KR102780431B1 - Apparatus for controlling distribution - Google Patents

Abstract

According to one embodiment of the present invention, a power distribution control device includes a communication circuit, a processor, and a memory, wherein the memory stores a generation-specific learning database in which, for each household in an apartment complex, load by time zone, information on household members indoors by time zone, weather information by time zone, and date information are associated, and the memory stores a power database in which the amount of power produced by time zone from a first power generation unit installed in the apartment complex and weather information by time zone are associated, wherein the first power generation unit includes a solar power generation unit and a wind power generation unit, and wherein the processor: generates a first artificial intelligence model based on the generation-specific learning data that outputs a predicted household member composition that is a household member composition expected to be indoors during a corresponding time zone when time zone, date, and weather forecast information are input; generates a second artificial intelligence model based on the generation-specific learning data that outputs a predicted load for a corresponding time zone when time zone, date, weather forecast information, and the predicted household member composition are input; generates a third artificial intelligence model based on the power database that outputs a predicted amount of power produced by the first power generation unit during a corresponding time zone when time zone weather forecast information is input; Based on the weather forecast information acquired through the above communication circuit, the predicted load for each generation is confirmed using the first artificial intelligence model and the second artificial intelligence model; based on the predicted load for each generation for each generation, the total predicted load for all generations is confirmed for each time zone; based on the third artificial intelligence model, the predicted power generation amount of the first power generation unit for each time zone is confirmed; based on the total predicted load, the predicted power generation amount of the first power generation unit, the generation unit price of the second power generation unit, and the rate of the external power supply source for each time zone, an operation schedule of a distribution unit that selectively supplies power from one of the first power generation unit, the second power generation unit, and the external power supply source to the apartment complex and a power generation schedule of the second power generation unit are calculated, wherein the second power generation unit includes a hydrogen fuel cell power generation unit; and the distribution unit is controlled according to the operation schedule.

Description

{APPARATUS FOR CONTROLLING DISTRIBUTION}

The present invention relates to a power distribution control device.

As abnormal weather phenomena such as summer heat waves and winter cold waves become more frequent and electronic products that satisfy various needs are developed, household power consumption is on the rise. In addition, as interest in sustainable development increases due to awareness of the climate crisis, the installation of solar power generators in apartment complexes is increasing, and the enforcement ordinance and enforcement regulations of the Building Act are being revised to allow the installation of wind power generators on the roofs of apartment complexes, and responding to the increasing demand for power using new and renewable energy has become a megatrend.

Meanwhile, in the electrical room of an apartment complex, a distribution panel is installed to receive external power supplied from the power company and distribute it to the load of each household. The distribution panel of an apartment complex is responsible for supplying electricity to multiple households, and since the voltage of the external power supplied from the power company is high, a safety accident occurring in the distribution panel can cause great damage, so a safety management device for stable distribution is necessary.

An object of the present invention is to provide a power distribution control device that enables energy to be supplied at minimal cost in an apartment complex that uses both renewable energy power generation facilities and external power sources.

Another object of the present invention is to provide a distribution safety management device for stable distribution, thereby preventing abnormalities occurring in a distribution panel from leading to a major accident at an early stage.

In addition, although the amount of electricity used in homes is on the rise, in the case of old apartment buildings, since facilities are often installed expecting a small amount of load, there are increasing cases of power outages occurring due to inability to respond appropriately when the amount of load increases rapidly. Another purpose of the present invention is to prevent power outages by providing managers with reports on the amount of load when necessary in the long-term trend of increasing load, thereby inducing managers to consider increasing the amount of power equipment.

According to one embodiment of the present invention, a power distribution control device includes a communication circuit, a processor, and a memory, wherein the memory stores a generation-specific learning database in which, for each household in an apartment complex, load by time zone, information on household members indoors by time zone, weather information by time zone, and date information are associated, and the memory stores a power database in which the amount of power produced by time zone from a first power generation unit installed in the apartment complex and weather information by time zone are associated, wherein the first power generation unit includes a solar power generation unit and a wind power generation unit, and wherein the processor: generates a first artificial intelligence model based on the generation-specific learning data that outputs a predicted household member composition that is a household member composition expected to be indoors during a corresponding time zone when time zone, date, and weather forecast information are input; generates a second artificial intelligence model based on the generation-specific learning data that outputs a predicted load for a corresponding time zone when time zone, date, weather forecast information, and the predicted household member composition are input; generates a third artificial intelligence model based on the power database that outputs a predicted amount of power produced by the first power generation unit during a corresponding time zone when time zone weather forecast information is input; Based on the weather forecast information acquired through the above communication circuit, the predicted load for each generation is confirmed using the first artificial intelligence model and the second artificial intelligence model; based on the predicted load for each generation for each generation, the total predicted load for all generations is confirmed for each time zone; based on the third artificial intelligence model, the predicted power generation amount of the first power generation unit for each time zone is confirmed; based on the total predicted load, the predicted power generation amount of the first power generation unit, the generation unit price of the second power generation unit, and the rate of the external power supply source for each time zone, an operation schedule of a distribution unit that selectively supplies power from one of the first power generation unit, the second power generation unit, and the external power supply source to the apartment complex and a power generation schedule of the second power generation unit are calculated, wherein the second power generation unit includes a hydrogen fuel cell power generation unit; and the distribution unit is controlled according to the operation schedule.

A power distribution control device according to one embodiment of the present invention predicts the load per generation and the power produced from the first power generation unit based on artificial intelligence, and controls the power distribution unit to use the power source that minimizes the cost by considering the power generation unit price of the second power generation unit and the power rate of an external power supply source, thereby enabling the use of renewable energy in an economical and efficient manner.

In addition, in a distribution control device according to an embodiment of the present invention, when there is a change in the number of household members in a generation, the load of the generation compared to the loads of other generations having the number of household members before the change and the load of the generation compared to the loads of other generations having the number of household members after the change are both considered to determine the predicted load of the generation, thereby making it possible to predict the load of the generation even during a period when the accuracy of artificial intelligence is low due to an insufficient amount of data.

In addition, the distribution control device according to one embodiment of the present invention determines the predicted load of a new generation by considering the load of the new generation compared to the load of other generations having the same number of members as the generation that has moved in, thereby making it possible to predict the load of the new generation even during a period when the accuracy of artificial intelligence is low due to insufficient data.

A power distribution safety management device according to one embodiment of the present invention monitors whether there is an abnormality inside a distribution panel by comparing an image acquired using a camera with an image acquired in a state where no abnormality has occurred, and determines a monitoring cycle so that monitoring occurs more frequently as the predicted load is higher, thereby enabling more monitoring during a period when the risk of an abnormality occurring is greater and damage in the event of an accident is greater, thereby effectively managing the safety of a power distribution system.

In addition, a power distribution safety management device according to an embodiment of the present invention periodically performs clustering for load, and when the center point of a cluster corresponding to the maximum load exceeds a specific value, generates a report showing a histogram of data included in the cluster having the largest center value over time and a change trend of the largest value among the cluster center values, and provides a report showing a histogram of data included in the cluster having the largest center value over time and a change trend of the largest value among the cluster center values to a manager device, thereby inducing the manager to consider installing more facilities, and preventing a power outage from occurring in an old apartment complex due to insufficient power facility capacity.

In addition, the power distribution safety management device according to one embodiment of the present invention can induce the manager of an apartment complex to react more quickly to load change trends by performing clustering by giving greater weight to recent data.

FIG. 1 illustrates a distribution control system including a distribution control device according to various embodiments of the present invention.
Figure 2 illustrates the structure of the learning database by generation.
FIG. 3 illustrates a method performed in a distribution control device according to various embodiments of the present invention.
FIG. 4 illustrates a first artificial intelligence model and a second artificial intelligence model generated in a distribution control device according to various embodiments of the present invention.
FIG. 5 is a block diagram of a distribution safety management device according to various embodiments of the present invention.
FIG. 6 illustrates an exemplary structure of a guide included in a power distribution safety management device according to various embodiments of the present invention.
FIG. 7 illustrates a method performed in a power distribution safety management device according to various embodiments of the present invention.

As used herein, the terms "comprises" and/or "comprising" do not exclude the presence or addition of one or more other components in addition to the components mentioned. It should be understood that the various embodiments of the present disclosure and the terms used therein are not intended to limit the technical features described in the present disclosure to specific embodiments, but include various modifications, equivalents, or substitutes of the embodiments. In connection with the description of the drawings, similar reference numerals may be used for similar or related components. The singular form of a noun corresponding to an item may include one or more of the items, unless the context clearly indicates otherwise. In the present disclosure, each of the phrases "A or B", "at least one of A and B", "at least one of A or B", "A, B, or C", "at least one of A, B, and C", and "at least one of A, B, or C" can include any one of the items listed together in the corresponding phrase, or all possible combinations thereof. Terms such as "first", "second", or "first" or "second" may be used merely to distinguish one component from another, and do not limit the components in any other respect (e.g., importance or order). When a component (e.g., a first component) is said to be "functionally" or "communicatively" coupled to another component (e.g., a second component), with or without the terms "functionally" or "communicatively", it means that the component can be coupled to the other component directly (e.g., wired), wirelessly, or through a third component.

Various embodiments of the present disclosure may be implemented as software including one or more instructions stored in a machine-readable storage medium (e.g., internal memory or external memory). For example, a processor of the device may call at least one instruction among the one or more instructions stored from the storage medium and execute it. This enables the device to operate to perform at least one function according to the at least one instruction called. The one or more instructions may include code generated by a compiler or code executable by an interpreter. The machine-readable storage medium may be provided in the form of a non-transitory storage medium. Here, 'non-transitory' only means that the storage medium is a tangible device and does not include a signal (e.g., electromagnetic waves), and this term does not distinguish between cases where data is stored semi-permanently and cases where it is stored temporarily in the storage medium.

According to one embodiment, the method according to various embodiments disclosed in the present specification may be provided as included in a computer program product. The computer program product may be traded between a seller and a buyer as a commodity. The computer program product may be distributed in the form of a machine-readable storage medium (e.g., a compact disc read only memory (CD-ROM)), or may be distributed online (e.g., downloaded or uploaded) via an application store (e.g., Play Store ^TM ) or directly between two user terminals (e.g., smart phones). In the case of online distribution, at least a part of the computer program product may be at least temporarily stored or temporarily generated in a machine-readable storage medium, such as a memory of a manufacturer's server, a server of an application store, or a relay server.

According to various embodiments, each component (e.g., a module or a program) of the above-described components may include a single or multiple entities, and some of the multiple entities may be separately arranged in other components. According to various embodiments, one or more components or steps of the above-described corresponding components may be omitted, or one or more other components or steps may be added. Alternatively or additionally, the multiple components (e.g., a module or a program) may be integrated into one component. In such a case, the integrated component may perform one or more functions of each of the multiple components identically or similarly to those performed by the corresponding component of the multiple components before the integration. According to various embodiments, the steps performed by the module, program or other component may be executed sequentially, in parallel, repeatedly, or heuristically, or one or more of the steps may be executed in a different order, omitted, or one or more other steps may be added.

FIG. 1 illustrates a distribution control system including a distribution control device according to various embodiments of the present invention. Referring to FIG. 1, the distribution control system may include a distribution control device (110), a distribution unit (130), a distribution safety management device (120), and a manager device (140).

The distribution control device (110) may include a communication circuit (111), a processor (112), and a memory (115). The communication circuit (111) may transmit information to or receive information from another electronic device, and the type of communication supported by the communication circuit (111) is not limited.

The processor (112) performs a calculation based on data received through the communication circuit (111) and/or data stored in the memory (115), and may transmit at least a part of the result of the calculation to another electronic device through the communication circuit (111) or store it in the memory (115). In this specification, the expression that the power distribution control device (110) performs a certain operation includes the meaning that the processor (112) performs a calculation by referring to the memory (115), or the processor (112) controls another component (e.g., the communication circuit (111)) of the power distribution control device (110) so that the corresponding operation is performed.

The processor (112) may include a data learning unit (113) and a data recognition unit (114). The data learning unit (113) may generate artificial intelligence models for predicting the load of a specific generation by time zone, as described below. The data recognition unit (114) may preprocess data and provide the preprocessed data to the data learning unit (113) for learning.

At least one of the data learning unit (113) and the data recognition unit (114) may be implemented in the form of a dedicated hardware chip for artificial intelligence, or may be implemented as part of an existing general-purpose processor (e.g., AP or CPU) or a graphics-only processor.

According to various embodiments, unlike the data learning unit (113) and the data recognition unit (114) shown in FIG. 1 as being included in the distribution control device (110), the data learning unit (113) and the data recognition unit (114) may be mounted on separate electronic devices, respectively.

In this case, the data learning unit (113) and the data recognition unit (114) are connected to each other by wire or wirelessly, so that model information generated in the data learning unit (113) can be provided to the data recognition unit (114), or data input to the data recognition unit (114) can be provided to the data learning unit (113) as additional learning data.

At least one of the data learning unit (113) and the data recognition unit (114) may be implemented as a software module. In this case, the software module may be stored in a non-transitory computer-readable recording medium. At least a part of the software module may be provided by an operating system (OS) or by a predetermined application.

The memory (115) can store a generation-specific learning database (116) and a power database (117). The structure of the generation-specific learning database (116) is illustrated in FIG. 2. Referring to FIG. 2, the generation-specific learning database (116) includes learning databases managed separately for each generation, and in the learning database of a specific generation, the load of the corresponding generation, information on household members living indoors, weather information, and date information are stored in association by time zone. Information on household members living indoors refers to household members who are registered as household members in the management system of the apartment complex and are recorded as being indoors. In order to identify which household members are indoors, the apartment complex can be equipped with an access control system for access control, and an authentication means (for example, a biometric means such as an ID card or a fingerprint) having a separate ID for each household member can be used. For example, if four IDs, ID1, ID2, ID3, and ID4, are issued to each household in unit 101, and if the household members corresponding to ID1, ID2, and ID3 enter the apartment complex through the access record and the household member corresponding to ID4 exits the apartment complex, but the household member corresponding to ID4 does not enter the apartment complex, the access control system of the apartment complex can determine that the household members corresponding to ID1, ID2, and ID3 among the household members of unit 101 are indoors and the household member corresponding to ID4 is outdoors. The distribution control device (110) can periodically obtain information on household members inside each household from the access control system of the apartment complex through the communication circuit (111).

As shown in Figure 2, weather information may include temperature, wind speed, and precipitation.

In the power database (117), the amount of power produced by the first power generation unit (131) by time zone and the weather information by time zone can be stored in association with each other.

The distribution unit (130) selectively supplies power from one of the first power generation unit (131), the second power generation unit (132), and the external power supply source (133) to the apartment complex. The operation schedule of the distribution unit (130) means information on at which time the distribution unit (130) supplies power from each of the first power generation unit (131), the second power generation unit (132), and the external power supply source (133) to the apartment complex. The operation schedule of the distribution unit (130) is calculated by the distribution control device (110), and the distribution control device (110) can control the distribution unit (130) so that the distribution unit (130) selectively supplies power from one of the first power generation unit (131), the second power generation unit (132), and the external power supply source (133) to the apartment complex according to the operation schedule.

The first power generation unit (131) includes a solar power generation unit and a wind power generation unit. The second power generation unit (132) includes a hydrogen fuel cell power generation unit. The difference between the first power generation unit (131) and the second power generation unit (132) lies in whether a fuel having a constant unit price, such as a hydrogen tank, is required for power generation. The power distribution control device (110) can receive the power generation unit price of the second power generation unit (132) from the management device (140) through the communication circuit (111) and store it in the memory (115). The power distribution control device (110) can receive rate information of an external power supply source (133) through the communication circuit (111) and store it in the memory (115). In general, the rate of an external power supply source (133) varies depending on the amount of power and also varies depending on the time zone.

The distribution control device (110) is configured to monitor the inside of the distribution panel, obtain the total predicted load for each time period for all households in the apartment complex from the distribution safety management device (120), and adjust the monitoring cycle inside the distribution panel based on this. If an abnormal situation occurs, the distribution control device (110) transmits an abnormality occurrence message to the management device (140).

FIG. 3 illustrates a method performed in a distribution control device according to various embodiments of the present invention.

At step 310, the processor (112) may generate a first artificial intelligence model that outputs a predicted generation member configuration, a second artificial intelligence model that outputs a predicted load, and a third artificial intelligence model that outputs a predicted power production amount of the first power generation unit. The first artificial intelligence model that outputs a predicted generation member configuration, the second artificial intelligence model that outputs a predicted load, and the third artificial intelligence model that outputs a predicted power production amount of the first power generation unit may be generated.

The first artificial intelligence model and the second artificial intelligence model will be described with reference to FIG. 4. As illustrated in FIG. 4, the processor (112) can generate the first artificial intelligence model and the second artificial intelligence model by learning the learning database of a specific generation included in the generation-specific learning database (116). When time zone, date, and weather forecast information are input, the first artificial intelligence model outputs a predicted household composition, which is a household composition expected to be indoors during the corresponding time zone. When time zone, date, weather forecast information, and predicted household composition are input, the second artificial intelligence model outputs a predicted load of the corresponding generation for the corresponding time zone. Since the learning database, the first artificial intelligence model, and the second artificial intelligence model are managed and generated by generation, a separate learning database is stored for each generation in the generation-specific learning database (116), and the processor (112) generates a separate first artificial intelligence model and a second artificial intelligence model for each generation.

The processor (112) can generate a third artificial intelligence model based on the power database (117). When time-based weather forecast information is input, the third artificial intelligence model outputs the expected power generation amount of the first power generation unit (131) for each time period. Since the first power generation unit (131) includes all solar power generation facilities and wind power generation facilities installed throughout the apartment complex, the processor (112) generates one third artificial intelligence model for the entire apartment complex.

Various machine learning techniques can be used to generate the first artificial intelligence model, the second artificial intelligence model, and the third artificial intelligence model. For example, at least one of a Recurrent Neural Network (RNN), a Convolution Neural Network (CNN), an Artificial Neural Network (ANN), and a Transformer model can be used for learning to generate the first artificial intelligence model, the second artificial intelligence model, and the third artificial intelligence model.

In step 320, the processor (112) can check the predicted load by generation using the first artificial intelligence model and the second artificial intelligence model. Specifically, the processor (112) obtains weather forecast information by time zone through the communication circuit (111), inputs the obtained weather forecast information, time zone, and date into the first artificial intelligence model to check the predicted household composition, and inputs the weather forecast information, time zone, date, and predicted household composition into the second artificial intelligence model to check the predicted load by generation for the corresponding generation. The processor (112) can check the predicted load by generation for each generation by repeating this process for all generations. The predicted load by generation is time-series data in which the predicted load is associated by time zone.

At step 330, the processor (112) can determine the total predicted load for all generations as the sum of the predicted loads per generation of all generations included in the apartment complex. Since the predicted load per generation is time-series data in which the predicted loads are related by time zone, the total predicted load is also time-series data in which the predicted loads are related by time zone.

At step 340, the processor (112) can check the expected power generation amount of the first power generation unit for each time zone by inputting time zone weather forecast information into the third artificial intelligence model.

At step 350, the processor (112) can calculate the operation schedule of the distribution unit (130) and the power generation schedule of the second power generation unit (132) based on the total predicted load, the expected power generation amount of the first power generation unit (131), the power generation unit price of the second power generation unit (132), and the rate of the external power supply source (133) by time zone. Power generation through the first power generation unit (131) does not incur any cost, but the amount of power generated through the first power generation unit (131) cannot be controlled, and in order to generate power through the second power generation unit (132), hydrogen must be supplied to the fuel cell from the hydrogen tank, so a constant power generation unit price occurs, but power generation can be controlled by regulating the hydrogen supply.

The processor (112) can calculate an operation schedule that enables power supply corresponding to the total predicted load while minimizing the sum of the power generation unit price of the second power generation unit (132) and the rate of the external power supply source (133). The processor (112) can obtain the SoC (state of charge) of the battery included in the first power generation unit (131) and the battery included in the second power generation unit (132) through the communication circuit (111) and consider this in calculating the operation schedule and power generation schedule.

At step 360, the processor (112) can control the distribution unit (130) according to the operation schedule of the distribution unit (130) calculated at step 350, and control the power production of the second power generation unit (132) according to the power generation schedule of the second power generation unit (132) calculated at step 350.

In some cases, there may be a generation in which the number of data included in the learning database is not sufficient to effectively predict the load using the first AI model and the second AI model at step 320. For example, this may be the case when the number of household members in a specific generation has changed or when a specific generation has moved in. In such cases, until a sufficient number of data is secured, rather than generating the first AI model and the second AI model with low accuracy using a small number of data, the predicted loads of other generations with the same conditions (generational balance and number of members) and sufficient data in the learning database may be used.

<In case of change in number of household members>

시간대에 제1 세대의 세대원 수가

명에서

명으로 변경되는 경우를 고려할 수 있다. 이 때, 프로세서(112)는

시간대와

+k 시간대 사이의 t 시간대에 대한 제1 세대의 예측 부하량 E(t)를 수학식 1과 같이 결정할 수 있다.

The number of members of the first generation in the time zone

In the name

It can be considered that the name is changed. At this time, the processor (112)

Time zone and

The predicted load E(t) of the first generation for t time periods between +k time periods can be determined as in mathematical expression 1.

는 상기 제1 세대와 넓이가 동일하고 세대원수가 c인 세대들 중 상기 제1 세대를 제외한 세대들의 t 시간대에 대한 예측 부하량의 평균이고,

는 상기 제1 세대와 넓이가 동일하고 세대원수가 c인 세대들 중 상기 제1 세대를 제외한 세대들의 t 시간대에서의 실제 부하량의 평균이고, R(t)는 제1 세대의 t 시간대에서의 실제 부하량이고, k는 미리 결정된 크기를 갖는 시간 간격으로서, 신뢰도 있는 제1 인공지능 모델 및 제2 인공지능 모델이 생성되기 위하여 데이터가 충분히 축적될 수 있는 시간 간격(예를 들어, 3개월)으로 결정될 수 있다. j는 미리 결정된 크기를 갖는 시간 간격(예를 들어, 1개월)이다.In mathematical expression 1,

is the average of the predicted load for time period t of the generations excluding the first generation among the generations with the same area as the first generation and with c as the number of generation members,

is the average of the actual loads of the generations excluding the first generation among the generations having the same area as the first generation and having c generation members, at time t, R(t) is the actual load of the first generation at time t, k is a time interval having a predetermined size, which can be determined as a time interval (e.g., 3 months) at which data can be sufficiently accumulated to generate reliable first artificial intelligence models and second artificial intelligence models. j is a time interval having a predetermined size (e.g., 1 month).

를 곱한 값이고, 해당 가중치는 거칠게 말하자면 다른 세대에 비해 전력을 많이 사용하는 정도를 의미한다. Looking at mathematical expression 1, E(t) is the weight of the average load of other generations with the number of generations after the change.

It is a value multiplied by , and the corresponding weight roughly means the degree to which power is used more than other generations.

는 세대원 수 변경 전에 동일 조건의 다른 세대들에 비해 부하가 높은 경향(h1)과 세대원 수 변경 후에 동일 조건의 다른 세대들에 비해 부하가 높은 경향(h2)의 가중평균이다. h1에 대한 가중치는 시간이 지날수록 감소하고, h2에 대한 가중치는 시간이 지날수록 높아지는데, 이는 시간이 지날수록 h2에 산입되는 데이터의 수가 많아져 h2의 값 자체의 신뢰도가 높아진다는 점을 반영한 것이다. weight

is a weighted average of the tendency to have a higher load than other generations under the same conditions before the change in the number of generation members (h1) and the tendency to have a higher load than other generations under the same conditions after the change in the number of generation members (h2). The weight for h1 decreases over time, and the weight for h2 increases over time, which reflects the fact that the number of data included in h2 increases over time, and thus the reliability of the value of h2 itself increases.

<In case of new transfer>

시간대에 세대원 수가

인 제1 세대가 신규 전입하는 경우를 생각할 수 있다. 이 경우, 프로세서(112)는

시간대와

+k 시간대 사이의 t 시간대에 대한 제1 세대의 예측 부하량 E(t)를 수학식 2와 같이 결정할 수 있다.

Number of household members in a time zone

We can consider a case where the first generation of people newly transfers. In this case, the processor (112)

Time zone and

The predicted load E(t) of the first generation for t time periods between +k time periods can be determined as in Equation 2.

를 곱한 값이다. 가중치

는 동일 조건을 갖는 다른 세대에 비해 전력을 많이 사용하는 정도를 반영한다.In mathematical expression 2, E(t) is the weighted average load of other generations with the same number of members and area as the generation that moved in.

is the value multiplied by the weight.

It reflects the degree to which a generation uses more electricity compared to other generations with the same conditions.

FIG. 5 is a block diagram of a distribution safety management device according to various embodiments of the present invention. Referring to FIG. 5, a distribution safety management device (120) may include a communication circuit (510), a processor (520), a memory (530), a guide (540), a camera driving unit (550), a camera (560), and a blocking driving unit (570). The communication circuit (510) may transmit information to or receive information from another electronic device, and the type of communication supported by the communication circuit (510) is not limited.

The processor (520) performs a calculation based on data received through the communication circuit (510) and/or data stored in the memory (530), and may transmit at least a part of the result of the calculation to another electronic device through the communication circuit (510) or store it in the memory (530). In this specification, the expression that the power distribution safety management device (120) performs a certain operation includes the meaning that the processor (520) performs a calculation by referring to the memory (530), or the processor (520) controls another component (e.g., the communication circuit (510)) of the power distribution safety management device (120) so that the corresponding operation is performed.

The processor (520) may include a data learning unit (521) and a data recognition unit (522). The data learning unit (521) may generate artificial intelligence models for predicting the load of a specific generation by time zone, as described below. The data recognition unit (522) may preprocess data and provide the preprocessed data to the data learning unit (521) for learning.

At least one of the data learning unit (521) and the data recognition unit (522) may be implemented in the form of a dedicated hardware chip for artificial intelligence, or may be implemented as part of an existing general-purpose processor (e.g., AP or CPU) or a graphics-only processor.

According to various embodiments, unlike the data learning unit (521) and the data recognition unit (522) shown in FIG. 1 as being included in the power distribution safety management device (120), the data learning unit (521) and the data recognition unit (522) may be mounted on separate electronic devices, respectively.

In this case, the data learning unit (521) and the data recognition unit (522) are connected to each other by wire or wirelessly, so that model information generated in the data learning unit (521) can be provided to the data recognition unit (522), or data input to the data recognition unit (522) can be provided to the data learning unit (521) as additional learning data.

At least one of the data learning unit (521) and the data recognition unit (522) may be implemented as a software module. In this case, the software module may be stored in a non-transitory computer-readable recording medium. At least a part of the software module may be provided by an operating system (OS) or by a predetermined application.

The memory (530) may include an image database (531), a load database (532), and a cluster information database (533). The image database (531) stores, in association, a first image captured when a camera (560) is positioned at an intersection on a guide (540) when no abnormality has occurred in the distribution panel, and coordinates of the intersection on the guide (540) where the camera (560) is positioned when each first image is captured.

The load database (532) may store the total load of the entire apartment complex and the time (date and time) at which the load appeared. The load included in the load database (532) refers to the historical load that actually occurred, not the predicted load.

In the cluster information database (533), the cluster information and the time of cluster information creation are stored when the cluster information described later is first created, and the update time and updated cluster information are cumulatively stored whenever the cluster information is updated thereafter.

Cluster information includes the number of clusters, the center of each cluster, the boundary of each cluster, and the monitoring frequency corresponding to each cluster. The boundary of a cluster means the boundary value of the interval for determining the total predicted load as corresponding to the cluster. For example, if there are three clusters, and the maximum value of the load data included in the cluster (C1) corresponding to the lowest load is 100 kWh, the minimum value of the load data included in the second cluster (C2) is 102 kWh and the maximum value is 200 kWh, and the minimum value of the load data included in the cluster (C3) corresponding to the highest load is 205 kWh, the upper boundary value of cluster C1 and the lower boundary value of cluster C2 can be determined as 101 kWh, which is the midpoint between 100 kWh and 102 kWh, and the upper boundary value of cluster C2 and the lower boundary value of cluster C3 can be determined as 202.5 kWh, which is the midpoint between 200 kWh and 205 kWh. Generalizing this, the boundary value between two adjacent clusters can be determined as the median of the maximum value of the load data included in the cluster corresponding to the smaller load among the two adjacent clusters and the minimum value of the load data included in the cluster corresponding to the larger load among the two adjacent clusters. In the example described above, if the total predicted load is 101 kWh or less, it can be determined to correspond to cluster C1, if the total predicted load is more than 101 kWh and less than or equal to 202.5 kWh, it can be determined to correspond to cluster C2, and if the total predicted load is more than 202.5 kWh, it can be determined to correspond to cluster C3.

The guide (540) may have a grid shape. The camera driving unit (550) is a means for providing power that is controlled by the processor (520) and can move the camera (560) along the guide (540). An exemplary structure of the guide (540) and the camera (560) is illustrated in FIG. 6. The processor (520) can control the camera driving unit (550) to temporarily stop the movement of the camera (560) when the camera (560) is located on an intersection point on the guide (540). The camera (560) can capture an image of the inside of the switchboard while stopped on the intersection point on the guide (540).

The blocking drive unit (570) is controlled by the processor (520) and is a means for providing power to disconnect the connection between the incoming wire and the wire inside the switchboard. The specific structure and the principle of the blocking operation of the blocking drive unit (570) are not limited as long as it is a means capable of disconnecting the connection between the incoming wire and the wire inside the switchboard by changing the physical structure of the connection portion connecting the incoming wire and the wire inside the switchboard.

The distribution safety management device (120) can be installed on the inside of the distribution panel door and can have an overall plate shape.

FIG. 7 illustrates a method performed in a power distribution safety management device according to various embodiments of the present invention.

In step 710, the processor (520) can obtain the total predicted load by time zone from the distribution control device (110) through the communication circuit (510). Here, the predicted load is the predicted load by time zone for all households in the apartment complex, and is a value calculated in step 330 of FIG. 3.

At step 720, the processor (520) can determine a monitoring schedule for the camera driver (550) based on the total predicted load per time zone. There is no limitation on the rule for determining the monitoring frequency from the total predicted load per time zone in the processor (520) as long as the monitoring frequency monotonically increases with respect to the total predicted load per time zone. For example, if the monitoring frequency corresponding to the total predicted load of a time zone t1 on a specific date is 5 times per hour and the monitoring frequency corresponding to the total predicted load of a time zone t2 is 10 times per hour, the processor (520) can determine a monitoring schedule such that monitoring occurs at a frequency of 5 times per hour in the time zone t1 and at a frequency of 10 times per hour in the time zone t2. Here, one round of monitoring means that the camera (560) captures an image once at every intersection (A1, A2, A3, A4, B1, B2, B3, B4, C1, C2, C3, C4, D1, D2, D3, and D4 in FIG. 6) on the guide (540).

In one embodiment, the processor (520) may perform clustering on the load database (532) to determine a rule for determining a monitoring schedule from the total predicted load by time zone, thereby determining load sections, and may set a monitoring frequency corresponding to each section. The processor (520) may first perform mean-shift clustering on the data included in the load database (532) to calculate the number of sections (n), and may perform k-means clustering with n as the number of clusters on the data included in the load database (532) to classify the data included in the load database (532) into n clusters. Thereafter, the processor (520) may calculate a center value and a boundary value for each cluster. The center value of the cluster may be obtained as a result of k-means clustering. The meaning and determination method of the boundary value of the cluster have been described above. Thereafter, the processor (520) may determine a monitoring frequency corresponding to each cluster. The monitoring frequency may be a value obtained by multiplying the center value of each cluster by a predetermined constant. For example, the monitoring frequency corresponding to a cluster with a center value of 50 kWh can be determined as 4 times per hour, and the monitoring frequency corresponding to a cluster with a center value of 100 kWh can be determined as 8 times per hour.

As time passes, the number of data stored in the load database (532) gradually increases, and according to an embodiment, cluster information updates through clustering targeting the entire data set including the additional data may occur periodically. That is, the processor (520) may periodically update the load database (532) and cluster information, and cumulatively store the updated cluster information and update time in the cluster information database (533).

Periodic cluster information updates can prompt the manager of the apartment complex to consider whether to expand the facility at an appropriate time. The processor (520) stores in memory a first value, which is the largest value among the cluster center values included in the initially stored cluster information, and a third value, which is the average value of the second values, which are predetermined based on the capacity of the power distribution facility, and if the largest value among the cluster center values included in the latest cluster information exceeds the third value, a report on the load can be transmitted to the manager device (140). The report can represent a histogram of data included in the cluster having the largest center value among the data included in the load database by occurrence time. In addition, the report can represent a change trend of the largest value among the cluster center values, for example, a change trend of the largest value among the cluster center values over the past five years.

에 대하여, T -

시점으로부터 T 시점 이전의 기간에 대한 데이터를

, T -2

시점으로부터 T -

시점 이전의 기간에 대한 데이터를

, T -3

시점으로부터 T -2

시점 이전의 기간에 대한 데이터를

, T -3

시점 이전의 기간에 대한 데이터를

라고 하면, 데이터

에 대한 가중치

는

로 결정되고,

일 수 있다. 프로세서(520)는 부하량 데이터베이스에 포함된 데이터에 가중치를 부여하고, 가중치가 부여된 데이터에 기초하여 클러스터 정보를 업데이트할 수 있다. 여기서

이 1년 이상의 값으로 정해지는 이유는 전력 사용량은 계절에 따른 변화가 크기 때문에 1년 미만의

을 설정하는 경우 계절에 따른 전력 사용량의 일시적 변화를 과도하게 반영할 수 있기 때문이다. 일 실시예에서,

은 1년의 정수배의 값을 가질 수 있다.In one embodiment, the processor (520) may assign greater weight to recent data among the data included in the load database when updating cluster information. In an environment where power usage changes over the long term according to trends in the use of electronic products, assigning greater weight to recent data has the advantage of better reflecting recent trends when deciding whether to expand facilities. For example, a time interval predefined as a weight assignment time point T and a value of one year or more

About, T -

Data for the period from point T to point T

, T -2

From point T -

Data for a period prior to a point in time

, T -3

T -2 from the point

Data for a period prior to a point in time

, T -3

Data for a period prior to a point in time

If you say so, data

Weight for

Is

is decided,

The processor (520) may assign weights to data included in the load database and update cluster information based on the weighted data. Here,

The reason why it is set to a value of more than one year is because electricity usage varies greatly depending on the season, so it is set to a value of less than one year.

This is because setting it may over-reflect temporary changes in power usage according to seasons. In one embodiment,

can have values that are multiples of a year.

Above, while the embodiments of the present invention have been described with reference to the attached drawings, it will be understood by those skilled in the art that the present invention may be implemented in other specific forms without changing the technical idea or essential features thereof. Therefore, it should be understood that the embodiments described above are exemplary in all respects and not restrictive.

Claims (5)

In a power distribution control device including a communication circuit, a processor, and a memory,
The above memory stores a generation-specific learning database that is associated with each household in an apartment complex, including load by time zone, household member information by time zone, weather information by time zone, and date information.
The above memory stores a power database related to the amount of power produced by each time zone from the first power generation unit installed in the apartment complex and the weather information by time zone - the first power generation unit includes a solar power generation unit and a wind power generation unit -,
The above processor:
Based on the above generational learning data, a first artificial intelligence model is created that outputs a predicted household composition, which is a household composition expected to be indoors during a given time zone, when time zone, date, and weather forecast information are input;
Based on the above generational learning data, a second artificial intelligence model is created that outputs a predicted load for a given time zone when time zone, date, weather forecast information, and the predicted generation composition are input;
Based on the above power database, when time-based weather forecast information is input, a third artificial intelligence model is generated that outputs the expected power generation amount of the first power generation unit for each time period;
Based on the weather forecast information obtained through the above communication circuit, the predicted load for each generation is confirmed using the first artificial intelligence model and the second artificial intelligence model;
Based on the above generational predicted load for each generation, the total predicted load for the entire generation is checked by time zone;
Based on the third artificial intelligence model, the expected power generation amount of the first power generation unit is confirmed for each time zone;
Based on the total predicted load, the predicted power generation amount of the first power generation unit, the generation unit price of the second power generation unit, and the rate of the external power supply source by time zone, an operation schedule of a distribution unit that selectively supplies power from one of the first power generation unit, the second power generation unit, and the external power supply source to the apartment complex and a power generation schedule of the second power generation unit are calculated, wherein the second power generation unit includes a hydrogen fuel cell power generation unit;
It is configured to control the distribution unit according to the above operation schedule,

The number of members of the first generation in the time zone

In the name

If the name is changed,
The above processor

Time zone and

+k The predicted load E(t) of the first generation for time periods t between time periods

is configured to decide,

is the average of the predicted load for time period t of the generations excluding the first generation among the generations with the same area as the first generation and with c as the number of generation members,

is the average of the actual load at time t of the generations excluding the first generation among the generations with the same area as the first generation and with c as the number of generation members,
R(t) is the actual load at time t of the first generation,
A distribution control device, where j and k are time intervals each having a predetermined size. delete In the first paragraph,

Number of household members in a time zone

In case of the first generation moving in,
The above processor

Time zone and

+k The predicted load E(t) of the first generation for time periods t between time periods

A distribution control device configured to determine. In the first paragraph,
The above processor is configured to receive rate information of the external power supply source through the communication circuit, a power distribution control device In the first paragraph,
The above weather information and the above weather forecast information, a distribution control device including temperature, wind speed, and precipitation.