WO2019225064A1

WO2019225064A1 - Weather prediction device, weather prediction method, and wind-power-generation output estimation device

Info

Publication number: WO2019225064A1
Application number: PCT/JP2019/003787
Authority: WO
Inventors: 征史深谷; 昌道中村
Original assignee: 株式会社日立製作所
Priority date: 2018-05-22
Filing date: 2019-02-04
Publication date: 2019-11-28
Also published as: JP2019203728A

Abstract

Provided are a weather prediction device, a weather prediction method, and a wind-power-generation output estimation device with which it is possible to highly accurately predict the state of wind speed at an evaluation point in various wind conditions in consideration of the effect of specific terrain. A weather prediction device characterized by being provided with: a global wind conditions analysis unit for predicting the wind speed/wind direction distribution in a space that includes an evaluation point from global wind conditions; a boundary condition extraction unit for analysis of local wind conditions, the boundary condition extraction unit extracting, from numerous wind speed/wind direction distributions obtained for each moment by the global wind conditions analysis unit, a representative wind speed distribution in the height direction at the evaluation point; a local wind conditions analysis unit for predicting the wind speed/wind direction distribution in the space that includes the evaluation point, where the representative wind speed distribution for each reference wind direction as obtained by the boundary condition extraction unit for analysis of local wind conditions is used as a boundary condition; and a correction coefficient calculation unit for obtaining a different correction coefficient for each reference wind direction by comparing the wind speed distribution at the evaluation point as predicted by the local wind conditions analysis unit and the representative wind speed distribution established by the boundary condition extraction unit.

Description

Weather prediction device, weather prediction method, and wind power generation output estimation device

The present invention relates to a weather forecasting device, a weather forecasting method, and a wind power generation output estimating device capable of estimating wind power generation output using weather prediction and prediction results regarding wind conditions.

Currently, the annual power generation of wind turbines is predicted based on the results of wind conditions analysis based on a weather prediction model. For example, wind turbines are analyzed in a large area using a numerical simulator such as the weather model WRF (Weather research and forecast) and regional weather model RAMS (Regional Atmospheric Modeling System) to cover approximately half of the Japanese archipelago. Predict changes in wind speed at the installation location.

However, in this global wind analysis, there is a limit to increasing the resolution of topography from the viewpoint of calculation time and the non-slip condition cannot be given on the ground surface, so the atmospheric boundary layer especially in the range from the ground surface to about 1000m above sea level. The prediction accuracy of the wind speed change is not enough.

Therefore, as a means of improving accuracy, the global wind analysis results are used as boundary conditions, and a nonlinear wind prediction model MASCOT (Microclimate Analysis System Complex Terrain) or Liam Compact RIAM-COMPACT (registered trademark: Research Institute for AppliedMappliedM). There is a method for predicting local wind conditions within a few kilometers around the wind turbine using another numerical simulator such as University, Computational Prediction of Airflow over Complex Terrain. In this local wind condition analysis, it is possible to increase the resolution of the topography and to provide a non-slip condition on the ground surface, so that it is possible to predict the wind speed change at the wind turbine installation position with high accuracy.

For example, Non-Patent Document 1 is known as a method of combining the global wind analysis and the local wind analysis described above. In Chapter 3 and Chapter 4 of Non-Patent Document 1, it is described that the result of the global wind analysis is statistically processed and used as a calculation condition for the local wind analysis.

According to the non-patent document 1, the direction of combining the global wind analysis and the local wind analysis is specified, but a specific statistical processing method that can realize this is described. Absent.

In order to obtain a highly accurate weather prediction system that can be realized by the above combination, for example, it is necessary to consider the processing capacity of the computer.

When estimating the annual power generation of a windmill, for example, a wind speed change every 10 minutes is calculated using a numerical simulator for global wind conditions analysis such as WRF and RAMS, and the obtained wind speed is converted into a wind turbine power curve (wind speed and power generation amount). To the amount of power generation via A predicted value of the annual power generation amount can be obtained by time integration of this power generation amount over one year. At that time, 60 (minutes) / 10 (minutes) x 24 (hours) x 365 (days) = 52560 different wind conditions are obtained. For all these cases, boundary conditions for local wind condition analysis are set. It is generally not realistic from the viewpoint of calculation time to extract and perform local wind analysis using, for example, MASCOT or RIAM-COMPACT.

Therefore, the different wind conditions in 52560 cases are classified into several typical wind conditions, and local wind conditions are analyzed using these as boundary conditions. Based on the relationship between the typical wind conditions and the wind speed at the wind turbine installation position, 52560 It is necessary to correct all the case wind conditions.

From the above, in the present invention, while considering the influence of detailed terrain, a weather prediction device, a weather prediction method, and the like that can predict the wind speed state of the evaluation point in various wind conditions by reflecting the detailed terrain, An object of the present invention is to provide a wind power generation output estimation device.

In view of the above, in the present invention, “a global wind state analysis unit that predicts the wind speed / wind direction distribution in the space including the evaluation point from the wind state in the global region and a large number of time points obtained by the global wind state analysis unit at each time. For each reference wind direction obtained by the boundary condition extraction unit for local wind analysis that extracts typical wind speed distribution in the height direction at the evaluation point from the wind speed and wind direction distribution, and the boundary condition extraction unit for local wind analysis The local wind condition analysis unit predicts the wind speed and direction distribution in the space including the evaluation point, and the wind speed distribution and boundary condition extraction unit predicted by the local wind condition analysis unit. The weather prediction apparatus includes a correction coefficient calculation unit that obtains a different correction coefficient for each reference wind direction by comparing the obtained representative wind speed distributions.

Further, in the present invention, “a wind power generation output estimation device using a weather prediction device, which calculates a long-term power generation amount by obtaining a wind condition in consideration of a correction coefficient with respect to a predicted value of a long-term wind condition. It is a wind power generation output estimation device characterized by adding a power generation output estimation unit.

Further, in the present invention, “a wind speed / wind direction distribution in a space including an evaluation point is predicted from a wind condition in a large area, and a number of wind speed / wind direction distributions obtained at each time are used to represent the height direction at the evaluation point. A typical wind speed distribution, using the typical wind speed distribution as a boundary condition, predicting the wind speed and wind direction distribution in the space including the evaluation point, and comparing the wind speed distribution at the evaluation point with the typical wind speed distribution for each reference wind direction The weather forecasting method is characterized by obtaining different correction coefficients.

Further, in the present invention, “a weather prediction method in which a calculation result of global wind analysis is used as a boundary condition for local wind analysis, and the calculation result of global wind analysis is corrected using the calculation result. The calculation results of the analysis are classified according to the reference wind direction at the reference elevation of the evaluation point, and for each reference wind direction, the average wind speed at the reference elevation is calculated as the reference wind speed. From the correlation, calculate the proportional coefficient with the highest correlation for each altitude, add the proportional coefficient to the reference wind speed to obtain the representative wind speed distribution in the height direction, and use this representative wind speed distribution as the boundary condition for local wind analysis "Weather forecasting method characterized by that".

Further, in the present invention, "the weather prediction method that uses the calculation result of the global wind analysis as a boundary condition of the local wind analysis, and corrects the calculation result of the global wind analysis using the calculation result,
The calculation results of the global wind condition analysis are classified according to the reference wind direction at the reference elevation at the evaluation point, and for each reference wind direction, the average wind speed at the reference elevation is calculated as the reference wind speed. From the correlation of wind speeds, calculate the proportional coefficient with the highest correlation for each altitude, and add the proportional coefficient to the reference wind speed to obtain the representative wind speed distribution in the height direction. The weather forecasting method is characterized by the boundary condition of

According to the present invention, it is possible to predict the wind speed state at the evaluation point in various wind conditions by reflecting the detailed topography while considering the influence of the detailed topography.

The block diagram which shows the structure of the weather prediction apparatus and wind power generation output estimation apparatus which concern on Example 1. FIG. The figure which shows an example of the result of having classified the wind speed distribution by the wind direction (reference | standard wind direction) of 16 directions in the reference | standard altitude of the evaluation point which concerns on Example 1. FIG. Wind speed _{V 0} which the reference altitude _{h 0,} shows the relationship between the wind speed _{V 1} in the elevations _{h 1.} Wind speed _{V 0} which the reference altitude _{h 0,} shows the relationship between the wind speed _{V 2} at altitude _{h 2.} The figure which shows the representative wind speed distribution 501 of the height direction which concerns on Example 1. FIG. The figure which shows the representative wind speed distribution 501 of a height direction, and the wind speed distribution 601 estimated using it as a boundary condition. The figure which shows another example of the result of having classified the wind speed distribution by the wind direction of 16 directions (reference | standard wind direction) in the reference | standard altitude of the evaluation point which concerns on Example 2. FIG. The figure which shows the representative wind speed distribution 801 of the height direction which concerns on Example 2. FIG.

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

FIG. 1 is a schematic block diagram showing a configuration example of a weather prediction apparatus and a wind power generation output estimation apparatus according to Embodiment 1 of the present invention. FIG. 1 shows a configuration as a weather prediction device, but a wind power generation output estimation device is a device in which a function for estimating an annual power generation amount of a windmill is added to the weather prediction device.

The weather forecasting device can be broadly classified into these functions: a global wind condition analysis unit 101, a boundary condition extraction unit 102 for local wind condition analysis, a local wind condition analysis unit 103, a local correction coefficient calculation unit 104, and a global wind condition analysis result. Each function of the correction unit 105 and the output unit 106 is configured.

Of these, the global wind condition analysis unit 101 uses numerical simulators of meso-meteorological models such as WRF and RAMS. For example, about half the area of the Japanese archipelago, the wind speed and wind direction distribution in the space including the evaluation point (the place where the windmill is planned to be installed) is predicted using the meteorological data and topographic data as input.

In addition, as the weather data handled here, GPV (Grid Point Value) which is weather forecast data at a plurality of times is input. As an example of GPV, weather forecast data at a time of every 3 hours at each grid point at intervals of 5 km can be mentioned. GPV is forecast data calculated by a numerical prediction device different from the prediction unit. The GPV can be obtained from, for example, a weather service support center in Japan. The GPV includes weather information indicating the weather at a certain point.

Also, as topographic data, for example, altitude data of 10 m intervals issued by the Geospatial Information Authority of Japan are input, but the global wind condition analysis unit 101 can only use altitude data of about 1 km intervals for calculation. Therefore, the calculation is based on rough terrain data. As a result of the calculation by the global wind condition analysis unit 101, a large number of wind speed / wind direction distributions in the space including the evaluation point are obtained for each time.

In the wind power generation output estimation apparatus according to the present invention, a process of finally estimating the annual power generation amount of the wind turbine is performed. At this time, for example, a numerical simulator for global wind condition analysis such as WRF and RAMS is used for 10 minutes. The wind speed change for each is calculated. However, since weather information such as wind speed in the past is measurement information in units of time (for example, 3 hours), it is necessary to estimate the wind speed every 10 minutes. For this reason, in the global wind condition analysis part 101, the process which estimates the wind speed for every 10 minutes is further performed by the time interpolation process.

In the boundary condition extraction unit 102 for local wind condition analysis, a representative representative of the height direction at the evaluation point (scheduled installation point of the windmill) is obtained from the large number of wind speed / wind direction distributions obtained at each time by the global wind condition analysis unit 101. A simple wind speed distribution.

Here, an example of a large number of wind speeds and wind direction distributions obtained at each time by the global wind condition analysis unit 101 is illustrated in FIG. 2, and an evaluation point (extracted by the boundary condition extraction unit 102 for local wind condition analysis) ( A typical wind speed distribution in the height direction at the installation point of the windmill is illustrated as 501 in FIG. In the following description, FIGS. 3 and 4 are used as supplemental explanatory materials for the concept of obtaining the wind speed distribution 501 in FIG. 5 from the large number of wind speeds and wind direction distributions obtained at each time by the global wind condition analysis unit 101 in FIG. It explains using.

FIG. 2 shows an example of the result of classifying the wind speed distribution according to the wind direction of 16 directions (reference wind direction) at the reference altitude of the evaluation point (scheduled wind turbine installation point). FIG. 2 is a schematic diagram of the wind speed distribution at the evaluation points obtained for each time by the global wind condition analysis unit 101, and shows an example in which the reference wind direction is north-northwest. Therefore, the wind speed distribution of FIG. 2 is obtained for every 16 azimuths. In the following description, the case of north-northwest will be described as a representative example, but the handling of other directions is the same.

In FIG. 2, the horizontal axis is the altitude and the vertical axis is the wind speed, the altitude h _{1 on the} vertical axis is the height at the evaluation point (where the wind turbine is planned to be installed), and the reference altitude h ₀ is the center position of the wind turbine installed at the evaluation point. height at altitude _{h 2} is over 1000m of evaluation points, for example.

Here, the wind speed distribution 201-210 of 10 cases is shown for the wind whose reference wind direction is north-northwest. The wind conditions vary greatly depending on the time. For example, in the wind speed distribution 210, the wind speed in the sky is smaller than the vicinity of the ground surface. However, overall, there is a tendency for wind speed to generally increase with increasing altitude. A graph as shown in FIG. 2 is obtained for each of 16 reference air directions.

In FIG. 2, the boundary condition extraction unit 102 for local wind condition analysis calculates the average wind speed of 10 cases at the reference altitude h ₀ and _sets this as the reference wind speed V _{0_ave} . Next, the relationship between the wind speed at the reference altitude h _{0 and} the wind speed at an altitude h _i (i = 1, 2, :) different from the reference altitude is graphed.

3, the wind speed V ₀ which the reference altitude h ₀ on the horizontal axis and the vertical axis represents the wind velocity V ₁ in the altitude h _1, is a diagram showing a relationship between them. Here, the altitude h ₁ is the height at the windmill installation point, the reference altitude h ₀ is the height at the center position of the windmill installed at the evaluation point, and the altitude h ₁ is relatively close to the reference altitude h ₀ . It shows that the correlation between V ₀ and V ₁ is high. Incidentally, the characteristics showing the correlation between V ₀ and V _1, when pulling the least squares straight line 301 passing through the origin, can be expressed by equation (1) using the proportionality factor K _1.
[Equation 1]
V ₁ = K ₁ × V ₀ (1)
Similarly, FIG. 4 shows the case of the wind speed V _{2 at} an altitude h ₂ having a large difference from the reference altitude h ₀ . Although the correlation between V ₀ and V ₂ is not necessarily high, when the least-squares line 401 passing through the origin is drawn with respect to the characteristic indicating the correlation between V ₀ and V ₂ , the proportional coefficient K ₂ is used to express Can be represented.
[Equation 2]
V ₂ = K ₂ × V ₀ (2)
Thus, the wind speed V _{i at an} altitude h _i (i = 1, 2, :) different from the reference altitude h ₀ can be expressed by the equation (3).
[Equation 3]
V _i = K _i × V ₀ (i = 1, 2, :) (3)
As a result, by adding the proportionality coefficient K _i to the reference wind speed V _{0_ave} , the representative wind speed V _Ri is obtained as shown in Equation (4), and a representative wind speed distribution 501 in the height direction as shown in FIG. 5 is obtained.
[Equation 4]
V _Ri = K _i × V 0 — _ave (i = 1, 2, :) (4)
This representative wind speed distribution V _Ri is set as a boundary condition in the local wind condition analysis unit 103. In the first embodiment, V ₀ and V _i are related using the least

square lines

301 and 401 passing through the origin. However, the expression representing the correlation between V ₀ and V _i is not necessarily a linear expression. It does not have to pass through the origin. That is, if the function f _i that more appropriately represents the correlation between V ₀ and V _i can be expressed as in equation (5), the representative wind speed V _Ri can be calculated as in equation (6).
[Equation 5]
V _i = f _i (V ₀ ) (i = 1, 2, :) (5)
[Equation 6]
V _Ri = f _i (V 0 — _ave ) (i = 1, 2, :) (6)
As described above, the representative wind speed distribution in the height direction is obtained for each of the 16 reference air directions.

The local wind condition analysis unit 103 uses a numerical simulator such as Riam-Compact or STAR-CCM + based on the Navier-Stokes equations. For example, for a region within a few km around the evaluation point, the representative wind speed distribution for each of the 16 directional reference wind directions obtained by the boundary condition extraction unit 102 for local wind condition analysis is used as the boundary condition. Predict wind speed and wind direction distribution. As the topographic data, for example, altitude data at intervals of 10 m issued by the Geospatial Information Authority of Japan are input. The local wind condition analysis unit 103 can increase the lattice resolution to about several meters as necessary, so that the calculation reflecting the more detailed terrain than the global wind condition analysis unit 101 is possible.

According to the processing of the local wind condition analysis unit 103, a calculation reflecting a detailed topography is possible. For this reason, in the numerical simulator used in the local wind condition analysis unit 103, a non-slip condition is given on the ground surface, and the wind speed on the ground surface is zero. On the other hand, since the non-slip condition is not given to the ground surface in the numerical simulator used in the global wind condition analysis unit 101, the wind speed of the representative wind speed distribution 501 on the ground surface is not zero.

FIG. 6 shows a representative wind speed distribution 501 in the height direction when the reference wind direction is north-northwest, and a wind speed distribution 601 predicted using this as a boundary condition. The representative wind speed distribution 501 in the height direction is obtained by the boundary condition extraction unit 102 for local wind condition analysis, and is shown in FIG. On the other hand, the wind speed distribution 601 predicted using the boundary condition is shown as a characteristic in which the wind speed increases from the wind speed 0 on the ground surface and gradually approaches the representative wind speed distribution 501 in the height direction. For this reason, the difference between the predicted wind speed distribution 601 and the representative wind speed distribution 501 tends to be large near the ground surface and small in the sky.

The correction coefficient calculation unit 104 compares the wind speed distribution 601 at the evaluation point predicted by the local wind condition analysis unit 103 with the representative wind speed distribution 501. Based on these comparison results in the correction coefficient calculation unit 104, the relationship between the representative wind speed distribution V _R (h) and the predicted wind speed distribution V _P (h) can be expressed by Expression (7). G (h) in (7) is a correction coefficient, and a different correction coefficient g (h) is obtained for each of the 16 reference wind directions.
[Equation 7]
V _P (h) = g (h) × V _R (h) (7)
The global wind condition analysis result correction unit 105 uses the correction coefficient g (h) described above to calculate the wind speed distribution V (h) in the height direction obtained by the global wind condition analysis unit 101 for each of 16 reference wind directions. Correction is performed as shown in equation (8).
[Equation 8]
V _mod (h) = g (h) × V (h) (8)
The output unit 106 outputs the wind speed distribution V _mod (h) corrected by the global wind condition analysis result correction unit 105. Examples of data output include recording on a recording medium, display on a display, transmission to an external device, and the like.

The configuration as the weather prediction apparatus shown in FIG. 1 is generally as described above. However, in order to realize a wind power generation output estimation apparatus, the global wind condition analysis result correction unit 105 further performs the following processing, for example. It will be a thing. This process is performed in a wind power generation output estimation unit (not shown) added to the global wind condition analysis result correction unit 105.

The processing in the wind power generation output estimation unit is performed every 10 minutes using a numerical simulator for global wind conditions analysis such as WRF and RAMS, and for the predicted value of wind conditions (wind speed, wind direction) for one year, The amount of power generation every 10 minutes based on the wind speed distribution V _mod (h) corrected for each wind direction is accumulated over the course of one year.

According to the first embodiment described above, the wind speed at the center of the wind turbine at the reference altitude h ₀ is considered in consideration of the non-slip condition even when the actual wind speed changes variously as indicated by 201 to 210 in FIG. By obtaining by correction, it is possible to cope with it easily. As a result, no matter what wind conditions occur during the estimation and prediction of the annual power generation, the wind speed at the center of the wind turbine at the reference altitude h ₀ is set by the correction coefficient g (h) at this height position. It becomes possible to simply calculate by calculating the power generation amount by using the multiplied value as a new wind speed.

The weather prediction apparatus according to the second embodiment of the present invention is different from the weather prediction apparatus according to the first embodiment in the calculation method of the representative wind speed distribution obtained by the boundary condition extraction unit 102 for local wind condition analysis.

This difference will be described with reference to FIG. FIG. 7 is a graph with the same vertical and horizontal axes as FIG. 2, but in this embodiment, not only the reference altitude h ₀ but also the average wind speed of 10 cases at an altitude h _i (i = 1, 2, :) different from the reference altitude. V _{i_ave} is calculated, and these are set as the representative wind speed V _{Ri in the} height direction as shown in equation (9).
[Equation 9]
V _Ri = V i — _ave (i = 1, 2, :) (9)
FIG. 8 shows a representative wind speed distribution 801 in the height direction thus obtained. The representative wind speed distribution 801 is different from the representative wind speed distribution 501 in the first embodiment, and the difference between the two is particularly large in the sky.

101: Global wind condition analysis unit, 102: Boundary condition extraction unit for local wind condition analysis, 103: Local wind condition analysis unit, 104: Correction coefficient calculation unit, 105: Global wind condition analysis result correction unit, 106: Output unit 201-210: Example of wind speed distribution in the height direction 301, 401: Correlation line of wind speed at the reference altitude and an altitude different from the reference altitude, 501: Example of representative wind speed distribution in the height direction, 601: Local wind Of wind speed distribution in the height direction predicted by state analysis, 801: Example of representative wind speed in the height direction

Claims

The global wind condition analysis unit predicts the wind speed and wind direction distribution in the space including the evaluation point from the wind conditions in the global region, and the evaluation point from the numerous wind speed and wind direction distributions obtained at each time by the global wind condition analysis unit. The boundary condition extraction unit for local wind condition analysis that extracts the representative wind speed distribution in the height direction at the boundary, and the representative wind speed distribution for each reference wind direction obtained by the boundary condition extraction unit for local wind condition analysis And local wind condition analysis unit for predicting wind speed and direction distribution in the space including the evaluation point, wind speed distribution at the evaluation point predicted by the local wind condition analysis unit and representative wind speed obtained by the boundary condition extraction unit A weather prediction apparatus comprising a correction coefficient calculation unit that obtains a different correction coefficient for each reference wind direction by comparing distributions.
The weather prediction device according to claim 1,
The typical wind speed distribution in the height direction at the evaluation point is the wind speed at an altitude different from the reference altitude, with the average wind speed at the reference altitude being the height that can be placed at the center position of the wind turbine planned to be installed at the evaluation point as the reference wind speed. And a reference wind speed distribution to obtain a representative wind speed distribution in the height direction.
The weather prediction device according to claim 1,
The typical wind speed distribution in the height direction at the evaluation point is the average wind speed at the reference altitude, which is the height that can be placed at the center position of the windmill scheduled to be installed at the evaluation point, and the average at an altitude different from the reference altitude. A weather prediction apparatus characterized by obtaining a representative wind speed distribution in a height direction from a wind speed and the reference wind speed.
The weather prediction device according to any one of claims 1 to 3,
A weather prediction apparatus comprising a global wind condition analysis result correction unit that corrects the wind speed distribution in the height direction obtained by the global wind condition analysis unit for each reference wind direction.
A wind power generation output estimation device using the weather prediction device according to any one of claims 1 to 4,
A wind power generation output estimation device that adds a wind power generation output estimation unit that calculates a wind generation amount in the long term by obtaining a wind condition in consideration of the correction coefficient with respect to a predicted value of a long-term wind condition.
Predict the wind speed and wind direction distribution in the space including the evaluation point from the wind conditions in the global area, and extract the representative wind speed distribution in the height direction at the evaluation point from the numerous wind speed and wind direction distributions obtained at each time. , Using the representative wind speed distribution as a boundary condition, predicting the wind speed / wind direction distribution in the space including the evaluation point, and differing for each reference wind direction by comparing the wind speed distribution at the evaluation point and the representative wind speed distribution A weather prediction method characterized by obtaining a correction coefficient.
A weather prediction method that uses the calculation result of the global wind analysis as the boundary condition of the local wind analysis and corrects the calculation result of the global wind analysis using the calculation result,
The calculation results of the global wind condition analysis are classified according to the reference wind direction at the reference elevation at the evaluation point, and for each reference wind direction, the average wind speed at the reference elevation is calculated as the reference wind speed. From the correlation of wind speeds, calculate the proportional coefficient with the highest correlation for each altitude, and add the proportional coefficient to the reference wind speed to obtain the representative wind speed distribution in the height direction. A weather forecasting method characterized in that the boundary condition is
A weather prediction method that uses the calculation result of the global wind analysis as the boundary condition of the local wind analysis and corrects the calculation result of the global wind analysis using the calculation result,
The calculation results of the global wind condition analysis are classified according to the reference wind direction at the reference altitude, and for each reference wind direction, the average wind speed at each elevation is calculated to obtain the representative wind speed distribution in the height direction. A weather prediction method characterized by using boundary conditions for analysis.