JP2010283139A - Production-of-electricity prediction device, and production-of-electricity prediction method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To avoid expansion of error of a predicted value of production of electricity by a photovoltaic power generation system even when weather forecast does not guess right. <P>SOLUTION: A production-of-electricity prediction device predicts production of electricity in a predetermined day of the photovoltaic power generation system. The device inputs a value provided by digitalizing a taken image of the western sky of an installation point where the photovoltaic power generation system is installed in a time period containing a time of sunset on the day before a predetermined day, and a value provided by digitalizing weather forecast of a predetermined day of a place containing the installation point, into a prediction formula derived based on an actual achievement value of the production of electricity, to calculate the production of electricity of the predetermined day. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a power generation amount prediction apparatus and a power generation amount prediction method for predicting a power generation amount of a solar power generation system.

Solar power generation is attracting attention as an infinite energy that does not depend on fossil fuels such as oil. Further, the value of solar power generation is expected to be further improved in the future due to the issue of CO ₂ emission rights and the introduction of the Renewable Portfolio Standard (RPS) system.

  On the other hand, a solar power generation system that performs solar power generation is a very unstable power generation facility whose power generation amount varies depending on weather conditions. Therefore, if the photovoltaic power generation system can predict the amount of power generation, the applicable range will be further expanded.

  Here, the power generation amount of the photovoltaic power generation system is substantially proportional to the amount of solar radiation, and the amount of solar radiation depends on the weather. Focusing on this point, for example, Patent Literature 1 discloses a technique for using a weather forecast as input information when predicting the power generation amount of a photovoltaic power generation system.

Japanese Patent No. 3984604

  When the power generation amount of the photovoltaic power generation system is predicted using only the weather forecast as input information as in the technique disclosed in Patent Document 1 described above, the error of the weather forecast is the error of the predicted value of the power generation amount as it is. End up. That is, when the weather forecast is not correct, there is a problem that an error in the predicted value of the power generation amount becomes large.

  The present invention provides a power generation amount prediction device and a power generation amount prediction method capable of avoiding an increase in an error in a prediction value of a power generation amount by a solar power generation system even when the weather forecast is not correct. With the goal.

In order to achieve the above object, the present invention provides:
A power generation amount prediction device for predicting a power generation amount on a predetermined day of a solar power generation system,
A value obtained by digitizing an image obtained by photographing the west sky of the installation point where the photovoltaic power generation system is installed in a time zone including the sunset time of the day before the predetermined day, and the predetermined of the area including the installation point The power generation amount of the predetermined day is calculated by inputting a numerical value of the weather forecast of the day of the day into a prediction formula derived based on the actual value of the power generation amount.

Moreover, it is a power generation amount prediction method in a power generation amount prediction device that predicts a power generation amount on a predetermined day of the solar power generation system,
A value obtained by digitizing an image obtained by photographing the west sky of the installation point where the photovoltaic power generation system is installed in a time zone including the sunset time of the day before the predetermined day, and the predetermined of the area including the installation point A power generation amount prediction process for calculating the power generation amount on the predetermined day by inputting a numerical value of the weather forecast of the day on a prediction formula derived based on the actual value of the power generation amount .

  According to the present invention, the power generation amount prediction device for predicting the power generation amount on a predetermined day of the solar power generation system is the installation point where the solar power generation system is installed in the time zone including the sunset time the day before the predetermined day. To the prediction formula derived based on the actual value of the power generation amount, the value obtained by digitizing the image of the west sky of the city and the value obtained by digitizing the weather forecast for a given day in the area including the installation point By inputting, the power generation amount on a predetermined day is calculated.

  Therefore, even when the weather forecast is not correct, it is possible to avoid an increase in the error of the predicted value of the power generation amount.

It is a block diagram which shows the structure of one Embodiment of the photovoltaic power generation amount prediction system to which the power generation amount prediction apparatus of this invention is applied. It is a figure for demonstrating operation | movement of the electric power generation amount prediction apparatus shown in FIG. FIG. 3 is a diagram for explaining an example of a method for calculating an image value by the image processing unit shown in FIG. 1 and FIG. 2, and (a) shows a case where an image value is calculated by dividing an image into horizontal stripes. FIG. 4B is a diagram illustrating a case where an image value is calculated by dividing an image into a grid, and FIG. 4C is a diagram illustrating a case where an image value is calculated using predetermined coordinate points in the image. . It is a figure for demonstrating the other example of the method by which the image process part shown in FIG.1 and FIG.2 calculates an image value. It is a flowchart for demonstrating the operation | movement which the electric power generation amount prediction apparatus shown in FIG.1 and FIG.2 estimates the electric power generation amount of a solar power generation system.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a block diagram showing a configuration of an embodiment of a photovoltaic power generation amount prediction system to which a power generation amount prediction apparatus of the present invention is applied.

  As shown in FIG. 1, the photovoltaic power generation amount prediction system of this embodiment includes a power generation amount prediction device 10, a solar power generation system 20, and a power conditioner 30.

  The photovoltaic power generation system 20 converts sunlight energy into electric power.

  The power conditioner 30 is connected to the power system 40 and the customer 50, converts the power generated by the solar power generation system 20, and supplies it to the power system 40 and the customer 50. Further, the power conditioner 30 measures the actual value of the power generation amount of the photovoltaic power generation system 20 for each time zone, and outputs the actual value of the measured power generation amount for each time zone to the power generation amount prediction device 10.

  The power generation amount prediction apparatus 10 includes a camera 11 that is a photographing unit, a control unit 12, a prediction unit 13, and a weather forecast reception unit 14.

  The camera 11 is fixed to face the west direction in order to photograph the west sky of the installation point where the photovoltaic power generation system 20 is installed. Hereinafter, the installation point where the solar power generation system 20 is installed is simply referred to as an installation point. The camera 11 shoots the sky in the west of the installation point when receiving a shooting instruction output from the control unit 12 and instructing to shoot an image. Then, the captured image is output to the control unit 12.

  The control unit 12 includes an imaging control unit 12-1 and an image processing unit 12-2.

  The imaging control unit 12-1 stores calendar information and position information of the installation point, and calculates the sunset time of the installation point based on them. The position information is, for example, latitude and longitude. Then, the imaging control unit 12-1 outputs an imaging instruction to the camera 11 in a time zone including the sunset time at the installation point. The time zone including sunset time is, for example, sunset time, 30 minutes before sunset time, 30 minutes after sunset time, and the like. In this case, the shooting instruction may be output at any one of these times, or the shooting instruction may be output at all these times. In Japan, it can be said that there is a high correlation between the state of the west sky in the time zone including the sunset time and the weather of the next day. This is because westerly winds blow over Japan, and the weather generally changes from west to east.

  The image processing unit 12-2 receives an image output from the camera 11. Then, an image value that is a value obtained by digitizing the received image is calculated. Then, the image processing unit 12-2 outputs the calculated image value to the prediction unit 13. The image received by the image processing unit 12-2 is expressed by RGB (Red, Green, Blue) values for each coordinate point constituting the image. The RGB value is a value indicating a color, and is a value based on three colors of red, green, and blue.

  Hereinafter, an operation in which the image processing unit 12-2 calculates an image value will be described in detail.

  FIG. 2 is a diagram for explaining the operation of the power generation amount prediction apparatus 10 shown in FIG.

  When receiving the image output from the camera 11, the image processing unit 12-2 converts the RGB value of the received image into an HSV (Hue, Saturation, Value) value as shown in FIG. The HSV value is a value indicating a color, and is a value based on hue, saturation, and lightness. Note that the image processing unit 12-2 does not have to convert an image into an HSV value. Then, the image processing unit 12-2 calculates the image value of the received image by a predetermined method.

  FIG. 3 is a diagram for explaining an example of a method by which the image processing unit 12-2 illustrated in FIGS. 1 and 2 calculates an image value. FIG. 3A illustrates an image obtained by dividing an image into horizontal stripes. The figure which shows the case where a value is calculated, (b) is the figure which shows the case where an image value is calculated by dividing | segmenting an image into a grid | lattice form, (c) is an image value calculated using the predetermined coordinate point in an image. It is a figure which shows the case where it does.

  In the case of a method of calculating an image value by dividing the received image 100 into a plurality of regions as shown in FIGS. 3A and 3B, the image processing unit 12-2 has the divided divided image 101-. The average value of the HSV values of the coordinate points constituting 1, 101-2 is calculated for each of the divided images 101-1, 101-2. Each of these average values becomes an image value. When the RGB value of the received image is not converted into the HSV value, the image processing unit 12-2 calculates the average value of the RGB values of the coordinate points constituting the divided divided image as the divided image 101-1. , 101-2.

  Here, the method of calculating the image value is not limited to the method of calculating by dividing the image 100 into a plurality of regions as described above. For example, as shown in FIG. 3C, HSV values or RGB values at a plurality of predetermined coordinate points 101-3 in the image 100 may be used as image values. Further, the image value may be calculated by the following method.

  FIG. 4 is a diagram for explaining another example of a method in which the image processing unit 12-2 illustrated in FIGS. 1 and 2 calculates an image value.

  In this method, as shown in FIG. 4, the weather is classified into a plurality of predetermined types. For example, “sunny”, “cloudy”, “rain”, “snow”, and the like. Then, the image is associated with any type of weather according to the distribution of the average value of the HSV values or the RGB values for each calculated divided image 101-2 (see FIG. 3B). Then, a type number that is a number indicating the associated type of weather is used as the image value. Here, the case where the image 100 is divided into a lattice shape has been described as an example. However, this method can be used even when the image 100 is divided into horizontal stripes as shown in FIG.

  Referring to FIG. 1 again, the weather forecast receiving unit 14 receives information indicating the weather forecast of the area including the installation point transmitted from the weather forecast provider 60. And the weather forecast value which is the value which digitized the received information is calculated for every time slot | zone. As weather forecast values, for example, 1 is set for sunny, 2 for cloudy, 3 for rain, and 4 for snow. Then, the calculated weather forecast value for each time zone is output to the prediction unit 13. The weather forecast provider 60 is, for example, the Japan Meteorological Agency or a weather information providing service company.

  The prediction unit 13 includes a predicted value calculation unit 13-1 and a database 13-2.

  The database 13-2 includes the actual value for each time zone of the amount of power generated by the photovoltaic power generation system 20 output from the power conditioner 30, the image value output from the image processing unit 12-2, and the weather forecast receiving unit 14. The weather forecast value for each time zone output from is stored.

  As shown in FIG. 2, the predicted value calculation unit 13-1 predicts the power generation amount of the solar power generation system 20 from the actual power generation value stored in the database 13-2, the image value, and the weather forecast value. The prediction formula for this is derived by multiple regression analysis. Specifically, the predicted value calculation unit 13-1 includes the actual value for each time zone of the power generation amount before the prediction day to predict the power generation amount and the weather forecast for each time zone before the day before the prediction target date. Perform multiple regression analysis based on the values. At this time, the power generation amount in the time zone m is set as an objective variable, and the weather forecast value and the image value in the time zone m are set as explanatory variables. The regression model in this case is shown in the following formula (1).

Y _m = β ₀ + β ₁ X ₁ + β ₂ X ₂ + β ₃ X ₃ +... + Β _n X _n + u (1)
In the above equation (1), Y _m is an objective variable, β ₀ is a constant term, X _{1 to} X _n are explanatory variables, β _{1 to} β _n are regression coefficients, and u is an error term. It is.

The predicted value calculation unit 13-1 obtains regression coefficients β _{1 to} β _n and a constant term β ₀ using the least square method, and derives a prediction formula for the power generation amount for each time zone. The presence of a plurality of explanatory variables and constant terms indicates that there are a plurality of image values calculated by the image processing unit 12-2, and the number of terms is the number of the image processing unit 12-. 2 differs depending on the method of calculating the image value.

  Then, as shown in FIG. 2, the prediction value calculation unit 13-1 outputs the weather forecast value in the time zone m of the prediction target day output from the weather forecast reception unit 14 and the prediction target day output from the control unit 12. Accepts the image value of the image taken on the previous day. Then, by inputting the received weather forecast value and image value into the derived prediction formula for the time zone m, a predicted value of the power generation amount in the time zone m of the prediction target day is calculated.

  Below, the operation | movement in which the electric power generation amount prediction apparatus 10 estimates the electric power generation amount of the solar power generation system 20 in the solar power generation amount prediction system comprised as mentioned above is demonstrated.

  FIG. 5 is a flowchart for explaining an operation in which the power generation amount prediction apparatus 10 shown in FIGS. 1 and 2 predicts the power generation amount of the solar power generation system 20.

  Here, in the database 13-2, the actual value for each time zone of the power generation amount before the previous day of the prediction target day, the weather forecast value for each time zone before the previous day of the prediction target day, and the prediction target day before It is assumed that image values of images taken by the camera 11 before the date are stored. Here, it is assumed that an image photographed by the camera 11 is divided into a lattice shape, and an average value of HSV values calculated for each divided divided image is used as an image value.

  First, the imaging control unit 12-1 of the control unit 12 calculates the sunset time on the day before the prediction target date of the installation point based on the calendar information and the installation point information. Then, the shooting control unit 12-1 outputs a shooting instruction to the camera 11 in a time zone including the sunset time on the day before the prediction target date.

  The camera 11 that has received the shooting instruction output from the shooting control unit 12-1 takes an image (step S1).

  Then, the camera 11 outputs the captured image to the image processing unit 12-2 of the control unit 12.

  The image processing unit 12-2 that has received the image output from the camera 11 converts the RGB value of the received image into an HSV value (step S2).

  Next, the image processing unit 12-2 divides the received image into a plurality of grid-like regions (step S3).

  Next, the image processing unit 12-2 calculates an average value of HSV values for each divided image (step S4).

  Then, the image processing unit 12-2 outputs the average value of the calculated HSV values for each divided image to the predicted value calculation unit 13-1 as an image value.

  In addition, the weather forecast receiving unit 14 receives information indicating the weather forecast for the prediction target day transmitted from the weather forecast provider 60. And the weather forecast value for every time slot | zone is calculated from the received information (step S5).

  Then, the weather forecast receiver 14 outputs the calculated weather forecast value for each time zone to the predicted value calculator 13-1.

  The prediction value calculation unit 13-1 includes the actual value for each time zone of the power generation amount before the day before the prediction target day, the weather forecast value for each time zone before the day before the prediction target day, and the day before the prediction target day. Image values of images previously captured by the camera 11 are acquired from the database 13-2. And the prediction formula of the time slot | zone m of a prediction object day is derived | led-out by multiple regression analysis from those acquired values (step S6).

  Next, the predicted value calculation unit 13-1 receives the image value output from the image processing unit 12-2 and the weather forecast value for each time zone output from the weather forecast receiving unit 14.

  Then, the predicted value calculation unit 13-1 converts the weather forecast value for the time zone m and the received image value among the weather forecast values for each received time zone into the prediction formula for the time zone m derived in step S6. Input (step S7).

  Thereby, the predicted value of the power generation amount in the time zone m of the prediction target day is calculated (step S8). By adding the predicted value of the power generation amount for each calculated time zone, the predicted value of the power generation amount on the prediction target day can be calculated.

  As described above, in this embodiment, the power generation amount prediction apparatus 10 digitizes an image obtained by photographing the west sky of the installation point where the photovoltaic power generation system 20 is installed in the time zone including the sunset time on the day before the prediction target date. Power generation amount on the prediction target day by inputting the calculated value and the numerical value of the weather forecast for a given day in the area including the installation point into the prediction formula derived based on the actual power generation amount Is calculated.

  Therefore, even when the weather forecast is not correct, it is possible to avoid an increase in prediction error.

  Strictly speaking, the weather forecast is not a weather forecast of the installation point where the solar cell system is installed. In this invention, since the image in the installation point in which the solar power generation system is installed is used, the weather in an installation point can be estimated more correctly. That is, it is possible to predict the power generation amount with higher accuracy than when only the weather forecast is used.

DESCRIPTION OF SYMBOLS 10 Power generation amount prediction apparatus 11 Camera 12 Control part 12-1 Imaging | photography control part 12-2 Image processing part 13 Prediction part 13-1 Predicted value calculation part 13-2 Database 14 Weather information receiving part 20 Solar power generation system 30 Power conditioner 40 Electricity system 50 Customer 60 Weather forecast provider 100 Image 101-1 and 101-2 Divided image 101-3 Predetermined coordinate point

Claims (10)

A power generation amount prediction device for predicting a power generation amount on a predetermined day of a solar power generation system,
A value obtained by digitizing an image obtained by photographing the west sky of the installation point where the photovoltaic power generation system is installed in a time zone including the sunset time of the day before the predetermined day, and the predetermined of the area including the installation point A power generation amount prediction device that calculates the power generation amount on the predetermined day by inputting a value obtained by digitizing the weather forecast of the day into a prediction formula derived based on the actual value of the power generation amount. In the power generation amount prediction apparatus according to claim 1,
An imaging unit that images the west sky of the installation point in a time zone including the sunset time of the installation point, and outputs the captured image;
An image processing unit that calculates an image value obtained by quantifying the image based on a color indicated by the image output from the photographing unit, and outputs the calculated image value;
A weather forecast receiving unit that receives information indicating the weather forecast of the area including the installation point, calculates a weather forecast value obtained by quantifying the weather forecast from the received information, and outputs the calculated weather forecast value When,
Deriving the prediction formula from the actual value of the power generation amount, the image value output from the image processing unit, and the weather forecast value output from the weather forecast receiving unit, the weather forecast of the predetermined day When the weather forecast value and the image value of the image taken on the previous day of the predetermined day are received, the received weather forecast value and the image value are input to the prediction formula, thereby the predetermined day. A power generation amount prediction apparatus comprising: a prediction unit that calculates the power generation amount. In the power generation amount prediction apparatus according to claim 2,
The weather forecast value is a numerical value of a weather forecast for a predetermined time zone,
The actual value of the power generation amount is the actual value of the power generation amount in the predetermined time zone,
The prediction unit derives the prediction formula of the predetermined time zone from the actual value of the power generation amount, the image value, and the weather forecast value, and the weather forecast value of the weather forecast of the predetermined day When the image value of the image taken on the day before the predetermined day is received, the received weather forecast value and the image value are input to the prediction formula of the predetermined time zone, thereby A power generation amount predicting device that calculates the power generation amount of each day for each time zone. In the power generation amount prediction apparatus according to claim 2 or claim 3,
The image processing unit divides the received image into a plurality of regions, calculates a plurality of values obtained by quantifying each of the plurality of divided images based on colors indicated by the divided plurality of divided images, A power generation amount prediction apparatus using the calculated plurality of values as the image value. In the power generation amount prediction apparatus according to claim 2 or claim 3,
The power generation amount prediction apparatus, wherein the image processing unit uses a value obtained by digitizing the image based on a color indicated by a predetermined coordinate point among coordinate points constituting the received image as the image value. In the power generation amount prediction apparatus according to any one of claims 2 to 5,
The image value is a power generation amount prediction device calculated based on RGB values. In the power generation amount prediction apparatus according to any one of claims 2 to 5,
The image value is a power generation amount prediction device calculated based on an HSV value. A power generation amount prediction method in a power generation amount prediction device that predicts a power generation amount on a predetermined day of a solar power generation system,
A value obtained by digitizing an image obtained by photographing the west sky of the installation point where the photovoltaic power generation system is installed in a time zone including the sunset time of the day before the predetermined day, and the predetermined of the area including the installation point A power generation amount prediction process for calculating the power generation amount on the predetermined day by inputting a numerical value of the weather forecast of the day on a prediction formula derived based on the actual value of the power generation amount Power generation forecasting method. The power generation amount prediction method according to claim 8,
The power generation amount prediction process is:
Processing of photographing the west sky of the installation point in a time zone including the sunset time of the installation point;
A process of calculating an image value obtained by digitizing the image based on a color indicated by the captured image;
A weather forecast value calculation process for receiving information indicating a weather forecast for a region including the installation point, and calculating a weather forecast value obtained by quantifying the weather forecast from the received information;
A prediction formula derivation process for deriving the prediction formula from the actual value of the power generation amount, the image value, and the weather forecast value;
The power generation amount of the predetermined day is calculated by inputting the weather forecast value of the weather forecast of the predetermined day and the image value of the image taken on the previous day of the predetermined day into the prediction formula. And a power generation amount prediction process. The power generation amount prediction method according to claim 9,
The weather forecast value is a numerical value of a weather forecast for a predetermined time zone,
The actual value of the power generation amount is the actual value of the power generation amount in the predetermined time zone,
The prediction formula deriving process is a process of deriving the prediction formula for the predetermined time zone from the actual value of the power generation amount, the image value, and the weather forecast value,
The power generation amount calculation process inputs the weather forecast value of the weather forecast on the predetermined day and the image value of an image captured on the day before the predetermined day to the prediction formula for the predetermined time zone. Thus, the power generation amount prediction method is a process of calculating the power generation amount on the predetermined day for each time zone.

Images