Predicted value calculation device, method and program for controlling predicted value calculation device

Abstract

To provide a prediction value calculation device, control method of the prediction value calculation device, and program thereof that obtain a prediction value of varying object data depending upon an impact of factor data.SOLUTION: In a prediction value calculation system 1000, a prediction value calculation device 100 obtaining a prediction value of varying object data depending upon factor data on a prescribed kind comprises: a relationship expression storage unit that stores a relationship expression indicative of a relationship between factor data and object data; a prediction object period reception unit that receives an input of information designating a prediction object period of calculating a prediction value of the object data; an object data calculation unit that calculates the object data in each period, using the factor data in a plurality of the same past periods corresponding to the prediction object period, and the relationship expression; a prediction value identification information reception unit that receives an input of prediction value identification information identifying the object data to be selected as a prediction value from the object data in each period; and a prediction value output unit that outputs the object data identified by the prediction value identification information as the prediction value of the object data in the prediction object period.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a predicted value calculation device, a control method for a predicted value calculation device, and a program. [Background technology]

[0002] Electric power companies control the amount of power generated by thermal power plants, nuclear power plants, etc. according to the daily demand for electricity. In order to create a supply-demand balance plan that plans the allocation of power generated at power plants where the amount of power generated can be controlled, such as thermal power plants and nuclear power plants, electric power companies forecast the amount of power demand for periods ranging from one month to one year.

[0003] Various technologies have been proposed for systems and methods for predicting such power demand (see, for example, Patent Document 1). [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2013-228891 A Summary of the Invention [Problem to be solved by the invention]

[0005] On the other hand, because electricity demand changes depending on various weather conditions such as temperature and solar radiation, accurate weather forecasting is important for accurately predicting electricity demand.

[0006] However, the meteorological data that affects electricity demand is diverse, including not only temperature and solar radiation but also precipitation, wind speed, and humidity, and this meteorological data changes in complex ways due to the effects of annual climate change.

[0007] Therefore, there is a demand for technology that can predict electricity demand taking into account such annual variations in weather data.

[0008] In addition to electricity demand, there are various other data that fluctuate due to weather data, such as clothing sales, the number of traffic accidents, the number of spectators at sporting events, and the number of patients with a particular disease. It would be preferable to be able to calculate predicted values ​​for these target data taking into account the effects of variability in weather data.

[0009] In addition to meteorological data, other data that affect specific target data include electricity market prices, and the number of employees and unemployment rates in specific fields such as agriculture and forestry. As with meteorological data, it is preferable to be able to calculate predicted values ​​for these factor data taking into account the effects of variability.

[0010] The present invention has been made in consideration of the above-mentioned problems, and has as an object to provide a predicted value calculation device, a control method for the predicted value calculation device, and a program that are capable of determining a predicted value taking into account the variance of factor data for target data that fluctuates under the influence of a predetermined type of factor data. [Means for solving the problem]

[0011] One means for solving the above problem is a prediction value calculation device for calculating a prediction value of target data that varies under the influence of a predetermined type of factor data, the device including a relational expression storage unit that stores a relational expression that expresses a relationship between the factor data and the target data, a prediction target period receiving unit that receives an input of information specifying a prediction target period for which a prediction value of the target data is to be calculated, and a prediction value calculation unit that receives an input of information specifying a prediction target period for which a prediction value of the target data is to be calculated, the prediction value calculation unit including a prediction value calculation unit that calculates a prediction value of the target data based on the prediction value, and a prediction value calculation unit that calculates a prediction value based on the prediction value. each The prediction system includes a target data calculation unit that calculates target data for each period using the factor data for that period and the relational equation, a predicted value identification information receiving unit that receives an input of predicted value identification information that identifies target data to be selected as the predicted value from the target data for each period, and a predicted value output unit that outputs the target data identified by the predicted value identification information as a predicted value of the target data for the prediction target period.

[0012] In addition, the problems and solutions disclosed in this application will be made clear by the description in the section of the preferred embodiment of the invention and the descriptions in the drawings. Effect of the Invention

[0013] For target data that fluctuates under the influence of a predetermined type of factor data, a predicted value can be obtained that takes into account the variance of the factor data. [Brief description of the drawings]

[0014] [Figure 1] FIG. 1 is an overall configuration diagram of a predicted value calculation system. [Diagram 2] FIG. 2 is a hardware configuration diagram of a predicted value calculation device. [Diagram 3] FIG. 2 is a hardware configuration diagram of a weather data providing device. [Figure 4] FIG. [Diagram 5] FIG. 11 is a diagram illustrating an example of a weather data table. [Figure 6] FIG. 1 is a diagram showing how the effects of long-term climate change are corrected. [Figure 7] FIG. 13 is a diagram illustrating an example of an air temperature correction table. [Figure 8] FIG. 1 is a diagram showing a breakdown of total power demand. [Figure 9] FIG. 1 is a diagram for explaining a method for calculating a hydroelectric power generation amount. [Figure 10] FIG. 2 is a diagram illustrating a functional configuration of a predicted value calculation device. [Figure 11] FIG. 13 is a diagram illustrating an example of an input screen for a prediction period. [Figure 12] FIG. 13 is a diagram showing a histogram of power demand. [Figure 13] FIG. 13 is a diagram showing an example of a display of a calculation result of an amount of power demand. [Figure 14] FIG. 13 is a diagram illustrating an example of a prediction scenario selection screen. [Figure 15] 13 is a flowchart showing an example of a process flow of a predicted value calculation device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] At least the following matters will become apparent from the description of this specification and the accompanying drawings. Hereinafter, the present invention will be described in detail based on an embodiment thereof with reference to the accompanying drawings.

[0016] ==Overall Configuration== 1 shows the overall configuration of a forecasted value calculation system 1000 according to one embodiment of the present invention. The forecasted value calculation system 1000 includes a forecasted value calculation device 100 and a meteorological data providing device 200. The forecasted value calculation device 100 and the meteorological data providing device 200 are connected to each other so as to be able to communicate with each other via a network 500 such as the Internet, a LAN (Local Area Network), or a telephone network.

[0017] The forecast value calculation device 100 is a computer that predicts target data that fluctuates under the influence of a predetermined type of factor data. As an example in this embodiment, the forecast value calculation device 100 predicts the power demand that fluctuates under the influence of meteorological data. In this embodiment, the forecast value calculation device 100 predicts the power demand in order to create a supply-demand balance plan for how to allocate the power generation amount in a power plant whose output can be controlled, such as a nuclear power plant or a thermal power plant.

[0018] Therefore, the power demand predicted by the predicted value calculation device 100 in this embodiment is not the total power demand (hereinafter also referred to as gross demand), which is the sum of the power consumed by power consumers, but the power demand (hereinafter also referred to as net demand) that should be borne by power plants that can control the output as described above.

[0019] A more detailed explanation will be given with reference to Figure 8. Figure 8 is a schematic diagram showing the breakdown of the total electricity demand (referred to as gross demand in Figure 8) by type of power generation. In other words, it is a diagram showing what types of power generation facilities are used to cover the gross demand. Figure 8 shows that the gross demand is mainly covered by electricity from solar power generation facilities, hydroelectric power plants, thermal power plants, and nuclear power plants.

[0020] The electricity demand met by power plants whose output can be controlled, such as nuclear power plants and thermal power plants, is net demand, and gross demand is determined by adding the electricity demand met by hydroelectric power plants and solar power generation facilities to this net demand.

[0021] The electricity demand covered by solar power generation facilities can be divided into PV self-consumption, PV surplus, and PV total purchase. The electricity demand for PV self-consumption and PV surplus corresponds to the amount of electricity generated by solar power generation facilities of customers who have contracts with electric power companies to purchase surplus electricity. The electricity demand for PV total purchase corresponds to the amount of electricity generated by solar power generation facilities of customers who have contracts with electric power companies to purchase all electricity.

[0022] Electricity demand for PV self-consumption represents electricity consumed by consumers who have signed a surplus purchase contract. For this reason, electricity demand for PV self-consumption does not appear at the sending end. Therefore, electricity demand excluding electricity demand for PV self-consumption from gross demand is also called sending end demand.

[0023] In addition, the demand for surplus PV electricity is equivalent to the amount of electricity remaining after subtracting the amount of electricity consumed by the consumers' own solar power generation facilities that have signed a surplus purchase contract.

[0024] The electricity demand for the full PV purchase is equivalent to the electricity generated by the solar power generation facilities of customers who have signed a full purchase contract.

[0025] Of course, the power demand predicted by the prediction value calculation device 100 may be not only the net demand, but also the sending end demand or the gross demand.

[0026] The weather data providing device 200 provides the forecasted value calculation device 100 with weather data that is used when the forecasted value calculation device 100 predicts the power demand.

[0027] In this embodiment, the meteorological data providing device 200 provides meteorological data on temperature, humidity, amount of solar radiation, and amount of precipitation for the past 50 years.

[0028] ==Hardware Configuration== Next, the hardware configuration of the forecasted value calculation device 100 and the weather data providing device 200 will be described with reference to FIGS.

[0029] <Prediction value calculation device> First, the hardware configuration of the predicted value calculation device 100 will be described.

[0030] The predicted value calculation device 100 includes a CPU (Central Processing Unit) 110 , a memory 120 , a communication device 130 , a storage device 140 , an input device 150 , an output device 160 , and a recording medium reading device 170 .

[0031] As shown in FIG. 4, the memory device 140 stores various programs and data such as a predicted value calculation device control program 710 executed by the predicted value calculation device 100, a first relational equation 600A, a second relational equation 600B, a third relational equation 600C, a weather data table 610, and a temperature correction table 620.

[0032] The predicted value calculation device control program 710 and various data stored in the storage device 140 are read into the memory 120 and executed or processed by the CPU 110, thereby implementing various functions of the predicted value calculation device 100.

[0033] Here, the storage device 140 is a non-volatile storage device such as a hard disk, a solid state drive (SSD), or a flash memory.

[0034] The weather data table 610 is a table that stores various weather data provided by the weather data providing device 200. The weather data table 610 is shown in Fig. 5. The weather data table 610 stores various weather data observation values ​​such as hourly temperature, humidity, amount of solar radiation, amount of precipitation, air pressure, and wind speed for the past several years (50 years in this embodiment). Of course, the weather data table 610 stores weather data for each target area for which the forecast value calculation device 100 forecasts the power demand.

[0035] The temperature correction table 620 is a table that stores correction coefficients for suppressing the effects of long-term climate change due to global warming, etc. The temperature correction table 620 is shown in Fig. 7. Fig. 6 shows how the temperature is corrected by the correction coefficients stored in the temperature correction table 620.

[0036] In this embodiment, the correction coefficient for each day from January 1st to December 31st is the difference between the average temperature for each year since 1970 and the present (latest year). In other words, by adding a correction coefficient determined according to the time of temperature observation to the observed value, it is possible to correct the temperature to what it would be if the observation time were the present (latest year). Of course, the correction coefficient may also be the ratio of the average temperature for each year since 1970 to the present (latest year). In this case, by multiplying the observed value by a correction coefficient according to the time of temperature observation, it is possible to correct the temperature to what it would be if the observation time were the present (latest year).

[0037] Note that the meteorological data used to correct the effects of long-term climate change due to global warming and the like is not limited to temperature, and may be other meteorological data such as precipitation and wind speed. In this case, the correction coefficient is determined according to the type of meteorological data and the time point at which the meteorological data was observed.

[0038] If the correction coefficient differs depending on the region for which the forecast value calculation device 100 forecasts the power demand, it is advisable to determine the correction coefficient for each region. In addition, the weather data table 610 may store weather data after the effects of long-term climate change have been corrected.

[0039] Next, the first relational expression 600A, the second relational expression 600B, and the third relational expression 600C will be described.

[0040] The first relational expression 600A, the second relational expression 600B, and the third relational expression 600C are relational expressions that express the relationship between predetermined types of weather data and the amount of electricity demand. In the following description, the first relational expression 600A, the second relational expression 600B, and the third relational expression 600C are collectively referred to as relational expression 600.

[0041] First, the first relational expression 600A will be described.

[0042] The first relational expression 600A represents the relationship between the temperature data, the humidity data, and the gross demand (total power demand). In this embodiment, the first relational expression 600A represents the relationship between the average temperature data for a day, the average humidity data for that day, and the gross demand for that day. The average temperature data and average humidity data for that day can be obtained by collecting the hourly temperature data and humidity data for that day stored in the weather data table 610 and calculating the average value. At this time, the temperature data can be corrected for long-term climate variations using a correction coefficient stored in the temperature correction table 620.

[0043] An example of the first relational expression 600A is shown in Equation (1).

[0044] lnD(i) = lnD0 + α1 × HMD(i) + α2×CDD(i)+α3×CDD(i) 2 + α4×HDD(i)+α5×HDD(i) 2 + α6×Week…(1) lnD(i) represents the natural logarithm of gross demand D on the i-th day. Additionally, lnD0 is the natural logarithm of the constant D0 determined by regression analysis. HMD is the daily average humidity, CDD is the cooling degree days, HDD is the heating degree days, and Week is data representing the day of the week and holidays. Additionally, α1 to α6 are coefficients determined by regression analysis.

[0045] Cooling degree days are calculated as "MAX(daily average temperature - A°C, 0)" and are the difference between the daily average temperature and A°C on days when the daily average temperature is A°C or higher. Similarly, heating degree days are calculated as "MAX(B°C - daily average temperature, 0)" and are the difference between the daily average temperature and B°C on days when the daily average temperature is B°C or lower.

[0046] The forecast value calculation device 100 performs a regression analysis using actual values ​​(temperature, humidity) of past weather data stored in a weather data table 610 shown in FIG. 5 and actual values ​​of power demand (not shown), and determines the constant D0 and coefficients α1 to α6 in equation (1) to obtain the first relational equation 600A.

[0047] When obtaining the first relational expression 600A by regression analysis, the predicted value calculation device 100 may perform the regression analysis using actual values ​​of meteorological data and power demand from a time period as close as possible to the power demand prediction period (i.e., closer to the present). For example, the regression analysis may be performed using actual values ​​of meteorological data and power demand from a recent past period (e.g., the most recent past year).

[0048] This embodiment allows the penetration rate and performance of various electric power equipment such as heating and cooling equipment, which affect the relationship between temperature and humidity and total electricity demand (i.e., the above-mentioned constant D0 and coefficients α1 to α6), to be closer to the situation expected for the forecast period (for example, the situation is different now than it was 10 years ago), making it possible to calculate a more accurate forecast value of total electricity demand.

[0049] Next, the second relational expression 600B will be described.

[0050] The second relational expression 600B represents the relationship between the solar radiation data and the amount of solar power generated by the solar power generation facility. In this embodiment, the second relational expression 600B represents the relationship between the average amount of solar radiation in one day and the amount of solar power generated on that day. An example of the second relational expression 600B is shown in Equation (2).

[0051] PV power generation (i) = (PV grid-connected equipment capacity (total capacity) + (1 - self-consumption rate) x PV grid-connected equipment capacity (surplus) x solar radiation (i) x power generation coefficient...(2) Here, PV power generation (i) represents the amount of solar power generation on the i-th day. PV power generation (i) includes the amount of power generation for PV self-consumption, PV surplus, and PV total purchase. The PV grid-connected equipment capacity (total) is a value indicating the power generation capacity such as the panel capacity of a solar power generation facility operating under a total purchase contract. The PV grid-connected equipment capacity (surplus) is a value indicating the power generation capacity such as the panel capacity of a solar power generation facility operating under a surplus power purchase contract. The self-consumption rate indicates the proportion of power that is self-consumed out of the amount of power generated by a solar power generation facility operating under a surplus power purchase contract. The solar radiation (i) is, for example, the total solar radiation on a horizontal surface on the i-th day. The power generation coefficient is a coefficient for calculating the amount of PV power generation.

[0052] As in the case of the first relational expression 600A, the forecasted value calculation device 100 performs a regression analysis using actual values ​​(solar radiation) of past weather data stored in the weather data table 610 shown in FIG. 5 and actual values ​​of solar power generation (not shown), and determines the power generation coefficient in equation (2), thereby obtaining the second relational expression 600B.

[0053] At this time, since the self-consumption rate changes over time, a value from as close to the forecast period for electricity demand as possible (i.e., a value from as close to the present as possible) is used.

[0054] Furthermore, when calculating the second relational expression 600B, the predicted value calculation device 100 may perform a regression analysis using actual values ​​of weather data (amount of solar radiation) from the past as close as possible to the prediction period of the power demand (i.e., the past closer to the present) and actual values ​​of the amount of solar power generation. For example, the regression analysis may be performed using actual values ​​of weather data from a specified period in the past (e.g., the past year) and actual values ​​of the amount of power demand. In this manner, the conditions such as the penetration rate and performance of solar power generation facilities, and the self-consumption rate, which affect the relationship between the amount of solar radiation and the amount of solar power generation, can be brought closer to the conditions expected for the prediction period, making it possible to calculate a more accurate predicted value of the amount of solar power generation.

[0055] Next, the third relational expression 600C will be described.

[0056] The third relational expression 600C expresses the relationship between the precipitation data and the amount of hydroelectric power generated by the hydroelectric power plant. The amount of hydroelectric power generated daily varies depending on the amount of water flowing into the power generation dam from the surrounding rivers on that day, but the amount of water flowing into the dam varies in a complex manner depending on the amount of precipitation on that day as well as past precipitation, the surrounding topography, and geology, so the third relational expression 600C is more complex than the first relational expression 600A and the second relational expression 600B. Examples of the third relational expression 600C are shown in equations (3) to (16). A diagram for explaining the third relational expression 600C is shown in FIG. 9.

[0057] Hydroelectric power generation (i) = α × Q(i) … (3) Here, hydroelectric power generation (i) represents the hydroelectric power generation on the i-th day. Q(i) represents the amount of water flowing into the dam on the i-th day. α represents the coefficient that represents the relationship between the amount of water Q(i) and hydroelectric power generation (i).

[0058] Q(i) is expressed as the following equation (4).

[0059] Q(i)=Q1(i)+Q2(i)+Q3(i)+Q4(i)+Q5(i) …(4) Here, Q1(i) and Q2(i) represent the amount of water that flows into the dam on the i-th day after rainfall in the forest surrounding the dam flows over the surface of the forest. Q3(i) represents the amount of water that is absorbed into the forest soil, emerges on the surface in a relatively short time, and flows into the dam on the i-th day. Q4(i) represents the amount of water that is absorbed into the forest soil, emerges on the surface over a longer period than Q3, and flows into the dam on the i-th day. Q5(i) represents the amount of water that is absorbed into the forest soil, emerges on the surface over an even longer period than Q4, and flows into the dam on the i-th day.

[0060] As shown in FIG. 9, Q1(i) to Q5(i) can be calculated as the amount of water flowing out from multiple tanks (four in the example shown in FIG. 9) arranged in series in the vertical direction.

[0061] First, we will explain Q1(i) and Q2(i). Q1(i) and Q2(i) can be calculated as the amount of water flowing out of the top tank, as shown in Figure 9 and equations (5) and (6).

[0062] Q1(i)=a1×I[S1(i)-Z1] …(5) Q2(i)=a2×I[S1(i)-Z2] …(6) Here, a1 and a2 are the outflow hole coefficients of the outflow holes formed on the side of the tank. Z1 and Z2 are the heights from the bottom of the tank to the outflow holes. S1(i) is the water depth in the tank on the i-th day. I[x] is a function such that I[x]=x when x>0 and I[x]=0 when x≦0.

[0063] In Figure 9 and equation (7), the amount of water absorbed by the soil from the forest is represented as g1(i).

[0064] g1(i) = b1 × S1(i) … (7) Here, b1 is the permeability coefficient of the outflow hole formed at the bottom of the top tank.

[0065] The water depth S1(i) of the tank described in equations (5) to (7) is expressed by the following equation (8).

[0066] S1(i)=S1(i-1)-{Q1(i-1)+Q2(i-1)+g1(i-1)} +P(i)-E(i) …(8) In other words, the tank's water depth S1(i) on day i can be calculated by subtracting the amount of water that flowed out during the previous day {Q1(i-1) + Q2(i-1) + g1(i-1)} from the tank's water depth S1(i-1) on the previous day, and adding the newly increased amount of water P(i) - E(i).

[0067] Here, P(i) represents the amount of precipitation on the i-th day, and E(i) represents the amount of water evaporated into the atmosphere.

[0068] Next, we will explain Q3(i). Q3(i) can be calculated as the amount of water flowing out from the second tank from the top, as shown in Figure 9 and equation (9).

[0069] Q3(i)=a3×I[S2(i)-Z3] …(9) Here, a3 is the outflow hole coefficient of the outflow hole formed on the side of the second tank, Z3 is the height from the bottom of the second tank to the outflow hole, and S2(i) is the water depth of this tank on the i-th day.

[0070] As shown in FIG. 9 and equation (10), the amount of water g1(i) that is absorbed by the soil from the forest and does not flow out but instead flows through the soil is represented by g2(i).

[0071] g2(i) = b2 × S2(i) … (10) Here, b2 is the seepage porosity coefficient of the outflow hole formed at the bottom of the second tank.

[0072] Moreover, the water depth S2(i) of the tank described in equations (9) and (10) is expressed as in the following equation (11).

[0073] S2(i)=S2(i-1)-{Q3(i-1)+g2(i-1)}+g1(i) …(11) In other words, the water depth S2(i) of the second tank on day i can be calculated by subtracting the amount of water that flowed out during the previous day {Q3(i-1)+g2(i-1)} from the water depth S2(i-1) of the tank on the previous day, and adding the newly increased amount of water g2(i).

[0074] Next, we will explain Q4(i). Q4(i) can be calculated as the amount of water flowing out from the third tank from the top, as shown in Figure 9 and equation (12).

[0075] Q4(i)=a4×I[S3(i)-Z4] …(12) Here, a4 is the outflow hole coefficient of the outflow hole formed on the side of the third tank, Z4 is the height from the bottom of the third tank to the outflow hole, and S3(i) is the water depth of this tank on the i-th day.

[0076] As shown in FIG. 9 and equation (13), the amount of water g2(i) flowing through the forest without flowing out and flowing directly through the soil is represented by g3(i).

[0077] g3(i) = b3 × S3(i) … (13) Here, b3 is the seepage porosity coefficient of the outflow hole formed at the bottom of the third tank.

[0078] Moreover, the water depth S3(i) of the tank described in equations (12) and (13) is expressed as in the following equation (14).

[0079] S3(i)=S3(i-1)-{Q4(i-1)+g3(i-1)}+g2(i) …(14) In other words, the water depth S3(i) of the third tank on day i can be calculated by subtracting the amount of water that flowed out during the previous day {Q4(i-1)+g3(i-1)} from the water depth S3(i-1) of the tank on the previous day, and adding the newly increased amount of water g2(i).

[0080] Next, we will explain Q5(i). Q5(i) can be calculated as the amount of water flowing out from the fourth tank from the top, as shown in Figure 9 and equation (15).

[0081] Q5(i)=a5×S4(i) …(15) where a5 is the outflow hole coefficient of the outflow hole formed in the side of the fourth tank, and S4(i) is the water depth of this tank on the i-th day.

[0082] Furthermore, the water depth S4(i) of the tank described in equation (15) is expressed as follows:

[0083] S4(i)=S4(i-1)-Q5(i-1)+g3(i) …(16) In other words, the water depth S4(i) of the fourth tank on day i can be calculated by subtracting the amount of water Q5(i-1) that flowed out during the previous day from the water depth S4(i-1) of the tank on the previous day, and adding the newly increased amount of water g3(i).

[0084] As in the case of the first relational expression 600A and the second relational expression 600B, the forecast value calculation device 100 performs a regression analysis using actual values ​​(precipitation) of past weather data stored in the weather data table 610 shown in FIG. 5 and actual values ​​of hydroelectric power generation (not shown), and determines various coefficients in equations (3) to (16), thereby deriving the third relational expression 600C.

[0085] When determining the third relational expression 600C, the forecast value calculation device 100 may perform a regression analysis using actual values ​​of meteorological data (amount of precipitation) from a time period as close as possible to the forecast period for the power demand (i.e., closer to the present) and actual values ​​of the amount of hydroelectric power generation. For example, the regression analysis may be performed using actual values ​​of meteorological data from a specific period in the most recent past (e.g., the past year) and actual values ​​of the amount of power demand. In this manner, conditions such as the state of forests and rivers and the performance of hydroelectric power plants, which affect the relationship between precipitation and the amount of hydroelectric power generation, can be brought closer to conditions expected for the forecast period, making it possible to calculate a more accurate forecast value for the amount of hydroelectric power generation.

[0086] Returning to FIG. 2, the predicted value calculation device control program 710 is a general term for programs for realizing the functions of the predicted value calculation device 100, and includes, for example, application programs and an OS (Operating System) that run on the predicted value calculation device 100, various libraries, etc.

[0087] The recording medium reading device 170 reads the predicted value calculation device control program 710 and various data recorded on a recording medium 800 such as an SD card, and stores them in the storage device 140 .

[0088] The communication device 130 transmits and receives the forecast value calculation device control program 710 and various data to and from the meteorological data providing device 200 and other computers (not shown) via the network 500 .

[0089] For example, the above-mentioned predicted value calculation device control program 710 and relational expression 600 can be stored in another computer, and the predicted value calculation device 100 can download the predicted value calculation device control program 710 and relational expression 600 from this computer.

[0090] The input device 150 includes various buttons and switches that accept commands and data input by the user, a keyboard, a touch sensor that detects the touch position on a touch panel display, an input interface such as a microphone, an acceleration sensor, a temperature sensor, a position detection sensor such as a GPS receiver or a compass, a camera, etc.

[0091] The output device 160 is an output user interface, such as a display device, a speaker, a vibrator, or lighting.

[0092] <Weather data providing device> Next, the meteorological data providing device 200 will be described.

[0093] 3, the meteorological data providing device 200 is configured to include a CPU 210, a memory 220, a communication device 230, a storage device 240, an input device 250, an output device 260, and a recording medium reading device 270. As described above, the hardware configuration of the meteorological data providing device 200 is basically the same as the hardware configuration of the forecast value calculation device 100. Therefore, a duplicated description of these hardware configurations will be omitted.

[0094] The storage device 240 stores the meteorological data providing device control program 720 executed by the meteorological data providing device 200, as well as various other programs and data.

[0095] The meteorological data providing device control program 720 and various data stored in the storage device 240 are read into the memory 220 and executed or processed by the CPU 210, thereby implementing various functions of the meteorological data providing device 200.

[0096] ==Functional Configuration== Fig. 10 shows a functional block diagram of a predicted value calculation device 100 according to this embodiment. The predicted value calculation device 100 has the functions of a relational expression storage unit 101, a prediction target period receiving unit 102, a target data calculation unit 103, a predicted value identification information receiving unit 104, and a predicted value output unit 105. These functions are realized by the hardware shown in Fig. 2 executing or processing a predicted value calculation device control program 710 according to this embodiment and various data.

[0097] <Relational equation storage section> The relational equation storage unit 101 stores relational equations that express the relationship between a predetermined type of weather data and the amount of power demand. In this embodiment, the relational equation storage unit 101 is embodied as a storage device 140. The relational equations that express the relationship between the predetermined type of weather data and the amount of power demand are embodied as a first relational equation 600A, a second relational equation 600B, and a third relational equation 600C.

[0098] <Forecast period reception section> The forecast period receiving unit 102 receives input of information specifying the forecast period for power demand. The operator can directly specify and input a specific forecast period, such as "January to December 2021" or "February to March 2021," or can select from a selection screen for the scenario expected period (forecast period) as shown in FIG. 11. In this case, for example, when the scenario expected period code "101" is selected, one year from the following April to March of the following year is specified.

[0099] When selecting the scenario expected period, in this embodiment, there are three options: "year," "2 months," and "month," but other periods such as "3 months" and "6 months" may also be used.

[0100] <Target Data Calculation Section> The target data calculation unit 103 calculates the power demand for each period by using a predetermined type of weather data for the same period of the past several years corresponding to the prediction target period and the relational expression 600.

[0101] For example, if the prediction target period is "April," the target data calculation unit 103 uses the weather data (temperature, humidity, solar radiation, precipitation) for April of each year for 51 years from 1970 to the latest year 2020, and the first relational expression 600A, the second relational expression 600B, and the third relational expression 600C to integrate the total power demand, photovoltaic power generation, and hydroelectric power generation for each day in April of each year for one month, and calculates the total total power demand, photovoltaic power generation, and hydroelectric power generation for one month in April of each year. Then, the target data calculation unit 103 calculates the power demand (net demand) for each April of each year for 51 years by subtracting the amount of photovoltaic power generation and the amount of hydroelectric power generation from the total power demand for each year.

[0102] The multiple electricity demands (net demands) calculated in this manner have variations according to the variations in the weather data for each year.

[0103] <Prediction value identification information reception unit> The predicted value specifying information receiving unit 104 receives input of predicted value specifying information for specifying the amount of power demand to be selected as a predicted value from the amount of power demand (April of each year) in each of the above-mentioned periods.

[0104] The predicted value specifying information may be information indicating, for example, a percentile value (for example, the 30th percentile value) of the power demand in each of the above-mentioned periods.

[0105] In this case, for example, as shown in Figure 13, the electricity demand for each April (each period) for 51 years from April 1970 to April 2020 is sorted in order of magnitude (for example, from smallest to largest) (Figure 13 shows only the first 6), and the electricity demand (or the electricity demand with the closest percentile value) whose position in the whole when counted from the smallest is the percentile value specified by the predicted value identification information (for example, the 30th percentile value) is identified as the predicted value.

[0106] Alternatively, the predicted value specifying information may be information indicating that the power demand closest to the average value of the power demand for each of the above-mentioned periods is to be set as the predicted value.

[0107] In this case, the average value of the power demand in April of each year for 51 years from April 1970 to April 2020 is compared with the power demand for each April of each year, and the power demand closest to the average value is identified as the predicted value. Note that Fig. 13 shows that the power demand closest to the average value is displayed with information indicating this (for example, a black circle) in the "Average value" column.

[0108] Alternatively, the predicted value specifying information may be information indicating that the most frequent value of the power demand in each of the above-mentioned periods is to be used as the predicted value.

[0109] In this case, the electricity demand for each April for 51 years from April 1970 to April 2020 is classified into multiple groups according to the value of the electricity demand, and the average value of the electricity demand within the group with the largest electricity demand is compared with each electricity demand within this group, and the electricity demand closest to this average value is identified as the predicted value.

[0110] The manner in which the power demand is classified into a plurality of groups is shown in Fig. 12. In Fig. 12, the group into which the largest amount of power demand is classified is labeled "mode."

[0111] Then, from among the power demands classified into the group indicated as the mode, the power demand closest to the average value of the power demands in this group is identified as the predicted value (mode).

[0112] FIG. 13 also shows that a "mode" column is provided, and information indicating that the amount of power demand identified as the mode (for example, a black circle mark) is displayed.

[0113] The operator can also directly specify the electricity demand for a specific year (for example, 2008) as a predicted value from the electricity demand for April of each of the 51 years.

[0114] Furthermore, the predicted value identification information receiving unit 104 can display a menu as shown in FIG. 14 so that the operator can select a method for selecting the power demand amount to be used as a predicted value.

[0115] <Prediction output section> The predicted value output unit 105 outputs the power demand identified by the predicted value identification information as a predicted value of the power demand for the prediction period.

[0116] This makes it possible to predict the power demand taking into account the variability of the meteorological data from year to year. That is, the power demand for each year is calculated using past weather data from several years that actually have variability, so the power demand values ​​reflect the variability of the meteorological data from year to year. Since the relational expression 600 used to calculate the power demand is the same, the calculated power demand takes into account the variability of the meteorological data.

[0117] By creating the relational expression 600 using past weather data and power demand as close as possible to the prediction target period, the power demand calculated using the relational expression 600 can be made closer to the power demand expected in the prediction target period. In this case, the power demand calculated by the target data calculation unit 103 can be regarded as the power demand taking into account the variation in the weather data in the prediction target period.

[0118] Then, by selecting the electricity demand for a year in which the trend in weather data variability is expected to be close to that of the forecast period from among the multiple electricity demands calculated from weather data from multiple years in the past, it is possible to obtain a forecast value for electricity demand that is in line with the predicted variability of weather data for the forecast period.

[0119] For example, if the temperature during the forecast period is expected to be higher than normal, the power demand can be calculated from the weather data of past years with similar trends. Alternatively, if it is difficult to predict the degree of variability in the weather data during the forecast period, multiple power demands calculated from the weather data of various past years can be obtained as forecast values, making it possible to know the degree to which the power demand during the forecast period may vary.

[0120] In addition, when the predicted value identification information is information indicating a percentile value, the predicted value output unit 105 compares the percentile value of the power demand for each period with the percentile value specified in the predicted value identification information, and outputs the power demand having the percentile value closest to the percentile value specified in the predicted value identification information as a predicted value.

[0121] In this manner, it is possible to obtain a forecast value of the power demand when it is assumed that the degree of variation in the power demand during the forecast period is a predetermined position (percentile value) in the whole. When the percentile value is 50%, the forecast value is the median.

[0122] In addition, when the predicted value identification information is information indicating that the power demand closest to the average value of the power demand for each of the above-mentioned periods is to be the predicted value, the predicted value output unit 105 compares the average value of the power demand for each period with the power demand for each period, and outputs the power demand closest to the average value as the predicted value.

[0123] In this manner, it is possible to obtain a predicted value of the power demand when it is assumed that the degree of variation in the power demand during the prediction period is average.

[0124] In addition, when the predicted value identification information indicates that the most frequent value of the power demand in each of the above-mentioned periods is to be the predicted value, the predicted value output unit 105 classifies the power demand in each period into a plurality of groups according to the value of the power demand, compares the average value of the power demand in the group into which the greatest power demand is classified with each power demand in that group, and outputs the power demand closest to this average value as the predicted value.

[0125] In this manner, it is possible to obtain a predicted value of the power demand in the prediction period assuming that the power demand is the most frequent value in the distribution of the variation.

[0126] Furthermore, when outputting the predicted value, the predicted value output unit 105 may also output the power demand for each period.

[0127] In this manner, it is possible to determine the predicted value while observing the overall variation in the power demand.

[0128] At this time, the predicted value output unit 105 may output the power demand for each of the above periods in chronological order together with information indicating the period (for example, "March to April 2010").

[0129] In this manner, it is possible to determine the predicted value while viewing past power demands in chronological order.

[0130] Furthermore, the predicted value output unit 105 may output the power demand for each of the above-mentioned periods together with information indicating the period, in order of the magnitude of the power demand.

[0131] In this manner, it is possible to determine the predicted value while viewing the past power demands in order of magnitude.

[0132] When outputting the predicted value, the predicted value output unit 105 may output a histogram displaying a value indicating the range of the power demand classified into each group, and the frequency of the power demand classified into each group, for each group obtained by classifying the power demand in each period according to the value of the power demand. Fig. 12 shows how the predicted value output unit 105 outputs the histogram.

[0133] In this manner, it is possible to determine the predicted value while observing the overall distribution state of the past power demand.

[0134] The predicted value of the power demand obtained as described above can be used to study the power generation plans of thermal power plants and nuclear power plants for the prediction period. In this case, for example, a power generation plan that minimizes the power generation cost for the prediction period is obtained by simulation using information such as fuel costs expected for the prediction period and power plant repair plans.

[0135] ==Process flow== Next, a control method for the predicted value calculation device 100 according to this embodiment will be described with reference to Fig. 15. Fig. 15 is a flowchart showing the procedure, and these steps are realized by the CPU 110 executing the predicted value calculation device control program 710 stored in the storage device 140 of the predicted value calculation device 100.

[0136] First, the forecast value calculation device 100 creates a relational expression 600 in advance using past weather data and power demand, and stores the relational expression 600 in the storage device 140. Then, the forecast value calculation device 100 accepts input of a forecast target period (S1000). At this time, the forecast value calculation device 100 displays a selection screen for a scenario assumed period (forecast target period) such as that shown in FIG. 11, for example, and prompts the operator to input the forecast target period. At this time, for example, if the scenario assumed period code "101" is selected, one year from the following April to March of the following year is specified as the forecast target period.

[0137] Next, the forecast value calculation device 100 calculates the power demand for each period by using a predetermined type of weather data for the same period over the past several years corresponding to the forecast target period and the relational expression 600 (S1010). As described above, the forecast value calculation device 100 calculates the gross demand (total power demand), the amount of solar power generation, and the amount of hydroelectric power generation for each period, and calculates the net demand (power demand) for each period by subtracting the amount of solar power generation and the amount of hydroelectric power generation from the gross demand for each period.

[0138] Then, the forecast value calculation device 100 receives input of forecast value specifying information that specifies the amount of power demand to be selected as a forecast value from the net demand calculated for each period, that is, a forecast scenario (S1020).

[0139] Then, the forecast value calculation device 100 outputs the power demand corresponding to the forecast scenario as a forecast value of the power demand for the forecast period (S1030).

[0140] Thereafter, if the operator inputs another forecast scenario, the processes of S1020 and S1030 are repeated again (S1040). With this embodiment, it becomes possible to consider the forecast value of the power demand during the forecast target period while assuming various variations in the weather data.

[0141] It should be noted that in S1040, if the operator selects to end the process, the predicted value calculation device 100 ends the process.

[0142] The predicted value calculation device 100, the control method and the program for the predicted value calculation device 100 according to this embodiment have been described above, and this embodiment makes it possible to predict the power demand taking into account the variability of the weather data from year to year. That is, first, the power demand for each year is calculated using weather data from several years in the past that actually has variability, so the values ​​of the power demand reflect the variability of the weather data from year to year. Then, the relational expression 600 used to calculate the power demand is the same, so the calculated power demand takes into account the variability of the weather data.

[0143] By creating the relational expression 600 using past weather data and power demand as close as possible to the prediction target period, the power demand calculated using the relational expression 600 can be made closer to the power demand expected in the prediction target period. In this case, the power demand calculated by the target data calculation unit 103 can be regarded as the power demand taking into account the variation in the weather data in the prediction target period.

[0144] Then, by selecting the electricity demand for a year in which the trend in weather data variability is expected to be close to that of the forecast period from among the multiple electricity demands calculated from weather data from multiple years in the past, it is possible to obtain a forecast value for electricity demand that is in line with the predicted variability of weather data for the forecast period.

[0145] For example, if the temperature during the forecast period is expected to be higher than normal, the power demand can be calculated from the weather data of past years with similar trends. Alternatively, if it is difficult to predict the degree of variability in the weather data during the forecast period, multiple power demands calculated from the weather data of various past years can be obtained as forecast values, making it possible to know the degree to which the power demand during the forecast period may vary.

[0146] The above-described embodiment is for the purpose of facilitating understanding of the present invention, and is not intended to limit the present invention. The present invention may be modified or improved without departing from the spirit of the present invention, and equivalents thereof are also included in the present invention.

[0147] For example, in the above embodiment, the net demand is calculated by subtracting the amount of photovoltaic power generation and the amount of hydroelectric power generation from the gross demand, but this is just one example, and the net demand may be calculated by subtracting the amount of photovoltaic power generation, the amount of hydroelectric power generation, and the amount of wind power generation from the gross demand. In this case, a relational equation is generated that expresses the relationship between the amount of wind power generation and meteorological data such as wind speed that affects the amount of wind power generation, and the amount of wind power generation is calculated by inputting predetermined meteorological data such as wind speed into this relational equation.

[0148] In addition, in the above embodiment, the forecast value calculation device 100 acquires weather data in advance from the weather data providing device 200 and stores it in the weather data table 610, but it may also be configured to acquire the weather data required for the calculation from the weather data providing device 200 each time a forecast value is calculated.

[0149] In addition to electricity demand, the predicted value calculation device 100 can calculate predicted values ​​for various target data that fluctuate due to the influence of weather data, such as clothing sales, the number of traffic accidents, the number of spectators at a sporting event, and the number of patients with a particular disease.

[0150] Furthermore, the predicted value calculation device 100 can use various factor data that affect the target data other than meteorological data to obtain a predicted value of the target data that varies under the influence of these factor data.

[0151] Examples of such combinations of factor data and target data include the amount of profit for a given period (e.g., one week) that fluctuates in response to electricity market prices, the amount of sales of a specific product that fluctuates in response to the number of employees in a specific field such as agriculture and forestry, and the number of lightning strikes and power outages.

[0152] In addition, when calculating the predicted value of the profit amount considering the variation of the electricity market price, the prediction value calculation device 100 uses the fourth relational expression 600D expressing the relationship between the weather data (temperature data, etc.) and the market price in addition to the above-mentioned first relational expression 600A to the third relational expression 600C. In this case, the target data calculation unit 103 calculates the market price of each period by using the weather data of the same period in the past corresponding to the prediction target period and the fourth relational expression 600D. Then, the target data calculation unit 103 calculates the profit amount of each period by inputting the power demand (net demand) of each period calculated separately using the first relational expression 600A to the third relational expression 600C and the market price of each period calculated using the fourth relational expression 600D to the power generation plan creation device 300 (not shown). The power generation plan creation device 300 is a computer communicably connected to the forecast value calculation device 100, and calculates the profit amount for each period by executing a program for calculating the profit amount for each period from the net demand for each period and the market price for each period. With this embodiment, it is possible to calculate a forecast value of the profit amount taking into account the fluctuation in the electricity market price that changes due to the influence of fluctuations in weather data.

[0153] In the above embodiment, the target data calculation unit 103 calculates the power demand for each period by using a predetermined type of weather data for the same period of the past several years corresponding to the prediction target period and the relational expression 600, but the past period corresponding to the prediction target period is not limited to the same period every year. For example, it may be the same period every month (e.g., the period from the 1st to the 10th of each month), the same period every day (e.g., the period from 9:00 to 12:00 every day), or the same period every week (e.g., the period from 10:00 to 12:00 every Wednesday).

[0154] In other words, the same past period corresponding to the prediction target period can be determined by specifying at what period the past factor data is to be used when the target data calculation unit 103 calculates the target data using the past factor data and the relational equation 600.

[0155] For example, suppose that the factor data is temperature data for each hour of each day and the target data is electricity demand for each hour of each day, and the period from 10:00 to 14:00 of the following day is specified as the prediction period.

[0156] At this time, for example, when information to calculate the power demand using temperature data for the same time period every day for the past month and relational expression 600 is input to the predicted value calculation device 100, the target data calculation unit 103 calculates the power demand for each period using temperature data for the same time period (10:00 to 14:00) every day for the past month corresponding to the prediction target period and relational expression 600. Alternatively, when information to calculate the power demand using temperature data for the same time period on the same day every year for the past 10 years and relational expression 600 is input, the target data calculation unit 103 calculates the power demand for each period using temperature data for the same time period (10:00 to 14:00) on the same day for the past 10 years corresponding to the prediction target period and relational expression 600.

[0157] The predicted value calculation device 100 may have a function of accepting a designation of the cycle in which past factor data is to be used to calculate the target data, or may have a configuration in which it is determined in advance that past factor data is to be used at a predetermined cycle (e.g., one year).

[0158] Alternatively, factor data calculated using the method described below may be used. As a calculation method, a larger number of sets of factor data than past records can be created by empirical simulation, which creates factor data based on past data, or simulation using a model estimated by regression analysis or the like. [Explanation of symbols]

[0159] 100 Prediction value calculation device 101 Relational equation storage section 102 Prediction Period Reception Department 103 Target Data Calculation Unit 104 Prediction value specific information reception unit 105 Prediction value output section 110 CPU 120 Memory 130 Communication Equipment 140 Storage device 150 Input Device 160 Output Device 170 Recording medium reader 200 Weather data providing device 300 Power generation planning device 500 Network 600 Relational Expressions 600A First Relation 600B Second Relation 600C Third Relation 600D Fourth Relation 610 Weather Data Table 620 Temperature Correction Table 710 Prediction value calculation device control program 720 Weather Data Providing Device Control Program 800 Recording media 1000 Prediction Calculation System

Claims

1. A prediction value calculation device for calculating a prediction value of target data that varies under the influence of a predetermined type of factor data, comprising: a relational expression storage unit that stores a relational expression that represents a relationship between the factor data and the target data; a prediction target period receiving unit that receives an input of information specifying a prediction target period for which the prediction value of the target data is to be calculated; a target data calculation unit that calculates target data for each of a plurality of past periods corresponding to the prediction target period by using the factor data for each of the past periods and the relational expression; a forecast value specification information receiving unit that receives input of forecast value specification information that specifies target data to be selected as the forecast value from the target data of each period; a forecast value output unit that outputs the target data identified by the forecast value identification information as a forecast value of the target data for the forecast target period; A prediction value calculation device comprising:

2. The prediction value calculation device according to claim 1 , the predicted value identification information is information indicating a percentile value of the target data for each of the periods; The predicted value output unit compares the percentile value of the target data for each period with the percentile value specified in the predicted value identification information, and outputs the target data having the percentile value closest to the percentile value specified in the predicted value identification information as the predicted value.

3. The prediction value calculation device according to claim 1 , the predicted value identification information is information indicating that the target data that is closest to an average value of the target data for each of the periods is to be the predicted value; The predicted value calculation device, wherein the predicted value output unit compares an average value of the target data with the target data for each period, and outputs the target data that is closest to the average value as the predicted value.

4. The prediction value calculation device according to claim 1 , the predicted value specification information is information indicating that the predicted value is a mode value of the target data for each of the periods; The predicted value output unit classifies the target data for each of the periods into a plurality of groups according to the values ​​of the target data, compares an average value of the target data in the group in which the most common target data is classified with each of the target data in the group, and outputs the target data that is closest to the average value as the predicted value.

5. The prediction value calculation device according to any one of claims 1 to 4, The predicted value output unit outputs target data for each of the periods when outputting the predicted value.

6. The prediction value calculation device according to claim 5 , The predicted value output unit outputs the target data for each of the periods in chronological order together with information indicating the period.

7. The prediction value calculation device according to claim 5 , The predicted value output unit outputs the target data for each of the periods together with information indicating the period, arranged in order of size of the target data.

8. The prediction value calculation device according to any one of claims 1 to 4, A prediction value calculation device in which, when outputting the prediction value, the prediction value output unit outputs, for each group obtained by classifying the target data for each of the periods according to the values ​​of the target data, a histogram displaying a value indicating the range of target data classified into each group and the frequency of the target data classified into each group.

9. The prediction value calculation device according to any one of claims 1 to 8, The predetermined type of factor data is meteorological data.

10. The prediction value calculation device according to claim 9, At least a part of the types of weather data among the predetermined types of weather data, A forecast value calculation device that corrects the observed values ​​of the weather data by performing calculations using a correction coefficient determined according to the type of weather data and the time point at which the weather data was observed, in order to mitigate the effects of long-term climate change.

11. A prediction value calculation device according to any one of claims 1 to 10, The target data is electricity demand.

12. A prediction value calculation device according to any one of claims 1 to 10, The target data is a profit amount.

13. The prediction value calculation device according to any one of claims 1 to 8, the predetermined type of factor data is meteorological data, The target data is an amount of electricity demand, The predetermined types of weather data include temperature data, humidity data, solar radiation data, and precipitation data, In the above relational expression, a first relational expression expressing a relationship between the temperature data, the humidity data, and a total power demand; a second relational expression expressing a relationship between the solar radiation data and the amount of solar power generated by the solar power generation facility; a third relational expression expressing a relationship between the precipitation data and the amount of hydroelectric power generated by the hydroelectric power plant; Includes: The target data calculation unit, a prediction value calculation device that calculates the power demand by subtracting the amount of solar power generation calculated using the solar radiation data and the second relational equation and the amount of hydroelectric power generation calculated using the precipitation data and the third relational equation from the total power demand calculated using the temperature data, the humidity data, and the first relational equation.

14. The prediction value calculation device according to any one of claims 1 to 8, the predetermined type of factor data is meteorological data, The target data is a profit amount, The predetermined types of weather data include temperature data, humidity data, solar radiation data, and precipitation data, In the above relational expression, a first relational expression expressing a relationship between the temperature data, the humidity data, and a total power demand; a second relational expression expressing a relationship between the solar radiation data and the amount of solar power generated by the solar power generation facility; a third relational expression expressing a relationship between the precipitation data and the amount of hydroelectric power generated by the hydroelectric power plant; a fourth relational expression expressing a relationship between the weather data and a market price; Includes: The target data calculation unit, calculating an amount of electricity demand by subtracting the amount of solar power generation calculated using the solar radiation data and the second relational expression and the amount of hydroelectric power generation calculated using the precipitation data and the third relational expression from the total amount of electricity demand calculated using the temperature data, the humidity data, and the first relational expression, and calculating the market price using the weather data and the fourth relational expression; A forecast value calculation device that inputs the calculated amount of power demand and the market price into a power generation plan creation device to calculate the profit amount.

15. The prediction value calculation device according to claim 13, A prediction value calculation device, wherein the temperature data is corrected by performing a calculation on the observed value of the temperature data using a correction coefficient determined according to the observation time point of the observed value in order to reduce the effects of long-term climate change.

16. A method for controlling a predicted value calculation device that calculates a predicted value of target data that varies under the influence of a predetermined type of factor data, comprising the steps of: The predicted value calculation device, storing a relational expression representing a relationship between the factor data and the target data; Accepting input of information designating a prediction target period for calculating a prediction value of the target data; calculating target data for each of a plurality of past periods corresponding to the prediction target period using the factor data and the relational expression; Accepting input of forecast value identification information that identifies target data to be selected as the forecast value from among the target data for each of the periods; A control method for a forecast value calculation device, which outputs the target data specified by the forecast value specifying information as a forecast value of the target data for the forecast target period.

17. On the computer, A function of storing a relational expression expressing a relationship between a predetermined type of factor data and target data that varies under the influence of the factor data; A function of receiving input of information specifying a prediction target period for calculating a prediction value of the target data; a function of calculating target data for each of a plurality of past periods corresponding to the prediction target period using the factor data and the relational expression; A function of receiving input of prediction value identification information that identifies target data to be selected as the prediction value from the target data of each of the periods; a function of outputting the target data identified by the prediction value identification information as a prediction value of the target data for the prediction target period; A program to achieve this.

Images