Information processing apparatus, information processing method, and program

Abstract

To accurately estimate power generation in an object time zone by using a plurality of estimation systems.SOLUTION: An information processing apparatus includes a label processing unit and a composite data generation unit. The label processing unit associates, for estimation data representing an estimated power generation amount for each time zone generated by the target power generation facility, a label representing a set of an estimation system used for calculation of the estimation data and a version corresponding to a time difference from an update time of the estimation data to a start time of the time zone. The synthetic data generation unit generates synthetic data representing the amount of power generated by the target power generation facility in the target time zone based on a plurality of pieces of target estimation data and a plurality of pieces of error data estimated in advance and corresponding to the plurality of pieces of target estimation data. Each of the plurality of pieces of target estimation data is estimation data estimating the amount of power generated by the target power generation facility in the target time zone, associated with a label different from that of the other pieces of target estimation data.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] An embodiment of the present invention relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] The amount of electricity generated by renewable energy sources, such as solar power generation and wind power generation, varies depending on weather conditions. Therefore, in order to efficiently utilize the electricity generated by renewable energy sources, it is necessary to predict the amount of electricity generated.

[0003] There are several methods for estimating the amount of power generated by renewable energy sources. For example, when adjusting supply and demand, an estimation method is used that estimates the amount of power generated based on the power generation record. For example, when bidding in advance in the market, an estimation method based on meteorological satellite images or a high-resolution meteorological model is used. For example, when bidding in the spot market, an estimation method based on a low-resolution meteorological model is used.

[0004] However, it is very difficult to use various estimation methods appropriately depending on the purpose. For example, if the estimation method to be used is fixed, it is difficult to deal with cases where no estimated data can be obtained because no performance data or meteorological data can be received. In addition, when multiple estimation methods are used in combination, the error due to the conditions of use is unknown, making it difficult to select the optimal estimation method. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] International Publication No. 2011 / 114752 [Patent Document 2] International Publication No. 2017 / 098827 [Patent Document 3] International Publication No. 2020 / 203854 Summary of the Invention [Problem to be solved by the invention]

[0006] The problem to be solved by the present invention is to accurately estimate the amount of power generation in a target time period by using a plurality of estimation methods. [Means for solving the problem]

[0007] The information processing device according to the embodiment includes a label processing unit and a synthetic data generating unit. The label processing unit associates, with estimation data representing an estimated amount of power generated by a target power generation facility for each time period, a label representing a combination of an estimation method used to calculate the estimation data and a version according to a time difference from an update time of the estimation data to a start time of the time period. The synthetic data generating unit generates synthetic data representing an amount of power generated by the target power generation facility in a target time period based on a plurality of target estimation data and a plurality of error data estimated in advance corresponding to the plurality of target estimation data. Each of the plurality of target estimation data is the estimation data estimating an amount of power generated by the target power generation facility in the target time period, and associated with the label different from other target estimation data. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram showing the configuration of an estimation device according to a first embodiment. [Diagram 2] 4 is a flowchart showing the flow of an estimation data acquisition process performed by the estimation device. [Diagram 3] 10 is a flowchart showing the flow of synthetic data generation processing performed by the estimation device. [Figure 4] FIG. 13 is a diagram showing an example of association of labels with estimated data. [Diagram 5] FIG. 13 is a diagram showing an example of a portion of an estimated data table. [Figure 6] FIG. 13 is a diagram showing an example of a portion of an error table. [Figure 7] FIG. 13 is a diagram showing an example of reading out estimated data. [Figure 8]FIG. 11 is a diagram showing an example of reading error data. [Figure 9] FIG. 13 is a diagram showing the configuration of an estimation device according to a second embodiment. [Figure 10] FIG. 2 is a diagram showing an example of a hardware configuration of an estimation apparatus. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, an embodiment of the present invention will be described with reference to the accompanying drawings.

[0010] (First embodiment) First, an estimation device 10 according to a first embodiment will be described. FIG. 1 is a diagram showing the configuration of an estimation device 10 according to the first embodiment.

[0011] The estimation device 10 estimates the amount of power generated by a target power generation facility during a target time period.

[0012] The target power generation facility is a renewable energy source that generates power using natural energy as an energy source. As an example, the target power generation facility is a solar power plant that generates power using solar energy as an energy source or a wind power plant that generates power using wind energy as an energy source. The target power generation facility may be of any scale. In the case of a solar power plant, the target power generation facility may be a string unit or a unit that is a total of multiple strings installed in the solar power plant. In the case of a wind power plant, the target power generation facility may be a wind power generator unit or a unit of multiple wind power generators installed in the wind power plant. In addition, the target power generation facility may be a unit that groups together multiple solar power plants that exist in an area where the amount of solar energy received is considered to be the same, or a unit that groups together multiple wind power plants that exist in an area where the amount of wind energy received is considered to be the same.

[0013] The target time period is a time period of a specified hour, 30 minutes, etc. on a specified date. For example, the target time period is a time period from 12:00 to 12:59 on a specified date. In addition, if the target power generation facility is a wind power plant, the target time period may be a time period longer than 1 hour or 30 minutes.

[0014] The estimation device 10 is realized, for example, by one or more information processing devices, and functions by the one or more information processing devices executing an estimation program. The estimation device 10 may also be realized by a server device or a cloud on a network. The estimation device 10 may also be realized by a hardware circuit such as a field-programmable gate array (FPGA) or an application specific integrated circuit (ASIC).

[0015] The estimation device 10 includes an estimated data acquisition unit 24, a label processing unit 26, an estimated data storage unit 28, an estimated data selection unit 30, an error table storage unit 32, an error data acquisition unit 34, and a synthetic data generation unit 36.

[0016] The estimated data acquisition unit 24 acquires estimated data representing an estimated amount of power generated by the target power generation facility for each time period. For example, the estimated data acquisition unit 24 may acquire estimated data calculated by an information processing device separate from the estimation device 10, or may receive estimated data from a server device or the like via a network.

[0017] The time period is a time period of one hour, 30 minutes, etc., starting from a specific time on a specific date. The estimated data acquisition unit 24 acquires a plurality of estimated data on power generation by the target power generation facility for one time period.

[0018] The multiple estimated data acquired for one time period differ in the estimation method or the time difference from the update time to the start time of the time period. In the case of solar power generation, the estimation methods include, for example, the TLM (telemetry model) method, which extrapolates from the most recent power generation results, the LFM (local model) method and the WFM (weather research and forecasting model) method, which use a frequently updated numerical weather model, and the SAT method, which uses meteorological satellite images.

[0019] The update time is the time when the estimated data is updated. For example, when the power generation amount for a one-hour time slot is estimated every hour, the update time is a predetermined minute (e.g., 00 minutes) in each hour. In addition, the start time of the time slot is 12:00 PM, for example, when the estimated data represents the estimated power generation amount for a one-hour time slot from 12:00 PM to 12:59 PM.

[0020] For example, the time difference between the update time and the start time of the time slot is 0 hours for estimated data updated at 12:00 PM and estimating the amount of power generation from 12:00 PM to 12:59 PM on the current day. The time difference between the update time and the start time of the time slot is 1 hour for estimated data updated at 12:00 PM and estimating the amount of power generation from 1:00 PM to 1:59 PM on the current day. The time difference is 6 hours for estimated data updated at 12:00 PM and estimating the amount of power generation from 6:00 PM to 6:59 PM on the current day.

[0021] The estimated data includes an estimated value of the power generation amount. Furthermore, information for identifying a time period and a time difference is added to the estimated data as metadata. The information for identifying a time period and a time difference includes, for example, an update time of the estimated data, an update interval at which the estimated data is updated, and an offset that is a time difference between when the estimated data is calculated and when the estimated data is updated.

[0022] The label processing unit 26 identifies a time period during which the amount of power represented by the estimated data is generated, based on the metadata added to the estimated data acquired by the estimated data acquisition unit 24. For example, the label processing unit 26 identifies the date, start time, and end time that represent the time period during which the amount of power represented by the estimated data acquired by the estimated data acquisition unit 24 is generated, based on the update time, update interval, and offset included in the metadata.

[0023] Furthermore, the label processing unit 26 identifies the version of the estimated data based on the metadata added to the estimated data acquired by the estimated data acquisition unit 24. The version indicates a value according to the time difference between the update time of the estimated data and the start time of the time period. In this embodiment, the version has a larger value as the time difference between the update time of the estimated data and the start time of the time period is longer.

[0024] When the estimated data is updated every hour and the time difference between the update time of the estimated data and the start time of the time period is 0 hours, i.e., when the estimated data is the latest, the version is, for example, "00". When the estimated data is updated every hour and the time difference between the update time of the estimated data and the start time of the time period is 1 hour, the version is, for example, "01". When the estimated data is updated every hour and the time difference between the update time of the estimated data and the start time of the time period is 6 hours, the version is, for example, "06".

[0025] Furthermore, when the estimated data is updated every 24 hours, and the time difference between the update time of the estimated data and the start time of the time period is 0 hours or more and less than 24 hours, the version is, for example, "00." When the estimated data is updated every 24 hours, and the time difference between the update time of the estimated data and the start time of the time period is 24 hours or more and less than 48 hours, the version is, for example, "01."

[0026] Furthermore, the label processing unit 26 generates a label representing a combination of the estimated method and the version for the estimated data acquired by the estimated data acquisition unit 24, and associates the generated label. In this embodiment, the label processing unit 26 generates a character string as a label by combining a character string representing the estimated method and a character string representing the version, and associates the label with the estimated data acquired by the estimated data acquisition unit 24. Then, the label processing unit 26 writes the estimated data into the estimated data storage unit 28 so that the time period and label can be identified.

[0027] The estimated data storage unit 28 stores a plurality of pieces of estimated data. The estimated data storage unit 28 stores each of the plurality of pieces of estimated data by identifying the label and the time period.

[0028] In this embodiment, the estimated data storage unit 28 stores an estimated data table that stores a plurality of estimated data. The estimated data table is described, for example, in a spreadsheet format including a plurality of cells. Each of the plurality of cells is specified by a row and a column. In this embodiment, in the estimated data table, the rows represent time periods and the columns represent labels. Each of the plurality of estimated data is stored in one of the plurality of cells that is specified by a time period and a label.

[0029] In addition, in the estimated data table, some cells in which estimated data should be stored may be blank cells, i.e., cells in which valid estimated data is not stored. For example, if estimated data cannot be acquired due to the occurrence of a problem that prevents reception of performance data or weather data, the estimated data table may have blank cells in which the estimated data that could not be acquired should be stored.

[0030] The estimated data selection unit 30 receives a target time period, which is a time period for estimating the amount of power generated by the target power generation facility. The estimated data selection unit 30 may receive the target time period from a user, or may periodically receive the target time period from another information processing device.

[0031] When a target time period is received, the estimated data selection unit 30 selects, from the estimated data storage unit 28, valid data from a group including estimated data that estimates the amount of power generation during the received target time period, as multiple target estimated data. For example, the estimated data selection unit 30 selects valid estimated data, i.e., estimated data included in cells that are not blank cells, from a group of multiple estimated data stored in the row of the target time period in the estimated data table, as multiple target estimated data. Thus, the estimated data selection unit 30 selects different numbers of multiple target estimated data depending on the target time period.

[0032] Then, the estimated data selection unit 30 acquires each of the selected plurality of target inferred data in a state in which the data can be identified by the label. That is, by designating a label, it is possible to identify the target inferred data associated with the designated label from among the plurality of target inferred data.

[0033] Each of the selected multiple target estimated data is estimated data that estimates the amount of power generated by the target power generation facility in the target time period and is associated with a label different from that of the other target estimated data. In other words, the multiple target estimated data does not include two or more estimated data associated with the same label.

[0034] For example, assume that the plurality of target inferred data includes N target inferred data (N is an integer equal to or greater than 2). In this case, a first target inferred data among the N target inferred data corresponds to a first label among N different labels. Also, an nth target inferred data among the N target inferred data (n is an integer equal to or greater than 1 and equal to or less than N) corresponds to the nth label among the N labels.

[0035] The estimated data selection unit 30 provides a plurality of target inference data acquired in a state in which each of the target inference data can be identified by a label to the synthetic data generation unit 36. Furthermore, the estimated data selection unit 30 provides a label set including a plurality of labels associated with the selected plurality of target inference data to the synthetic data generation unit 36 ​​and the error data acquisition unit 34. For example, when the plurality of target inference data includes N pieces of target inference data, the label set includes N labels.

[0036] The error table storage unit 32 stores a pre-generated error table. The error table stores error data representing a pre-estimated error included in the estimated data for each time period and each label.

[0037] In this embodiment, the error table is written in a spreadsheet format including a plurality of cells. Each of the plurality of cells is specified by a row and a column. In the error table, for example, the rows represent time periods and the columns represent labels. Each of the plurality of error data is stored in one of the plurality of cells, which is specified by a time period and a label.

[0038] Also, for example, in the case of solar power generation, the error data for the time slots with the same start time on different nearby days are the same or similar. Therefore, the error table may store error data for each time slot within a day, for example, for each month or season. For example, the error table may store monthly average error data for each time slot within a day, for example, for each month.

[0039] The error data for each time period and each label can be generated based on, for example, actual data representing actual values ​​of the amount of power generated in each time period in the target power generation facility in the past, and estimated data calculated in the past. For example, the error data is calculated based on an error value obtained by subtracting the actual data from the estimated data for each time period.

[0040] In this embodiment, the error data for each time period and each label includes a weighting coefficient. The weighting coefficient is expressed by a real number greater than 0. The weighting coefficient has a larger value as the error estimated to be included in the estimated data is larger.

[0041] For example, the error data for each time period and each label may include a bias value in addition to the weighting coefficient. The bias value is expressed by a real number. The bias value has a positive sign when the sign of the error estimated to be included in the estimated data is positive, and has a negative sign when the sign of the error estimated to be included in the estimated data is negative. In addition, the larger the absolute value of the error, the larger the absolute value of the bias value.

[0042] The error data acquisition unit 34 receives the label set acquired from the estimation data selection unit 30. Furthermore, the error data acquisition unit 34 accepts the specified target time period.

[0043] When the target time period and the label set are received, the error data acquisition unit 34 acquires from the error table a plurality of error data corresponding to a plurality of labels included in the label set for a time period corresponding to the received target time period. In this case, the error data acquisition unit 34 acquires the plurality of error data in a state in which each of the error data can be identified by the label. In other words, by specifying a label, it is possible to identify the error data associated with the specified label from the plurality of error data acquired by the error data acquisition unit 34.

[0044] Therefore, the multiple error data acquired by the error data acquisition unit 34 correspond one-to-one to the multiple target estimation data selected by the estimation data selection unit 30. For example, assume that the multiple target estimation data includes N target estimation data. In this case, the first target estimation data of the N target estimation data corresponds to the first error data of the N error data. Also, the nth target estimation data of the N target estimation data corresponds to the nth error data of the N error data.

[0045] Then, the error data acquisition unit 34 provides the plurality of error data thus acquired in a state in which each of the error data can be identified by the label to the composite data generation unit 36 ​​.

[0046] The synthetic data generation unit 36 ​​acquires a plurality of pieces of target estimation data from the estimation data selection unit 30. The synthetic data generation unit 36 ​​also acquires a plurality of pieces of target estimation data from the error data acquisition unit 34. The plurality of pieces of target estimation data and the plurality of pieces of error data correspond one-to-one. The synthetic data generation unit 36 ​​can also specify a first label to identify a first target estimation data among the plurality of pieces of target estimation data and a first error data among the plurality of pieces of error data.

[0047] The composite data generator 36 generates composite data representing the amount of power generated by the target power generation facility in the target time period, based on the plurality of acquired target estimation data and the plurality of acquired error data.

[0048] For example, the synthetic data generating unit 36 ​​divides each of the plurality of target estimation data by a value corresponding to a weighting coefficient included in the corresponding error data among the plurality of error data. For example, the synthetic data generating unit 36 ​​divides the first target estimation data among the plurality of target estimation data by a value corresponding to a weighting coefficient included in the first error data among the plurality of error data. For example, the value corresponding to the weighting coefficient is the square of the weighting coefficient.

[0049] Then, the synthetic data generating unit 36 ​​calculates a value obtained by summing the multiple target estimation data divided by a value corresponding to a weighting coefficient included in the corresponding error data among the multiple error data, and multiplying the sum by a standardization coefficient as synthetic data. The standardization coefficient is a value that sets the influence of the value corresponding to the weighting coefficient to 1. For example, the standardization coefficient is the reciprocal of the sum of the reciprocals of the values ​​corresponding to the weighting coefficients.

[0050] As a result, the composite data generating unit 36 ​​can perform magnification correction on each of the plurality of target estimation data so that the smaller the corresponding error data is, the greater the influence is, and the larger the corresponding error data is, the less the influence is, and calculate a weighted average of the plurality of target estimation data after magnification correction.The composite data generating unit 36 ​​can then output the weighted average thus calculated as composite data.

[0051] Furthermore, the synthetic data generating unit 36 ​​may subtract a bias value included in the corresponding error data from each of the plurality of target estimation data before dividing the plurality of error data by the weighting coefficient. In this way, when each of the plurality of error data includes a bias component, the synthetic data generating unit 36 ​​can correct the bias component of each of the plurality of target estimation data.

[0052] Then, the composite data generating unit 36 ​​outputs the calculated composite data as the amount of power generation estimated to be generated by the target power generation facility in the target time period.

[0053] 2 is a flowchart showing the flow of an estimated data acquisition process performed by the estimation device 10. The estimation device 10 acquires the estimated data according to the flow shown in FIG.

[0054] First, in S11, the estimation device 10 determines whether or not the estimation data has been acquired. If the estimation data has not been acquired (No in S11), the estimation device 10 waits in S11 until the estimation data is acquired. If the estimation data has been acquired (Yes in S11), the estimation device 10 advances the process to S12.

[0055] In S12, the estimation device 10 identifies a time period during which the amount of power generation represented by the acquired estimated data is generated by the target power generation facility. Then, in S13, the estimation device 10 identifies a label representing a combination of the estimation method used to calculate the acquired estimated data and a version according to the time difference from the update time of the acquired estimated data to the start time of the time period. Then, in S14, the estimation device 10 associates the acquired estimated data with the identified time period and the identified label, and stores the acquired estimated data in an estimated data table. Then, after completing the process of S14, the estimation device 10 returns the process to S11 and waits for acquisition of the next estimated data.

[0056] 3 is a flowchart showing the flow of a synthetic data generation process performed by the estimation device 10. The estimation device 10 generates synthetic data according to the flow shown in FIG.

[0057] First, in S21, the estimation device 10 determines whether or not the target time period has been accepted. If the target time period has not been accepted (No in S21), the estimation device 10 waits in S21 until the target time period is accepted. If the target time period has been accepted (Yes in S21), the estimation device 10 advances the process to S22.

[0058] Next, in S22, the estimation device 10 acquires valid data from among a group of multiple estimated data stored in the row of the target time period in the estimated data table as multiple target estimated data. Next, in S23, the estimation device 10 acquires multiple labels associated with the multiple target estimated data. Next, in S24, the estimation device 10 acquires multiple error data corresponding to multiple label rows from among the multiple error data stored in the row of the target time period in the error table.

[0059] Next, in S25, the estimation device 10 generates composite data representing the amount of power generated by the target power generation facility in the target time zone based on the multiple target estimation data and the multiple error data. Next, in S26, the estimation device 10 outputs the calculated composite data as the amount of power generation estimated to be generated by the target power generation facility in the target time zone.

[0060] After completing the process of S26, the estimation device 10 returns the process to S21 and waits for acceptance of the next target time slot.

[0061] 4 is a diagram illustrating an example of association of labels with estimated data. The estimation device 10 acquires, for example, estimated data calculated by the TLM method, estimated data calculated by the LFM method, and estimated data calculated by the WRF method.

[0062] The estimated data calculated by the TLM method is updated, for example, every hour. At the same update time, the estimation device 10 obtains estimated data calculated by the TLM method for seven different time periods from 0 hours, 1 hour, ..., 6 hours after the update time.

[0063] When the estimation device 10 receives estimated data obtained by estimating the amount of power generation in a time period 0 hours after the update time using the TLM method, the estimation device 10 generates a character string of "TLM" as information for identifying the estimation method, generates a character string of "00" as a version, and generates a character string combining "TLM" and "00" as a label. When the estimation device 10 receives estimated data obtained by estimating the amount of power generation in a time period 1 hour after the update time using the TLM method, the estimation device 10 generates a character string of "TLM" as information for identifying the estimation method, generates a character string of "01" as a version, and generates a character string combining "TLM" and "01" as a label. When the estimation device 10 receives estimated data obtained by estimating the amount of power generation in a time period 6 hours after the update time using the TLM method, the estimation device 10 generates a character string of "TLM" as information for identifying the estimation method, generates a character string of "06" as a version, and generates a character string combining "TLM" and "06" as a label.

[0064] The estimated data calculated by the LFM method is updated, for example, every hour. At the same update time, the estimation device 10 obtains estimated data calculated by the LFM method for eleven different time periods from 0 hours, 1 hour, ..., 10 hours after the update time.

[0065] When the estimation device 10 receives estimated data obtained by estimating the amount of power generation in a time period 0 hours after the update time using the LFM method, the estimation device 10 generates a character string of "LFM" as information for identifying the estimation method, generates a character string of "00" as a version, and generates a character string combining "LFM" and "00" as a label. When the estimation device 10 receives estimated data obtained by estimating the amount of power generation in a time period 1 hour after the update time using the LFM method, the estimation device 10 generates a character string of "LFM" as information for identifying the estimation method, generates a character string of "01" as a version, and generates a character string combining "LFM" and "01" as a label. When the estimation device 10 receives estimated data obtained by estimating the amount of power generation in a time period 10 hours after the update time using the LFM method, the estimation device 10 generates a character string of "LFM" as information for identifying the estimation method, generates a character string of "10" as a version, and generates a character string combining "LFM" and "10" as a label.

[0066] The estimated data calculated by the WRF method is updated, for example, every 24 hours. The estimation device 10 acquires estimated data calculated by the WRF method for (7 x 24) types of time slots, from the current day to six days later, and from 0 hours to 23 hours after the update time of each day, at the same update time.

[0067] When the estimation device 10 receives estimation data that estimates the amount of power generation in a time period from 0 hours to 23 hours after the update time on the day using the WRF method, the estimation device 10 generates a character string of "WRF" as information for identifying the estimation method, generates a character string of "00" as a version, and generates a character string combining "WRF" and "00" as a label. When the estimation device 10 receives estimation data that estimates the amount of power generation in a time period from 0 hours to 23 hours after the update time one day later using the WRF method, the estimation device 10 generates a character string of "WRF" as information for identifying the estimation method, generates a character string of "01" as a version, and generates a character string combining "WRF" and "01" as a label. When the estimation device 10 receives estimation data that estimates the amount of power generation in a time period from 0 hours to 23 hours after the update time six days later using the WRF method, the estimation device 10 generates a character string of "WRF" as information for identifying the estimation method, generates a character string of "06" as a version, and generates a character string combining "WRF" and "06" as a label.

[0068] FIG. 5 is a diagram showing an example of a portion of an estimated data table written in a spreadsheet format.

[0069] The estimation device 10 associates the generated label with each of the multiple estimated data at each update time, and then stores each of the multiple estimated data in a cell in the row of the corresponding time period and the column of the corresponding label in the estimated data table.

[0070] For example, when the update time is 12:00 on Apr. 12, 2023, the estimation device 10 acquires estimated data associated with labels TLM00 to TLM06, estimated data associated with labels LFM00 to LFM10, and estimated data associated with labels WRF00 to WRF06 before the update time. In this case, as shown in Fig. 5, the estimation device 10 stores each estimated data in a cell in a row of a corresponding time period and a column identified by the label in the estimated data table.

[0071] When the estimating device 10 is unable to acquire some of the estimated data, it leaves the cells in which the estimated data that could not be acquired should be stored blank.

[0072] FIG. 6 is a diagram showing an example of a portion of an error table written in a spreadsheet format.

[0073] The Error Table is written in a spreadsheet format where columns are identified by labels, so that the columns in the Error Table have a one-to-one correspondence with the columns in the Estimated Data Table.

[0074] In the error table, rows are identified by hourly time periods over a one-year period. However, in the example of Fig. 6, rows in the error table are not identified by day, but by time periods from 0:00 to 23:00 on a monthly basis. Note that rows in the error table may be identified by three-month or seasonal units instead of by month.

[0075] The error table includes a weighting factor (σ) and a bias value (β) for each of a number of cells identified by a time period and a label.

[0076] The weighting factor (σ) is, for example, the mean absolute error (MAE) or the root mean square error (RMSE), and the bias value (β) is, for example, the mean bias error (MBE).

[0077] For example, the error (err(m,d,h,p,v)) is defined as shown in equation (1).

number

[0078] pred(m,d,h,p,v) represents estimated data. meas(m,d,h) represents actual data. m represents month. d represents day. h represents time period in 1-hour units. p represents estimation method. v represents version.

[0079] That is, the error is represented by a value obtained by subtracting the actual value from the estimated data.

[0080] In this case, the mean absolute error (MAE) is expressed by, for example, equation (2).

number

[0081] D is a set of numbers representing the days in the month identified by m. |D| is the number of days in the month identified by m. * indicates the average value for each month.

[0082] Moreover, the root mean square error (RMSE) is expressed by, for example, equation (3).

number

[0083] Moreover, the mean bias error (MBE) is expressed by, for example, equation (4).

number

[0084] For example, the error table stores, in each of a plurality of cells identified by time periods and labels, a weighting factor (σ) calculated by formula (2) or formula (3) based on the power generation performance and estimated data for the past year in the target power generation facility, and a bias value (β) calculated by formula (4). Note that the weighting factor (σ) and the bias value (β) may be calculated by other calculation formulas, not limited to formulas (2) to (4).

[0085] FIG. 7 is a diagram showing an example of reading out estimated data from the estimated data table.

[0086] The estimation device 10 accepts a target time period (t). When the target time period (t) is accepted, the estimation device 10 acquires, as a plurality of target estimated data, a plurality of estimated data stored in a row of a time period corresponding to the target time period (t) in the estimated data table. In this case, the estimation device 10 acquires a plurality of target estimated data, excluding cells in which no valid estimated data is stored (e.g., blank cells in FIG. 7).

[0087] Then, the estimation device 10 acquires each of the acquired plurality of pieces of target inference data in a state in which the label can be identified. For example, the estimation device 10 acquires each of the acquired plurality of pieces of target inference data in a state in which the label can be identified. C The time is obtained by substituting (t) where t represents the date and time of the target time period. C The subscript C of is a string representing a label, and y c Specify one of the multiple values ​​included in (t).

[0088] Furthermore, the estimation device 10 generates a label set including a plurality of labels associated with a plurality of target estimation data. COLS(t) represents the label set. COLS(t) includes a set of character strings representing the labels.

[0089] For example, in the example of FIG. 7, the estimation device 10 accepts a period from 12:00 to 12:59 on April 12, 2023 as the target time period (t). In the example of FIG. 7, the estimation device 10 receives a plurality of target estimation data (y c As the label set (COLS(t)), the estimation device 10 acquires estimated data of the column of TLM00, estimated data of the column of TLM01, estimated data of the column of TLM02, estimated data of the column of TLM04, estimated data of the column of TLM05, estimated data of the column of TLM06, estimated data of the column of LFM00, estimated data of the column of LFM01, estimated data of the column of LFM03, estimated data of the column of LFM04, etc. from 12:00 to 12:59. In the example of Fig. 7, the estimation device 10 acquires a set including TLM00, TLM01, TLM02, TLM04, TLM05, TLM06, LFM00, LFM01, LFM03, LFM04, etc.

[0090] FIG. 8 is a diagram showing an example of reading out error data from the estimated data table.

[0091] After generating the label set (COLS(t)), the estimation device 10 identifies multiple cells from rows in the error table identified by the month component (t.month) and hour component (t.hour) of the target time period (t) and multiple columns identified by multiple labels included in the label set (COLS(t)). Then, the estimation device 10 acquires a weighting coefficient (σ) and a bias value (β) in the error data from each of the identified multiple cells. In this case, the estimation device 10 acquires the weighting coefficient (σ) and the bias value (β) in a state in which the labels can be identified. For example, the estimation device 10 stores each of the acquired multiple weighting coefficients (σ) in an array σ C Substitute (t.month, t.hour) to obtain σ C The subscript C in (t.month, t.hour) is a string representing the label, and σ C The estimation device 10 also identifies one of the multiple values ​​included in (t.month, t.hour). C Substitute (t.month, t.hour) to obtain β C The subscript C of (t.month, t.hour) is a string representing a label, and β C Identifies one of multiple values ​​contained in (t.month, t.hour).

[0092] For example, in the example of FIG. 8, the estimation device 10 generates a σ 2 , which includes weighting coefficients (σ) in the row from 12:00 to 12:59 (t.hour) in April (t.month) and in the columns TLM00, TLM01, TLM02, TLM04, TLM05, . . . C8, the estimation device 10 obtains a β C Get (t.month, t.hour).

[0093] Then, the estimation device 10 calculates y c (t), σ C (t.month, t.hour), β C Based on (t.month, t.hour) and COLS(t), the synthetic data (y MIX Specifically, the estimation device 10 calculates the synthetic data (y MIX Calculate (t).

number

[0094] Here, N(t) represents a standardization coefficient and is expressed by equation (6).

number

[0095] By performing the calculation shown in the formula (5), the estimation device 10 obtains a plurality of target estimation data (y c (t)) is weighted by the weighting coefficient (σ C The estimation device 10 can divide the error data by a value corresponding to the weighting coefficient (σ C The sum of multiple target estimation data divided by the value corresponding to (t.month, t.hour) is multiplied by the normalization coefficient (N(t)) and the sum is used as the synthetic data (y MIX (t)).

[0096] In addition, in formula (5), the weighting coefficient (σ C The value according to (t.month, t.hour) is the square of the weighting coefficient (σ C (t.month, t.hour) 2 ) The normalization coefficient (N(t)) is a value that makes the influence of the value according to the weighting coefficient 1. For example, the normalization coefficient is the weighting coefficient (σ C The reciprocal of the sum of the reciprocals of the values ​​corresponding to (t.month, t.hour).

[0097] In addition, in the above equation (5), the estimation device 10 uses a weighting coefficient (σ C Before dividing the value according to (t.month, t.hour), the bias value (β C Instead, the estimation device 10 subtracts the weighting factor (σ C Before dividing the value according to (t.month, t.hour), the bias value (β C There is no need to subtract (t.month, t.hour).

[0098] By performing the above-mentioned processing, the estimation device 10 can synthesize a plurality of estimation data with different estimation methods and versions for a target time period by performing a weighted average so as to minimize the errors contained therein. As a result, the estimation device 10 according to the present embodiment can estimate the power generation amount for the target time period with high accuracy using a plurality of estimation data.

[0099] Second embodiment Next, an estimation device 10 according to a second embodiment will be described. Since the estimation device 10 according to the second embodiment has substantially the same functions and configurations as those of the first embodiment, the components having substantially the same functions and configurations are denoted by the same reference numerals, and detailed descriptions thereof will be omitted except for the differences.

[0100] FIG. 9 is a diagram showing the configuration of an estimation device 10 according to the second embodiment.

[0101] The estimation device 10 according to the second embodiment further includes an initialization unit 42, a performance data acquisition unit 44, and an update unit 46 in addition to the components included in the first embodiment.

[0102] Before the estimation device 10 is introduced as a device for estimating the amount of power generated by the target power generation facility, the initialization unit 42 initializes the error table stored in the error table storage unit 32. For example, the initialization unit 42 generates error data for each time zone and label based on past performance data and past estimated data of the amount of power generated by the target power generation facility, and stores the error data in the error table. In this case, the initialization unit 42 may generate average error data for each time zone and label on a monthly basis, and store the average error data in the error table shown in FIG. 6.

[0103] The initialization unit 42 may also initialize the error table based on error parameters representing multiple pieces of error data for each time period and label in a power generation facility other than the target power generation facility. In this case, the initialization unit 42 acquires error parameters representing multiple pieces of error data normalized by power generation capacity for each time period and label in a power generation facility other than the target power generation facility. The initialization unit 42 then corrects the acquired error parameters according to the capacity of the target power generation facility to generate multiple pieces of error data to be stored in the error table.

[0104] In addition, the multiple error data included in the error parameter may not satisfy all or part of the multiple error data for each time period and each label to be stored in the error table. That is, the multiple error data included in the error parameter may not include error data for some or all months or time periods, or may not include error data for some or all labels. In such a case, the initialization unit 42 may generate a regression equation or the like based on the multiple error data included in the error parameter, and generate all of the multiple error data for each time period and each label to be stored in the error table.

[0105] In addition, the initialization unit 42 may generate multiple error data to be stored in the error table by combining multiple error data generated based on past actual data and past estimated data of the amount of power generated by the target power generation equipment with error parameters.

[0106] The estimation device 10 of the second embodiment is equipped with such an initialization unit 42, and thus can estimate the amount of power generated by the target power generation facility during the target time period immediately after the estimation device 10 is introduced as a device for estimating the amount of power generated by the target power generation facility.

[0107] After the estimation device 10 is introduced as a device for estimating the amount of power generated by the target power generation facility, the performance data acquisition unit 44 acquires performance data representing the amount of power generated by the target power generation facility.

[0108] The update unit 46 acquires estimated data for a time period in which the actual data acquired by the actual data acquisition unit 44 indicates the amount of power generated, from the estimated data table stored in the estimated data storage unit 28. For example, when the actual data acquisition unit 44 acquires actual data indicating the actual value of the amount of power generated in the target power generation facility in a first time period, the update unit 46 acquires first estimated data, which is the estimated data for the first time period. Furthermore, the update unit 46 calculates actual error data based on the actual data indicating the actual value of the amount of power generated in the target power generation facility in the first time period and the first estimated data, which is the estimated data for the first time period. Then, the update unit 46 updates the error data corresponding to the label associated with the first estimated data for the first time period in the error table, based on the actual error data.

[0109] For example, when the weighting coefficient (σ) is the mean absolute error (MAE), the update unit 46 updates the weighting coefficient (σ=MAE) in the error table based on the equation (7).

number

[0110] Furthermore, when the weighting coefficient (σ) is the root mean square error (RMSE), the update unit 46 updates the weighting coefficient (σ=RMSE) in the error table based on equation (8).

number

[0111] Furthermore, when the bias value (β) is the mean bias error (MBE), the update unit 46 updates the bias value (β=MBE) in the error table based on the equation (9).

number

[0112] δ is a predetermined coefficient. For example, when error data for each time period on a monthly basis is stored in the error table, δ is 2 / (1+30)≈0.065. This allows the update unit 46 to update the weighting coefficient (σ) to a 30-day moving average value.

[0113] The estimation device 10 of the second embodiment is equipped with such an actual data acquisition unit 44 and an update unit 46, and therefore, after the estimation device 10 is introduced as a device for estimating the amount of power generated by the target power generation facility, it is possible to update the error table so that the amount of power generated by the target power generation facility can be accurately estimated.

[0114] In addition, in the second embodiment, the estimated data selection unit 30 may acquire multiple target estimated data by other methods instead of acquiring valid data from a group including estimated data that estimates the power generation amount for the target time period from the estimated data storage unit 28 as multiple target estimated data.

[0115] For example, the estimated data selection unit 30 may acquire, as the plurality of target estimated data, a data group having the newest version or a version newer than a predetermined value from a group including a plurality of estimated data that estimate the amount of power generation in the received target time period, from the estimated data storage unit 28. Note that in this case as well, the estimated data selection unit 30 acquires valid data, i.e., data included in cells that are not blank, as the target estimated data.

[0116] The estimation device 10 of the second embodiment can thus accurately estimate the amount of power generation in a time period close to the update time, for example, by acquiring a group of data with the newest version or a version newer than a predetermined value as multiple target estimation data.

[0117] In the second embodiment, the estimated data selection unit 30 selects a first group including estimated data that estimates the amount of power generation in the received target time period from the estimated data storage unit 28. Then, the error data acquisition unit 34 selects a second group including error data corresponding to the labels associated with the estimated data included in the first group from the error table stored in the error table storage unit 32. Then, the error data acquisition unit 34 acquires a predetermined number or a predetermined ratio of data groups from the second group starting from the data group with the smallest error as multiple error data. Then, the estimated data selection unit 30 selects data groups from the first group that are associated with labels corresponding to the multiple error data as multiple target estimated data.

[0118] The estimation device 10 of the second embodiment can accurately estimate the amount of power generation by acquiring, as multiple target estimated data, multiple estimated data corresponding to a predetermined number or a predetermined percentage of error data, starting from the data with the smallest error, from a first group including estimated data that estimates the amount of power generation for the received target time period.

[0119] (Hardware configuration) Fig. 10 is a diagram showing an example of a hardware configuration of the estimation device 10 according to each embodiment. The estimation device 10 is realized by a computer having a hardware configuration as shown in Fig. 10, for example. The estimation device 10 includes a CPU (Central Processing Unit) 301, a RAM (Random Access Memory) 302, a ROM (Read Only Memory) 303, an operation input device 304, a display device 305, a storage device 306, and a communication device 307. These units are connected to each other via a bus.

[0120] The CPU 301 is a processor that executes arithmetic processing, control processing, etc. according to a program. The CPU 301 uses a predetermined area of ​​the RAM 302 as a working area and executes various processes in cooperation with programs stored in the ROM 303, the storage device 306, etc.

[0121] The RAM 302 is a memory such as a Synchronous Dynamic Random Access Memory (SDRAM), and functions as a work area for the CPU 301. The ROM 303 is a memory that stores programs and various information in a non-rewritable manner.

[0122] The operation input device 304 is an input device such as a mouse, a keyboard, etc. The operation input device 304 receives information input by a user as an instruction signal, and outputs the instruction signal to the CPU 301.

[0123] The display device 305 is a display device such as an LCD (Liquid Crystal Display), etc. The display device 305 displays various information based on a display signal from the CPU 301.

[0124] The storage device 306 is a device that writes and reads data to a semiconductor storage medium such as a flash memory, or a magnetically or optically recordable storage medium, etc. The storage device 306 writes and reads data to the storage medium in response to control from the CPU 301. The communication device 307 communicates with external devices via a network in response to control from the CPU 301.

[0125] The program executed by the computer has a modular configuration including an estimated data acquisition module, a label processing module, an estimated data selection module, an error data acquisition module, and a synthetic data generation module. The program may further include an initialization module, a performance data acquisition module, and an update module.

[0126] This program is deployed on the RAM 302 and executed by the CPU 301 (processor), causing the computer to function as the estimated data acquisition unit 24, the label processing unit 26, the estimated data selection unit 30, the error data acquisition unit 34, and the composite data generation unit 36. This program may also cause the computer to function as the initialization unit 42, the performance data acquisition unit 44, and the update unit 46. Note that the estimated data acquisition unit 24, the label processing unit 26, the estimated data selection unit 30, the error data acquisition unit 34, the composite data generation unit 36, the initialization unit 42, the performance data acquisition unit 44, and the update unit 46 may be partially or entirely configured as hardware circuits. The RAM 302 and the storage device 306 also function as the estimated data storage unit 28 and the error table storage unit 32.

[0127] In addition, the program executed by the computer is provided as a file in a format that can be installed on the computer or in a format that can be executed by the computer, and is recorded on a computer-readable recording medium such as a CD-ROM, a flexible disk, a CD-R, or a DVD (Digital Versatile Disk).

[0128] The program may be stored on a computer connected to a network such as the Internet and provided by being downloaded via the network. The program may be provided or distributed via a network such as the Internet. The program executed by the estimation device 10 may be provided by being pre-installed in the ROM 303 or the like.

[0129] Although the embodiments of the present invention have been described above, the above-mentioned embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be implemented in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These novel embodiments and their modifications are included in the scope and spirit of the invention, and are included in the scope of the invention and its equivalents described in the claims.

[0130] (Additional Note) The above-described embodiments can be summarized as the following technical solutions.

[0131] [Technical proposal 1] a label processing unit that associates, with estimated data that indicates an estimated amount of power generated by a target power generation facility for each time period, a label that indicates a combination of an estimation method used to calculate the estimated data and a version according to a time difference from an update time of the estimated data to a start time of the time period; a synthetic data generation unit that generates synthetic data representing an amount of power generated by the target power generation facility in a target time period based on a plurality of target estimation data and a plurality of error data that are estimated in advance and correspond to the plurality of target estimation data; Equipped with Each of the plurality of target estimated data is the estimated data that estimates the amount of power generated by the target power generation facility in the target time period, and is associated with the label that is different from the other target estimated data. Information processing device.

[0132] [Technical proposal 2] A first error data among the plurality of error data corresponds to a first target estimation data among the plurality of target estimation data; The first error data represents a pre-estimated error included in the estimation data of the estimation method and the version included in the label associated with the first target estimation data during a time period corresponding to the target time period. An information processing device according to technical proposal 1.

[0133] [Technical proposal 3] Each of the plurality of error data includes a weighting coefficient; The synthetic data generation unit sums the plurality of target estimation data divided by a value corresponding to the weighting coefficient included in the corresponding error data among the plurality of error data, and calculates the sum as the synthetic data by multiplying the sum by a standardization coefficient such that an influence of the value corresponding to the weighting coefficient is 1. An information processing device according to Technical Proposal 2.

[0134] [Technical proposal 4] The normalization coefficient is the reciprocal of the sum of the reciprocals of the values ​​corresponding to the weighting coefficients. An information processing device according to technical proposal 3.

[0135] [Technical proposal 5] each of the plurality of error data includes a bias value; The synthetic data generating unit subtracts the bias value included in the corresponding error data from each of the plurality of target estimation data before dividing the value according to the weighting coefficient. An information processing device according to technical proposal 3 or 4.

[0136] [Technical proposal 6] Further comprising an error table storage unit that stores the error table, The error table stores error data representing a pre-estimated error included in the estimated data for each time period and for each label. An information processing device described in any one of technical proposals 2 to 5.

[0137] [Technical proposal 7] The error table is further configured to be initialized based on error parameters representing the plurality of error data for each time period and each label in a power generation facility different from the target power generation facility. An information processing device according to Technical Proposal 6.

[0138] [Technical proposal 8] The initialization unit initializes the error table based on a past actual value of the amount of power generation for each of the time periods in the target power generation facility and the error parameter. An information processing device according to Technical Proposal 7.

[0139] [Technical proposal 9] and an update unit that calculates actual error data based on actual data representing an actual value of the amount of power generated in the target power generation facility during a first time slot and first estimated data which is the estimated data during the first time slot, and updates error data corresponding to the label associated with the first estimated data during the first time slot in the error table based on the actual error data. An information processing device according to Technical Proposal 6.

[0140] [Technical proposal 10] an estimated data storage unit that stores the estimated data by identifying the label and the time period; an estimated data selection unit that receives the target time period and selects, from the estimated data storage unit, valid data from a group including the estimated data that estimates the amount of power generation during the received target time period, as the plurality of target estimated data; an error data acquisition unit that acquires the plurality of error data corresponding to the plurality of acquired target estimation data from the error table; Further equipped An information processing device according to any one of technical proposals 6 to 9.

[0141] [Technical proposal 11] an estimated data storage unit that stores the estimated data by identifying the label and the time period; an estimated data selection unit that receives the target time period and selects, from the estimated data storage unit, a data group having the newest version from a group including a plurality of pieces of estimated data that estimate the amount of power generation during the received target time period; and an error data acquisition unit that acquires the plurality of error data corresponding to the plurality of acquired target estimation data from the error table; Further equipped An information processing device according to any one of technical proposals 6 to 9.

[0142] [Technical proposal 12] an estimated data storage unit that stores the estimated data by identifying the label and the time period; an estimated data selection unit that receives the target time period and selects, from the estimated data storage unit, a first group including the estimated data that estimates the amount of power generation for the received target time period; an error data acquisition unit that selects a second group from the error table, the second group including error data corresponding to the label associated with the estimated data included in the first group, and acquires a predetermined number or a predetermined ratio of data groups from the second group starting from the data group having the smallest error as the plurality of error data; Further equipped with The estimation data selection unit selects, from the first group, a data group to which the labels corresponding to the plurality of error data are associated, as the plurality of target estimation data. An information processing device according to any one of technical proposals 6 to 9.

[0143] [Technical proposal 13] The target power generation facility generates electricity using natural energy as an energy source. An information processing device according to any one of technical proposals 1 to 11.

[0144] [Technical proposal 14] An information processing method for estimating an amount of power generated by a target power generation facility in a target time period by an information processing device, comprising: The information processing device, Associating estimated data representing an estimated amount of power generated by the target power generation facility for each time period with a label representing a combination of an estimation method used to calculate the estimated data and a version corresponding to a time difference from an update time of the estimated data to a start time of the time period; generating synthetic data representing an amount of power generated by the target power generation facility in the target time period based on a plurality of target estimation data and a plurality of error data corresponding to the plurality of target estimation data and estimated in advance; Each of the plurality of target estimated data is the estimated data that estimates the amount of power generated by the target power generation facility in the target time period, and is associated with the label that is different from the other target estimated data. Information processing methods.

[0145] [Technical proposal 15] A program that causes an information processing device to function as an estimation device that estimates an amount of power generated by a target power generation facility in a target time period, The information processing device, a label processing unit that associates, with estimated data that indicates an estimated amount of power generated by the target power generation facility for each time period, a label that indicates a combination of an estimation method used to calculate the estimated data and a version according to a time difference from an update time of the estimated data to a start time of the time period; a synthetic data generation unit that generates synthetic data representing an amount of power generated by the target power generation facility in the target time period based on a plurality of target estimation data and a plurality of error data that are estimated in advance and correspond to the plurality of target estimation data; and make it work. Each of the plurality of target estimated data is the estimated data that estimates the amount of power generated by the target power generation facility in the target time period, and is associated with the label that is different from the other target estimated data. program. [Explanation of symbols]

[0146] 10 Estimation device 24 Estimation Data Acquisition Section 26 Label processing section 28 Estimation data storage unit 30 Estimation data selection section 32 Error table storage unit 34 Error data acquisition section 36 Synthetic Data Generation Unit 42 Initialization section 44 Performance Data Acquisition Department 46 Update section

Claims

1. a label processing unit that associates, with estimated data that indicates an estimated amount of power generated by a target power generation facility for each time period, a label that indicates a combination of an estimation method used to calculate the estimated data and a version according to a time difference from an update time of the estimated data to a start time of the time period; a synthetic data generation unit that generates synthetic data representing an amount of power generated by the target power generation facility in a target time period based on a plurality of target estimation data and a plurality of error data that are estimated in advance and correspond to the plurality of target estimation data; Equipped with Each of the plurality of target estimated data is the estimated data that estimates the amount of power generated by the target power generation facility in the target time period, and is associated with the label that is different from the other target estimated data. Information processing device.

2. A first error data among the plurality of error data corresponds to a first target estimation data among the plurality of target estimation data; The first error data represents a pre-estimated error included in the estimation data of the estimation method and the version included in the label associated with the first target estimation data during a time period corresponding to the target time period. The information processing device according to claim 1 .

3. Each of the plurality of error data includes a weighting coefficient; The synthetic data generation unit sums up the plurality of target estimation data divided by a value corresponding to the weighting coefficient included in the corresponding error data among the plurality of error data, and calculates the sum as the synthetic data by multiplying the sum by a standardization coefficient such that an influence of the value corresponding to the weighting coefficient is 1. The information processing device according to claim 2 .

4. The normalization coefficient is the reciprocal of the sum of the reciprocals of the values ​​corresponding to the weighting coefficients. The information processing device according to claim 3 .

5. each of the plurality of error data includes a bias value; The synthetic data generation unit subtracts the bias value included in the corresponding error data from each of the plurality of target estimation data before dividing the value according to the weighting coefficient. The information processing device according to claim 3 .

6. Further comprising an error table storage unit that stores the error table, The error table stores error data representing a pre-estimated error included in the estimated data for each time period and for each label. The information processing device according to claim 2 .

7. The error table is further configured to be initialized based on error parameters representing the plurality of error data for each time period and each label in a power generation facility different from the target power generation facility. The information processing device according to claim 6.

8. The initialization unit initializes the error table based on a past actual value of the amount of power generation for each of the time periods in the target power generation facility and the error parameter. The information processing device according to claim 7.

9. The system further includes an update unit that calculates actual error data based on actual data representing an actual value of the amount of power generated in the target power generation facility during a first time slot and first estimated data which is the estimated data during the first time slot, and updates error data corresponding to the label associated with the first estimated data during the first time slot in the error table based on the actual error data. The information processing device according to claim 6.

10. an estimated data storage unit that stores the estimated data by identifying the label and the time period; an estimated data selection unit that receives the target time period and selects, from the estimated data storage unit, valid data from a group including the estimated data that estimates the amount of power generation during the received target time period, as the plurality of target estimated data; an error data acquisition unit that acquires the plurality of error data corresponding to the plurality of acquired target estimation data from the error table; Further equipped The information processing device according to claim 6.

11. an estimated data storage unit that stores the estimated data by identifying the label and the time period; an estimated data selection unit that receives the target time period and selects, from the estimated data storage unit, a data group having the newest version from a group including a plurality of pieces of estimated data that estimate the amount of power generation during the received target time period; and an error data acquisition unit that acquires the plurality of error data corresponding to the plurality of acquired target estimation data from the error table; Further equipped The information processing device according to claim 6.

12. an estimated data storage unit that stores the estimated data by identifying the label and the time period; an estimated data selection unit that receives the target time period and selects, from the estimated data storage unit, a first group including the estimated data that estimates the amount of power generation in the received target time period; an error data acquisition unit that selects a second group from the error table, the second group including error data corresponding to the label associated with the estimated data included in the first group, and acquires a predetermined number or a predetermined ratio of data groups from the second group starting from the data group having a smaller error as the plurality of error data; Further equipped with The estimation data selection unit selects, from the first group, a data group to which the labels corresponding to the plurality of error data are associated, as the plurality of target estimation data. The information processing device according to claim 6.

13. The target power generation facility generates electricity using natural energy as an energy source. The information processing device according to claim 1 .

14. An information processing method for estimating an amount of power generated by a target power generation facility in a target time period by an information processing device, comprising: The information processing device, Associating estimated data representing an estimated amount of power generated by the target power generation facility for each time period with a label representing a combination of an estimation method used to calculate the estimated data and a version corresponding to a time difference from an update time of the estimated data to a start time of the time period; generating synthetic data representing an amount of power generated by the target power generation facility in the target time period based on a plurality of target estimation data and a plurality of error data corresponding to the plurality of target estimation data and estimated in advance; Each of the plurality of target estimated data is the estimated data that estimates the amount of power generated by the target power generation facility in the target time period, and is associated with the label that is different from the other target estimated data. Information processing methods.

15. A program that causes an information processing device to function as an estimation device that estimates an amount of power generated by a target power generation facility in a target time period, The information processing device, a label processing unit that associates, with estimated data that indicates an estimated amount of power generated by the target power generation facility for each time period, a label that indicates a combination of an estimation method used to calculate the estimated data and a version according to a time difference from an update time of the estimated data to a start time of the time period; a synthetic data generation unit that generates synthetic data representing an amount of power generated by the target power generation facility in the target time period based on a plurality of target estimation data and a plurality of error data that are estimated in advance and correspond to the plurality of target estimation data; and make it work. Each of the plurality of target estimated data is the estimated data that estimates the amount of power generated by the target power generation facility in the target time period, and is associated with the label that is different from the other target estimated data. program.

Images