WO2017115686A1 - Forecasting system, information processing device, forecasting method, and forecasting program - Google Patents

Abstract

A forecasting system is provided with: a sun-direction-cloudiness calculation means 101 for calculating sun-direction cloudiness on the basis of the degree of cloudiness of cells when the sky above a specified location is divided so as to include two or more cells along the height direction, the sun-direction cloudiness being the degree of cloudiness in the direction of the sun at a designated first location; and a forecasting means 102 for forecasting a forecasted value using the sun-direction cloudiness, the forecasted value being a category pertaining to solar energy. The degree of cloudiness of the cells may be the degree of cloudiness of the cells when the surface of the ground including the first location, together with the sky above said surface, is divided so as to include two or more cells along the horizontal direction in addition to including two or more cells along the height direction. It is permissible for the sun-direction-cloudiness calculation means 101 to calculate the sun-direction cloudiness on the basis of the degree of cloudiness of a cell, from among cells of such description, through which a straight line connecting the first location and the sun passes.

Description

Prediction system, information processing apparatus, prediction method, and prediction program

The present invention relates to a prediction system, an information processing apparatus, a prediction method, and a prediction program for predicting values of items relating to solar energy.

Renewable energy is spreading. In particular, solar energy is widely used as an energy source for obtaining heat and electric power. Accordingly, it is required to predict items related to renewable energy with high accuracy.

For example, a power generation company that supplies power based on renewable energy (for example, a specific-scale electric power company or an independent power generation company) predicts the amount of power generated based on renewable energy and makes a power generation plan. However, when the actual power generation amount becomes lower than the predicted value, the power generation company, for example, has to receive power supply from another power generation company, resulting in an economic loss.

By the way, solar energy is greatly affected by the weather. For this reason, it is difficult to predict items related to solar energy with high accuracy. In addition, although the amount of electric power generation based on solar energy, the amount of solar radiation, etc. are mentioned as an item regarding solar energy, This item is not limited to these.

As a method for predicting the amount of solar radiation, which is one of the items related to solar energy, there is a method for predicting the amount of sunlight using the cloud amount directly above the point to be predicted. However, the clouds that affect the amount of solar radiation are not limited to those located directly above. Thus, the method using the cloud amount directly above has a problem that a highly accurate predicted value cannot be obtained.

Therefore, Patent Document 1 discloses a block through which a straight line connecting a predicted point and the sun passes at a shooting altitude when the amount of solar radiation is predicted using information on clouds extracted from a satellite image obtained by capturing a region including the predicted point. A method for assigning a weight that gives the largest weight to the cloud information is described.

More specifically, the method described in Patent Document 1 applies to a block crossed by a straight line connecting a predicted point and the sun at a predetermined altitude (for example, 5000 m above sea level) that is a cloud image data position on three-dimensional coordinates. By assigning the largest weight, the information on the cloud at the altitude and in the sun direction as viewed from the predicted point is reflected in the predicted value most. However, the technique described in Patent Document 1 does not take into consideration the influence of the cloud amount at an altitude other than a predetermined altitude on the amount of solar radiation at the predicted point. In the first place, it is difficult to determine whether a cloud in a captured image is a cloud located at a predetermined altitude only from the captured image. As described above, the method described in Patent Document 1 has a problem that the influence of clouds in the solar direction cannot be correctly reflected in the predicted value.

FIG. 15 is an explanatory diagram showing the horizontal distance between a certain point (Tokyo: 139 degrees 45 minutes east longitude, 35 degrees 41 minutes north latitude) and a cloud located in the solar direction when viewed from the point, for each cloud altitude. . FIG. 15 shows the horizontal distance between the point and clouds at each altitude when the elevation angle φ in the sun direction as viewed from the point is 30 degrees, 45 degrees, 60 degrees, and 80 degrees. It is shown. According to FIG. 15, when the elevation angle φ in the solar direction is relatively low (for example, 30 degrees) as at 12:00 of the winter solstice, the cloud in the lower layer (for example, altitude 0.2 km) and the upper layer (for example, altitude 15) It can be seen that the cloud at a distance of .85 km) is separated by 27 km or more in horizontal distance. Here, as shown in FIG. 16, the elevation angle φ in the sun direction refers to an angle in the height direction of the sun in a three-dimensional space when the reference point is the origin. Further, hereinafter, the horizontal angle θ in the solar direction means that the angle of the solar position on the horizontal plane when viewed from the reference point is 0 degrees south.

JP 2013-253851 A

雲 Clouds of various altitudes can exist on the straight line connecting the predicted point and the sun. In order to correctly reflect the influence of the cloud on the straight line in the predicted value, height information is required as cloud information. Further, not only information on clouds at one altitude (cloud amount, etc.) but also information on clouds at two or more altitudes that may exist on the straight line is necessary. The method described in Patent Document 1 treats the height of the cloud at a predetermined one altitude, so that the influence of the cloud actually in the solar direction is not reflected in the predicted value. There is a problem that the amount of solar radiation cannot be accurately predicted, such as being reflected in the predicted value as if there is a cloud not in the sun.

Therefore, an object of the present invention is to provide a prediction system, an information processing apparatus, a prediction method, and a prediction program that can accurately predict items related to solar energy.

The prediction system according to the present invention is based on the cloud amount of each cell when the sky above a specific point is divided so as to include two or more cells in the height direction. A solar direction cloud amount calculating means for calculating a certain solar direction cloud amount, and a predicting means for predicting a value of a prediction target, which is an item relating to solar energy at the first point, using the solar direction cloud amount. .

The information processing apparatus according to the present invention is based on the cloud amount of each cell when the sky above a specific point is divided so as to include two or more cells in the height direction. The solar direction cloud amount calculation means for calculating the solar direction cloud amount is provided.

In addition, the prediction method according to the present invention is based on the cloud amount of each cell when the information processing apparatus divides the sky above a specific point so as to include two or more cells in the height direction. A solar cloud amount that is a cloud amount in the solar direction at the point is calculated, and a prediction target value that is an item relating to solar energy at the first point is predicted using the solar cloud amount.

In addition, the prediction program according to the present invention causes the computer to perform the calculation at the designated first point based on the cloud amount of each cell when the sky above the specific point is divided so as to include two or more cells in the height direction. A process of calculating a solar cloud amount that is a cloud amount in the solar direction and a process of predicting a prediction target value that is an item relating to solar energy at the first point using the solar cloud amount are performed. .

According to the present invention, items related to solar energy can be accurately predicted.

It is a block diagram which shows the example of the prediction system of 1st Embodiment. 3 is a block diagram illustrating a configuration example of a cloud amount calculation unit 4. FIG. It is explanatory drawing which shows the example of the locus | trajectory of the sun seen from a certain point. It is explanatory drawing which shows a response | compatibility with the lattice point in GPV data, and the straight line showing the sun direction in an estimated point. It is explanatory drawing which shows the example of the passage distance in the cell of the straight line showing the sun direction. It is a flowchart which shows an example of operation | movement of 1st Embodiment. It is explanatory drawing which shows a response | compatibility with the lattice point in the GPV data of 2nd Embodiment, and the straight line showing the sun direction in an estimated point. It is explanatory drawing which shows the example of the passage distance in the cell of the straight line showing the sun direction. It is a block diagram which shows the example of the prediction system of 3rd Embodiment. It is explanatory drawing which shows the example of the cell used as the calculation object of solar direction cloudiness. It is explanatory drawing which shows the outline of the calculation method of LNB. It is explanatory drawing which compares and compares the actual solar radiation amount, the predicted value of the solar radiation amount by 1st Embodiment, and the solar radiation amount estimated using the cloud amount right above. It is a block which shows the outline | summary of the prediction system by this invention. It is a block diagram which shows the structural example of the information processing apparatus by this invention. It is explanatory drawing which shows the horizontal distance of a certain point and the cloud located in a solar direction for every altitude of a cloud. It is explanatory drawing which shows the example of the elevation angle (phi) of the sun direction, and horizontal angle (theta).

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

In the following embodiments, an example in which the item related to solar energy is the amount of solar radiation and the prediction system of the present invention predicts the amount of solar radiation in the future will be described as an example. That is, the case where the prediction target is the amount of solar radiation will be described as an example. However, the prediction target is not limited to the amount of solar radiation, and may be an item relating to solar energy. For example, the prediction target may be the amount of power generated based on solar energy. In addition, as other examples of the prediction target, for example, various things such as an increase in the temperature of a building and an influence on agricultural products can be considered.

Embodiment 1. FIG.
FIG. 1 is a block diagram illustrating an example of a prediction system according to the first embodiment of this invention. A prediction system 1 illustrated in FIG. 1 includes a prediction condition acquisition unit 2, a performance value acquisition unit 3, a cloud amount calculation unit 4, a learning unit 5, and a prediction unit 6.

The prediction condition acquisition unit 2 acquires information on a prediction point and a prediction time that is a time to be predicted as a prediction condition. The prediction condition acquisition unit 2 includes, for example, a user interface and a network interface, and may acquire position information of the predicted point and information (time information) of the predicted time via these interfaces. Note that the predicted point and the predicted time are the point and time at which the prediction system 1 can acquire a cloud amount or a forecast value of an amount correlated with the cloud amount at the predicted time in a predetermined range including the predicted point. There is no particular limitation. The predicted time may be one time point in the future, such as 12:00 on the year, month, day of the year, or a future time, such as from 12:00 to 13:00 on the day of the year, month, month, and day. It may be a period including a plurality of time points. In that case, an average value or a sum value of a plurality of time points included in the period may be a value in the period. A time-series range that can be set as the predicted time may be determined in advance.

The actual value acquisition unit 3 acquires the actual value of the prediction target (in this example, the amount of solar radiation). The actual value acquisition unit 3 may acquire, as the actual value, a prediction target value actually observed at a certain point at a plurality of past times corresponding to the predicted time, for example. Here, the plurality of past times corresponding to the predicted time may be, for example, the same time (hour) as the predicted time in each day for a predetermined period past the predicted time. In addition to the same time as the predicted time, the time before and after the predicted time may be included. Further, the point from which the actual value is acquired may be, for example, a predicted point or any other point.

The cloud amount calculation unit 4 calculates the cloud amount in the solar direction (hereinafter referred to as solar direction cloud amount) at the specified point and at the specified time for learning and prediction. For example, the cloud amount calculation unit 4 may calculate the solar cloud amount at the point and time at which the actual value acquisition unit 3 acquired the actual value in accordance with an instruction from the prediction condition acquisition unit 2 as learning data. . The cloud amount in the solar direction calculated by the cloud amount calculation unit 4 is associated with the actual measurement value of the prediction target at the same time at the same point, and provided to the learning unit 5 described later.

Also, the cloud amount calculation unit 4 calculates the cloud amount in the solar direction (predicted value) at the prediction time of the prediction point as the prediction data.

The learning unit 5 uses the past prediction target actual value acquired by the actual value acquisition unit 3 and the past solar direction cloud amount correlated with the actual result value calculated by the cloud amount calculation unit 4 to determine whether the prediction target and the solar Learn the relationship with directional cloud cover. The method of learning (machine learning) by the learning unit 5 is not particularly limited. For example, a support vector machine, a neural network, regression analysis, or the like can be used.

The prediction unit 6 is a learning result by the learning unit 5, which is information indicating the relationship between the prediction target and the solar direction cloud amount, and the solar direction cloud amount (predicted value) at the prediction time of the prediction point calculated by the cloud amount calculation unit 4. Based on this, the value of the prediction target (in this example, the amount of solar radiation) at the prediction time of the prediction point is predicted.

In addition, when performing the learning and prediction described above, information (for example, temperature, humidity, etc.) other than the solar directional cloud amount that affects the prediction target can be further used as an explanatory variable.

Next, the solar direction cloud amount calculation method of this embodiment will be described. FIG. 2 is a block diagram illustrating a configuration example of the cloud amount calculation unit 4. 2 includes a solar direction calculation unit 41, a three-dimensional cloud amount information acquisition unit 42, and a solar direction cloud amount calculation unit 43.

The solar direction calculation unit 41 calculates the angle information of the sun at the designated point and time. As the sun angle information, the sun direction calculation unit 41 may calculate, for example, the elevation angle φ and the horizontal angle θ in the sun direction when the sun is viewed from a specified point at a specified time.

FIG. 3 is an explanatory diagram showing a one-day trajectory of the sun as seen from a reference point. Note that the X, Y, and Z axes in the upper graph in FIG. 3 correspond to the X, Y, and Z axes in FIG. 16. In addition, the horizontal axis of each graph in the lower stage represents time [hour]. The vertical axis of the lower left graph represents the elevation angle φ [degree] in the solar direction, and the vertical axis of the lower right graph represents the horizontal angle θ [degree] in the solar direction.

The three-dimensional cloud amount information acquisition unit 42 acquires three-dimensional GPV (Grid Point Value) data of forecast values related to weather. Here, the GPV data includes not only at least two divided regions in the horizontal direction but also at least two layers in the height direction as cells constituting the grid. Further, the GPV data includes at least a cloud amount at each grid point or an amount correlated with the cloud amount as a forecast value related to weather. In general, grid points in GPV data are points that represent values in each cell included in the grid when the area to be distributed is divided into a predetermined grid (two-dimensional or three-dimensional grid) (for example, the center) ). Although depending on the calculation method of the forecast value of GPV data, generally, the forecast value related to the weather at each grid point indicated by the GPV data is considered to be uniform within the cell to which the grid point corresponds.

The three-dimensional cloud amount information acquisition unit 42, for example, converts a three-dimensional space including a predetermined range of the ground surface including a specified point and the sky thereof into two or more divided regions in the horizontal direction and two or more layers in the height direction. Three-dimensional GPV data including at least a cloud amount at a specified time or a value having an amount correlated with the cloud amount at each lattice point of the grid when the grid is divided so as to have a grid may be acquired. Here, the upper limit of the sky may be set to a predetermined height at which clouds are generally generated (for example, 13 km above sea level).

The GPV data may be, for example, weather forecast data periodically distributed by a weather forecast data distribution company if the predicted location is Japan. Specific examples of weather forecast data include, for example, upper cloud cover, middle cloud cover, and cloud cover at each grid point obtained by dividing the ground surface in the range of 22.4 to 47.6 degrees north latitude and 120 to 150 degrees east longitude into 5 km meshes. GPV data called Meso ソ Spectral Model (MSM), which includes 39 hours of forecast values of hourly intervals of the lower cloud cover, can be mentioned. In addition to MSM, you can also use GPV data such as a local numerical forecast model (LFM) or a global numerical forecast model (Global Spectral Model; GMS) depending on the location and time you want to predict. It is. The lattice size is not limited to a 5 km mesh, and may be, for example, a 2 km mesh, a 1 km mesh, or a finer mesh. Further, the GPV data acquired by the three-dimensional cloud amount information acquisition unit 42 is not limited to existing weather forecast data. The three-dimensional cloud amount information acquisition unit 42 may create the GPV data by calculation.

The solar direction cloud amount calculation unit 43 is designated based on the position information of the designated point, the sun angle information calculated by the solar direction calculation unit 41, and the GPV data acquired by the three-dimensional cloud amount information acquisition unit 42. Calculate the cloud amount in the solar direction at the point and time.

FIG. 4 is an explanatory diagram showing an example of lattice points to which GPV data corresponds, together with an example of the solar direction at the predicted point. FIG. 4A shows an example of lattice points in the horizontal direction and an example of the sun direction (horizontal angle θ) at the predicted point. FIG. 4B shows an example of lattice points in the height direction and an example of the sun direction (elevation angle φ) at the predicted point. In FIG. 4, black circles represent points (lattice points) to which information (forecast values) of cloud amounts or amounts correlated with the cloud amounts are given. FIG. 4 shows only some lattice points, but there is one lattice point per cell. That is, in this example, there is a cloud amount or a forecast value having an amount correlated with the cloud amount for each cell.

As shown in FIG. 4, the cell through which a straight line representing the solar direction passes from the position information of the predicted point, the angle information of the sun, and the position information of each cell (including the width information in the horizontal direction and the height direction). Can be specified.

Note that when latitude and longitude are developed on a two-dimensional plane, distortion occurs in the angle and distance. In FIG. 4, it is assumed that the distortion is small in a sufficiently narrow range where a cloud can be confirmed, and an orthogonal coordinate system is created using equirectangular projection. In addition to this, various representations using a general map projection method are possible as a coordinate system for calculating a straight line representing the sun direction. Even if the projection is not perpendicular to the parallels and meridians, the position of the predicted point, the angle information of the sun, and the position information of each cell ( Cell including a straight line representing the solar direction can be specified.

First, the solar direction cloud amount calculation unit 43 calculates a straight line representing the solar direction among the cells in the three-dimensional space corresponding to the GPV data from the position information of the designated point and the angle information of the sun at the designated time. Identify the cell that passes through. When the solar direction cloud amount calculation unit 43 specifies a cell through which the straight line passes, the solar direction cloud amount calculation unit 43 calculates the solar direction cloud amount based on the cloud amount in the specified cell. The cloud amount in each cell is obtained, for example, by referring to a forecast value of the cloud amount at a lattice point in GPV data or by calculating from an amount correlated with the cloud amount.

For example, the sun direction cloud amount calculation unit 43 may add the cloud amount in all cells through which the straight line passes to be the sun direction cloud amount at a specified time at a specified point. Moreover, the solar direction cloud amount calculation part 43 may calculate the solar direction cloud amount at the designated time of the designated point further using the passing distance of the straight line in the cell. FIG. 5 is an explanatory diagram showing an example of the passage distance of the straight line in a certain cell. In FIG. 5, r _i represents a passing distance in a straight cell i representing the sun direction.

The solar direction cloud amount calculation unit 43 calculates the cloud amount of the cell ^layer i passing through a straight line representing the solar direction at the time t at the designated point, c ^layer _i (t), and the intra-cell passing distance of the cell r ^layer _i (t ), The solar cloud amount C ^layer (t) for each layer (lower layer, middle layer, upper layer) may be obtained using the following equation (1). Here, layer is an identifier for identifying a layer, and in equation (1), u = upper layer, m = middle layer, and l = upper layer. In addition, i is an identifier for identifying a cell through which a straight line representing the sun direction passes at a designated point in time t in a layer indicated by the layer.

In equation (1), the weighted sum of the amount of cloud in the cell was obtained for each layer, with the passage distance in the cell of the straight line representing the sun direction as a weight, for each layer. Above, the result is divided by the sum of the passage distances of the straight lines in the layer to normalize the solar cloud amount for each layer, but the normalization process may be omitted. Moreover, in Formula (1), although the solar direction cloud amount was calculated | required for every layer, what added these is good also as final solar direction cloud amount. However, it is preferable to obtain the cloud amount for each layer because the weight for the cloud amount can be given to each layer in the learning process.

In the present embodiment, the prediction condition acquisition unit 2 and the actual value acquisition unit 3 are realized by, for example, a CPU that operates according to a program and various information input means. The cloud amount calculation unit 4, the learning unit 5, and the prediction unit 6 are realized by a CPU or the like that operates according to a program, for example. FIG. 1 shows an example in which the cloud amount calculation unit 4 calculates both the solar direction cloud amount at the past time as the learning data and the solar direction cloud amount at the future time as the prediction data. The prediction system may separately include a processing unit that calculates a solar cloud amount at a past time as learning data and a processing unit that calculates a solar cloud amount at a future time as prediction data.

Next, the operation of this embodiment will be described with reference to FIG. FIG. 6 is a flowchart showing an example of the operation of the present embodiment. In the example illustrated in FIG. 6, first, the prediction condition acquisition unit 2 acquires a prediction condition (step S <b> 101). The prediction condition acquisition unit 2 acquires at least information on the position of the prediction point and the prediction time as the prediction condition.

Next, the actual value acquisition unit 3 acquires a past predicted target actual value (irradiation amount) as learning data (step S102). The actual value acquisition unit 3 may acquire, for example, the amount of solar radiation S (t) at a plurality of past times t corresponding to the predicted time at the predicted point.

Here, the time t applied to the learning data may be the same time on each day for a predetermined number of days (for example, one month) past the predicted time. For example, when the predicted time is 10:00 on April 3, 2012, the above time t may be the following set of times.

t = [2012-03-03 10:00,
2012-03-04 10:00,
. . . ,
2012-04-02 10:00]

Next, the cloud amount calculation unit 4 calculates the solar direction cloud amount at the point and time at which the actual value acquisition unit 3 acquired the actual value (step S103).

The cloud amount calculation unit 4 uses, for example, the above formula (1), and the solar direction cloud amounts C ^u (t), C ^m (t), and C ^l (for each layer at a plurality of past times t of the prediction point. t) may be calculated.

Next, the learning unit 5 determines the prediction target and the sun based on the past prediction target actual value obtained in step S102 and the solar cloud amount at the same point and time as the actual value obtained in step S103. The relationship with the amount of directional clouds is learned (step S104: learning process).

The learning unit 5, for example, calculates the solar cloud amounts C ^u (t), C ^m (t), and C ^l (t) for each layer at a plurality of past times t at the prediction point obtained by the cloud amount calculation unit 4. Machine learning may be performed using the solar radiation amount S (t) at the predicted point at time t as the explanatory variable, and the function F as expressed by the following equation (2) may be obtained.

S (t) = F (C ^u (t), C ^m (t), C ^l (t)) (2)

Next, the cloud amount calculation unit 4 calculates (estimates) the cloud amount in the solar direction at the predicted time of the predicted point using the predicted value at the predicted time indicated by the GPV data (step S105).

For example, when the predicted time is t ′, the cloud amount calculation unit 4 uses the GPV data including the predicted value at the time t ′ and uses the solar direction cloud amount C ^u (t ′), for each layer at the predicted time point t ′. C ^m (t ′) and C ^l (t ′) may be calculated using the above equation (1).

Next, based on the learning result obtained in step S104 and the solar directional cloud amount calculated in step S105, the prediction unit 6 calculates the prediction target value (predicted value of solar radiation amount) at the prediction time of the prediction point. Calculate (step S106: prediction process).

For example, when the above function F is obtained as a learning result, the prediction unit 6 uses the predicted point calculated by the cloud amount calculation unit 4 for the function F and the solar direction cloud amount C ^u () for each layer at the prediction time t ′. The solar radiation amount S (t ′) at the prediction point and the prediction time may be obtained by substituting t ′), C ^m (t ′), and C ^l (t ′).

Finally, the prediction unit 6 outputs a prediction result (step S107).

As described above, according to the present embodiment, the designation is made based on the cloud amount in each cell included in the three-dimensional grid divided so as to include at least two cells in the height direction as well as the horizontal direction. Since the cloud amount in the solar direction at the designated time at the designated point is calculated and used for learning and prediction, prediction with high accuracy can be performed.

FIG. 6 illustrates an example in which the learning unit 5 performs learning after acquiring the prediction condition. However, the learning unit 5 performs learning at a plurality of times at predetermined points, for example. It is also possible to generate data for learning (a pair of the actual value and the cloud amount in the solar direction) and perform learning, and store the learning result. In such a case, the prediction unit 6 can perform prediction using the learning result of the closest condition after acquiring the prediction condition. Note that the learning unit 5 may perform learning by dividing learning data for each point, season, or time period, or may perform learning without dividing learning data.

Further, in the above, an example in which the set of the same time (hour) as the predicted time is set as the time t of the learning data has been shown, but the learning unit 5 adds the time t before and after the time t. The times t−1 and t + 1 may be set as the time of learning data. For example, the learning unit 5 may obtain a function F represented by the following expression (3) by opportunity learning. Even if the straight line representing the solar direction differs in the time zone and the total passing distance through the cell in the sky is different, if the cloud amount in the solar direction is standardized, learning is not affected by the difference. Can do.

S (t) = F (C ^u (t−1), C ^m (t−1), C ^l (t−1),
C ^u (t), C ^m (t), C ^l (t),
C ^u (t + 1), C ^m (t + 1), C ^l (t + 1)) (3)

For example, when it is desired to perform prediction at 10:00 on April 3, 2012, the learning unit 5 may use the following data at time t for learning.

t = [2012-03-03 9:00, same day 10:00, same day 11:00,
2012-03-04 9:00, same day 10:00, same day 11:00,
. . . ,
2012-04-02 9:00, same day 10:00, same day 11:00]

In addition to this, for example, the upper, middle and lower cloud cover directly above the target point, the temperature, humidity, theoretical solar radiation, extraneous solar radiation, maximum cloud top height, cloud base height, etc. May be added to

In the present invention, the “cloud amount” is not particularly limited as long as it is a cloud-related amount. The cloud amount may be, for example, a cloud generation probability, a cloud thickness, a wet number, a cloud area, or the like in addition to the cloud amount [%] used for GPV data.

Further, the cloud amount calculation unit 4 may correct the calculated solar cloud amount based on the state of eclipse and the presence or absence of shielding (buildings or trees) at each point. For example, it is possible to assume that the amount of clouds is infinite during a solar eclipse or when there is a shield.

Embodiment 2. FIG.
Next, a second embodiment will be described with reference to the drawings. In the second embodiment, it is assumed that the GPV data does not include the cloud amount of each lattice point, and includes a correlated amount in the cloud amount of each lattice point.

The configuration of the prediction system of this embodiment may be the same as that of the first embodiment shown in FIG. Hereinafter, the same parts as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted.

Next, a cloud amount calculation method according to this embodiment will be described. The cloud amount calculation unit 4 of the present embodiment calculates the solar direction cloud amount at a specified time at a specified point using information correlated with the cloud amount of each lattice point included in the GPV data. The cloud amount calculation unit 4 of the present embodiment uses a three-dimensional grid whose height direction is divided by the atmospheric pressure surface. Then, the cloud amount calculation unit 4 calculates the temperature and humidity of each lattice point in the grid (two-dimensional grid) on the ground surface as the amount corresponding to the cloud amount in each cell included in such a three-dimensional grid, and the target region. Using the relative humidity at each atmospheric pressure surface in the sky, the humidity number at an altitude corresponding to each atmospheric pressure surface at each lattice point of the two-dimensional grid is obtained.

In the present embodiment, a configuration example of the cloud amount calculation unit 4 will be described using the cloud amount calculation unit 4 shown in FIG.

FIG. 7 is an explanatory diagram showing a correspondence between lattice points in the GPV data of the present embodiment and straight lines representing the sun direction at the predicted points. As shown in FIG. 7, in this embodiment, pseudo lattice points in the height direction are set by using the air pressure surface indicated by the GPV data as a cell division (division standard) in the height direction. The horizontal cell division may be the same as in the first embodiment. That is, it may be determined by the value of latitude and longitude. In addition, in FIG. 7, although the width | variety of the height direction of the cell which points to the same atmospheric pressure surface shows the example which does not change with points, the width of the height direction of the cell which points to the same atmospheric pressure surface is on the ground surface It may be different for each point (lattice point in a two-dimensional grid defined on the ground surface). Also, for example, within the range through which the straight line representing the solar direction passes, assuming that the height of the air pressure surface does not change much depending on the point, the amount of calculation is reduced by applying the height of each air pressure surface at the predicted point to other points. May be. In that case, as shown in FIG. 7, the width in the height direction of the cells pointing to the same atmospheric pressure surface does not change depending on the point. On the other hand, it is also possible to obtain the height of the air pressure surface for each point and use it as a cell division in the height direction above the point. In that case, the width in the height direction of the cell pointing to the same atmospheric pressure surface may vary depending on the point. Even if the height direction width of cells pointing to the same barometric surface is different depending on the point, if the three-dimensional space is divided into cells without exception and the three-dimensional grid is configured, the position information of the predicted point and From the angle information of the sun and the position information of each cell (including information on the width in the horizontal direction and the height direction), the cell through which the straight line representing the solar direction passes can be specified.

In the present embodiment, the three-dimensional cloud amount information acquisition unit 42 divides the ground surface including the designated point so as to include two or more divided regions, and the temperature, humidity, and GPV data including at least the atmospheric pressure and the relative humidity on each atmospheric pressure surface at a specified time above the ground surface is acquired. The GPV data of this embodiment does not have to be exact three-dimensional GPV data. That is, the three-dimensional cloud amount information acquisition unit 42 adds the forecast value (at least temperature, humidity, and atmospheric pressure) related to the weather at each lattice point of the two-dimensional grid obtained by dividing the ground surface including the designated point in the horizontal direction. What is necessary is just to acquire the GPV data containing the predicted value of the relative humidity in two or more different atmospheric pressure surfaces above the ground surface. Here, the air temperature, humidity, and air pressure on the ground surface at each lattice point on the ground surface and the relative humidity on each air pressure surface correspond to an amount correlated with the cloud amount at each lattice point in the three-dimensional space.

Further, for example, if the GPV data includes altitude information of each atmospheric pressure surface, the solar direction cloud amount calculation unit 43 uses the altitude information to determine the position of each cell in the height direction above the target area. It may be determined. As an example, the solar direction cloud amount calculation unit 43 sets the sky above the target region so that the height corresponding to the altitude of each atmospheric pressure surface where relative humidity is indicated in the GPV data is located at the center of each cell in the height direction. You may divide into the above layers.

Or, as in the following listed equation (4), it is also possible to determine the altitude h _i of each pressure surface relative humidity is shown in GPV data.

Here, _tg is the ground temperature at the designated point, and _pg is the ground pressure at the designated point. Moreover, what is necessary is just to input the atmospheric pressure which wants to obtain the altitude for p.

Further, the following equation (5) is an example of a method for obtaining the cell width th _i in the height direction using the obtained altitude of the atmospheric pressure surface.

In this way, the solar direction cloud amount calculation unit 43 can divide the three-dimensional space including the ground surface including the designated point and the space above it so as to include two or more layers in the height direction. Then, the solar direction cloud amount calculation unit 43 uses the ground surface temperature (air temperature), the humidity, the atmospheric pressure, and the relative humidity at each atmospheric pressure surface, which are correlated with the cloud amount at each lattice point, to specify the cloud amount. The cloud amount in the solar direction at the designated time of the designated point may be obtained.

The solar direction cloud amount calculation unit 43, for example, for each cell through which a straight line representing the solar direction passes based on the temperature T and the relative humidity RH at the designated time t in each cell (each point and each atmospheric pressure surface). The dew point temperature Td in the cell may be obtained, and the wet number H corresponding to the cloud amount in the cell may be obtained from the obtained dew point temperature Td. As a method for obtaining the dew point temperature Td from the temperature T and the relative humidity RH, for example, the following equation (6) can be used. Further, as a method for obtaining the wet number H from the temperature T and the dew point temperature Td, for example, the following equation (7) can be used.

Wet number H = T-Td (7)

The solar direction cloud amount calculation unit 43 may obtain the wet number H ^j _i (t) of the cell through which the straight line passes as the cloud amount c ^j _i (t) in the cell, using, for example, the equation (7). . Here, j represents an identifier for identifying a pressure surface, and i represents an identifier for identifying a cell on the pressure surface. Further, the solar cloud amount calculation unit 43 may further calculate the cloud amount c ^j _i (t) in the cell by performing the following calculation from the obtained wet number H ^j _i (t).

c ^j _i (t) = max (3-H ^j _i (t), 0) (8)

Formula (8) is a calculation formula for excluding the influence of a wet number (a value greater than 3) that is generally considered to be free of clouds.

When the cloud amount in this cell (wet number in this example) is obtained for each cell through which the straight line representing the solar direction passes, the solar direction cloud amount calculation unit 43 uses the same method as in the first embodiment, The solar cloud amount may be calculated based on the cloud amount.

The solar direction cloud amount calculation unit 43 may use, for example, the sum of the cloud amounts of cells passing through a straight line representing the solar direction as the solar direction cloud amount, and further, the passage distance of the straight line in each cell The cloud amount may be weighted, and the sum of the weights may be used to calculate the solar direction cloud amount. Similarly to the first embodiment, the solar cloud amount calculation unit 43 may normalize the solar cloud amount by dividing by the sum of the passing distances. FIG. 8 is an explanatory diagram illustrating an example of a passing distance in a straight cell representing the sun direction in the present embodiment. For example, when the cloud amount at each point and each atmospheric pressure surface is obtained from the GPV data acquired by the three-dimensional cloud amount information acquisition unit 42, the solar direction cloud amount calculation unit 43 uses the following formula (9) to calculate the sun for each layer. The directional cloud amount may be obtained.

Here, layer is an identifier for identifying the atmospheric pressure surface, and corresponds to the above j. C ^layer (t) is a cloud amount in the solar direction on the atmospheric pressure surface indicated by layer at time t of the designated point. Further, c ^layer _i (t) is a cloud amount at time t in the cell indicated by i on the atmospheric pressure surface indicated by layer. Further, r ^layer _i (t) is a passing distance of the straight line at time t in the cell indicated by i on the atmospheric pressure surface indicated by layer.

The learning unit 5 of the present embodiment, for example, the solar cloud amounts C ¹ (t), C ² (t),... For each layer (for each atmospheric pressure surface) at a plurality of past times t obtained by the cloud amount calculation unit 4. ^.. , C ^j_max (t) is an explanatory variable, and the solar radiation amount S (t) at the predicted point at time t is machined to be an explanatory variable, and a function F as expressed by the following equation (10) is performed. You may get Note that j_max is the number of air pressure surfaces (number of divisions).

S (t) = F (C ¹ (t), C ² (t),..., C ^j_max (t)) (10)

In addition, for example, when the function F is obtained as a learning result, the prediction unit 6 uses the predicted point calculated by the cloud amount calculation unit 4 for the function F and the solar cloud amount C for each layer at the prediction time t ′. Substituting ¹ (t ′), C ² (t ′),..., C ^j_max (t ′), the solar radiation amount S (t ′) at the predicted point and the predicted time t ′ may be obtained.

As described above, according to the present embodiment, even when the cloud amount for each layer is not explicitly included in the GPV data, the height direction is determined using the amount correlated with the cloud amount included in the GPV data. The cloud amount in each cell divided so as to include two or more cells can be obtained. In particular, according to the method of this embodiment that divides the height direction based on the atmospheric pressure surface, cells in the height direction can be subdivided based on the atmospheric pressure that has a high correlation with the cloud amount, so that more accurate learning is possible. And can make predictions. Other points are the same as in the first embodiment.

Embodiment 3. FIG.
Next, a third embodiment will be described with reference to the drawings. The third embodiment calculates the solar directional cloud amount at the predicted time of the predicted point based on the cloud amount in the cell through which the straight line representing the solar direction at the predicted time of the predicted point passes without using the learning data. The value of the prediction target (in this example, the amount of solar radiation) at the predicted time of the point is predicted.

FIG. 9 is a block diagram illustrating an example of the prediction system of the third embodiment. As shown in FIG. 9, the prediction system 1 of this embodiment includes a prediction condition acquisition unit 2, a cloud amount calculation unit 4a, and a prediction unit 6a. In addition, the same code | symbol is attached | subjected about the component similar to said embodiment, and description is abbreviate | omitted.

In this embodiment, the cloud amount calculation unit 4a calculates the cloud amount in the solar direction at a specified point and at a specified time for prediction. The cloud amount calculation unit 4a calculates, for example, the solar direction cloud amount (prediction value) at the prediction time of the prediction point as the prediction data. The solar cloud amount information calculated in the present embodiment includes the cloud amount in each cell through which a straight line representing the solar direction passes or the solar cloud amount for each layer.

Further, the prediction unit 6a determines the value of the prediction target (in this example, the amount of solar radiation) at the prediction time of the prediction point based on the solar cloud amount (prediction value) at the prediction time of the prediction point calculated by the cloud amount calculation unit 4a. Predict.

The prediction unit 6a may predict the value of the prediction target at the prediction time of the prediction point, for example, by weighting the solar cloud amount for each layer calculated by the cloud amount calculation unit 4a. For example, the prediction unit 6a calculates the cloud amount in each cell through which a straight line representing the solar direction included in the solar direction cloud amount (predicted value) at the prediction time of the prediction point passes, in the height direction position of the cell in the three-dimensional grid ( In other words, weighting may be performed based on the height of the grid to predict the value of the prediction target at the prediction time of the prediction point. At this time, the prediction unit 6a may give a smaller weight to the cloud amount of a cell having a higher height in the three-dimensional grid. This is because the high-rise clouds are thin and the low-rise clouds are dark, so that clouds at higher positions are more likely to transmit sunlight than clouds at lower positions.

The prediction unit 6a calculates, for example, a weighted sum of the sun direction cloud amount for each layer according to the assigned weight, and represents a relationship between a predetermined weighted sum of the sun direction cloud amount for each layer and a prediction target, The value of the prediction target at the prediction time of the prediction point may be calculated by substituting the calculated weighted sum.

As described above, according to the present embodiment, based on the solar cloud amount obtained for each layer, a predetermined weight is applied to the solar cloud amount, and then the prediction target value at the prediction time of the prediction point is calculated. Therefore, it is possible to appropriately reflect the influence on the predicted value of clouds in different solar directions depending on the height. Therefore, the item regarding solar energy can be accurately predicted.

In each of the above embodiments, all the cells that pass through the straight line representing the solar direction are used for calculating the amount of cloud. For example, as shown in FIG. 10, between the maximum cloud top height and the cloud bottom height, You may exclude the cell of the altitude which is not included (refer to the shaded part) from the calculation target of the cloud amount. In addition, as the maximum cloud top altitude, a buoyancy zero altitude (Level of Neutral bouyancy; LNB) can be used. Further, as the cloud bottom height, a lifted solidification point height (LiftedftCondensation Level; LCL) can be used.

As a calculation method of LCL and LNB, for example, a generally known method of obtaining from an equivalent temperature level, or a method of obtaining from the ground temperature T and dew point temperature Td may be used.

For example, LCL = 125 + (T−Td) may be set. Further, based on the LCL obtained in this way, LNB may be obtained using an emmagram. FIG. 11 is an explanatory diagram showing an outline of an LNB calculation method. The solar direction cloud amount calculation unit 43 may calculate the LNB by a method as shown in FIG. 11, for example.

That is, first, the state of the temperature at each altitude (atmospheric pressure) is plotted on an emagram and a state curve is created. Then, using the altitude indicated by the ground surface pressure (the ground surface height) as a starting altitude, the plot point of the pressure and temperature is raised on the dry insulation line, and when reaching the altitude (atmosphere) corresponding to the LCL, it is raised on the wet insulation line. The intersection point where the state curve finally exceeds the wet heat insulation line is defined as LNB.

FIG. 12 is an explanatory diagram showing a comparison between the actual solar radiation amount, the predicted value of the solar radiation amount according to the first embodiment, and the solar radiation amount predicted using the cloud amount directly above. In FIG. 12, the horizontal axis represents the elapsed time when 0:00 of a certain date and time is set to 0, and the vertical axis represents the amount of sunlight. The data shown in FIG. 12 is data when the predicted time is set at one hour intervals in a period of three days in mid-January. From FIG. 12, it can be seen that the method according to the first embodiment can obtain a predicted value closer to the actual value than the method using the cloud amount directly above. In addition, there is a steep change in the actual amount of solar radiation around 11:00 on the first day, but the cloud that was not in the sun direction moved at a high speed so that it overlapped with the sun at around 11:00. It is assumed that the reason is that a shade was created at the predicted point. Thus, although the clouds that affect the prediction point can change every moment, according to the method of predicting the amount of solar radiation based on the amount of clouds in the solar direction having the greatest influence as in the present invention, such a It can be seen that it is possible to accurately follow any change.

As shown in FIG. 15, in Japan, the distance between the immediately above and the solar direction tends to increase in winter when the elevation angle in the solar direction is smaller. On the other hand, in the summer when the elevation angle in the solar direction is large, the distance between directly above and the solar direction is small, and even at an altitude of 15 km at the time of south-central time, it is only less than 3 km. For this reason, the cloud amount calculation method used for learning and prediction may be switched according to the season and the elevation angle in the sun direction. In other words, when the predicted time is a predetermined season such as summer, or when the elevation angle in the solar direction at the predicted time is greater than or equal to a predetermined angle, calculation is performed using the cloud amount directly above as the cloud amount used for learning and prediction. You may shorten the time which takes. On the other hand, when the predicted time is a predetermined season such as winter or when the elevation angle in the solar direction at the predicted time is less than a predetermined angle, the cloud amount in the solar direction may be used as the cloud amount used for learning and prediction. .

Next, the outline of the present invention will be described. FIG. 13 is a block diagram showing an outline of a prediction system according to the present invention. As shown in FIG. 13, the prediction system according to the present invention includes solar direction cloud amount calculation means 101 and prediction means 102.

The solar direction cloud amount calculation means 101 (for example, the cloud amount calculation unit 4, the solar direction cloud amount calculation unit 43) calculates the cloud amount of each cell when the sky above a specific point is divided so as to include two or more cells in the height direction. Based on this, the solar cloud amount that is the cloud amount in the solar direction at the designated first point is calculated. The specific point may be, for example, predetermined grid points on the ground surface including the designated first point.

The prediction unit 102 (for example, the prediction unit 6) predicts the value of the prediction target, which is an item related to solar energy at the first point, using the solar direction cloud amount calculated by the solar direction cloud amount calculation unit 101.

By providing such a configuration, it is possible to calculate the amount of cloud in the solar direction that has a particularly large impact on items related to solar energy in consideration of the information on the altitude of the cloud, so that the value of the item can be accurately predicted. it can.

FIG. 14 is a block diagram showing a configuration example of the information processing apparatus according to the present invention. As shown in FIG. 14, the information processing apparatus according to the present invention includes the solar cloud amount calculation unit 101 described above. As already explained, the amount of cloud in the solar direction that has a particularly large impact on items related to solar energy can be determined in consideration of the altitude information of the cloud, so it is effective for all applications that require such cloud amount. Information can be provided.

In addition, each said embodiment can be described also as the following additional remarks.

(Supplementary note 1) Each of the three-dimensional grids when the ground surface including the first point and the sky above are divided so as to include two or more cells in the horizontal direction and two or more cells in the height direction. Solar direction cloud amount calculating means for calculating a solar direction cloud amount that is a cloud amount in the solar direction at the first point, based on the cloud amount in the cell through which a straight line connecting the first point and the sun of the cell passes. A prediction system comprising: a prediction unit that predicts a value of a prediction target that is an item related to solar energy at the first point using the amount of cloud in the solar direction.

(Supplementary Note 2) Based on the solar direction cloud amount calculated as the learning data and the actual value of the prediction target at the same point and time as when the solar direction cloud amount as the learning data was calculated, the prediction target and the solar direction cloud amount The solar direction cloud amount calculation unit calculates, as learning data, the solar direction cloud amount at a plurality of past times corresponding to the predicted time of the first point that is the predicted point. At the same time, as the prediction data, the solar cloud amount at the prediction time of the first point which is the prediction point is calculated, and the prediction unit converts the learning result by the learning unit and the solar cloud amount calculated as the prediction data. The prediction system according to supplementary note 1, wherein the prediction target value at the prediction time of the first point that is the prediction point is predicted based on the prediction point.

(Supplementary Note 3) Three-dimensional when a three-dimensional space including the ground surface including the first point and the sky above is divided so as to include two or more cells in the horizontal direction and two or more cells in the height direction. 3D cloud amount information acquisition means for acquiring GPV data including a forecast value of a cloud amount at a specified time or an amount correlated with the cloud amount at each grid point of the grid, and the solar direction cloud amount calculation means uses GPV data The prediction system according to Supplementary Note 1 or Supplementary Note 2, wherein a solar cloud amount that is a cloud amount in the solar direction at a specified time at a first point is calculated.

(Supplementary note 4) The prediction system according to supplementary note 3, wherein the GPV data is the amount of cloud in the sky for each lattice point on the ground surface and includes the amount of clouds in the upper, middle, and lower layers.

(Supplementary note 5) The GPV data includes temperature, humidity and atmospheric pressure at each of the lattice points on the ground surface, and relative humidity at two or more atmospheric pressure planes above the ground surface as quantities correlated with the cloud amount. The prediction system described in.

(Supplementary Note 6) The solar cloud amount calculation means is based on the temperature, humidity, and atmospheric pressure at each of predetermined lattice points on the ground surface included in the GPV data, and relative humidity at two or more atmospheric pressure surfaces above the ground surface. The 3D grid is set by dividing the ground surface including the first point and the sky above it by a grid point on the ground surface and two or more air pressure surfaces above the ground surface, and setting the 3D grid. Among the cells included in the cell, the wet number in the cell is calculated as the amount of cloud in the cell through which the straight line connecting the first point and the sun passes, and the sun at the first point is calculated based on the calculated wet number The prediction system according to appendix 5, which calculates a solar cloud amount that is a cloud amount in the direction.

(Supplementary note 7) A three-dimensional grid obtained by dividing the ground surface including the designated first point and the sky above it so as to include two or more cells in the horizontal direction and two or more cells in the height direction. Solar direction cloud amount calculation means for calculating a solar direction cloud amount that is a cloud amount in the solar direction at the first point based on a cloud amount in a cell through which a straight line connecting the first point and the sun passes An information processing apparatus comprising:

(Supplementary note 8) The information processing apparatus according to supplementary note 7, wherein the three-dimensional grid has a height direction divided based on a length of distance or a magnitude of atmospheric pressure.

(Supplementary Note 9) The information processing apparatus divides the ground surface including the designated first point and the sky so as to include two or more cells in the horizontal direction and two or more cells in the height direction. Based on the cloud amount in the cell through which the straight line connecting the first point and the sun passes among each cell included in the three-dimensional grid, the solar direction cloud amount that is the cloud amount in the solar direction at the first point is calculated. A prediction method characterized by predicting a prediction target value, which is an item relating to solar energy at the first point, using the calculated cloud amount in the solar direction.

(Supplementary note 10) The prediction method according to supplementary note 9, wherein the three-dimensional grid is divided in the height direction based on the length of the distance or the magnitude of the atmospheric pressure.

(Supplementary Note 11) When the computer divides the ground surface including the designated first point and the sky above it so as to include two or more cells in the horizontal direction and two or more cells in the height direction. A process of calculating a solar cloud amount that is a cloud amount in the solar direction at the first point based on a cloud amount in a cell through which a straight line connecting the first point and the sun passes among the cells included in the three-dimensional grid; The prediction program for performing the process which estimates the value of the prediction object which is an item regarding the solar energy in a 1st point using the calculated solar direction cloud amount.

(Supplementary note 12) The prediction program according to supplementary note 11, wherein the three-dimensional grid is divided in the height direction based on the length of the distance or the magnitude of the atmospheric pressure.

Although the present invention has been described with reference to the present embodiment and examples, the present invention is not limited to the above-described embodiment and examples. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

This application claims priority based on Japanese Patent Application No. 2015-256402 filed on Dec. 28, 2015, the entire disclosure of which is incorporated herein.

The present invention is not limited to the use of predicting items related to solar energy, but can be suitably applied to a use of performing some calculation using the cloud amount in the solar direction.

DESCRIPTION OF SYMBOLS 1 Prediction system 2 Prediction condition acquisition part 3 Actual value acquisition part 4, 4a Cloud cover calculation part 41 Solar direction calculation part 42 Three-dimensional cloud cover information acquisition part 43 Solar direction cloud cover calculation part 5 Learning part 6, 6a Prediction part 101 Solar direction cloud cover calculation Means 102 Prediction means

Claims (12)

1. Based on the cloud amount of each cell when the sky above a specific point is divided so as to include two or more cells in the height direction, the cloud amount in the solar direction at the designated first point is calculated. Solar cloud amount calculation means;
   A prediction system comprising: prediction means for predicting a value to be predicted, which is an item relating to solar energy at the first point, using the solar cloud amount.
2. The cloud amount of each cell is determined by dividing the ground surface including the first point and the sky above each cell so as to include two or more cells in the horizontal direction and two or more cells in the height direction. Cloud cover,
   The prediction system according to claim 1, wherein the solar direction cloud amount calculation unit calculates the solar direction cloud amount based on a cloud amount of a cell through which a straight line connecting the first point and the sun passes among the cells.
3. The solar cloud amount calculation unit calculates the solar cloud amount based on a cloud amount of a cell passing through a straight line connecting the first point and the sun and a passing distance of the straight line in the cell. The prediction system described in.
4. The prediction system according to claim 1, wherein the division in the height direction is performed based on a distance length.
5. The prediction system according to any one of claims 1 to 3, wherein the division in the height direction is performed based on a magnitude of atmospheric pressure.
6. The cloud amount of each cell is determined by dividing the ground surface including the first point and the sky above it by a predetermined lattice point on the ground surface and two or more atmospheric pressure surfaces above the ground surface. Cloud cover,
   The solar cloud amount calculation means calculates the humidity in the region corresponding to the cell based on the temperature, humidity, and atmospheric pressure at each of the lattice points and the relative humidity at each of the atmospheric pressure surfaces as the cloud amount of the cell. The prediction system according to any one of claims 1 to 5, wherein the solar cloud amount is calculated based on the calculated wet number.
7. The cloud amount of each cell is the cloud amount of each cell when the ground surface including the first point and the sky above are divided into two or more cells in the horizontal direction and divided into two or more cells in the height direction. And
   The solar direction cloud amount calculation means is a cell through which a straight line connecting the first point and the sun passes through each of the cells, and based on the cloud amount of a cell within the range of the maximum cloud top height and the cloud bottom height, The prediction system according to any one of claims 1 to 6, wherein the solar cloud amount is calculated.
8. The solar direction cloud amount calculation means calculates the solar direction cloud amount after giving a weight to the cloud amount of each cell based on a position in a height direction. The prediction system described in Crab.
9. The prediction system according to claim 8, wherein the weight is smaller as the position in the height direction is higher.
10. Based on the cloud amount of each cell when the sky above a specific point is divided so as to include two or more cells in the height direction, the cloud amount in the solar direction at the designated first point is calculated. An information processing apparatus comprising solar direction cloud amount calculation means.
11. Information processing device
    Based on the cloud amount of each cell when the sky above a specific point is divided to include two or more cells in the height direction, the cloud amount in the solar direction at the designated first point is calculated. ,
    A prediction method that predicts a prediction target value, which is an item related to solar energy at the first point, using the cloud amount in the solar direction.
12. On the computer,
    Based on the cloud amount of each cell when the sky above a specific point is divided so as to include two or more cells in the height direction, the cloud amount in the solar direction at the designated first point is calculated. Processing,
    The prediction program for performing the process which estimates the value of the prediction object which is an item regarding the solar energy in the said 1st point using the said solar direction cloud amount.

Images