquisition unit 61 inputs the measurement information MI acquired by the measurement information acquisition unit 11 at the timing when the accident information is acquired by the accident information acquisition unit 12 (that is, the accident measurement information AMI) to the accident classification model 63, Get cause information. The result information acquisition unit 30 is information acquired from a patrolperson of the distribution facility EQ, and acquires result information indicating the cause of the accident in association with the accident information. The accident history information storage unit 100 stores the accident history information 101 in which the result information acquired by the result information acquisition unit 30 is associated with the information indicating the timing at which the accident indicated by the result information occurs. The trend determination unit 21 determines whether the state of the distribution facility EQ has changed based on the measurement information MI. The target information storage unit 80 includes target information 81 in which the determination result of the tendency determination unit 21, measurement information MI, and information indicating the timing at which the measurement information MI is acquired (in this example, the measurement date MD) are associated with each other. Remembered. The updated model generation unit 43 is an updated accident classification model 55 used in place of the accident classification model 63 based on the target information 81, the target accident information 91, and the accident history information 101 acquired or generated at the same time. Is generated and stored in the model storage unit 62.

  Here, when the power system equipment and the surrounding environment of the power system equipment have changed, it may be difficult to acquire appropriate cause information in the accident classification model 63 that has been used conventionally. The accident cause classification apparatus 1 according to the present embodiment determines whether or not the power system facility and the surrounding environment of the power system have changed, that is, whether or not the state of the distribution facility EQ has changed, and the state of the distribution facility EQ has changed. In this case, the updated accident classification model 55 used in place of the accident classification model 63 can be generated.

  Further, in the accident cause classification device 1 of the present embodiment, learning is performed by machine learning so that the tendency determination model storage unit 22 outputs the input measurement information MI based on the measurement information MI acquired in the past. The trend determination model 23 is stored, and the trend determination unit 21 inputs the measurement information MI acquired by the measurement information acquisition unit 11 to the trend determination model 23, and outputs the output information and the input measurement information MI. Based on the difference, it is determined whether the state of the distribution facility EQ has changed. Thereby, the accident cause classification device 1 of the present embodiment can determine whether or not the state of the distribution facility EQ has changed by a simple method.

  Further, in the accident cause classification apparatus 1 of the present embodiment, the accident frequency determination unit 41 is based on the target accident information 91 and the target information 81, and the distribution information EQ including the measurement information MI in the target information 81 has at least an accident. It is determined whether it has occurred twice. The update model generation unit 43 generates the updated accident classification model 55 when the determination result of the accident number determination unit 41 indicates that the accident has occurred at least twice. Here, even if the trend determining unit 21 determines that the state of the distribution facility EQ has changed, the teacher information (in this example, the measurement information is used when the updated accident classification model 55 is generated). MI and accident measurement information AMI) may not be sufficiently collected. In order to generate the updated accident classification model 55, for example, it is required to collect accident measurement information AMI acquired at the time of an accident that has occurred at least twice for a certain distribution facility EQ. The accident cause classification device 1 of the present embodiment sufficiently collects teacher information for generating the updated accident classification model 55 when the determination result of the accident frequency determination unit 41 indicates that the accident has occurred at least twice. After that, the updated accident classification model 55 can be generated.

  Further, in the accident cause classification device 1 of the present embodiment, the learning data selection unit 42 is associated with the determination result of the tendency determination unit 21 indicating that the state of the distribution facility EQ is changing in the target information 81. The measurement information MI and accident measurement information AMI acquired after the measurement date MD from which the measurement information MI is acquired are selected. The update model generation unit 43 generates the updated accident classification model 55 based on the measurement information MI and the accident measurement information AMI selected by the learning data selection unit 42. Here, when the state of the distribution facility EQ is changed, the measurement information MI and the accident measurement information AMI acquired before the change occurs are used as teacher information for generating the updated accident classification model 55. May not be preferred. The accident cause classification device 1 of the present embodiment generates the updated accident classification model 55 based on the measurement information MI and the accident measurement information AMI acquired after a change has occurred in the distribution equipment EQ, so that the current distribution equipment EQ The updated accident classification model 55 can be generated according to the state.

  In the accident cause classification apparatus 1 of the present embodiment, the target accident information 91 used for generating the updated accident classification model 55 among the target accident information 91 stored in the target accident information storage unit 90 is already learned. Stored as target accident information. Further, the updated model generation unit 43 generates the updated accident classification model 55 based on the target accident information 91 other than the target accident information 91 that is already learned target accident information among the target accident information 91. The accident cause classification apparatus 1 of the present embodiment uses the target accident information 91 already used for learning the updated accident classification model 55, that is, the target accident information 91 in which the data characteristics are already reflected in the updated accident classification model 55. The updated accident classification model 55 can be generated by excluding it.

  Further, in the accident cause classification device 1 of the present embodiment, the adaptation determination unit 52 inputs the target accident information 91 (accident measurement information AMI) to the updated accident classification model 55, and generates the cause information and the cause information acquisition unit. Based on the cause information acquired by 61, it is determined whether or not the updated accident classification model 55 can be applied. According to the accident cause classification apparatus 1 of the present embodiment, it is determined whether the updated accident classification model 55 is similar to the accuracy of the accident classification model 63 or improved than the accuracy of the accident classification model 63, and the generated updated Whether or not the accident classification model 55 is appropriate can be determined.

  In the accident cause classification apparatus 1 of the present embodiment, the adaptive range selection unit 51 selects accident measurement information AMI to be input to the updated accident classification model 55. The adaptation determination unit 52 inputs the accident measurement information AMI of the distribution equipment EQ selected by the adaptation range selection unit 51 to the updated accident classification model 55 generated by the update model generation unit 43, and generates the cause information and the cause information Based on the cause information acquired by the acquisition unit 61, it is determined whether the updated accident classification model 55 can be applied. As a result, the accident cause classification apparatus 1 according to the present embodiment applies the updated accident classification model 55 to the distribution equipment EQ (accident measurement information AMI) within a range where the cause information can be acquired by the updated accident classification model 55. can do.

  Moreover, in the accident cause classification device 1 of the present embodiment, the adaptive range selection unit 51 changes the distribution equipment EQ to be selected when the determination result of the adaptation determination unit 52 indicates that the updated accident classification model 55 is not adapted. To do. Thereby, the accident cause classification device 1 of the present embodiment can more appropriately select a range in which cause information can be acquired by the updated accident classification model 55.

  Further, in the accident cause classification apparatus 1 of the present embodiment, the updating unit 53 determines that the updated accident classification model 55 is the accident classification when the determination result of the adaptation determination unit 52 indicates that the updated accident classification model 55 is applied. The model 63 is stored as the model 63. As a result, the accident cause classification apparatus 1 of the present embodiment has the updated accident classification model 55 when the accuracy of the updated accident classification model 55 is equal to or higher than the accuracy of the accident classification model 63. 55 can be used as the accident classification model 63.

  In addition, although the display control part 70 demonstrated above the case where the image IM1 which shows the cause information acquired from the cause information acquisition part 61 was displayed on the terminal device TM1, it is not restricted to this. The display control unit 70 may be configured to display information other than the cause information on the terminal device TM1. FIG. 11 is a diagram illustrating an example of a period from when the state of the distribution facility EQ is changed to when the updated accident classification model 55 is generated. As shown in FIG. 4, the distribution facility EQ5 is determined to have a change in state on the measurement date MD “2017/11/7”. Further, as shown in FIG. 5, the distribution facility EQ5 has a “ground fault” accident on the accident occurrence dates AD “2017/11/15” and “2017/11/30”. As described above, the accident cause classification device 1 does not generate the updated accident classification model 55 until two or more accidents occur in the same distribution facility EQ. Therefore, as shown in FIG. 11, while it is determined that the state of the distribution facility EQ5 has changed in “2017/11/7”, until “2017/11/30” when the second accident occurs. The accident classification model 63 that does not correspond to the change of the distribution facility EQ outputs the cause information. For example, in the patrol of the distribution equipment EQ after “2017/11/7”, the patrolist preferably knows that a change has occurred in the distribution equipment EQ. FIG. 12 is a diagram illustrating another example of the image IM2 that the display control unit 70 displays on the terminal device TM1. When the target information 81 includes a trend determination result indicating that a change has occurred in the distribution facility EQ, the display control unit 70 includes the terminal device TM1 after the measurement date MD associated with the trend determination result. An image IM2 including a message (message MS shown) indicating that a change has occurred in the distribution facility EQ is displayed.

  Thereby, in the accident cause classification device 1 of the present embodiment, the display control unit 70 displays the display device (this device) based on the cause information acquired by the cause information acquisition unit 61 and the determination result determined by the tendency determination unit 21. In one example, the display of information related to the distribution facility EQ displayed on the terminal device TM1) (in this example, the message MS) is controlled. As a result, the accident cause classification apparatus 1 of the present embodiment allows the patrol staff to check the distribution facility EQ even during the period from when the state of the distribution facility EQ changes until the updated accident classification model 55 is generated. It can be shown that the state is changing.

  In the above description, the target information 81 includes the measurement date MD indicating the date on which the measurement information MI is acquired. However, the present invention is not limited to this. As long as the target information 81 includes information indicating the timing at which the measurement information MI is acquired, any information may be included instead of the measurement date MD. Moreover, although the case where the accident occurrence date AD indicating the date when the accident indicated by the result information is included is included in the accident history information 101 has been described, it is not limited thereto. The target accident information 91 and the accident history information 101 may include any information in place of the accident occurrence date AD as long as the information indicating the timing of occurrence of the accident indicated by the result information is included. The information indicating the timing may be, for example, information indicating the time or information indicating the date. In this case, the updated model generation unit 43 generates the updated accident classification model 55 based on the measurement information MI and the accident measurement information AMI acquired at the same time as the accident information.

  According to at least one embodiment described above, the accident cause classification apparatus 1 includes a model storage unit 62, a measurement information acquisition unit 11, an accident information acquisition unit 12, a result information acquisition unit 30, and an accident history information storage unit. 100, a tendency determination unit 21, a target information storage unit 80, a target accident information storage unit 90, a cause information acquisition unit 61, and an update model generation unit 43. The model storage unit 62 generates the cause of the accident based on the cause information indicating the cause of the accident that occurred in the distribution facility EQ and the measurement information MI obtained by measuring the voltage applied to the distribution facility EQ or the current flowing through the distribution facility EQ. The accident classification model 63 learned by machine learning is stored so as to output cause information indicating The measurement information acquisition unit 11 acquires measurement information MI indicating a measurement result obtained by measuring a voltage applied to the distribution facility EQ or a current flowing through the distribution facility EQ. The accident information acquisition unit 12 acquires accident information indicating whether an accident has occurred in the power distribution equipment EQ. The result information acquisition unit 30 is information acquired from a patrolperson of the power distribution facility EQ, and acquires result information indicating the cause of the accident. The accident history information storage unit 100 stores the accident history information 101 in which the result information acquired by the result information acquisition unit 30 is associated with the date when the result information is acquired. The trend determination unit 21 determines whether the state of the distribution facility EQ has changed based on the measurement information MI. The target information storage unit 80 stores target information 81 in which the determination result of the tendency determination unit 21 is associated with the measurement information MI and the measurement date MD from which the measurement information MI is acquired. The target accident information storage unit 90 includes target accident information 91 in which measurement information MI acquired in response to acquiring accident information indicating that an accident has occurred in the power distribution equipment EQ and the accident occurrence date AD are associated with each other. Remember. The cause information acquisition unit 61 inputs the target accident information 91 to the accident classification model 63 and generates cause information. The update model generation unit 43 generates an updated accident classification model 55 used in place of the accident classification model 63 based on the target information 81, the target accident information 91, and the accident history information 101.

  Although several embodiments of the present invention have been described, these embodiments are presented by way of example and are not intended to limit the scope of the invention. These embodiments can be implemented in various other forms, and various omissions, replacements, and changes can be made without departing from the spirit of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are also included in the invention described in the claims and the equivalents thereof.

DESCRIPTION OF SYMBOLS 1 ... Accident cause classification device, 10 ... Acquisition function part, 11 ... Measurement information acquisition part, 12 ... Accident information acquisition part, 20 ... Trend determination function part, 21 ... Trend determination part, 22 ... Trend determination model storage part, 23 ... Trend determination model, 30 ... result information acquisition unit, 40 ... relearning function unit, 41 ... accident frequency determination unit, 42 ... learning data selection unit, 43 ... update model generation unit, 50 ... model update function unit, 51 ... adaptive range Selection part 52 ... Adaptive judgment part 53 ... Update part 54 ... Update model storage part 55 ... Updated accident classification model 56 ... Update range information 60 ... Cause information acquisition function part 61 ... Cause information acquisition part 62 ... Model storage unit, 63 ... Accident classification model, 64 ... Range information, 70 ... Display control unit, 80 ... Target information storage unit, 81 ... Target information, 90 ... Target accident information storage unit, 91 ... Target accident information, 100 ... Accident history information Department, 101 ... accident history information

Claims (10)

A model storage unit that stores an accident classification model that is learned by machine learning and that outputs cause information when measurement information is input;
A measurement information acquisition unit that acquires measurement information indicating a result of measurement of power in the target facility;
An accident information acquisition unit for acquiring accident information indicating whether an accident has occurred in the target facility;
A cause information acquisition unit for acquiring the cause information by inputting the measurement information acquired by the measurement information acquisition unit at the timing when the accident information is acquired by the accident information acquisition unit to the accident classification model;
A result information acquisition unit that is information acquired from a patrolperson of the target facility, and acquires result information indicating the cause of the accident in association with the accident information;
An accident history information storage unit that stores accident history information in which the result information acquired by the result information acquisition unit and information indicating the timing at which the accident indicated by the result information is associated;
Based on the measurement information, a tendency determination unit that determines whether the state of the target facility has changed,
A target information storage unit that stores target information in which the determination result of the tendency determination unit, the measurement information, and information indicating the timing at which the measurement information is acquired are associated;
A target accident information storage unit for storing target accident information in which the measurement information acquired in response to acquiring the accident information and information indicating the timing at which the accident occurred are associated;
Based on the target information acquired or generated at the same time, the target accident information, and the accident history information, an updated accident classification model used instead of the accident classification model is generated, and the model storage unit An update model generation unit to be stored in
An accident cause classification device comprising: A trend determination model storage unit that stores a trend determination model learned by machine learning so as to output the input measurement information based on the measurement information acquired in the past,
The trend determination unit
The measurement information acquired by the measurement information acquisition unit is input to the trend determination model, and whether the state of the target facility has changed based on the difference between the output information and the input measurement information Determine
The accident cause classification device according to claim 1. Based on the target accident information and the target information, the target equipment that includes the measurement information in the target information further includes an accident number determination unit that determines whether the accident has occurred at least twice.
The updated model generation unit generates the updated accident classification model when the determination result of the accident number determination unit indicates that the accident has occurred at least twice.
The accident cause classification device according to claim 1 or 2. Among the target information, the measurement information acquired after the timing at which the measurement information associated with the determination result of the tendency determination unit indicating that the state of the target facility has changed is acquired, and the A learning data selection unit for selecting target accident information,
Further comprising
The updated model generation unit generates the updated accident classification model based on the measurement information selected by the learning data selection unit.
The accident cause classification device according to any one of claims 1 to 3. Of the target accident information stored in the target accident information storage unit, the target accident information used for generation of the updated accident classification model is stored as learned target accident information,
The updated model generation unit generates the updated accident classification model based on the target accident information other than the target accident information that is already learned target accident information among the target accident information.
The accident cause classification device according to any one of claims 1 to 4. The target accident information is input to the updated accident classification model, and whether or not the updated accident classification model can be applied is determined based on the generated cause information and the cause information acquired by the cause information acquisition unit. The accident cause classification device according to any one of claims 1 to 5, further comprising an adaptive determination unit. An adaptive range selection unit that selects the range corresponding to the position where the target equipment exists, the range to which the updated accident classification model is applied,
Further comprising
The adaptive determination unit inputs the target accident information of the target equipment existing in the range selected by the adaptive range selection unit into the updated accident classification model updated by the update model generation unit, and generates the target accident information Based on the cause information and the cause information acquired by the cause information acquisition unit, determine whether the updated accident classification model can be applied,
The accident cause classification device according to claim 6. The adaptation range selection unit changes the range to be selected when the determination result of the adaptation determination unit indicates that the updated accident classification model is not adapted.
The accident cause classification device according to claim 7. The update part which memorize | stores the said updated accident classification model in the said model memory | storage part as the said accident classification model when the determination result of the said adaptation determination part shows that the said updated accident classification model is adapted. The accident cause classification device according to any one of claims 1 to 8. The display control part which controls the display of the information about the target equipment displayed on a display device based on the cause information acquired by the cause information acquisition part and the determination result determined by the tendency determination part. The accident cause classification device according to any one of claims 1 to 9.

Images