JP2023167067A - Fault point survey supporting system and fault point survey supporting program - Google Patents

Abstract

To appropriately support the survey of a fault point in the event of an accident in a distribution system.SOLUTION: A fault point survey supporting system comprises a distribution map information system 3 for storing facility information including positions of distribution facilities, a customer reception system 7 for acquiring abnormality information that includes an abnormal event including spark from the distribution facilities and a place of occurrence from the outside, and a survey supporting computer 8 which acquires the distribution facilities corresponding to an accident section identified by a distribution automated system 2 from the distribution map information system 3 and detects the distribution facility having a high degree of priority in survey from among the corresponding distribution facilities based on survey information including the abnormality information acquired by the customer reception system 7.SELECTED DRAWING: Figure 1

Description

The present invention relates to a fault point search support system and a fault point search support program that support fault point search when an accident occurs in a power distribution system.

Conventionally, when an accident occurs on a power distribution line, restoration work relies heavily on the experience of on-site workers, and the time it takes to locate the accident point varies depending on the experience and ability of the worker. Furthermore, in recent years, due to personnel reductions and a reduction in the number of new employees hired, the aging of workers has progressed, and the transfer of skills to younger workers has not been carried out properly. - There is a risk that abilities will not be inherited properly.

Against this background, by utilizing the map, equipment, accident, and work information contained in the system, on-site workers can determine and instruct effective exploration routes to explore the accident site, and determine procedures for operating equipment at the site. An accident point exploration support system is known that automatically determines and gives instructions (for example, see Patent Document 1). This system determines the accident location exploration policy using accident history information that records accident occurrence information and past accident location exploration results, map information that includes the location of power distribution equipment, and information on work performed before the accident. The system includes an exploration route determining means and an exploration procedure determining means for generating work instructions for accident point exploration.

Japanese Patent Application Publication No. 2002-152932

By the way, when a power distribution line accident occurs, abnormalities such as sparks and fallen trees often occur around the accident point, and such abnormalities may be witnessed by nearby residents. Furthermore, it has been experienced that abnormalities occur in power distribution equipment due to wind, rain, snow accumulation, etc., leading to power distribution line accidents. However, although the system described in Patent Document 1 determines the exploration route by taking past accident point exploration results into consideration, it does not take such unusual information or weather information into account, so it is difficult to determine an appropriate exploration route. In other words, it is difficult to properly support exploration.

SUMMARY OF THE INVENTION Therefore, an object of the present invention is to provide a fault point search support system and a fault point search support program that can appropriately support fault point search when an accident occurs in a power distribution system.

In order to solve the above problem, the invention of claim 1 provides an equipment information storage means for storing equipment information including the location of the power distribution equipment, and an abnormality information including abnormal events including sparks from the power distribution equipment and their occurrence locations. An abnormality information acquisition means acquired from the outside, and the distribution equipment corresponding to the fault section identified by the distribution automation system are acquired from the equipment information storage means, and based on exploration information including the abnormality information acquired by the abnormality information acquisition means. This is a fault point exploration support system, comprising: a fault point estimating means for determining a power distribution facility having a high exploration priority from among the corresponding power distribution facilities.

The invention according to claim 2 is the fault point exploration support system according to claim 1, which includes an inspection result storage means for storing past inspection results for the power distribution equipment, and the fault point estimation means includes the fault point detection support system including the inspection results. The present invention is characterized in that the power distribution equipment having a high exploration priority is determined based on exploration information.

The invention according to claim 3 is the accident point exploration support system according to claim 1 or 2, further comprising a weather information acquisition means for acquiring weather information at the time of occurrence of an accident identified by the power distribution automation system, and the accident point estimation The means is characterized in that the power distribution equipment having a high exploration priority is determined based on the exploration information including the weather information.

The invention according to claim 4 is the accident point exploration support system according to claims 1 to 3, further comprising an accident information storage means for storing past accident information that is information regarding accidents that have occurred in the past regarding power distribution equipment, The estimating means is characterized in that it determines the power distribution equipment having a high exploration priority based on the exploration information including the past accident information.

The invention according to claim 5 is the accident point exploration support system according to claims 1 to 4, in which the accident point estimating means, when the power distribution equipment corresponding to the accident section and the exploration information are input, selects the exploration priority. It is characterized by using a learning model for fault point estimation that is machine learned based on past performance data so that power distribution equipment with a high degree of failure is output.

The invention of claim 6 provides a computer with an equipment information storage means for storing equipment information including the location of the power distribution equipment, and an abnormality information storage unit for acquiring abnormal information including an abnormal event including a spark from the power distribution equipment and its occurrence location from outside. information acquisition means and the power distribution equipment corresponding to the fault section identified by the distribution automation system from the equipment information storage means, and based on the exploration information including the abnormality information acquired by the abnormality information acquisition means, This is a fault point exploration support program that functions as a fault point estimation means for identifying power distribution equipment with a high exploration priority from among power distribution equipment.

According to a seventh aspect of the invention, in the fault point exploration support program according to claim 6, the computer functions as an inspection result storage means for storing past inspection results for power distribution equipment, and the fault point estimation means The present invention is characterized in that a power distribution facility having a high exploration priority is determined based on the exploration information including the results.

The invention according to claim 8 is the accident point exploration support program according to claim 6 or 7, which causes the computer to function as a weather information acquisition means for acquiring weather information at the time of occurrence of an accident specified by the power distribution automation system, The fault point estimating means is characterized in that it determines the power distribution equipment having a high exploration priority based on the exploration information including the weather information.

The invention of claim 9 is the accident point exploration support program according to claims 6 to 8, which causes the computer to function as an accident information storage means for storing past accident information that is information about accidents that have occurred in the past regarding power distribution equipment. The fault point estimating means is characterized in that the fault point estimation means determines the power distribution equipment having a high exploration priority based on the exploration information including the past accident information.

The invention of claim 10 is the fault point exploration support program according to claims 6 to 9, wherein when the fault point estimation means receives the power distribution equipment corresponding to the fault section and the exploration information, the fault point estimation means selects the exploration priority. It is characterized by using a learning model for fault point estimation that is machine learned based on past performance data so that power distribution equipment with a high degree of failure is output.

According to the inventions recited in claims 1 and 6, abnormality information including an abnormal event including a spark from the power distribution equipment and its occurrence location from among the power distribution equipment corresponding to the fault section identified by the power distribution automation system.・Based on the exploration information, power distribution equipment with a high exploration priority is determined. That is, since the power distribution equipment that should be investigated preferentially is determined based on the location where an abnormality such as a spark has occurred, it is possible to appropriately support the investigation of the fault point. In addition, since power distribution equipment with high exploration priority is automatically determined without relying on human judgment, it can be determined stably without variation even if experienced personnel are absent or technology has not been inherited. becomes possible.

According to the inventions recited in claims 2 and 7, exploration priority is determined based on past inspection results and exploration information for power distribution equipment from among the power distribution equipment that corresponds to the fault section identified by the power distribution automation system. The distribution equipment with the highest value is identified. In other words, the power distribution equipment that should be investigated preferentially is determined based on abnormal conditions of the power distribution equipment discovered during past inspections, so it is possible to appropriately support the investigation of fault points.

According to the inventions recited in claims 3 and 8, exploration priority is determined based on meteorological information and exploration information at the time of occurrence of the accident from among the power distribution equipment corresponding to the accident section specified by the power distribution automation system. Expensive power distribution equipment is determined. In other words, the power distribution equipment that should be investigated preferentially is determined based on weather information such as rainfall, wind speed, and snowfall at the time of the accident, making it possible to appropriately support the investigation of the accident point.

According to the inventions described in claims 4 and 9, exploration priority is determined based on past accident information and exploration information regarding power distribution equipment from among the power distribution equipment corresponding to the accident section specified by the power distribution automation system. Expensive power distribution equipment is determined. In other words, the power distribution equipment that should be investigated preferentially is determined based on information on similar accidents that have occurred in the past, so it is possible to appropriately support the investigation of the accident point.

According to the inventions recited in claims 5 and 10, power distribution equipment with a high exploration priority is output using the learning model for fault point estimation that has been machine learned. It becomes possible to make a more appropriate calculation.

1 is a schematic configuration diagram showing an accident point exploration support system according to an embodiment of the present invention. 2 is a schematic configuration block diagram of an exploration support computer of the accident point exploration support system of FIG. 1. FIG. 2 is a first flowchart showing a support flow by the accident point exploration support system of FIG. 1. FIG. 2 is a second flowchart showing a support flow by the accident point exploration support system of FIG. 1; 3 is a third flowchart showing a support flow by the accident point exploration support system of FIG. 1. FIG. 3 is a fourth flowchart showing a support flow by the accident point exploration support system of FIG. 1. FIG. 3 is a fifth flowchart showing a support flow by the accident point exploration support system of FIG. 1. FIG. FIG. 3 is a functional block diagram showing a schematic configuration of a learning model for accident point estimation of the exploration support computer of FIG. 2. FIG.

The present invention will be described below based on the illustrated embodiments.

FIG. 1 is a schematic configuration diagram showing an accident point exploration support system 1 according to an embodiment of the present invention. This fault point exploration support system 1 is a system that supports the detection of fault points by electric power companies when an accident occurs in a power distribution system, and mainly includes a power distribution automation system 2 and a power distribution map information system (equipment information storage means). 3, a patrol inspection system (inspection result storage means) 4, a weather information system (weather information acquisition means) 5, an accident management system (accident information storage means) 6, and a customer reception system (anomaly information acquisition means) 7 and an exploration support computer (accident point estimating means) 8 that can freely communicate with these.

Power distribution automation system 2 has the same configuration as off-the-shelf/existing power distribution automation systems, and monitors the power distribution system and controls switches, and when an accident or power outage occurs, it identifies the power outage section/area on the power distribution system. It also generates power outage information, such as the occurrence of a power outage and the restoration of power supply, based on the results of monitoring and control. In addition, when an accident occurs, it outputs the time of occurrence, line code, line name, utility pole number in the power outage section, zero-sequence voltage value (V0 level), number of V0 times, classification of ground fault or short circuit, etc.

The power distribution map information system 3 is a system that manages and stores information (equipment information) regarding each power distribution facility. In the case of utility poles, it memorizes the pole number, age, etc.

The patrol inspection system 4 is a system that manages and stores past inspection results for each power distribution facility, including defects and deterioration found during past patrol inspections, repaired contents, etc. (including whether or not a repair ticket has been issued and repairs vote type) is stored for each power distribution facility. For example, occurrence of contact with a tree, defective electric wire, defective insulator, occurrence of rust, etc. are stored.

The weather information system 5 is a system that manages and stores weather information from moment to moment, and may be an external/third party system that provides weather information, or may acquire only desired weather information from an external system. , may be a storage system. Furthermore, a weather monitoring system installed individually on a utility pole or the like by an electric power company may be included, and in this case, it becomes possible to obtain more detailed weather information around the installation. Then, weather information at the time of occurrence of the accident specified by the power distribution automation system 2 is acquired from the weather information system 5 via a communication unit (weather information acquisition means) 83 of the exploration support computer 8, which will be described later.

The accident management system 6 is a system that manages and stores past accident information, which is information regarding accidents that have occurred in the past regarding power distribution equipment, and includes output information from the power distribution automation system 2 at the time of past accidents, weather information, and finally finalized information. It memorizes the accident point, cause of the accident, etc. Furthermore, the entire power distribution system is divided into a plurality of jurisdiction areas, and past accident information is managed and stored for each jurisdiction area.

The customer reception system 7 is a system that accepts, manages, and stores abnormal information regarding power distribution equipment and power distribution systems from the outside. Memorize the location, date and time of occurrence, etc. Here, in this embodiment, abnormality information is received from customer M of the electric power company via telephone, e-mail, etc., but abnormality information can be received and acquired from any person, not just customer M. It's okay. Alternatively, an image showing an abnormal state of the power distribution equipment may be received via the smartphone M1 of the consumer M, and this image may be analyzed to determine an abnormal event in a fault point estimation task 84 described later.

The exploration support computer 8 is a computer for determining which power distribution equipment should be explored preferentially when an accident occurs, and as shown in FIG. , an accident point estimation task (accident point estimation means) 84, a learning task 85,
It includes a learning model 86 for estimating accident points, a record database 87 for estimating accident points, and a central processing unit 88 that controls these.

The input unit 81 is an interface for inputting various information, commands, etc., and inputs a start command for the accident point estimation task 84, etc. The display unit 82 is a display that displays various data and information, and displays the results of the accident point estimation task 84 and the like. The communication unit 83 is an interface for communicating with the outside via the Internet network, telephone communication network, etc., and acquires information from the systems 2 to 7.

The fault point estimation task 84 acquires from the distribution map information system 3 the power distribution equipment corresponding to the fault area/outage area identified by the power distribution automation system 2, and prioritizes exploration among the corresponding power distribution equipment based on the exploration information. This is a task program that identifies power distribution equipment with high performance. In other words, when an accident occurs in the power distribution system, it is activated automatically or by inputting a start command, and the distribution automation system 2 collects information (exploration information) that should be referenced when exploring among the power distribution equipment included in the accident area. ), determine which power distribution facilities should be explored preferentially. Here, the exploration information includes abnormality information, inspection results, weather information, and past accident information.

Specifically, as shown in Figures 3 to 7, first, the distribution automation system 2 collects information such as the time of accident occurrence, line code, line name, utility pole number in the power outage section, V0 level, V0 number, and whether it is a ground fault or short circuit. The classification etc. are acquired (step S1). Next, information regarding the power distribution equipment corresponding to the acquired line code, line name, and utility pole number of the power outage section is extracted and acquired from the power distribution map information system 3 (step S2). As a result, an extraction file F1 containing "Hamamoto No. 1 switch", "Hamamoto No. 2 transformer", etc. is generated.

Here, V0 is a symbol representing the ground fault strength (1 to 3), and when V0 = 3, it is a complete ground fault, and there is a high possibility that it will occur around a well-grounded device. When V0=1, it is not a complete ground fault, so it is predicted that trees will come into contact with the high voltage line, cracks in the insulator, etc. Furthermore, if the frequency/number of ground faults occurs multiple times, it can be assumed that some kind of obstacle or electric wire is touching or separating due to wind and rain. In this way, the cause of the accident and the location of the accident can be estimated based on the V0 level and the number of V0 times, and priority points are given in consideration of the V0 level and the number of V0 times in step S13, which will be described later. .

Next, based on the past inspection results stored in the patrol inspection system 4, an exploration priority is assigned to each power distribution facility in the extraction file F1 (step S3). In other words, in the equipment information of extraction file 1, a search is made for utility poles for which defects or repairs have been memorized (repair slips have been issued) based on past inspection results, and priority points (flags) are assigned to utility pole numbers for which repair slips have been issued. ). At this time, priority points are given according to preset conditions, since the possibility of a distribution line accident differs depending on the nature of the defect or repair. For example, add as follows.

Tree contact: 0.5
Defective wire: 0.5
Insulator defect: 0.5
Rust generation equipment (progress rate: 50% or more): 0.5
Rust generating equipment (progress rate: black rust): 1.0
As a result, an extraction file F2 is generated in which priority points are given to the corresponding equipment/utility pole numbers. Here, by comparing the aerial photographs of the accident area, it is determined that there are trees within a predetermined distance (for example, 20 m) at right angles to a linear range connecting adjacent pillars with a line, and that there is no possibility of tree contact or collapse. It may be set that it can be assumed.

Next, weather information at the time of the accident is acquired from the weather information system 5 (step S4).
). At this time, if individual weather monitoring systems are installed on utility poles, etc., weather information will be obtained from the weather monitoring systems around the power outage area or accident location.

Then, if there was lightning at the time of the accident (“Yes” in step S5), an exploration priority is assigned to utility poles on which lightning arresters, transformers, etc. are installed that are likely to fail due to lightning (step S6). . That is, when there is lightning, grounded equipment is often the cause of failure, so equipment such as lightning arresters and transformers are prioritized. At this time, the added value of the priority points is adjusted according to the V0 level (ground fault strength). For example, the priority points to be added are adjusted according to the strength of V0 as follows.

If V0=1, additional priority points: 0.5
When V0=2, additional priority points: 1.0
If V0=3, additional priority points: 1.5
As a result, an extraction file F3 is generated in which additional points (underlined in the file F3) are added to the corresponding equipment/utility pole numbers.

In addition, if the wind speed at the time of the accident was equal to or higher than a predetermined value (for example, 15 m) (if "Yes" in step S7), the equipment/equipment installed on the utility pole or the defective state of the utility pole (whether there is a repair ticket or not) ) etc., there is a risk of failure due to wind pressure, so exploration priorities are assigned to these utility poles (step S8). In other words, if the wind speed is strong, it is assumed that the edge wire of the equipment will break or the wire will break due to contact with a tree. Furthermore, if the arms are rusted, priority points will be given because the arms may break due to wind pressure and the wires may get mixed up. For example, add priority points as follows.

If there is no repair ticket: Telephone pole with transformer installed: 0.5
Telephone pole with grounded lightning arrester: 0.5
Area where trees are close together: 0.5
If there is a repair ticket, the repair ticket type is tree contact: 1.5
Repair ticket type is arm rust (small): 0.5
Repair ticket type is arm rust (medium): 1.0
Repair ticket type is arm rust (large): 1.5
Repair ticket type is electric wire: 1.0
As a result, an extraction file F4 is generated in which additional points (underlined in the file F4) are added to the corresponding equipment/utility pole numbers. Here, the extraction file F4 shows a case where there is a repair ticket and the repair ticket type is tree contact.

In addition, if the snowfall at the time of the accident was more than a predetermined value (for example, 10 cm) (if "Yes" in step S9), depending on the installation location, type, and defective state of the power distribution equipment (presence or absence of a repair ticket), Since there is a risk of failure due to the weight of snow, exploration priorities are assigned to these power distribution facilities (step S10). That is, when there is snowfall, priority points are given because there is a risk of trees collapsing due to the snowfall, wires breaking due to the weight of the snow, etc. For example, add priority points as follows.

If there is no repair ticket, electric wire specified for mountainous areas with a thickness of 25 sq or less: 0.5
Telephone pole designated for mountainous areas: 0.5
Area where trees are close together: 0.5
If there is a repair ticket, the repair ticket type is tree contact: 1.5
Repair ticket type is arm rust: 1.0
Repair ticket type: Wire spacing is narrow: 1.0
As a result, an extraction file F5 is generated in which additional points (underlined in the file F5) are added to the corresponding equipment/utility pole numbers. Here, the extraction file F5 shows a case where there is a repair ticket and the repair ticket type is tree contact.

Furthermore, if the amount of rainfall at the time of the accident was more than a predetermined value (for example, 15 mm/h) (if "Yes" in step S11), depending on the type of power distribution equipment and its defective state (presence of repair ticket), etc. Since there is a risk of failure due to rainwater, exploration priorities are assigned to these power distribution facilities (step S12). In other words, when there is a lot of rainfall, priority points are given because landslides can cause trees to collapse and rainwater can enter rusty switch bodies. For example, add priority points as follows.

If there is no repair ticket: Telephone pole with switch installed: 0.5
Area where trees are close together: 0.5
If there is a repair ticket, the repair ticket type is tree contact: 1.0
Repair ticket type is switch rust (small): 1.0
Repair ticket type is switch rust (medium or higher): 1.5
As a result, an extraction file F6 is generated in which additional points (underlined in the file F6) are added to the corresponding equipment/utility pole numbers. Extracted file F6 shows a case where there is a repair ticket and the repair ticket type is tree contact. Here, priority points may be given to utility poles in which humidity is high due to rain or the like, and a transformer with a small load relative to the capacity is installed, as there is a risk of burnout.

Next, there is a risk of failure depending on the zero-sequence voltage value, number of occurrences of V0, and ground fault/short circuit classification obtained from the distribution automation system 2, as well as the type of distribution equipment and defective state (presence or absence of a repair ticket). , allocates exploration priorities to those power distribution facilities (step S13). For example, add priority points as follows.

"If the number of V0 times is 3 or more and there is a ground fault"
Without repair ticket Locations where trees are close together: 0.5
Telephone pole with equipment installed: 0.5
If there is a repair ticket, areas where trees are close together: 1.0
Bad PC cutout: 1.0
"In case of short circuit"
Without repair ticket Locations where trees are close together: 0.5
Wire thickness 25sq: 0.5
If there is a repair ticket, areas where trees are close together: 1.0
Poor distance between electric wires: 1.0
As a result, an extraction file F7 is generated in which additional points (underlined in file F7) are added to the corresponding equipment/utility pole numbers. Here, the extraction file F7 shows a case where there is a short circuit and a repair ticket.

Next, based on the information on the power distribution equipment in the accident section acquired in step S2 and the weather information acquired in step S4, the cause of the accident that occurred under the same conditions and the location of the accident equipment are determined from the past accident information in the accident management system 6. information is extracted (step S14). That is, the cause of the accident that occurred on the same track and weather conditions (with an error of ±10%) as the accident that occurred this time and the location of the accident equipment are extracted from the accident management system 6 and used in the next step and thereafter. Here, past accident information within the jurisdiction area including the relevant accident section is targeted for extraction.

For example, the following information is extracted and the extraction results, such as the cause of the accident, are stored in an extraction file (step S15).

Cause of accident Crosstalk due to contact with trees (probability 60%)
Equipment failure (probability 30%)
Ground fault caused by birds, animals or snakes (10% probability)
Accident equipment location Location with less than 25 square meters of electric wire (probability 70%)
Switch pillar (20%)
Transformer pillar (10%)
Specific accident section Utility pole number: Between 1 and 5 (probability 30%)
Next, based on the extracted cause of the accident, an exploration priority is assigned to the power distribution equipment in the extraction file F1 (step S16). For example, add priority points as follows.

Cause of accident Crosstalk due to contact with trees (probability 60%)
Equipment failure (probability 30%)
Ground fault caused by birds, animals or snakes (10% probability)
In this case, the priority level is increased for the utility poles in extraction file 1 that meet the following conditions. At this time, the weight is added based on the probability weight and the defective state (presence or absence of a repair ticket).

If there is no repair ticket: Total additional points 1.5
Tree approach point: 0.9 (weight 60%)
Equipment installation pillar: 0.45 (weight 30%)
Electric wire 25sq or less: 0.15 (weight 10%)
If there is a repair ticket: Total additional points 2.0
Tree contact: 1.2 (weight 60%)
Equipment failure: 0.6 (weight 30%)
Defective wire (25sq or less): 0.2 (weight 10%)
As a result, an extraction file F8 is generated in which additional points (underlined in file F8) are added to the corresponding equipment/utility pole numbers. Here, if there is a high probability of ground faults caused by birds, animals, or snakes based on past accident information, there is basically a risk that similar ground faults will occur on all utility poles. There is a risk of ground fault,'' added the comment. Note that snakes and other animals often appear around the rainy season when the humidity is high and birds make nests on telephone poles (March to June), so the probability of encountering snakes is high during this period.

Further, in the same manner as the cause of the accident in step S16, an exploration priority is assigned to the power distribution equipment in the extraction file F1 based on the accident equipment location extracted in step S15 (step S17).

Next, based on the specific accident section extracted in step S15, an exploration priority is assigned to the power distribution equipment in the extraction file F1 (step S18). That is, when a specific accident section is extracted in step S15 (when the same accident section as in the past is specified), priority points are given to utility poles in that section. However, a maximum number of points to be awarded (for example, a maximum of 2.0) is set. For example, if the area between utility pole numbers 1 to 5 is extracted as a specific accident section and the weight/probability is 30%, additional points will be given to the corresponding equipment/utility pole number as shown in the underline in file F9.

Next, based on the weather information (with an error of ±10%) acquired in step S4 and the information on accidents that occurred recently, the accident management system 6 selects information on past accidents that occurred under the same conditions from all areas covered by the accident management system 6. Information on the cause of the accident and the location of the accident equipment is extracted (step S19). In other words, we identify the causes of accidents that have occurred in recent years from statistics on accident information in all areas under our jurisdiction, and identify causes of accidents that may occur in common across all jurisdictions and throughout the company (such as deterioration over time, equipment failure of a specific manufacturer, etc.). Causes of accidents at specific facilities (years, etc.) will be extracted and used to assign exploration priorities.

For example, the following information is extracted and the accident cause extraction results (company-wide) are stored in an extraction file (step S20).

Accident cause Poor tree contact probability 45%
20% probability of defective transformer made by ○○
15% probability of defective transformer made by ○○
10% probability of ground fault caused by birds, animals or snakes
Other probability 10%
Next, exploration priorities are assigned to the power distribution equipment in the extraction file F1 based on the extracted cause of the accident, etc. (step S21). For example, add priority points as follows.

45% probability that the cause of the accident and the location of the accident was poor contact with the tree.
20% probability of defective transformer made by ○○
15% probability of defective transformer made by ○○
10% probability of ground fault caused by birds, animals or snakes
Other probability 10%
in the case of,
No repair ticket: Total additional points: 1.5
Tree approach point: 0.7 (weight 45%)
Defective transformer made by ○○: 0.3 (weight 20%)
Defective switch made by ○○: 0.2 (weight 15%)
Wire thickness 25sq or less: 0.15 (weight 10%)
Others: 0.15 (weight 10%)
With repair ticket: Total additional points 2.0
Tree approach point: 0.9 (weight 45%)
Defective transformer made by ○○: 0.4 (weight 20%)
Defective switch made by ○○: 0.3 (weight 15%)
Wire thickness 25sq Defective: 0.2 (weight 10%)
Others: 0.2 (weight 10%)
As a result, an extraction file F10 is generated in which additional points (underlined in the file F10) are added to the corresponding equipment/utility pole numbers.

Next, in the vicinity of the occurrence of the accident, the customer reception system 7 receives abnormality information in the accident section from the customer M, and in the case of information directly connected to the accident event (wire breakage, sparks, information on fallen utility poles, etc.) (step S22 If "Yes"), the priority order of the corresponding power distribution equipment is set as the highest priority, and the extraction file F11 is generated with the priority order assigned. On the other hand, if there is no abnormality information from the customer M, an extraction file F12 is generated by assigning priorities to the extraction file F10 in descending order of priority points.

Here, the priority order and priority points may be changed depending on the abnormal event or occurrence location of the abnormal information. For example, one week before the accident, you may notice that there is a crow's nest on the utility pole, there are trees growing around the pole, or the arm of the utility pole is bent. In the case of an event that is likely to be triggered, predetermined priority points may be given instead of the highest priority.

Such a fault point estimation task 84 performs machine learning based on past performance data so that when the power distribution equipment and exploration information corresponding to the accident section are input, the power distribution equipment with a high exploration priority is output. A learning model 86 for estimating accident points is used. This accident point estimation learning model 86 is created by the learning task 85.

That is, the learning task 85 uses past performance data recorded and accumulated in the accident point estimation performance database 87 to create a learning model 86 for accident point estimation using a known machine learning algorithm such as a neural network. This fault point estimation performance database 87 is based on the power distribution equipment corresponding to the fault section and exploration information as input information, and the power distribution facilities with high exploration priority determined by experts and experts in fault point exploration and information analysis. This is a database that records and accumulates performance data including. In addition, past performance data is created based on distribution equipment and exploration information that corresponds to the actual accident section, and distribution equipment that has a high exploration priority determined by experts and experts in accident point exploration and information analysis. This includes data created during pre-training, as well as data created during pre-training.

As shown in FIG. 8, this learning task 85 uses machine learning/deep learning using a neural network, and is based on the performance data recorded in the accident point estimation performance database 87. Create a neural network with the power distribution equipment and exploration information to be used as the input layer, the power distribution equipment with high exploration priority as the output layer, and the analysis process from the input layer to the output layer as the middle layer. Then, the learning task 85 uses the performance data of the accident point estimation learning model 86 as learning data to learn various parameters in the intermediate layer. That is, the learning task 85 performs learning of various parameters in the intermediate layer so that power distribution equipment with a high exploration priority is appropriately selected and output based on the power distribution equipment corresponding to the accident section and the exploration information.

As described above, according to this fault point detection support system 1, abnormal events including sparks from power distribution equipment and their occurrence locations are included in the power distribution equipment corresponding to the fault section identified by the power distribution automation system 2. Based on the abnormality information and exploration information, power distribution equipment with a high exploration priority is determined. That is, since the power distribution equipment that should be investigated preferentially is determined based on the location where an abnormality such as a spark has occurred, it is possible to appropriately support the investigation of the fault point. In addition, since power distribution equipment with high exploration priority is automatically determined without relying on human judgment, it can be determined stably without variation even if experienced personnel are absent or technology has not been inherited. becomes possible.

Further, from among the power distribution equipment corresponding to the fault section identified by the power distribution automation system 2, power distribution equipment with a high exploration priority is determined based on past inspection results and exploration information for the power distribution equipment. In other words, the power distribution equipment that should be investigated preferentially is determined based on abnormal conditions of the power distribution equipment discovered during past inspections, so it is possible to appropriately support the investigation of fault points.

Further, from among the power distribution facilities corresponding to the accident section specified by the power distribution automation system 2, power distribution facilities with a high exploration priority are determined based on weather information and exploration information at the time of occurrence of the accident. In other words, the power distribution equipment that should be investigated preferentially is determined based on weather information such as rainfall, wind speed, and snowfall at the time of the accident, making it possible to appropriately support the investigation of the accident point.

Further, from among the power distribution equipment corresponding to the accident section specified by the power distribution automation system 2, power distribution equipment with a high exploration priority is determined based on past accident information and exploration information regarding the power distribution equipment. In other words, the power distribution equipment that should be investigated preferentially is determined based on information on similar accidents that have occurred in the past, so it is possible to appropriately support the investigation of the accident point.

Moreover, since the power distribution equipment with a high exploration priority is output using the learning model 86 for fault point estimation that has been machine learned, it becomes possible to more appropriately determine the power distribution equipment that should be explored preferentially.

Although the embodiments of this invention have been described in detail above, the specific configuration is not limited to this embodiment, and even if there are changes in the design within the scope of the gist of this invention, Included in invention. For example, the causes of accidents, priority points, etc. described above are merely examples and are not limited to these, of course.

On the other hand, the above-described accident point exploration support system 1 and exploration support computer 3 may be configured by installing the following accident point exploration support program into a general-purpose computer. That is, the computer is used as an equipment information storage means (power distribution map information system 3) that stores equipment information including the location of power distribution equipment, and an inspection result storage means (tour inspection system 4) that stores past inspection results for power distribution equipment. , a weather information acquisition means (weather information system 5) that acquires weather information at the time of an accident identified by the power distribution automation system 2, and an accident information storage unit that stores past accident information that is information about accidents that have occurred in the past regarding power distribution equipment. A storage means (accident management system 6), an abnormality information acquisition means (customer reception system 7) that acquires abnormal information including abnormal events including sparks from power distribution equipment and their occurrence locations from outside, and a power distribution automation system 2. The power distribution equipment corresponding to the specified accident section is acquired from the equipment information storage means, and exploration priority is determined from among the relevant power distribution equipment based on exploration information including abnormality information, inspection results, weather information, and past accident information. The fault point estimation means functions as a fault point estimation means (fault point estimation task 84) that determines the distribution equipment with high exploration priority. A learning model 86 for estimating an accident point that is machine-learned based on past performance data is used so as to be output.

1 Fault point exploration support system 2 Power distribution automation system 3 Power distribution map information system (equipment information storage means)
4 Patrol inspection system (inspection result storage means)
5 Weather information system (weather information acquisition means)
6 Accident management system (accident information storage means)
7 Customer reception system (means for acquiring abnormality information)
8 Exploration support computer (fault point estimation means)
84 Accident point estimation task (accident point estimation means)
85 Learning task 86 Learning model for accident point estimation 87 Performance database for accident point estimation

Claims (10)

equipment information storage means for storing equipment information including the location of the power distribution equipment;
an abnormality information acquisition means for externally acquiring abnormality information including an abnormal event including a spark from the power distribution equipment and its occurrence location;
The power distribution equipment corresponding to the fault section identified by the power distribution automation system is acquired from the equipment information storage means, and based on the exploration information including the abnormality information acquired by the abnormality information acquisition means, the power distribution equipment is selected from among the corresponding power distribution equipment. Fault point estimation means for determining power distribution equipment with high exploration priority;
An accident point exploration support system characterized by comprising: Equipped with an inspection result storage means for storing past inspection results for the power distribution equipment,
The fault point estimating means determines the power distribution equipment having a high exploration priority based on the exploration information including the inspection results.
The accident point exploration support system according to claim 1. comprising a weather information acquisition means for acquiring weather information at the time of occurrence of an accident identified by the power distribution automation system,
The fault point estimating means determines the power distribution equipment having a high exploration priority based on the exploration information including the weather information.
The accident point exploration support system according to claim 1 or 2, characterized in that: Equipped with an accident information storage means for storing past accident information that is information regarding accidents that have occurred in the past regarding power distribution equipment,
The accident point estimating means determines the power distribution equipment having a high exploration priority based on the exploration information including the past accident information.
The accident point exploration support system according to any one of claims 1 to 3. The fault point estimating means performs machine learning based on past performance data so that when the power distribution equipment corresponding to the fault section and the exploration information are input, the power distribution equipment with a high exploration priority is output. Using a learning model for accident point estimation,
The accident point exploration support system according to any one of claims 1 to 4. computer,
equipment information storage means for storing equipment information including the location of the power distribution equipment;
an abnormality information acquisition means for externally acquiring abnormality information including an abnormal event including a spark from the power distribution equipment and its occurrence location;
The power distribution equipment corresponding to the fault section identified by the power distribution automation system is acquired from the equipment information storage means, and based on the exploration information including the abnormality information acquired by the abnormality information acquisition means, the power distribution equipment is selected from among the corresponding power distribution equipment. Fault point estimation means for determining power distribution equipment with high exploration priority;
An accident point exploration support program that is characterized by functioning as a. causing the computer to function as an inspection result storage means for storing past inspection results for power distribution equipment;
The fault point estimating means determines the power distribution equipment having a high exploration priority based on the exploration information including the inspection results.
The accident point exploration support program according to claim 6. causing the computer to function as a weather information acquisition means for acquiring weather information at the time of occurrence of an accident identified by the power distribution automation system,
The fault point estimating means determines the power distribution equipment having a high exploration priority based on the exploration information including the weather information.
The accident point exploration support program according to claim 6 or 7, characterized in that: causing the computer to function as an accident information storage means for storing past accident information that is information regarding accidents that have occurred in the past regarding power distribution equipment;
The accident point estimating means determines the power distribution equipment having a high exploration priority based on the exploration information including the past accident information.
The accident point exploration support program according to any one of claims 6 to 8. The fault point estimating means performs machine learning based on past performance data so that when the power distribution equipment corresponding to the fault section and the exploration information are input, the power distribution equipment with a high exploration priority is output. Using a learning model for accident point estimation,
The accident point exploration support program according to any one of claims 6 to 9.