KR102700731B1 - System and Method for managing Power Distribution Facility - Google Patents

Abstract

The method for managing power distribution facilities of the present invention is a method for managing a plurality of power distribution facilities installed in a power supply system and participating in power supply, comprising: a step of collecting failure history data regarding a faulty power distribution facility that has stopped operating among the plurality of power distribution facilities and replacement history data regarding a replacement power distribution facility that has been replaced before its operation has stopped among the plurality of power distribution facilities; a step of estimating a life probability distribution by manufacturer based on the failure history data; a step of calculating a remaining life probability distribution of the replacement power distribution facility after replacement based on the life probability distribution; a step of estimating the remaining life of the replacement power distribution facility based on the remaining life probability distribution after replacement; a step of generating failure estimation data based on the replacement history data and the estimated remaining life of the replacement power distribution facility; and a step of estimating the remaining life of an active power distribution facility that is in operation among the plurality of power distribution facilities based on the failure estimation data.

Description

{System and Method for managing Power Distribution Facility}

The present invention relates to a method for managing a plurality of power distribution facilities installed in a power supply system.

In general, in order to supply the generated power stably, various types of distribution facilities must be installed. For example, the generated power is stably supplied to the demand side through various types of distribution facilities such as transformers and circuit breakers.

If power distribution equipment assets stop operating due to a breakdown, the power supply will be disrupted, which will not only cause inconvenience to ordinary households but may also result in huge economic losses as the power supply to industrial sites is cut off.

Therefore, for institutions responsible for power supply, it is necessary to prevent power supply interruptions by replacing distribution facility assets with new products at an appropriate time.

However, since it is necessary to have countless types and numbers of distribution facility assets to ensure a stable power supply, there is a problem in that the management cost of distribution facility assets increases when disposing of distribution facility assets that still have a long lifespan left and replacing them with new distribution facility assets.

When calculating the remaining life of existing distribution equipment assets, there was difficulty in accurately predicting the life due to the lack of factors affecting the life.

Therefore, a method for managing distribution facility assets that can supply power stably without power supply interruption while reducing management costs of distribution facility assets is needed.

The present invention relates to a method for managing power distribution facilities, which estimates the remaining lifespan with high accuracy for each of a plurality of power distribution facilities with different durability installed in a power supply system and calculates the optimal replacement time.

According to one feature of the present invention, a method for managing a plurality of distribution facilities installed in a power supply system and involved in power supply comprises the steps of: collecting failure history data regarding a faulty distribution facility that has stopped operating among the plurality of distribution facilities and replacement history data regarding a replacement distribution facility that has been replaced before its operation has stopped among the plurality of distribution facilities; estimating a life probability distribution by manufacturer based on the failure history data; calculating a remaining life probability distribution of the replacement distribution facility after replacement based on the life probability distribution; estimating the remaining life of the replacement distribution facility based on the remaining life probability distribution after replacement; generating failure estimation data based on the replacement history data and the estimated remaining life of the replacement distribution facility; and estimating the remaining life of an active distribution facility that is in operation among the plurality of distribution facilities based on the failure estimation data.

The above failure history data may include a manufacturer name, an installation date, and a failure date, and the above replacement history data may include a manufacturer name, an installation date, a replacement date, and measurement values for a plurality of characteristics measured between the installation date and the replacement date.

The above-mentioned plurality of characteristics may include at least one of an average load and an operating temperature of the replacement distribution equipment.

The step of estimating the life probability distribution by manufacturer based on the above failure history data can calculate the life of the failed power distribution equipment by the difference between the installation date and the failure date included in the failure history data, and approximate the life probability distribution as a Weibull distribution based on the calculated life.

The step of calculating the probability distribution of the remaining life after replacement of the above-mentioned replacement distribution equipment may be performed by using the manufacturer's life probability distribution corresponding to the condition that the life of the above-mentioned distribution equipment is longer than the difference between the replacement date and the installation date as the probability distribution of the remaining life after replacement of the above-mentioned replacement distribution equipment.

The step of estimating the remaining life of the above replacement distribution equipment can estimate the remaining life based on the probability distribution of the remaining life after replacement through random sampling of the above replacement distribution equipment.

The step of estimating the remaining life of the above replacement distribution equipment can estimate the remaining life based on the probability distribution of the remaining life after replacement through rejection sampling for the above replacement distribution equipment.

The step of estimating the remaining life of the above replacement distribution equipment can use the normal distribution as a suggested distribution used in rejection sampling.

The step of generating the above failure estimation data may include: a step of estimating a failure date of the replacement distribution equipment by adding the estimated remaining life of the replacement distribution equipment to the replacement date included in the replacement history data; and a step of generating failure estimation data including the manufacturer name of the replacement distribution equipment, the installation date, the estimated failure date, and measurement values for a plurality of characteristics measured between the installation date and the replacement date.

The step of estimating the remaining life of the above active distribution equipment includes: a step of performing machine learning using learning data in which the manufacturer name and the measured value included in the failure estimation data are input data and the difference between the estimated failure date and the installation date is output data to create a life estimation model of the active distribution equipment; a step of estimating the expected life by inputting the manufacturer name and the measured value of the active distribution equipment into the life estimation model; and a value obtained by deducting the elapsed period since the installation of the active distribution equipment from the estimated expected life can be estimated as the remaining life of the active distribution equipment.

According to another feature of the present invention, a method for managing a plurality of distribution facilities installed in a power supply system and involved in power supply comprises the steps of: collecting failure history data on a faulty distribution facility that has stopped operating among the plurality of distribution facilities and replacement history data on a replacement distribution facility that has been replaced before its operation has stopped among the plurality of distribution facilities; estimating a life probability distribution by manufacturer based on the failure history data; calculating a remaining life probability distribution of the replacement distribution facility after replacement based on the life probability distribution; estimating the remaining life of the replacement distribution facility based on the remaining life probability distribution after replacement; generating failure estimation data based on the replacement history data and the estimated remaining life of the replacement distribution facility; estimating the remaining life of an active distribution facility that is in operation among the plurality of distribution facilities based on the failure estimation data; and generating a health index for evaluating the health of the plurality of distribution facilities.

The step of generating the above soundness index can determine the weights of each of a plurality of features using a Lasso sparseness analysis algorithm, and select a feature having a weight greater than a threshold value among the plurality of features as a valid feature.

The step of generating the above soundness index may assign a reference score to each of a plurality of valid characteristics based on the degree to which the valid characteristic affects the remaining life.

The step of generating the above soundness index can set a plurality of intervals to which the measurement value can belong for each of the plurality of valid characteristics.

The step of generating the above health index can assign a score to each corresponding valid characteristic of the plurality of sections, which is used as a health index for evaluating the health of the plurality of distribution facilities.

The above failure history data may include a manufacturer name, an installation date, and a failure date, and the above replacement history data may include a manufacturer name, an installation date, a replacement date, and measurement values for a plurality of characteristics measured between the installation date and the replacement date.

The above-mentioned plurality of characteristics may include at least one of an average load and an operating temperature of the replacement distribution equipment.

The step of estimating the life probability distribution by manufacturer based on the above failure history data can calculate the life of the failed power distribution equipment by the difference between the installation date and the failure date included in the failure history data, and approximate the life probability distribution as a Weibull distribution based on the calculated life.

The step of calculating the probability distribution of the remaining life after replacement of the above-mentioned replacement distribution equipment may be performed by using the manufacturer's life probability distribution corresponding to the condition that the life of the above-mentioned distribution equipment is longer than the difference between the replacement date and the installation date as the probability distribution of the remaining life after replacement of the above-mentioned replacement distribution equipment.

The step of estimating the remaining life of the above replacement distribution equipment can estimate the remaining life based on the probability distribution of the remaining life after replacement through random sampling of the above replacement distribution equipment.

The present invention has the effect of reducing costs while stably supplying power by calculating the optimal replacement time for each of a plurality of power distribution facilities with different durability installed in a power supply system.

FIG. 1 is a block diagram illustrating a power distribution facility management system according to one embodiment.
FIG. 2 is an example diagram of a life probability distribution estimated using failure history data according to one embodiment.
Figure 3 is an example diagram for estimating the remaining life of replacement distribution equipment based on the life probability distribution of Figure 2.
Figure 4 is a flowchart illustrating a method for managing power distribution facilities according to one embodiment.
Figure 5 is a flowchart detailing the steps for generating fault estimation data of Figure 4.
Figure 6 is a flowchart detailing the steps for estimating the remaining life of the active distribution equipment of Figure 4.
Figure 7 is a flowchart explaining a distribution facility management method according to another embodiment.
Figure 8 is a flowchart detailing the steps for generating the soundness index of Figure 7.

Hereinafter, the embodiments disclosed in this specification will be described in detail with reference to the attached drawings, and identical or similar components will be given the same or similar drawing reference numerals, and redundant descriptions thereof will be omitted. The suffixes "module" and/or "part" used for components in the following description are given or used interchangeably only for the convenience of writing the specification, and do not have distinct meanings or roles in themselves. In addition, when describing the embodiments disclosed in this specification, if it is determined that a specific description of a related known technology may obscure the gist of the embodiments disclosed in this specification, the detailed description thereof will be omitted. In addition, the attached drawings are only intended to facilitate easy understanding of the embodiments disclosed in this specification, and the technical ideas disclosed in this specification are not limited by the attached drawings, and should be understood to include all modifications, equivalents, and substitutes included in the spirit and technical scope of the present invention.

Terms that include ordinal numbers, such as first, second, etc., may be used to describe various components, but the components are not limited by the terms. The terms are used only to distinguish one component from another.

When it is said that a component is "connected" or "connected" to another component, it should be understood that it may be directly connected or connected to that other component, but that there may be other components in between. On the other hand, when it is said that a component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.

In this application, it should be understood that terms such as “comprises” or “has” are intended to specify the presence of a feature, number, step, operation, component, part or combination thereof described in the specification, but do not exclude in advance the possibility of the presence or addition of one or more other features, numbers, steps, operations, components, parts or combinations thereof.

FIG. 1 is a block diagram illustrating a power distribution facility management system according to one embodiment, FIG. 2 is an example diagram of a life probability distribution estimated using failure history data according to one embodiment, and FIG. 3 is an example diagram for estimating the remaining life of a replacement power distribution facility based on the life probability distribution of FIG. 2.

Referring to Fig. 1, a distribution facility management system (1) includes a plurality of distribution facilities (10; 1, 2, ..., n) and a distribution facility management device (20). The distribution facility management system (1) can be operated by a power supply organization that supplies power to power consumers in a wide area.

The distribution equipment (10) may be installed in a power supply system and may be composed of various equipment involved in power supply. For example, the distribution equipment (10) may be composed of a switch, a transformer, etc. Each of the plurality of distribution equipment (10; 1, 2, ..., n) illustrated in Fig. 1 may be manufactured by multiple manufacturers and may have different durability.

The distribution facility management device (20) collects various types of data on a plurality of distribution facilities (10; 1, 2, …, n), performs machine learning on the collected data as learning data to generate a life estimation model, and estimates the remaining life of an active distribution facility (10) currently in operation among the plurality of distribution facilities (10; 1, 2, …, n) based on the generated life estimation model. The distribution facility management device (20) can determine an optimal replacement time based on the remaining life of the active distribution facility (10). For example, the distribution facility management device (20) can generate a health index for evaluating the health of the plurality of distribution facilities (10; 1, 2, …, n) based on the life estimation model.

First, the power distribution facility management device (20) collects the manufacturer name, installation date, and failure date of a faulty power distribution facility (10) among a plurality of power distribution facilities (10; 1, 2, ..., n) as failure history data and stores it in the first database. Since the failure history data includes the actual failure date, it can correspond to uncensored data.

The power distribution facility management device (20) collects the manufacturer name, installation date, replacement date, and measurement values for multiple features measured between the installation date and the replacement date for replacement power distribution facilities (10; 1, 2, ..., n) that were replaced before a failure occurs as replacement history data and stores them in a second database. The multiple features may include at least one of average load and operating temperature, but are not limited thereto, and may include internal operating variables and external environmental variables of the power distribution facility (10). Since the replacement history data does not include data after the replacement date, it may correspond to censored data.

The power distribution facility management device (20) can collect the manufacturer's name, installation date, and measurement values for multiple characteristics measured periodically/non-periodically after the installation date of each of multiple power distribution facilities (10) installed and operating in a power supply system and store them in a second database.

Next, the power distribution facility management device (20) can estimate the life probability distribution for each manufacturer based on the failure history data.

Referring to Fig. 2, the power distribution facility management device (20) can estimate the life probability distribution of a specific manufacturer by using the Weibull distribution as a probability model for the life probability distribution. Specifically, the power distribution facility management device (20) can calculate the life of a failed power distribution facility (10) by the difference between the installation date and the failure date included in the failure history data, and can approximate the life probability distribution to the Weibull distribution based on the calculated life.

The power distribution facility management device (20) can approximate the life probability distribution of each manufacturer's faulty power distribution facility (10) for a plurality of power distribution facilities (10; 1, 2, ..., n) as an exponential Weibull distribution based on the calculated life of each manufacturer's faulty power distribution facility (10). Specifically, the power distribution facility management device (20) can approximate the life probability distribution of each manufacturer as an exponential Weibull distribution by determining three parameters of the following mathematical expression 1 representing the exponential Weibull distribution.

[Mathematical formula 1]

Next, the distribution facility management device (20) can calculate the remaining life probability distribution after replacement of the replacement distribution facility (10) based on the life probability distribution. For example, the distribution facility management device (20) can calculate the remaining life probability distribution after replacement of the replacement distribution facility (10) based on the manufacturer's life probability distribution corresponding to the condition that the life of the distribution facility (10) is longer than the difference between the replacement date and the installation date.

Specifically, the power distribution facility management device (20) can calculate the conditional probability distribution according to the following mathematical expression 2 as the probability distribution of the remaining life after replacement. Here, s corresponds to the replacement date.

[Mathematical formula 2]

Next, the distribution facility management device (20) can estimate the remaining life of the replacement distribution facility (10) based on the probability distribution of the remaining life after replacement.

Referring to FIG. 3, the distribution facility management device (20) can calculate the remaining life probability distribution after replacement from the life probability distribution, and estimate the remaining life of the replacement distribution facility (10) based on the calculated remaining life probability distribution after replacement. Since the life probability distribution is estimated by manufacturer, the distribution facility management device (20) can be based on the life probability distribution of the distribution facility (10) manufactured by the same manufacturer as the replacement distribution facility (10) whose remaining life is to be estimated.

For example, the distribution facility management device (20) can estimate the remaining life of the replacement distribution facility (10) based on the probability distribution of the remaining life after replacement through rejection sampling. The distribution facility management device (20) can use a normal distribution as a suggested distribution used in rejection sampling.

For example, the power distribution facility management device (20) can estimate the remaining life based on the probability distribution of the remaining life after replacement through random sampling for the replacement power distribution facility (10). However, the present invention is not limited thereto, and the remaining life can be estimated based on the probability distribution of the remaining life after replacement through various types of sampling methods.

Next, the distribution facility management device (20) can generate failure estimation data based on the replacement history data and the estimated remaining life of the replacement distribution facility (10).

Specifically, the distribution facility management device (20) can estimate the failure date of the replacement distribution facility (10) by adding the estimated remaining life of the replacement distribution facility (10) to the replacement date included in the replacement history data. For example, the distribution facility management device (20) can estimate the failure date of the replacement distribution facility (10) by adding the estimated remaining life and the replacement date by manufacturers with the same or similar durability.

Thereafter, the distribution facility management device (20) generates failure estimation data including the manufacturer name of the replacement distribution facility (10), the installation date, the estimated failure date, and measurement values for multiple characteristics measured between the installation date and the replacement date. The failure estimation data includes the estimated failure date of the replacement distribution facility (10), so it can correspond to uncensored data.

Next, the distribution facility management device (20) estimates the remaining life of an active distribution facility (10) currently in operation among a plurality of distribution facilities (10; 1, 2, ..., n) based on the failure estimation data.

Specifically, the power distribution facility management device (20) performs machine learning using learning data that uses the manufacturer name and measurement values included in the failure estimation data as input data and the difference between the estimated failure date and the installation date as output data to create a life estimation model of the active power distribution facility (10).

Thereafter, the distribution facility management device (20) inputs the manufacturer name and measurement values of the active distribution facility (10) into the life estimation model to estimate the expected life. The distribution facility management device (20) can estimate the remaining life of the active distribution facility (10) by subtracting the period of time since the installation of the active distribution facility (10) from the estimated expected life. For example, the distribution facility management device (20) can generate a replacement message for the active distribution facility (10) if the remaining life of the active distribution facility (10) is less than a predetermined reference period.

Next, the distribution facility management device (20) can generate a health index for evaluating the health of a plurality of distribution facilities (10; 1, 2, …, n).

Specifically, the power distribution facility management device (20) selects a feature that has a significant effect on the remaining life from among a plurality of features as a valid feature. For example, the power distribution facility management device (20) may determine the weights of each of the plurality of features using a Lasso Regression algorithm, and select a feature having a weight greater than a threshold value from among the plurality of features as a valid feature.

Thereafter, the distribution facility asset management device (20) can assign a reference score to each of the plurality of valid characteristics based on the degree to which the valid characteristic affects the remaining life. For example, the distribution facility asset management device (20) can determine the reference scores of the plurality of valid characteristics so that the reference score of the valid characteristic has a value proportional to the weight of the corresponding valid characteristic, and the sum of the reference scores of each of the plurality of valid characteristics becomes 100.

Thereafter, the distribution facility asset management device (20) can set multiple sections to which the measurement value can belong for each valid characteristic. For example, the distribution facility asset management device (20) can set multiple sections to which the measurement value can belong for each of the multiple valid characteristics by using a widely known change point detection algorithm.

Thereafter, the distribution facility asset management device (20) can assign a score lower than the reference score of the corresponding valid characteristic to each of the plurality of sections. The score assigned to each of the plurality of sections can be used as a health index for evaluating the health of the plurality of distribution facilities (10; 1, 2, ..., n).

For example, among a plurality of distribution facilities (10; 1, 2, ..., n), for an active distribution facility (10) currently in operation, a section to which a plurality of measured values for a plurality of valid characteristics belong can be determined, and a value obtained by adding up the scores corresponding to the determined section can be determined as the health index of the active distribution facility (10).

Specifically, when the health index of the active distribution equipment (10) is relatively low, it can be determined that the remaining life of the active distribution equipment (10) is relatively long, and when the health index of the active distribution equipment (10) is relatively high, it can be determined that the remaining life of the active distribution equipment (10) is relatively short.

FIG. 4 is a flowchart explaining a method for managing power distribution equipment according to one embodiment, FIG. 5 is a flowchart explaining in detail a step for generating fault estimation data of FIG. 4, FIG. 6 is a flowchart explaining in detail a step for estimating the remaining life of active power distribution equipment of FIG. 4, FIG. 7 is a flowchart explaining a method for managing power distribution equipment according to another embodiment, and FIG. 8 is a flowchart explaining in detail a step for generating a health index of FIG. 7.

First, the distribution facility management device (20) collects failure history data regarding a faulty distribution facility that has stopped operating among a plurality of distribution facilities and replacement history data regarding a replacement distribution facility that has been replaced before the operation has stopped among the plurality of distribution facilities (S100).

For example, the power distribution facility management device (20) collects the manufacturer name, installation date, and failure date of a faulty power distribution facility (10) among a plurality of power distribution facilities (10; 1, 2, ..., n) as failure history data and stores it in the first database. Since the failure history data includes the actual failure date, it can correspond to uncensored data.

For example, the power distribution facility management device (20) collects the manufacturer name, installation date, replacement date, and measurement values for multiple features measured between the installation date and the replacement date for replacement power distribution facilities (10; 1, 2, ..., n) that were replaced before a failure occurs as replacement history data and stores them in a second database. The multiple features may include at least one of an average load and an operating temperature, but are not limited thereto, and may include internal operating variables and external environmental variables of the power distribution facility (10). Since the replacement history data does not include data after the replacement date, it may correspond to censored data.

The power distribution facility management device (20) can collect the manufacturer's name, installation date, and measurement values for multiple characteristics measured periodically/non-periodically after the installation date of each of multiple power distribution facilities (10) installed and operating in a power supply system and store them in a second database.

Next, the power distribution facility management device (20) can estimate the life probability distribution by manufacturer based on the failure history data (S200).

Referring to Fig. 2, the power distribution facility management device (20) can estimate the life probability distribution of a specific manufacturer by using the Weibull distribution as a probability model for the life probability distribution. Specifically, the power distribution facility management device (20) can calculate the life of a failed power distribution facility (10) by the difference between the installation date and the failure date included in the failure history data, and can approximate the life probability distribution to the Weibull distribution based on the calculated life.

The power distribution facility management device (20) can approximate the life probability distribution of each manufacturer's faulty power distribution facility (10) for a plurality of power distribution facilities (10; 1, 2, ..., n) as an exponential Weibull distribution based on the calculated life of each manufacturer's faulty power distribution facility (10). Specifically, the power distribution facility management device (20) can approximate the life probability distribution of each manufacturer as an exponential Weibull distribution by determining three parameters of the following mathematical expression 1 representing the exponential Weibull distribution.

[Mathematical formula 1]

Next, the distribution facility management device (20) can calculate the remaining life probability distribution after replacement of the replacement distribution facility (10) based on the life probability distribution. For example, the distribution facility management device (20) can calculate the remaining life probability distribution after replacement of the replacement distribution facility (10) based on the manufacturer's life probability distribution corresponding to the condition that the life of the distribution facility (10) is longer than the difference between the replacement date and the installation date (S300).

Specifically, the power distribution facility management device (20) can calculate the conditional probability distribution according to the following mathematical expression 2 as the probability distribution of the remaining life after replacement. Here, s corresponds to the replacement date.

[Mathematical formula 2]

Next, the power distribution facility management device (20) can estimate the remaining life of the replacement power distribution facility (10) based on the probability distribution of the remaining life after replacement (S400).

Referring to FIG. 3, the distribution facility management device (20) can calculate the remaining life probability distribution after replacement from the life probability distribution, and estimate the remaining life of the replacement distribution facility (10) based on the calculated remaining life probability distribution after replacement. Since the life probability distribution is estimated by manufacturer, the distribution facility management device (20) can be based on the life probability distribution of the distribution facility (10) manufactured by the same manufacturer as the replacement distribution facility (10) whose remaining life is to be estimated.

For example, the distribution facility management device (20) can estimate the remaining life of the replacement distribution facility (10) based on the probability distribution of the remaining life after replacement through rejection sampling. The distribution facility management device (20) can use a normal distribution as a suggested distribution used in rejection sampling.

For example, the power distribution facility management device (20) can estimate the remaining life based on the probability distribution of the remaining life after replacement through random sampling for the replacement power distribution facility (10). However, the present invention is not limited thereto, and the remaining life can be estimated based on the probability distribution of the remaining life after replacement through various types of sampling methods.

Next, the distribution facility management device (20) can generate failure estimation data based on the replacement history data and the estimated remaining life of the replacement distribution facility (10) (S500).

Specifically, the distribution facility management device (20) can estimate the failure date of the replacement distribution facility (10) by adding the estimated remaining life of the replacement distribution facility (10) to the replacement date included in the replacement history data (S510). For example, the distribution facility management device (20) can estimate the failure date of the replacement distribution facility (10) by adding the estimated remaining life and the replacement date by manufacturers with the same or similar durability.

Thereafter, the power distribution facility management device (20) generates failure estimation data including the manufacturer name of the replacement power distribution facility (10), the installation date, the estimated failure date, and measurement values for multiple characteristics measured between the installation date and the replacement date. Since the failure estimation data includes the estimated failure date of the replacement power distribution facility (10), it can correspond to uncensored data (S520).

Next, the distribution facility management device (20) estimates the remaining life of the currently operating active distribution facility (10) among the plurality of distribution facilities (10; 1, 2, ..., n) based on the failure estimation data (S600).

Specifically, the power distribution facility management device (20) performs machine learning using learning data that uses the manufacturer name and measurement values included in the failure estimation data as input data and the difference between the estimated failure date and the installation date as output data to create a life estimation model of the active power distribution facility (10) (S610).

Thereafter, the distribution facility management device (20) inputs the manufacturer name and measurement values of the active distribution facility (10) into the life estimation model to estimate the expected life (S620). The distribution facility management device (20) can estimate the remaining life of the active distribution facility (10) by subtracting the elapsed period of time since the installation of the active distribution facility (10) from the estimated expected life (S630). For example, the distribution facility management device (20) can generate a replacement message for the active distribution facility (10) if the remaining life of the active distribution facility (10) is less than a predetermined reference period.

Next, referring to Fig. 7, the distribution facility management device (20) can generate a health index for evaluating the health of a plurality of distribution facilities (10; 1, 2, ..., n) (S700).

Referring to Fig. 8, the power distribution facility management device (20) selects a feature that has a significant effect on the remaining life from among a plurality of features as a valid feature. For example, the power distribution facility management device (20) may determine the weights of each of the plurality of features using a Lasso Regression algorithm, and select a feature having a weight greater than a threshold value from among the plurality of features as a valid feature (S710).

Thereafter, the distribution facility asset management device (20) can assign a reference score to each of the plurality of valid characteristics based on the degree to which the valid characteristic affects the remaining life (S720). For example, the distribution facility asset management device (20) can determine the reference scores of the plurality of valid characteristics so that the reference score of the valid characteristic has a value proportional to the weight of the corresponding valid characteristic, and the sum of the reference scores of each of the plurality of valid characteristics becomes 100.

Thereafter, the power distribution facility asset management device (20) can set multiple sections to which the measurement value can belong for each valid characteristic (S730). For example, the power distribution facility asset management device (20) can set multiple sections to which the measurement value can belong for each of the multiple valid characteristics by using a widely known change point detection algorithm.

Thereafter, the distribution facility asset management device (20) can assign a score lower than the reference score of the corresponding valid characteristic to each of the plurality of sections (S740). The score assigned to each of the plurality of sections can be used as a health index for evaluating the health of the plurality of distribution facilities (10; 1, 2, ..., n).

For example, among a plurality of distribution facilities (10; 1, 2, ..., n), for an active distribution facility (10) currently in operation, a section to which a plurality of measured values for a plurality of valid characteristics belong can be determined, and a value obtained by adding up the scores corresponding to the determined section can be determined as the health index of the active distribution facility (10).

Specifically, when the health index of the active distribution equipment (10) is relatively low, it can be determined that the remaining life of the active distribution equipment (10) is relatively long, and when the health index of the active distribution equipment (10) is relatively high, it can be determined that the remaining life of the active distribution equipment (10) is relatively short.

Although the embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements made by those skilled in the art in the art to which the present invention pertains also fall within the scope of the present invention.

1: Distribution facility management system
10: Distribution equipment
20: Distribution facility management device

Claims (20)

A method for managing multiple distribution facilities installed in a power supply system and involved in power supply, performed in a distribution facility management system operated by a power supply agency that supplies power to power demanders in a wide area.
A step of collecting failure history data regarding a faulty distribution facility that has stopped operating among the plurality of distribution facilities and replacement history data regarding a replacement distribution facility that has been replaced before the operation has stopped among the plurality of distribution facilities;
A step of estimating the life probability distribution by manufacturer based on the above failure history data;
A step of calculating a probability distribution of the remaining life after replacement of the replacement distribution equipment based on the probability distribution of the life;
A step of estimating the remaining life of the replacement distribution equipment based on the probability distribution of the remaining life after replacement;
A step of generating failure estimation data based on the above replacement history data and the estimated remaining life of the replacement distribution equipment; and
A step of estimating the remaining life of an active distribution facility in operation among the plurality of distribution facilities based on the above fault estimation data is included.
The above failure history data is,
Include the manufacturer's name, installation date, and failure date.
The above replacement history data is,
Includes manufacturer name, installation date, replacement date, and measurements for multiple characteristics measured between the installation date and replacement date.
The above plural characteristics are,
Including at least one of the average load and operating temperature of the above replacement distribution equipment,
The step of estimating the life probability distribution by manufacturer based on the above failure history data is as follows.
The life of the above faulty distribution equipment is calculated by the difference between the installation date and the failure date included in the failure history data, and the life probability distribution is approximated by the Weibull distribution based on the calculated life.
The step of calculating the probability distribution of the remaining life after replacement of the above replacement distribution equipment is as follows.
The manufacturer's life probability distribution corresponding to the condition that the life of the above distribution equipment is longer than the difference between the replacement date and the installation date is the probability distribution of the remaining life after replacement of the above replacement distribution equipment.
The step of estimating the remaining life of the above replacement distribution equipment is:
A method for managing power distribution equipment, characterized in that the remaining life is estimated based on the probability distribution of the remaining life after replacement through rejection sampling for the above replacement power distribution equipment.
delete delete delete delete In the first paragraph,
The step of estimating the remaining life of the above replacement distribution equipment is:
A method for managing power distribution equipment, characterized in that the remaining life is estimated based on a probability distribution of the remaining life after replacement through random sampling for the above replacement power distribution equipment.
delete In the first paragraph,
The step of estimating the remaining life of the above replacement distribution equipment is:
A method for managing power distribution facilities, characterized in that a normal distribution is used as a proposed distribution used in rejection sampling. In the first paragraph,
The step of generating the above fault estimation data is:
A step for estimating the failure date of the replacement distribution equipment by adding the estimated remaining life of the replacement distribution equipment to the replacement date included in the replacement history data; and
A method for managing power distribution equipment, comprising the step of generating failure estimation data including the manufacturer name of the replacement power distribution equipment, the installation date, the estimated failure date, and measurement values for multiple characteristics measured between the installation date and the replacement date. In Article 9,
The step of estimating the remaining life of the above active distribution equipment is:
A step of creating a life estimation model of the active distribution equipment by performing machine learning using learning data in which the manufacturer name and measurement values included in the above failure estimation data are input data and the difference between the estimated failure date and the installation date is output data;
A step for estimating the expected life by inputting the manufacturer name and measurement values of the above active distribution equipment into the life estimation model; and
A distribution facility management method for estimating the remaining life of the active distribution facility by deducting the period of time since the installation of the active distribution facility from the estimated expected lifespan. ◈Claim 11 was abandoned upon payment of the registration fee.◈ A method for managing multiple distribution facilities installed in a power supply system and involved in power supply, performed in a distribution facility management system operated by a power supply agency that supplies power to power demanders in a wide area.
A step of collecting failure history data regarding a faulty distribution facility that has stopped operating among the plurality of distribution facilities and replacement history data regarding a replacement distribution facility that has been replaced before the operation has stopped among the plurality of distribution facilities;
A step of estimating the life probability distribution by manufacturer based on the above failure history data;
A step of calculating a probability distribution of the remaining life after replacement of the replacement distribution equipment based on the probability distribution of the life;
A step of estimating the remaining life of the replacement distribution equipment based on the probability distribution of the remaining life after replacement;
A step of generating failure estimation data based on the above replacement history data and the estimated remaining life of the replacement distribution equipment;
A step of estimating the remaining life of an active distribution facility in operation among the plurality of distribution facilities based on the above fault estimation data; and
Including a step of generating a health index for evaluating the health of multiple distribution facilities,
The above failure history data is,
Include the manufacturer's name, installation date, and failure date.
The above replacement history data is,
Includes manufacturer name, installation date, replacement date, and measurements for multiple characteristics measured between the installation date and replacement date.
The above plural characteristics are,
Including at least one of the average load and operating temperature of the above replacement distribution equipment,
The step of estimating the life probability distribution by manufacturer based on the above failure history data is as follows.
The life of the above faulty distribution equipment is calculated by the difference between the installation date and the failure date included in the failure history data, and the life probability distribution is approximated by the Weibull distribution based on the calculated life.
The step of calculating the probability distribution of the remaining life after replacement of the above replacement distribution equipment is as follows.
The manufacturer's life probability distribution corresponding to the condition that the life of the above distribution equipment is longer than the difference between the replacement date and the installation date is the probability distribution of the remaining life after replacement of the above replacement distribution equipment.
The step of estimating the remaining life of the above replacement distribution equipment is:
A method for managing power distribution equipment, characterized in that the remaining life is estimated based on the probability distribution of the remaining life after replacement through rejection sampling for the above replacement power distribution equipment. ◈Claim 12 was abandoned upon payment of the registration fee.◈ In Article 11,
The steps for generating the above health index are:
A method for managing power distribution facilities, comprising the step of determining weights of each of a plurality of features using a Lasso sparseness analysis algorithm and selecting a feature having a weight greater than a threshold value among the plurality of features as an effective feature. ◈Claim 13 was waived upon payment of the registration fee.◈ In Article 12,
The steps for generating the above health index are:
A method for managing power distribution equipment, further comprising the step of assigning a reference score to each of a plurality of valid characteristics based on the degree to which the valid characteristics affect the remaining life. ◈Claim 14 was abandoned upon payment of the registration fee.◈ In Article 13,
The steps for generating the above health index are:
A method for managing power distribution facilities further comprising the step of setting a plurality of intervals to which measurement values can belong for each of the plurality of valid characteristics. ◈Claim 15 was abandoned upon payment of the registration fee.◈ In Article 14,
The steps for generating the above health index are:
A method for managing power distribution facilities, further comprising the step of assigning a score used as a health index for evaluating the health of the plurality of power distribution facilities to each corresponding valid characteristic of the plurality of sections. delete delete delete delete ◈Claim 20 was abandoned upon payment of the registration fee.◈ In Article 11,
The step of estimating the remaining life of the above replacement distribution equipment is:
A method for managing power distribution equipment, characterized in that the remaining life is estimated based on a probability distribution of the remaining life after replacement through random sampling for the above replacement power distribution equipment.

Images