WO2022024960A1 - 日射量補正方法、日射量補正装置、コンピュータプログラム、モデル、モデル生成方法及びモデル提供方法 - Google Patents
日射量補正方法、日射量補正装置、コンピュータプログラム、モデル、モデル生成方法及びモデル提供方法 Download PDFInfo
- Publication number
- WO2022024960A1 WO2022024960A1 PCT/JP2021/027457 JP2021027457W WO2022024960A1 WO 2022024960 A1 WO2022024960 A1 WO 2022024960A1 JP 2021027457 W JP2021027457 W JP 2021027457W WO 2022024960 A1 WO2022024960 A1 WO 2022024960A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- solar radiation
- radiation amount
- model
- amount
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/12—Sunshine duration recorders
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/10—Devices for predicting weather conditions
Definitions
- the present invention relates to a solar radiation correction method, a solar radiation correction device, a computer program, a model, a model generation method, and a model provision method.
- Photovoltaic power generation which is one of the renewable energies, has the advantage of not generating greenhouse gases, and its widespread use is expected. It is necessary to predict the amount of solar radiation for the construction plan and actual operation of the solar power plant.
- Patent Document 1 at a point where the amount of solar radiation is actually measured, the measured value of the amount of solar radiation is compared with the estimated value of the amount of solar radiation obtained by a predetermined formula using the meteorological satellite image data related to the point.
- a system that adjusts the parameters of the calculation formula so that the error between the estimated value and the measured value is less than the allowable amount, and estimates the amount of solar radiation at any point near the measured point by the calculation formula using the adjusted parameters is disclosed. Has been done.
- the present invention has been made in view of such circumstances, and is a solar radiation correction method, a solar radiation correction device, a computer program, a model, a model generation method, and a solar radiation correction method capable of providing solar radiation amount data having a small deviation from the measured value.
- the purpose is to provide a model provision method.
- the solar radiation amount correction method acquires solar radiation amount data, acquires meteorological data, and corrects the acquired solar radiation amount data based on the acquired meteorological data.
- the solar radiation correction device is based on the first acquisition unit for acquiring solar radiation amount data, the second acquisition unit for acquiring meteorological data, and the meteorological data acquired by the second acquisition unit. It is provided with a correction unit for correcting the solar radiation amount data acquired by the first acquisition unit.
- the computer program according to the embodiment of the present invention causes a computer to acquire solar radiation amount data, acquire meteorological data, and correct the acquired solar radiation amount data based on the acquired meteorological data.
- the model according to the embodiment of the present invention is generated by machine learning using the solar radiation amount data and the meteorological data as input variables and the corrected solar radiation amount data as output variables.
- the solar radiation amount data and the meteorological data are acquired, the corrected solar radiation amount data is acquired, the solar radiation amount data and the meteorological data are used as input variables, and the corrected solar radiation amount data is used as an input variable.
- the model providing method stores a plurality of different models generated by machine learning using the solar radiation amount data and the meteorological data as input variables and the corrected solar radiation amount data as output variables, and a plurality of them. It accepts the selection of the required weather information service provider from the meteorological information service providers of the above, and provides a model corresponding to the selected meteorological information service provider from the plurality of different models.
- FIG. 1 is a schematic diagram showing an example of the configuration of the solar radiation amount correction system of the present embodiment.
- the solar radiation correction system includes a server 50 as a solar radiation correction device.
- the server 50 can access the meteorological data DB 61 and the solar radiation data DB 62.
- the server 50 is connected to a communication network 1 such as the Internet.
- the management server 100 of each of the plurality of weather information service providers is connected to the communication network 1.
- Each management server 100 can access the solar radiation amount DB 110.
- the communication network 1 is connected to a plurality of weather information users or terminal devices 10 of each user.
- the terminal device 10 is, for example, a personal computer (PC), a tablet, a smartphone, or the like.
- the meteorological information service provider is a company that provides a meteorological database globally, and provides users with solar radiation amount data using its own prediction model.
- the solar radiation amount DB 110 the solar radiation amount data generated by using the prediction model is recorded.
- the management server 100 can access the solar radiation amount DB 110, read the required solar radiation amount data, and provide the read solar radiation amount data to the user.
- a meteorological information user is a company that uses solar radiation data and related information related to the solar radiation data.
- the weather information user can receive the solar radiation amount data and the related information related to the solar radiation amount data from the server 50 by using the terminal device 10.
- FIG. 2 is a block diagram showing an example of the configuration of the server 50.
- the server 50 includes a control unit 51 that controls the entire server, an insolation amount data acquisition unit 52, a weather data acquisition unit 53, a storage unit 54, a first correction unit 55, an insolation amount-related information providing unit 56, and a second correction unit 57. And a learning processing unit 58.
- the first correction unit 55 and the second correction unit 57 are also collectively referred to as a correction unit.
- the second correction unit 57 includes an input data generation unit 571 and a model unit 572.
- the model unit 572 is composed of a semiconductor memory, a hard disk, or the like, and stores a model (learned model) generated by machine learning.
- the learning processing unit 58 includes a learning data generation unit 581, a model unit 582, and a parameter determination unit 583.
- the model unit 582 is composed of a semiconductor memory, a hard disk, or the like, and stores a model before machine learning.
- the model generated by performing machine learning in the learning processing unit 58 can be stored in the model unit 572.
- the model unit 582 may store a model in the middle of machine learning, a model for re-learning, and a model that has been trained.
- the learning processing unit 58 is not an indispensable configuration, and may be provided in another server that performs learning processing.
- the server 50 may be configured by a plurality of servers, and the functions may be distributed among the servers.
- the first correction unit 55 and the second correction unit 57 may be configured to include only one of them.
- the control unit 51 can be configured by a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and the like.
- the storage unit 54 is composed of a semiconductor memory, a hard disk, or the like, and can store required data such as data obtained as a result of processing in the server 50.
- the solar radiation amount data acquisition unit 52 has a communication function on the communication network 1 and can acquire solar radiation amount data from the management server 100 of the weather information service provider.
- FIG. 3 is a schematic diagram showing an example of solar radiation amount data.
- the solar radiation amount data can be hourly data of the solar radiation amount over a required period at one or a plurality of points.
- the amount of solar radiation is the amount of radiant energy that a unit area receives from the sun in a unit time, and the unit can be expressed as [kWh / m 2 / h].
- the amount of solar radiation can be, for example, the amount of total solar radiation on a horizontal plane obtained by measuring the amount of solar radiation from the entire sky.
- the required period may be either a past period, a past-to-future period, or a future period. As shown in FIG.
- the solar radiation amount data includes the solar radiation amount data D1 at the point L1, the solar radiation amount data D2 at the point L2, the solar radiation amount data D3 at the point L3, the solar radiation amount data D4 at the point L4, and the point L5. It can be configured as the solar radiation amount data D5 at, the solar radiation amount data D6 at the point L6, and so on. Further, the points L2, L4, L5 and the like are also observation points of the amount of solar radiation, and the measured value of the amount of solar radiation can also be acquired.
- the meteorological data acquisition unit 53 can access the meteorological data DB 61 and acquire meteorological data from the meteorological data DB 61.
- the meteorological data for example, data observed at a meteorological observatory, data provided by a meteorological forecasting company, or the like can be used.
- the meteorological data can be stored in the meteorological data DB 61 in advance. Further, the meteorological data may be acquired at any time and stored in the meteorological data DB 61.
- Meteorological data can include physical quantities caused by clouds and physical quantities that have a correlation with clouds, and includes not only directly observable data but also processed data of observed data.
- FIG. 4 is an explanatory diagram showing an example of meteorological data.
- the meteorological data includes, for example, temperature, temperature difference, humidity difference, cloud amount, humidity logarithm, humidity number, sensible temperature difference, pressure pressure, pressure constant, direct solar radiation amount, scattered solar radiation amount, and the like. Is done.
- the meteorological data may be on a daily basis or on an hourly basis.
- the temperature can be the daily average temperature data.
- the temperature difference can be the temperature difference data within one day.
- the humidity difference can be the humidity difference data within one day.
- the cloud cover is the ratio of clouds to the entire sky, and can be used as daily average cloud cover data.
- the logarithm of humidity can be the logarithm data of the daily average humidity represented by ⁇ 1 / log (humidity) ⁇ .
- the dew point can be the daily average data expressed by (temperature-dew point temperature).
- the dew point may be the difference between the temperature at a certain latitude and altitude and the dew point temperature at the latitude and altitude.
- the physical characteristic value which can calculate the dew point may be used instead of the dew point.
- the sensible temperature difference difference in Heat Index, difference in heat index
- the barometric pressure can be the barometric pressure data at the altitude level of the daily average.
- the predetermined values a1 and the constant a2 are not limited to the examples in the figure.
- the direct solar radiation amount is the amount of solar radiation from only the range of the sun's photosphere in the entire sky, and can be used as the daily integrated direct solar radiation amount data.
- the scattered solar radiation amount is the amount of solar radiation from a range other than the photosphere of the sun in the entire sky, and can be used as daily integrated scattered solar radiation amount data.
- the solar radiation amount correction may be performed only by the first correction unit 55, or the solar radiation amount correction may be performed only by the second correction unit 57, and the first correction unit 55 and the second correction unit 57 may be performed. Insolation correction may be performed in combination.
- the first correction unit 55 can correct the acquired solar radiation amount data based on the acquired meteorological data.
- the acquired solar radiation data does not fully consider the effects of clouds, rain, fog, etc. on the solar radiation, but by correcting the acquired solar radiation data using meteorological data, clouds, rain, etc. It is possible to consider the influence caused by fog and provide solar radiation amount data with a small deviation from the measured value. Further, when the acquired solar radiation amount data is a future prediction, the future solar radiation amount data can be corrected by acquiring the future meteorological data.
- FIG. 5 is an explanatory diagram showing an example of the solar radiation amount correction method by the first correction unit 55.
- the correction method C1 can be performed as appropriate.
- the solar radiation amount data and the meteorological data provided by the weather information service provider may include an error due to a time lag with the measured value, which affects the accuracy of the solar radiation amount correction other than the correction method C1.
- the first correction unit 55 acquires the solar radiation amount data before correction
- the first correction unit 55 corrects the time lag of the solar radiation amount data.
- the first correction unit 55 weighted averages the unit time data (for example, hourly data) of the amount of solar radiation, and shifts the time axis by the required time.
- the required time can be, for example, 20 minutes before, but is not limited to this.
- the weight of the weighted average is determined according to, but is not limited to, the ratio of the difference from the hourly data before and after the shift time.
- the first correction unit 55 corrects the solar radiation amount data before the correction.
- the barometric pressure constant is ⁇ 112- [atmospheric pressure at sea level] / 10 ⁇
- the barometric pressure constant threshold value can be, for example, 1. The lower the sea level pressure, the larger the pressure constant.
- the atmospheric pressure constant is equal to or higher than the atmospheric pressure constant threshold value, the surface portion becomes a low pressure system, air flows in from a portion having a higher atmospheric pressure, and as a result, an updraft is generated and clouds are generated. Therefore, it is necessary to consider the influence of clouds, and it is considered necessary to correct the acquired solar radiation amount data.
- the correction of the amount of solar radiation data can be corrected by dividing the total amount of solar radiation in the horizontal plane by the power of a (atmospheric pressure constant).
- the exponent a > 1.
- the first correction unit 55 corrects the solar radiation amount data before the correction.
- the atmospheric pressure at the altitude above sea level is below the atmospheric pressure threshold (for example, 101 kPa)
- the surface portion becomes a low pressure
- the coefficient b may be appropriately changed according to the supply source of the solar radiation amount data.
- the first correction unit 55 corrects the solar radiation amount data before the correction.
- the humidity is low because the dew point is large, and it is considered that the influence of clouds does not need to be considered.
- the humidity is high because the dew point is small, and it is considered that rain or fog is occurring in the area, and the acquired solar radiation amount data needs to be corrected. Be done.
- the above-mentioned correction methods C1 to C4 may be performed at the same time if the correction conditions are satisfied, but only a part of the correction methods may be performed.
- the correction methods C1 and C2 have a greater effect of reducing the deviation from the measured values than the correction methods C3 and C4, the correction methods C1 and C2 may be performed, and the correction method C1 may be used to further improve the deviation.
- ⁇ C4 may be performed.
- the second correction unit 57 can correct the acquired solar radiation amount data based on the acquired meteorological data.
- the acquired solar radiation data does not fully consider the effects of clouds, rain, fog, etc. on the solar radiation, but by correcting the acquired solar radiation data using meteorological data, clouds, rain, etc. It is possible to consider the influence caused by fog and provide solar radiation amount data with a small deviation from the measured value. Further, when the acquired solar radiation amount data is a future prediction, the future solar radiation amount data can be corrected by acquiring the future meteorological data.
- the second correction unit 57 includes an input data generation unit 571 and a model unit 572.
- the model part and the model are synonymous, but the model part also has the meaning of an aggregate of a plurality of models.
- the model unit 572 is a machine-learned model, and can be configured by, for example, a neural network.
- the model unit 572 formulates the relationship between the input variable and the output variable, and is generated by machine learning using the solar radiation amount data and the meteorological data as the input variable and the corrected solar radiation amount data as the output variable. ..
- the model unit 572 can output the corrected solar radiation amount data.
- FIG. 6 is a schematic diagram showing a first example of the daily integrated solar radiation amount correction method by the second correction unit 57.
- the model 572a is composed of a neural network, includes an input layer, an intermediate layer, and an output layer, and the parameters (weight, bias, etc.) of the neural network are determined by machine learning.
- the input data generation unit 571 is an input including the daily integrated horizontal plane total solar radiation amount as solar radiation amount data, the daily average humidity number as meteorological data, the daily average ⁇ 1 / log (humidity) ⁇ , and the daily average temperature. Data is generated and input to model 572a.
- the model 572a can output the corrected solar radiation amount data (daily integrated horizontal plane total solar radiation amount) as output data. As a result, the influence caused by clouds can be taken into consideration, and it is possible to provide solar radiation amount data having a small deviation from the measured value.
- the corrected solar radiation amount data can be obtained.
- the dew point data as an input variable of the model 572a, it is possible to consider the influence of rain, fog, etc. on the amount of solar radiation.
- the corrected solar radiation amount data can be obtained.
- the logarithmic data of the humidity data as the input variable of the model 572a
- the influence of the cloud on the amount of solar radiation can be considered.
- the corrected solar radiation amount data can be obtained by inputting the cloud amount data into the model 572a.
- the cloud cover data as an input variable of the model 572a
- the influence of the cloud on the amount of solar radiation can be considered.
- the solar radiation amount data input to the model 572a can be uncorrected solar radiation amount data acquired from the weather information service provider, but is not limited to this.
- the corrected solar radiation amount data to which all or part of the correction methods C1 to C4 may be applied by the first correction unit 55 described above may be used. That is, the corrected solar radiation amount data is included as an input variable of the model 572a.
- the correction of the solar radiation amount data may be performed based on the meteorological data, or the solar radiation amount data may be weighted averaged in units of time to correct the time lag.
- the model 572a further corrects the corrected solar radiation amount data, so that the accuracy of the correction can be further improved.
- FIG. 7 is a schematic diagram showing the relationship between the input data and the output data.
- the time information t1, t2, t3, t4, t5, ... May be a time unit or a day unit.
- t1 to t24 represent 24 hours
- t1 to t30 represent 30 days.
- the corrected solar radiation amount data dt1'in the time information t1 is output.
- the correction period (required period) to be corrected when correcting the solar radiation amount data may be several hours, several days, several months, or several years.
- the amendment period may be a past period, a future period, or a period from the past to the future.
- the model 572a can output the prediction data of the solar radiation amount data after correction.
- future forecast data may be used as the meteorological data.
- the solar radiation amount data is in the hour unit and the meteorological data is in the daily unit, the meteorological data having the same value in the daily unit may be assigned to each of the time information t1 to t24.
- FIG. 8 is a schematic diagram showing a second example of the daily integrated solar radiation amount correction method by the second correction unit 57.
- the difference from the first example shown in FIG. 6 is that the meteorological data in the input data includes the humidity difference within one day, the daily integrated scattered solar radiation amount, the daily integrated direct solar radiation amount, the daily average cloud amount, and the daily temperature difference, 1 The difference in sensible temperature during the day is included.
- the meteorological data to be input may be a part of the daily humidity difference, the daily integrated scattered solar radiation amount, the daily integrated direct solar radiation amount, the daily average cloud amount, the daily temperature difference, and the daily sensible temperature difference. ..
- the input meteorological data includes a combination of temperature data and dew point temperature data, a combination of temperature data and wind-cooled temperature data, a combination of temperature data and sensible temperature data, or temperature data and heat index data. Combinations may be included. As a result, it is possible to consider further effects caused by clouds and the like, and it is possible to provide solar radiation amount data in which the deviation from the measured value is further improved.
- FIG. 9 is a schematic diagram showing a third example of the daily integrated solar radiation amount correction method by the second correction unit 57.
- the difference from the first example shown in FIG. 6 is that instead of outputting the corrected solar radiation amount data (daily integrated horizontal plane total solar radiation amount) itself, the daily correction amount for the uncorrected solar radiation amount data ( The point is to output the difference) obtained by subtracting the amount of solar radiation before correction from the amount of solar radiation after correction.
- the corrected solar radiation amount data can be recorded in the solar radiation amount data DB 62.
- model generation method (learning method) will be explained.
- the learning processing unit 58 includes a learning data generation unit 581, a model unit 582, and a parameter determination unit 583.
- the model unit 582 stores the model before machine learning.
- the model unit 582 may store a model during machine learning or scheduled to be relearned.
- the model can be composed of, for example, a neural network including an input layer, an intermediate layer and an output layer. In the model before machine learning, the parameters of the neural network are undecided.
- FIG. 10 is a schematic diagram showing an example of a method for generating the model 582a.
- the learning processing unit 58 can generate a model in which the uncorrected solar radiation amount data and the meteorological data are input variables and the corrected solar radiation amount data is an output variable.
- the learning data generation unit 581 uses the daily integrated horizontal plane total solar radiation as solar radiation data, the daily average dew point as meteorological data, the daily average ⁇ 1 / log (humidity) ⁇ , and the daily average.
- Data for training including an identifier that identifies the source of the temperature and solar radiation data is generated and input to the model 582a.
- the parameter determination unit 583 is a neural network so that the difference between the learning output data (daily integrated horizontal plane total solar radiation amount) output by the model 582a and the measured value of the daily integrated horizontal plane total solar radiation amount as teacher data becomes small. Parameters (weight wij, bias blm) are adjusted. For the difference between the training output data and the teacher data, the objective function (or loss function) may be determined in advance, and the parameters may be adjusted so that the value of the objective function becomes smaller.
- the parameter determination unit 583 can determine the parameter when the difference between the learning output data and the teacher data is the smallest.
- the model generated by the generation method shown in FIG. 10 corresponds to the model shown in FIG.
- the model generation method shown in FIG. 8 uses the daily humidity difference, daily integrated scattered solar radiation amount, daily integrated direct solar radiation amount, daily average cloud cover, daily temperature difference, and daily sensible temperature as input data for learning. Just add the difference.
- the trained model may be re-learned, but it may be re-learned from the beginning using the initial parameters. For example, when the model is trained for each specific trigger, it can be trained from the beginning using the past training data. Specific triggers include, for example, when a meteorological satellite is updated.
- Examples of the machine learning method include linear regression, ridge regression, Lasso regression, elastic net, etc. as linear models, and k-neighborhood method, regression tree, random forest, etc. as non-linear models, in addition to neural networks. , Gradient boosting, support vector regression, projection tracking regression, Gaussian process regression, etc. A non-linear regression model is preferable to a linear regression model.
- the model generated by machine learning does not have to be uniform. Since the solar radiation amount data provided by the weather information service provider is generated by the weather information service provider using its own prediction model, the data characteristics of the solar radiation amount data predicted according to the prediction model used. May also be different. Therefore, it is possible to generate a model unique to each weather information service provider.
- FIG. 11 is an explanatory diagram showing the correspondence between the weather information service provider and the machine-learned model.
- the model 572a can be used for the solar radiation amount data provided by the company A
- the model 572b can be used for the solar radiation amount data provided by the company B
- Model 572c can be used for the solar radiation amount data provided by Company C.
- the models 572a, 572b, 572c, ... Can be stored in the model unit 572 in association with the weather information service provider.
- the second correction unit 57 uses the model corresponding to the selected meteorological information service provider to obtain the solar radiation.
- Quantitative data can be corrected.
- the solar radiation amount data provided by company A can be corrected by using the model 572a for company A.
- the solar radiation amount data provided by company B may be corrected using the model 572b for company B
- the solar radiation amount data provided by company C may be corrected using the model 572c for company C. can.
- the model is generated and used in association with the weather information service provider, but the model may be generated and used in association with the prediction model used for the prediction of the solar radiation amount data. ..
- FIG. 12 is a flowchart showing a first example of the solar radiation amount correction process by the server 50.
- the control unit 51 acquires the total amount of horizontal solar radiation for the required period (correction target period) at the required point from the weather information service provider (S11), and the weather at the required point or its vicinity for the required period. Acquire data (S12). If the meteorological data does not exist at the required time point, the meteorological data at the point closest to the required point or the point where the meteorological condition is close to the required point may be acquired.
- the control unit 51 corrects the time lag of the total solar radiation in the horizontal plane according to the correction method C1 (S13), and determines whether or not the barometric pressure constant is equal to or higher than the barometric pressure constant threshold value (S14).
- the barometric pressure constant is equal to or higher than the barometric pressure constant threshold value (YES in S14)
- the total solar radiation on the horizontal plane is divided by a predetermined number and corrected as in the correction method C2 (S15).
- the control unit 51 performs the process of step S16 described later.
- the control unit 51 determines whether or not the sea level pressure is below the barometric pressure threshold value (S16), and when the sea level level pressure is below the barometric pressure threshold value (YES in S16), the entire horizontal plane is as described in the correction method C3. It is corrected by multiplying the amount of solar radiation by a predetermined number b (S17). When the atmospheric pressure at the altitude above sea level is not less than the atmospheric pressure threshold value (NO in S16), the control unit 51 performs the process of step S18 described later.
- the control unit 51 determines whether or not the dew point is less than the dew point threshold value (S18), and if the dew point is less than the dew point threshold value (YES in S18), the horizontal plane all-sun solar radiation as in the correction method C3. The amount is multiplied by a predetermined number c to correct (S19), the corrected horizontal total solar radiation amount is output (S20), and the process is terminated.
- the control unit 51 performs the process of step S20.
- steps S13, steps S14 to S15, steps S16 to S17, and steps S18 to S19 may be performed entirely or only partially.
- FIG. 13 is a flowchart showing a second example of the solar radiation amount correction process by the server 50.
- FIG. 13 shows a correction process using the second correction unit 57.
- the control unit 51 accepts the selection of the weather information service provider (S31), and selects the machine-learned model corresponding to the selected weather information service provider from the model unit 572 (S32).
- the control unit 51 acquires the total amount of solar radiation on the horizontal plane for the required period (correction target period) at the required point from the selected meteorological information service provider (S33), and the required point at the required point or its vicinity. Acquire the meteorological data of the period (S34).
- the control unit 51 inputs the horizontal plane total solar radiation amount and meteorological data into the selected model, corrects the horizontal plane total solar radiation amount (S35), outputs the corrected horizontal plane total solar radiation amount (S36), and performs processing. finish.
- the solar radiation amount correction may be performed by performing only the processing shown in FIG. 12, or the solar radiation amount correction may be performed by performing only the processing shown in FIG. Further, both solar radiation amount correction processings shown in FIGS. 12 and 13 may be performed.
- FIG. 14 is a flowchart showing an example of model generation processing by the server 50.
- the control unit 51 reads a model for formulating the relationship between the input variable and the output variable from the model unit 582 (S41), and sets initial values of parameters (weight, bias, etc.) (S42).
- the control unit 51 acquires the horizontal total solar radiation amount of the target period at the target point from the weather information service provider (S43), and acquires the meteorological data of the target period at the target point (S44).
- the target point can be a point where the measured value can be obtained by observing the total amount of solar radiation on the horizontal plane, and a plurality of points are preferable.
- the control unit 51 acquires actual measurement data of the total solar radiation of the horizontal plane during the target period at the target point (S45), inputs the total solar radiation of the horizontal plane and the meteorological data to the model as input variables, and outputs the horizontal plane of the model.
- the parameters weight, bias, etc. are adjusted so that the value of the objective function is minimized by using the total amount of solar radiation and the measured data (S46).
- the control unit 51 determines whether or not the value of the objective function is within the permissible range (S47), and if the value of the objective function is not within the permissible range (NO in S47), the processing after step S46 is continued. When the value of the objective function is within the allowable range (YES in S47), the control unit 51 stores the generated model in the model unit 572 (S48), and ends the process.
- FIG. 15 is an explanatory diagram showing an example of a correction result of the horizontal total solar radiation amount by the first correction unit 55.
- Three evaluation points are listed: Choshi, Tokyo, and Utsunomiya.
- the upper graph shows the monthly cumulative amount of solar radiation [kWh / m 2 / month], and shows the amount of solar radiation before correction, the amount of solar radiation after correction, and the measured values of the amount of solar radiation (Japan Meteorological Agency data).
- the lower table shows the amount of solar radiation before correction and the amount of solar radiation after correction on a yearly basis (2018, 2019) in comparison with the measured values of the amount of solar radiation (Japan Meteorological Agency data). As shown in FIG.
- the difference between the corrected amount of solar radiation and the measured value of the amount of solar radiation is smaller than the difference between the amount of solar radiation before correction and the measured value of the amount of solar radiation.
- FIG. 16 is an explanatory diagram showing an example of a correction result of the horizontal total solar radiation amount by the second correction unit 57.
- Three evaluation points are listed: Miyazaki, Nagasaki, and Saga.
- the upper graph shows the monthly cumulative amount of solar radiation [kWh / m 2 / month], and shows the amount of solar radiation before correction, the amount of solar radiation after correction, and the measured values of the amount of solar radiation (Japan Meteorological Agency data).
- the lower table shows the amount of solar radiation before correction and the amount of solar radiation after correction on a yearly basis (2018, 2019) in comparison with the measured values of the amount of solar radiation (Japan Meteorological Agency data). As shown in FIG.
- the difference between the corrected amount of solar radiation and the measured value of the amount of solar radiation is smaller than the difference between the amount of solar radiation before correction and the measured value of the amount of solar radiation.
- FIG. 17 is an explanatory diagram showing an example of the correction result of the slope solar radiation amount by the second correction unit 57.
- the amount of solar radiation on the slope can be obtained by using a known means such as METPV-11 based on the total amount of solar radiation on the horizontal plane and the inclination angle of the slope.
- the evaluation point is Mobara, where the solar power plant is installed.
- the graph on the left shows the monthly cumulative amount of solar radiation [kWh / m 2 / month], and shows the amount of solar radiation before correction, the amount of solar radiation after correction, and the measured values of the amount of solar radiation.
- the table on the right shows the amount of solar radiation before correction and the amount of solar radiation after correction on a yearly basis from 2014 to 2019, comparing them with the measured values of the amount of solar radiation (Japan Meteorological Agency data).
- the slope angle is 10 degrees.
- the difference between the corrected amount of solar radiation and the measured value of the amount of solar radiation is smaller than the difference between the amount of solar radiation before correction and the measured value of the amount of solar radiation. It can be seen that the accuracy can be improved and it can be suitably used for calculating the amount of power generation of a photovoltaic power plant.
- the server 50 of this embodiment can be used to provide various services related to the amount of solar radiation.
- the services that the server 50 can provide will be described.
- FIG. 18 is a schematic diagram showing an example of a model provision method by the server 50.
- the server 50 stores a plurality of different models generated by machine learning in the model unit 572.
- the server 50 can provide the model for the company A to the management server 100 of the company A.
- the management server 100 of the company A can predict the amount of solar radiation by correcting the solar radiation amount data recorded in the solar radiation amount DB 110 using the provided model.
- the server 50 can provide the model for the company B to the management server 100 of the company B.
- the management server 100 of the company B can use the provided model to correct the solar radiation amount data recorded in the solar radiation amount DB 110 and predict the solar radiation amount. Further, when the request for the model for company C is received from the company M, the server 50 can provide the model for the company C to the management server 100 of the company M. The management server 100 of the company M can predict the amount of solar radiation by correcting the solar radiation amount data acquired from the company C by using the provided model.
- the meteorological information service provider provides the solar radiation data provision service using its own prediction model, it can be supported to improve the accuracy of the provided solar radiation data. Further, when the weather information service provider is provided with the solar radiation amount data from another person, the accuracy of the provided solar radiation amount data can be improved.
- the solar radiation amount-related information providing unit 56 can generate related information related to the solar radiation amount, and can provide the related information to, for example, a weather information user.
- the weather information user can display the related information by using the terminal device 10.
- FIG. 19 is a schematic diagram showing a first display example of the amount of solar radiation provided by the server 50.
- the terminal device 10 can display the solar radiation amount screen 200.
- the user can select a required point on the map screen 201.
- a required point can be selected on the point list screen 202.
- the user can further select a required period (for example, OO year to OO year, OO month to OO month, OO day to OO day, etc.) in the period selection area 203.
- the point and period selected by the terminal device 10 are transmitted to the server 50 together with the request for related information.
- the server 50 can provide the amount of solar radiation data at the selected point and period.
- the solar radiation amount data can be displayed on the solar radiation amount screen 204.
- points S1 and S2 are selected, and the transition of the solar radiation amount data at the points S1 and S2 is displayed.
- FIG. 20 is a schematic diagram showing a second display example of the amount of solar radiation provided by the server 50.
- the terminal device 10 can display the solar radiation amount screen 210.
- the solar radiation amount data at four different points can be displayed on each of the solar radiation amount screens 211, 212, 213, and 214, and the reliability of the solar radiation amount data can also be displayed.
- the reliability may be calculated in advance by the server 50 based on the loss rate of the meteorological data used when correcting the solar radiation amount data and the coefficient of determination (R 2 ) used as an evaluation index of the model.
- the user can grasp the amount of solar radiation at a plurality of points at the same time, and can also grasp the reliability of the amount of solar radiation data.
- FIG. 21 is a schematic diagram showing a third display example of the amount of solar radiation provided by the server 50.
- the terminal device 10 can display the solar radiation amount screen 220.
- the transition of the solar radiation amount at the required point is displayed in different display modes.
- the transition indicated by reference numeral 221 indicates high accuracy
- the transition indicated by reference numeral 223 indicates low accuracy
- the transition indicated by reference numeral 222 indicates medium accuracy.
- the user can immediately grasp how accurate the accuracy is in which period when there is a difference in accuracy in the transition of the solar radiation amount data.
- FIG. 22 is a schematic diagram showing a display example of the predicted power generation amount of the power plant provided by the server 50.
- the terminal device 10 can display the predicted power generation amount screen 230 of the power plant. On the predicted power generation amount screen 230, the user can select a required power plant on the power plant list screen 231. The user can further select a prediction period (for example, OO day to OO day) in the period selection area 232.
- the power plant and forecast period selected by the terminal device 10 are transmitted to the server 50 together with the request for related information.
- the server 50 can obtain the solar radiation amount data at the selected power plant and the predicted period.
- the server 50 can provide the calculated power generation amount to the terminal device 10.
- the terminal device 10 can display the transition of the power generation amount of the selected power plant (power plants P1 and P2 in the example of the figure) and the average daily power generation amount on the power generation amount display screen 233.
- FIG. 23 is a schematic diagram showing a display example of power sales revenue of a power plant provided by the server 50.
- the terminal device 10 can display the power sales revenue screen 240 of the power plant.
- the user can select the target power plant on the power plant list screen 241.
- the user can further select a target period (for example, OO year) in the period selection area 242.
- the power plant and target period selected by the terminal device 10 are transmitted to the server 50 together with the request for related information.
- the server 50 can obtain the solar radiation amount data in the selected power plant and the target period.
- the solar radiation amount related information providing unit 56 can provide the terminal device 10 with the solar radiation amount, the system capacity, the loss coefficient, the annual power generation amount, the power selling price, and the power selling income in the selected power plant and the target period.
- the system capacity and the loss coefficient can be stored in the storage unit 54 in advance in association with the power plant.
- the annual power generation amount can be calculated by the formula ⁇ system capacity x daily solar radiation amount x loss coefficient x 365 ⁇ .
- the income from selling electricity can be calculated by the formula ⁇ annual power generation amount x selling price ⁇ . Revenue can be calculated by subtracting expenses from electricity sales revenue.
- the terminal device 10 can display the amount of solar radiation, the system capacity, the loss coefficient, the amount of annual power generation, the selling price, and the selling income in the display areas 243, 244, 245, 246, 247, and 248, respectively.
- the accuracy is equivalent to that of the snow-free period. Therefore, it is possible to predict the amount of insolation on the slope during the snowfall period. Specifically, when a certain value of snow cover and snowfall is recognized, the error of the corrected solar radiation is reduced by correcting the albedo (numerical value is a specific value) when predicting the slope solar radiation. be able to.
- Albedo is the reflectance of solar radiation, and the ratio of reflected radiation to incident radiation is called reflectance.
- the albedo the main substance that calibrates the ground surface, is as follows.
- the correction of the amount of solar radiation based on the albedo change is not limited to the case of snow cover, and can be performed according to the albedo of the ground surface object.
- This embodiment can be applied to the agricultural field. For example, by providing the relationship between the past amount of solar radiation and the yield of agricultural products to farmers, groups such as agricultural cooperatives, and agricultural businesses, and by providing the forecast of the amount of solar radiation, the yield of agricultural products can be predicted in advance. In addition, by providing the amount of solar radiation in the future, it can be used for adjusting the harvest time and predicting the spread of diseases of agricultural products.
- the solar radiation amount correction method of the present embodiment acquires solar radiation amount data, acquires meteorological data, and corrects the acquired solar radiation amount data based on the acquired meteorological data.
- the solar radiation correction device of the present embodiment has a first acquisition unit for acquiring solar radiation amount data, a second acquisition unit for acquiring meteorological data, and the first acquisition unit based on the meteorological data acquired by the second acquisition unit. It is provided with a correction unit that corrects the solar radiation amount data acquired by the acquisition unit.
- the computer program of the present embodiment causes a computer to acquire solar radiation amount data, acquire meteorological data, and correct the acquired solar radiation amount data based on the acquired meteorological data.
- the solar radiation correction method acquires solar radiation data and meteorological data.
- the solar radiation amount data can be acquired from, for example, a management server operated by a weather information service provider.
- the solar radiation amount data can be hourly data of the solar radiation amount over a required period at one or a plurality of points.
- the amount of solar radiation is the amount of radiant energy that a unit area receives from the sun in a unit time, and the unit can be expressed as [kWh / m 2 / h].
- the amount of solar radiation can be, for example, the amount of total solar radiation on a horizontal plane obtained by measuring the amount of solar radiation from the entire sky.
- the required period may be either a past period, a past-to-future period, or a future period.
- meteorological data for example, data observed at a meteorological observatory, data provided by a meteorological forecasting company, or the like can be used.
- Meteorological data can include physical quantities caused by clouds and physical quantities that have a correlation with clouds, and includes not only directly observable data but also processed data of observed data.
- the solar radiation amount correction method corrects the acquired solar radiation amount data based on the acquired meteorological data.
- the acquired solar radiation data does not fully consider the effects caused by clouds, but by correcting the acquired solar radiation data using meteorological data, it is possible to consider the effects caused by clouds and actually measure it. It is possible to provide solar radiation amount data with a small deviation from the value.
- the solar radiation amount correction method of the present embodiment corrects the acquired solar radiation amount data by using a model using the solar radiation amount data and the meteorological data as input variables.
- the solar radiation amount correction method corrects the acquired solar radiation amount data using a model that uses solar radiation amount data and meteorological data as input variables. That is, the model uses the solar radiation amount data and the meteorological data as input variables and the corrected solar radiation amount data as output variables, and formulates the relationship between the input variables and the output variables.
- the corrected solar radiation amount data may be a difference (difference between the corrected solar radiation amount data and the uncorrected solar radiation amount data) with respect to the acquired solar radiation amount data (sunlight amount data before correction).
- Models can be generated by machine learning. As a result, when the solar radiation amount data and the meteorological data are input to the model, the corrected solar radiation amount data can be obtained.
- the input variable of the model includes the dew point data, and the meteorological data including the dew point data is acquired.
- the solar radiation amount correction method includes the dew point data as an input variable of the model, and by inputting the dew point data into the model, the corrected solar radiation amount data can be obtained.
- Dew point is the difference between the temperature at a certain latitude and hardness and the dew point temperature at that latitude and altitude. The higher the dew point temperature, the higher the humidity, and the smaller the dew point, the higher the humidity.
- the dew point may be the difference between the air temperature at a certain altitude and the dew point temperature at the altitude.
- the input variable of the model includes the logarithmic data of the humidity data, and the meteorological data including the logarithmic data of the humidity data is acquired.
- the solar radiation correction method includes the logarithmic data of humidity data as an input variable of the model, and by inputting the logarithmic data of humidity data into the model, the corrected solar radiation amount data can be obtained.
- the logarithmic data of the humidity data can be represented by, for example, ⁇ 1 / log (humidity) ⁇ .
- the input variables of the model are temperature data, temperature difference data, humidity difference data or perceived temperature difference data, cloud amount data, direct solar radiation amount data, temperature data and dew point in a predetermined period.
- Temperatur difference data temperature difference data
- humidity difference data or perceived temperature difference data cloud amount data
- direct solar radiation amount data combination of temperature data and dew point temperature data, combination of temperature data and wind-cooled temperature data, combination of temperature data and perceived temperature data
- meteorological data including at least one of the scattered solar radiation data.
- the solar radiation correction method uses temperature data, temperature difference data, humidity difference data or perceived temperature difference data, cloud amount data, direct solar radiation amount data, combination of temperature data and dew point temperature data, and temperature as input variables of the model. It includes at least one of a combination of data and wind-cooled temperature data, a combination of temperature data and perceived temperature data, a combination of temperature data and heat index data, and scattered solar radiation amount data.
- the predetermined period can be, for example, one day.
- the temperature data in the predetermined period is the average temperature in one day
- the temperature difference data in the predetermined period is the temperature difference in one day
- the humidity difference data in the predetermined period is the humidity difference in one day. The same applies to other data in a predetermined period.
- temperature data, temperature difference data, humidity difference data or perceived temperature difference data, cloud amount data direct solar radiation amount data, combination of temperature data and dew point temperature data, combination of temperature data and wind cooling temperature data, By inputting at least one of the combination of the temperature data and the perceived temperature data, the combination of the temperature data and the heat index data, and the scattered solar radiation amount data, the corrected solar radiation amount data can be obtained.
- the influence of clouds can be considered.
- the input variable of the model includes the solar radiation amount data corrected based on the acquired meteorological data, the corrected solar radiation amount data is acquired, and the solar radiation amount data is further obtained. to correct.
- the solar radiation amount correction method includes the solar radiation amount data corrected based on the acquired meteorological data as an input variable of the model.
- the meteorological data may include at least one of dew point data and sea level barometric pressure data.
- the correction of the solar radiation amount data may be performed based on the meteorological data, or the solar radiation amount data may be weighted averaged in units of time to correct the time lag.
- the model further corrects the corrected solar radiation amount data, so that the accuracy of the correction can be further improved.
- the solar radiation correction method of the present embodiment acquires meteorological data including dew point data, and if the acquired dew point data is less than a predetermined dew point threshold, the acquired solar radiation amount data is corrected.
- the solar radiation correction method acquires meteorological data including humidity data, and if the acquired humidity data is less than a predetermined humidity threshold, the acquired solar radiation data is corrected.
- the dew point exceeds the dew point threshold, the humidity is low because the dew point is large, and it is considered that the influence of clouds does not need to be considered.
- the dew point is less than the dew point threshold, the humidity is high because the dew point is small, and it is considered that rain or fog is occurring in the area, and the acquired solar radiation amount data needs to be corrected. Be done.
- the solar radiation amount correction method of the present embodiment acquires meteorological data including the atmospheric pressure data at the altitude above sea level, and corrects the acquired solar radiation amount data when the acquired atmospheric pressure data at the altitude above sea level is equal to or less than a predetermined atmospheric pressure threshold.
- the solar radiation correction method acquires meteorological data including sea level barometric pressure data, and if the acquired barometric pressure data is below a predetermined barometric pressure threshold, the acquired solar radiation amount data is corrected.
- a predetermined barometric pressure threshold for example, 101 kPa
- the surface part becomes a low pressure system, air flows in from the part with higher atmospheric pressure, and as a result, an updraft is generated and clouds are generated. It is necessary to consider the effect, and it is considered that the acquired solar radiation amount data needs to be corrected.
- meteorological data including sea level barometric pressure data is acquired, and when the difference between the predetermined value and the acquired sea level level barometric pressure data is equal to or greater than the predetermined difference threshold value, the acquired solar radiation amount is obtained. Correct the data.
- the solar radiation correction method acquires meteorological data including sea level barometric pressure data, and if the difference between the predetermined value and the acquired sea level level barometric pressure data is equal to or greater than the predetermined difference threshold, the acquired solar radiation amount data is corrected.
- the atmospheric pressure constant ⁇ predetermined value a1- [atmospheric pressure at sea level] / constant a2 ⁇ , and the above-mentioned difference can be expressed by the atmospheric pressure constant.
- the lower the sea level pressure the larger the pressure constant.
- the barometric pressure constant is equal to or higher than the barometric pressure constant threshold (corresponding to the differential threshold)
- the surface part becomes a low pressure system, air flows in from the higher pressure part, and as a result, an updraft is generated and clouds are generated.
- the correction of the amount of solar radiation data can be corrected by dividing the amount of solar radiation data by the power of a (atmospheric pressure constant).
- the exponent a > 1.
- the solar radiation correction method of this embodiment acquires solar radiation data from a weather information service provider.
- the solar radiation correction method obtains solar radiation data from the weather information service provider.
- Each weather information service provider provides solar radiation data using its own forecasting model. Therefore, when correcting the solar radiation amount data, the correction method can be changed for each weather information service provider.
- the solar radiation correction method of the present embodiment accepts the selection of the required meteorological information service provider from a plurality of meteorological information service providers, and uses the model corresponding to the selected meteorological information service provider. Correct the acquired solar radiation amount data.
- the solar radiation correction method accepts the selection of the required meteorological information service provider from multiple meteorological information service providers, and uses the model corresponding to the selected meteorological information service provider to obtain the acquired solar radiation amount data. to correct.
- the weather information service providers are company A, company B, and company C
- the solar radiation amount data provided by company A can be corrected by using the model for company A.
- the solar radiation amount data provided by the company B can be corrected by using the model for the company B
- the solar radiation amount data provided by the company C can be corrected by using the model for the company C.
- the solar radiation amount data can be appropriately corrected.
- the model of this embodiment is generated by machine learning using the solar radiation amount data and the meteorological data as input variables and the corrected solar radiation amount data as output variables.
- the model is generated by machine learning with the solar radiation data and meteorological data as input variables and the corrected solar radiation data as output variables.
- the model is a formulation of the relationship between the input variable and the output variable, and when the solar radiation amount data and the meteorological data are input, the corrected solar radiation amount data can be output.
- the solar radiation amount data and the meteorological data are acquired, the corrected solar radiation amount data is acquired, the solar radiation amount data and the meteorological data are used as input variables, and the corrected solar radiation amount data is used.
- the model generation method uses the acquired solar radiation amount data and meteorological data as input variables, and generates a model using the corrected solar radiation amount data as an output variable.
- the model can be configured with a neural network (including an input layer, an intermediate layer, and an output layer).
- the acquired solar radiation amount data and meteorological data are used as learning input data, and the measured values of the solar radiation amount data are used as teacher data.
- input data for training is input to the model, it is possible to generate the model by adjusting the parameters (weight, bias) of the neural network so that the difference between the solar radiation amount data output by the model and the teacher data becomes small. can.
- the model providing method of the present embodiment stores a plurality of different models generated by machine learning using the solar radiation amount data and the meteorological data as input variables and the corrected solar radiation amount data as an output variable, and stores a plurality of different weather information. It accepts the selection of the required weather information service provider from the service providers, and provides the model corresponding to the selected weather information service provider from the plurality of different models.
- the model provision method stores multiple different models generated by machine learning. It accepts the selection of the required meteorological information service provider from multiple meteorological information service providers, and provides a model corresponding to the selected meteorological information service provider. Assuming that the weather information service providers are company A and company B, when the request for the model for company A is received from company A, the model for company A can be provided to company A. Similarly, when a request for a model for company B is received from company B, the model for company B can be provided to company B. As a result, when the weather information service provider provides the solar radiation data provision service using its own prediction model, it is possible to support the accuracy of the provided solar radiation data to be improved.
Landscapes
- Environmental & Geological Engineering (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Atmospheric Sciences (AREA)
- Biodiversity & Conservation Biology (AREA)
- Ecology (AREA)
- Environmental Sciences (AREA)
- Air Conditioning Control Device (AREA)
- Air-Conditioning For Vehicles (AREA)
- Photovoltaic Devices (AREA)
Abstract
Description
積雪(新、乾) 0.75~0.9
積雪(旧、湿) 0.4~0.6
なお、アルベド変化に基づく日射量の補正は、積雪がある場合に限定されるものではなく、地表面物体のアルベドに応じて行うこともできる。
10 端末装置
50 サーバ
51 制御部
52 日射量データ取得部
53 気象データ取得部
54 記憶部
55 第1補正部
56 日射量関連情報提供部
57 第2補正部
571 入力データ生成部
572 モデル部
58 学習処理部
581 学習データ生成部
582 モデル部
583 パラメータ決定部
61 気象データDB
62 日射量データDB
100 管理サーバ
110 日射量DB
Claims (16)
- 日射量データを取得し、
気象データを取得し、
取得した気象データに基づいて、取得した日射量データを補正する、
日射量補正方法。 - 日射量データ及び気象データを入力変数とするモデルを用いて、取得した日射量データを補正する、
請求項1に記載の日射量補正方法。 - 前記モデルの入力変数は、湿数データを含み、
湿数データを含む気象データを取得する、
請求項2に記載の日射量補正方法。 - 前記モデルの入力変数は、湿度データの対数データを含み、
湿度データの対数データを含む気象データを取得する、
請求項2又は請求項3に記載の日射量補正方法。 - 前記モデルの入力変数は、所定期間での気温データ、気温差データ、湿度差データ又は体感温度差データ、雲量データ、直達日射量データ、気温データと露点温度データの組み合わせ、気温データと風冷温度データの組み合わせ、気温データと体感温度データの組み合わせ、気温データと熱指数データの組み合わせ、及び散乱日射量データの少なくとも一つを含み、
所定期間での気温データ、気温差データ、湿度差データ又は体感温度差データ、雲量データ、直達日射量データ、気温データと露点温度データの組み合わせ、気温データと風冷温度データの組み合わせ、気温データと体感温度データの組み合わせ、気温データと熱指数データの組み合わせ、及び散乱日射量データの少なくとも一つを含む気象データを取得する、
請求項2から請求項4のいずれか一項に記載の日射量補正方法。 - 前記モデルの入力変数は、取得した気象データに基づいて補正された日射量データを含み、
補正された日射量データを取得して、さらに日射量データを補正する、
請求項2から請求項5のいずれか一項に記載の日射量補正方法。 - 湿数データを含む気象データを取得し、
取得した湿数データが所定の湿数閾値未満の場合、取得した日射量データを補正する、
請求項1に記載の日射量補正方法。 - 海抜レベルの気圧データを含む気象データを取得し、
取得した海抜レベルの気圧データが所定の気圧閾値以下の場合、取得した日射量データを補正する、
請求項1又は請求項7に記載の日射量補正方法。 - 海抜レベルの気圧データを含む気象データを取得し、
所定値と取得した海抜レベルの気圧データとの差分が所定の差分閾値以上の場合、取得した日射量データを補正する、
請求項1、請求項7又は請求項8のいずれか一項に記載の日射量補正方法。 - 気象情報サービス事業者から日射量データを取得する、
請求項1から請求項9のいずれか一項に記載の日射量補正方法。 - 複数の気象情報サービス事業者の中から所要の気象情報サービス事業者の選択を受け付け、
選択された気象情報サービス事業者に対応する前記モデルを用いて、取得した日射量データを補正する、
請求項2から請求項6のいずれか一項に記載の日射量補正方法。 - 日射量データを取得する第1取得部と、
気象データを取得する第2取得部と、
前記第2取得部で取得した気象データに基づいて前記第1取得部で取得した日射量データを補正する補正部と
を備える日射量補正装置。 - コンピュータに、
日射量データを取得し、
気象データを取得し、
取得した気象データに基づいて、取得した日射量データを補正する、
処理を実行させるコンピュータプログラム。 - 日射量データ及び気象データを入力変数とし、補正後の日射量データを出力変数とする機械学習により生成されたモデル。
- 日射量データ及び気象データを取得し、
補正後の日射量データを取得し、
前記日射量データ及び気象データを入力変数とし、前記補正後の日射量データを出力変数とするモデルを生成する、
モデル生成方法。 - 日射量データ及び気象データを入力変数とし、補正後の日射量データを出力変数とする機械学習により生成された複数の異なるモデルを記憶し、
複数の気象情報サービス事業者の中から所要の気象情報サービス事業者の選択を受け付け、
前記複数の異なるモデルのうち、選択された気象情報サービス事業者に対応するモデルを提供する、
モデル提供方法。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2022540271A JP7451715B2 (ja) | 2020-07-31 | 2021-07-26 | 日射量補正方法、日射量補正装置、コンピュータプログラム、モデル、モデル生成方法及びモデル提供方法 |
US18/006,737 US20230273344A1 (en) | 2020-07-31 | 2021-07-26 | Insolation correction method, insolation correction device, recording medium, model, model generating method, and model providing method |
EP21848577.9A EP4191290A1 (en) | 2020-07-31 | 2021-07-26 | Solar radiation amount correction method, solar radiation amount correction device, computer program, model, model generation method, and model providing method |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2020130447 | 2020-07-31 | ||
JP2020-130447 | 2020-07-31 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022024960A1 true WO2022024960A1 (ja) | 2022-02-03 |
Family
ID=80035655
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2021/027457 WO2022024960A1 (ja) | 2020-07-31 | 2021-07-26 | 日射量補正方法、日射量補正装置、コンピュータプログラム、モデル、モデル生成方法及びモデル提供方法 |
Country Status (4)
Country | Link |
---|---|
US (1) | US20230273344A1 (ja) |
EP (1) | EP4191290A1 (ja) |
JP (1) | JP7451715B2 (ja) |
WO (1) | WO2022024960A1 (ja) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR102490858B1 (ko) * | 2022-07-05 | 2023-01-20 | 강릉원주대학교산학협력단 | 인공지능 기반 태양 복사량 및 대기 투과도 결정 장치 및 방법 |
WO2023248484A1 (ja) * | 2022-06-22 | 2023-12-28 | 日本電信電話株式会社 | 全天日射量推定装置、全天日射量学習装置、全天日射量推定方法、及び全天日射量推定プログラム |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH11211560A (ja) | 1998-01-21 | 1999-08-06 | Fujitsu Ltd | 衛星画像データを用いた日射量推定システム、方法及び記録媒体 |
JP2013057476A (ja) * | 2011-09-09 | 2013-03-28 | Toshiba Corp | Pmv推定装置およびそのプログラム |
KR20130031732A (ko) * | 2011-09-21 | 2013-03-29 | 엘지전자 주식회사 | 일사량 추정을 위해 전천사진으로부터 운량을 계산하는 방법, 상기 계산한 운량을 이용한 태양광 발전량 예측 장치 |
JP2016066183A (ja) * | 2014-09-24 | 2016-04-28 | 地域エネルギー株式会社 | 情報処理装置、情報処理方法および情報処理プログラム |
WO2016121202A1 (ja) * | 2015-01-30 | 2016-08-04 | 日本電気株式会社 | 予測装置、予測方法、及び、プログラム |
-
2021
- 2021-07-26 WO PCT/JP2021/027457 patent/WO2022024960A1/ja active Application Filing
- 2021-07-26 EP EP21848577.9A patent/EP4191290A1/en active Pending
- 2021-07-26 US US18/006,737 patent/US20230273344A1/en active Pending
- 2021-07-26 JP JP2022540271A patent/JP7451715B2/ja active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH11211560A (ja) | 1998-01-21 | 1999-08-06 | Fujitsu Ltd | 衛星画像データを用いた日射量推定システム、方法及び記録媒体 |
JP2013057476A (ja) * | 2011-09-09 | 2013-03-28 | Toshiba Corp | Pmv推定装置およびそのプログラム |
KR20130031732A (ko) * | 2011-09-21 | 2013-03-29 | 엘지전자 주식회사 | 일사량 추정을 위해 전천사진으로부터 운량을 계산하는 방법, 상기 계산한 운량을 이용한 태양광 발전량 예측 장치 |
JP2016066183A (ja) * | 2014-09-24 | 2016-04-28 | 地域エネルギー株式会社 | 情報処理装置、情報処理方法および情報処理プログラム |
WO2016121202A1 (ja) * | 2015-01-30 | 2016-08-04 | 日本電気株式会社 | 予測装置、予測方法、及び、プログラム |
Non-Patent Citations (1)
Title |
---|
OKE SHINICHIRO, HIROYUKI MURATA, HIROFUMI TAKIKAWA, TATEKI SAKAKIBARA: "A Forecasting-Error Reduction of Daily Total Insolation by Integration of Forecasted Daily Solar Irradiance Curve", JOURNAL OF JAPAN SOLAR ENERGY SOCIETY, vol. 33, no. 2, 1 January 2007 (2007-01-01), pages 53 - 53, XP055890772 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2023248484A1 (ja) * | 2022-06-22 | 2023-12-28 | 日本電信電話株式会社 | 全天日射量推定装置、全天日射量学習装置、全天日射量推定方法、及び全天日射量推定プログラム |
KR102490858B1 (ko) * | 2022-07-05 | 2023-01-20 | 강릉원주대학교산학협력단 | 인공지능 기반 태양 복사량 및 대기 투과도 결정 장치 및 방법 |
Also Published As
Publication number | Publication date |
---|---|
JP7451715B2 (ja) | 2024-03-18 |
US20230273344A1 (en) | 2023-08-31 |
EP4191290A1 (en) | 2023-06-07 |
JPWO2022024960A1 (ja) | 2022-02-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Panamtash et al. | A copula-based Bayesian method for probabilistic solar power forecasting | |
Duarte et al. | NASA/POWER and DailyGridded weather datasets—how good they are for estimating maize yields in Brazil? | |
Persson et al. | Multi-site solar power forecasting using gradient boosted regression trees | |
Boata et al. | Functional fuzzy approach for forecasting daily global solar irradiation | |
Gharbia et al. | Spatially distributed potential evapotranspiration modeling and climate projections | |
Marquez et al. | Forecasting of global and direct solar irradiance using stochastic learning methods, ground experiments and the NWS database | |
WO2022024960A1 (ja) | 日射量補正方法、日射量補正装置、コンピュータプログラム、モデル、モデル生成方法及びモデル提供方法 | |
Paulescu et al. | Structured, physically inspired (gray box) models versus black box modeling for forecasting the output power of photovoltaic plants | |
Calanca et al. | Application of long-range weather forecasts to agricultural decision problems in Europe | |
US20170228743A1 (en) | Crop forecasting with incremental feature selection and spectrum constrained scenario generation | |
KR102002548B1 (ko) | 영농형 태양광 발전 예측시스템 및 이를 이용한 발전량 및 수익 예측 방법 | |
JP2019213299A (ja) | エネルギー運用装置及び方法並びにシステム | |
JP7210338B2 (ja) | 予測装置、予測方法及びコンピュータプログラム | |
Pepe et al. | Model estimation for solar generation forecasting using cloud cover data | |
CN116415704A (zh) | 一种基于多数据融合与同化的区域精量灌溉方法及系统 | |
JP7104609B2 (ja) | 数値解析装置及び方法 | |
JP6600033B2 (ja) | 気象予測補正装置、気象予測補正方法及びプログラム | |
Castellvi | A new simple method for estimating monthly and daily solar radiation. Performance and comparison with other methods at Lleida (NE Spain); a semiarid climate | |
Gomez-Quiles et al. | Price and resource-related uncertainty in the estimation of the revenue of a wind farm | |
Nizharadze et al. | Learning the Gap in the Day-Ahead and Real-Time Locational Marginal Prices in the Electricity Market | |
Gemignani et al. | Solar radiation synthetic sequences and bidding strategies for energy auctions | |
Jewson et al. | Communicating properties of changes in lagged weather forecasts | |
JP7514148B2 (ja) | 情報処理装置、情報処理方法、情報処理システム及びコンピュータプログラム | |
Croley | PROBABILISTIC GREAT LAKES HYDROLOGY OUTLOOKS 1 | |
Tsanakas et al. | State‐of‐Play and Emerging Challenges in Photovoltaic Energy Yield Simulations: A Multi‐Case Multi‐Model Benchmarking Study |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21848577 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2022540271 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2021848577 Country of ref document: EP |
|
ENP | Entry into the national phase |
Ref document number: 2021848577 Country of ref document: EP Effective date: 20230228 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |