JP2022188698A - Wind state prediction device and method for predicting wind state - Google Patents

Abstract

To allow a precise prediction of the wind state of a year from wind state measurement data for a short period of time in a wind state prediction target site.SOLUTION: A wind state prediction device 100 includes: a function acquisition unit 112 for acquiring a local wind state function, which shows a local wind state of a prediction target point, and a large-area wind state function, which shows a large-area wind state; and a wind state prediction unit 113 for correcting the large-area wind state function by using a parameter in the local wind state function, and predicting the wind state of the prediction target site by using the corrected large-area wind state function.SELECTED DRAWING: Figure 1

Description

TECHNICAL FIELD The present invention relates to a wind condition prediction device and a wind condition prediction method for highly accurately predicting wind conditions at a prediction target point.

When selecting a construction site for a wind power plant, it is necessary to install a wind condition measuring instrument and measure the wind condition (how the wind blows, for example, the distribution of wind speed and direction) for one year in order to predict the amount of power generated. Installation of a wind condition measuring instrument requires a lot of cost, so it is desirable to accurately determine in advance a location with good wind conditions. Currently, annual wind conditions are predicted based on global wind condition forecast data published by meteorological organizations, etc. , the prediction error tends to increase.

Patent Documents 1 and 2 are disclosed for such problems. The weather prediction device described in Patent Literature 1 executes a local wind condition simulation based on global wind condition prediction data, and predicts wind conditions in consideration of the influence of topography and the like. Local wind resource simulations are not performed under all conditions, but are performed for representative wind resources, and annual average wind speeds are predicted from the results.

In the wind condition prediction system described in Patent Document 2, by correcting global wind condition prediction data based on short-term measurement data measured by a Doppler lidar or the like, which is easier to install than a wind condition measuring instrument, The purpose is to predict the annual wind speed considering local topography and other conditions. Calculate the time correlation between the global wind resource forecast data and the measured data in the period when the wind speed was measured, and use the correlation coefficient to correct the global wind resource forecast data for the unmeasured period, resulting in an annual Calculate the average wind speed of

Japanese Patent Application Laid-Open No. 2019-203728 JP 2020-159725 A

The weather prediction device of Patent Literature 1 can predict in detail the wind conditions in the vicinity of the prediction target point by local simulation. However, it is difficult to improve the accuracy of the annual average wind speed prediction value because the accuracy of the local simulation decreases due to the error between the global wind condition data and the actual wind condition.

In the wind condition prediction system described in Patent Document 2, it is possible to grasp errors in global wind condition data based on short-term wind speed measurement data. However, there is often a difference between the error in the observed period and the error in the unobserved period, and it is often difficult to improve accuracy only by mathematical correction based on time correlation.

The present invention has been made in view of such a background, and a wind condition prediction device and a wind condition prediction device that enable highly accurate prediction of annual wind conditions from short-term wind condition measurement data at a wind condition prediction target point. An object of the present invention is to provide a method for predicting wind conditions.

In order to solve the above-described problems, the wind resource prediction device according to the present invention acquires a local wind resource function that indicates the local wind resource at a prediction target point, and a global wind resource function that indicates a global wind resource. and a wind resource prediction unit that corrects the global wind resource function using parameters included in the local wind resource function and predicts the wind resource at the prediction target point using the corrected global wind resource function. and a part.

ADVANTAGE OF THE INVENTION According to the present invention, it is possible to provide a wind resource prediction device and a wind resource prediction method that enable a yearly wind resource prediction with high accuracy from short-term wind resource measurement data at a wind resource prediction target point. . Problems, configurations, and effects other than those described above will be clarified by the following description of the embodiments.

1 is a functional block diagram of a wind condition prediction device according to this embodiment; FIG. It is a graph which shows the relationship of the altitude and wind speed which concern on this embodiment. 4 is a flow chart of wind condition prediction processing according to the present embodiment. 9 is a flow chart of wind condition prediction processing according to a modification of the present embodiment. It is a rose map which shows the appearance frequency distribution of the wind direction and wind speed concerning the modification of this embodiment. 9 is a flow chart of wind condition prediction processing according to a modification of the present embodiment. 9 is a flow chart of wind condition prediction processing according to a modification of the present embodiment.

≪Overview of the wind forecast device≫
A wind condition prediction device in a form (embodiment) for carrying out the present invention will be described below. The wind resource prediction device acquires a wind resource profile, which is the distribution of wind speed in the direction of altitude, from short-term wind resource measurement data at a prediction target point, based on a wind resource physical model near the ground surface. Next, the wind condition prediction device corrects the wind condition profile in the global wind condition based on this wind condition profile to predict the annual wind condition.

The wind profile is determined by topography, and the influence of topography does not change even during unobserved periods. Therefore, even a correction based on a wind profile estimated from short-term measurements is effective throughout the year. Therefore, it is possible to predict wind conditions with high accuracy. The wind conditions include average wind speed, appearance rate of wind speed/wind direction, average wind speed by wind direction, wind condition profile, and so on.

<<Overall configuration of the wind condition prediction device>>
FIG. 1 is a functional block diagram of a wind condition prediction device 100 according to this embodiment. Wind condition prediction device 100 is a computer, and includes control unit 110 , storage unit 120 , and input/output unit 180 . User interface devices such as a display, a keyboard, and a mouse are connected to the input/output unit 180 . The input/output unit 180 may include a communication device and be capable of transmitting/receiving data to/from another device. Also, a media drive may be connected to the input/output unit 180 so that data can be exchanged using a recording medium.

The storage unit 120 includes storage devices such as ROM (Read Only Memory), RAM (Random Access Memory), and SSD (Solid State Drive). Storage unit 120 stores global wind condition data 130 , local wind condition data 140 , and program 121 . The program 121 includes a description of the procedure for wind condition prediction processing (see FIG. 3), which will be described later.

≪Global Wind Data≫
The global wind condition data 130 is meteorological data called GPV (Grid Point Value) issued by meteorological organizations such as the Meteorological Agency and analysis results obtained by global meteorological analysis. A WRF (Weather Research and Forecasting model) or the like is used in global meteorological analysis, and calculation is performed with a resolution of several hundred meters to several kilometers. In global wind condition analysis as well, weather data issued by meteorological organizations are used for boundary conditions and initial conditions, but the resolution is several kilometers to several tens of kilometers.

≪Local wind condition data≫
The local wind condition data 140 is data measured by a wind condition measuring instrument, a Doppler lidar, or the like in the vicinity of the prediction target point, data calculated by local wind condition analysis considering terrain undulations, and the like. To measure the wind conditions, a wind condition measuring instrument is used, which is a stick-shaped structure several tens of meters high and has sensors for measuring wind speed and direction. In addition, Doppler lidar, which measures wind speed and direction by using the reflection of laser beams from aerosols in the atmosphere, is used. These measuring instruments can measure the altitude direction of the prediction target point and the wind conditions in its vicinity.

While the global wind condition data 130 is data for a long period of time (second period) on a yearly basis, the local wind condition data 140 is for short-term (first period) observation, for example, several days to several months. data based on Also, although the local wind condition data 140 has been observed for a long period of time on a yearly basis, some data may be missing due to malfunction of equipment, etc., and usable data may be available for a short period of time.

In the analysis of local wind conditions, computational fluid dynamics is performed using local terrain data recorded at a resolution of several meters to several tens of meters above sea level issued by organizations such as the Geospatial Information Authority of Japan and global meteorological data such as GPV as analysis conditions. Techniques such as analysis are used. When performing computational fluid dynamics analysis, a three-dimensional computational grid is generated with a resolution that enables grasping of changes in wind conditions caused by local topographical changes at prediction target points.

In local wind analysis, initial conditions and boundary conditions are determined from global wind data. At this time, the value of the global wind condition data itself may be used, or a value that has undergone some processing may be used. For example, in the calculation grid of the global wind condition data 130, the data of the grid points closest to the prediction target point may be used, or the average value of neighboring grid points may be used. You may let Also, the data period may be limited.
The local wind condition analysis may be a technique for analyzing a steady state, or may be a non-steady analysis for grasping changes over time. The local wind condition data obtained in this manner includes detailed changes in the wind condition near the ground surface at the prediction target point, which are not included in the global wind condition data.

≪Control section≫
The control unit 110 includes a CPU (Central Processing Unit) and includes a data acquisition unit 111 , a function acquisition unit 112 , and a wind condition prediction unit 113 . The data acquisition unit 111 acquires the global wind condition data and the local wind condition data at the prediction target point, and stores them in the global wind condition data 130 and the local wind condition data 140, respectively.

≪Control section: function acquisition section≫
The function acquisition unit 112 acquires a function indicating wind conditions from the global wind condition data 130 and the local wind condition data 140 . For example, the function acquisition unit 112 acquires a function indicating a wind condition profile (distribution of wind speed in the altitude direction) for each wind direction. Below, the function indicating the wind condition profile is also referred to as the wind condition profile function. Further, the wind condition profile function obtained from the global wind condition data 130 is a global wind condition profile function (global wind condition function), and the wind condition profile function obtained from the local wind condition data 140 is a local wind condition profile function (local wind condition function ).

FIG. 2 is a graph 210 showing the relationship between altitude and wind speed according to this embodiment. It can be seen that the wind speed decreases as it approaches the ground surface, and the wind speed increases as the altitude increases. There is a theoretical relationship between altitude and wind speed, for example, the logarithmic law. Specifically, the wind speed is proportional to the logarithm obtained by dividing the altitude by the surrounding ground surface roughness, and the friction speed, which is a parameter indicating the strength of the wind turbulence (see equation (1) below). In addition, zero surface displacement may be considered for altitude (see equation (2) described later). In addition, it may be a function including temperature.

Wind speed = proportional coefficient x friction speed x ln (altitude/surface roughness) (1)
Wind speed = proportional coefficient × friction speed × ln ((altitude - zero plane displacement) / surface roughness) (2)

As shown in equations (1) and (2), the wind profile (wind profile function) when following the logarithmic law is a logarithmic function that includes parameters such as friction velocity, surface roughness, zero surface displacement, and temperature. Although shown, it may be shown by other functions such as power functions and polynomial functions. Alternatively, the wind profile function may be a function (map) showing the distribution of wind speed with respect to altitude using a statistical method. Alternatively, the wind profile function may be a function (map) calculated by combining a plurality of formulas such as machine learning techniques including neural networks.

The wind condition profile function acquired by the function acquisition unit 112 is determined from parameters such as the Obukhov length, heat flux, etc. in addition to the above-described parameters such as friction velocity, ground surface roughness, zero surface displacement, and temperature. contains the parameters to be used. These parameters include parameters that are determined by the local topography of the prediction target point, particularly the topography on the windward side, or that are greatly influenced by the topography. Such parameters are unique to the prediction target point. For example, surface roughness and zero plane displacement are determined by windward topography.

The function acquisition unit 112 obtains the wind condition profile function type (logarithmic function, polynomial function, neural network, etc.) at the prediction target point from the local wind condition data 140, and the local (neighboring) terrain according to this type. parameters are obtained for each wind direction. The function acquisition unit 112 also acquires from the global wind condition data 130 the type of the wind condition profile function in the vicinity of the prediction target point and the parameters corresponding to this type for each wind direction.

≪Control section: Wind condition prediction section≫
The wind condition prediction unit 113 corrects the function indicating the wind condition of the global wind condition data 130 using the function indicating the wind condition of the local wind condition data 140, and calculates the annual wind condition at the prediction target point. For example, the wind condition prediction unit 113 corrects the global wind condition profile function using a parameter that is included in the local wind condition profile function and determined by the terrain near the prediction target point.

For example, when the wind resource profile function is expressed in the form of equation (2), the wind resource prediction unit 113 obtains the ground surface roughness and zero plane displacement obtained from the global wind resource data 130 from the local wind resource data 140. The global wind profile function is corrected by replacing the ground surface roughness and the zero plane displacement. Using this corrected global wind condition profile function, the wind condition prediction unit 113 calculates the annual wind condition (wind speed) at the prediction target point. In other words, the parameters included in the local wind resource profile function (local wind resource function) are used to correct the global wind resource profile function (global wind resource function), and the corrected global wind resource profile function is used to determine the forecast target Predict wind conditions (annual wind conditions) at a location.

<<Wind Prediction Processing>>
FIG. 3 is a flowchart of wind condition prediction processing according to the present embodiment.
In step S11, the wind condition prediction unit 113 divides the global wind condition data 130 and the local wind condition data 140 by wind direction. The width of the wind direction section is, for example, a predetermined angle (30°, 22.5°, etc.).

In step S12, the wind condition prediction unit 113 starts a process of repeating steps S13 and S14 for each wind direction divided in step S11.
In step S<b>13 , the function acquisition unit 112 acquires wind condition profile functions of the global wind condition data 130 and the local wind condition data 140 . The function acquisition unit 112 selects, for example, a logarithmic function as the function, and calculates the proportional coefficient, frictional velocity, ground surface roughness, and zero surface displacement as parameters.

In step S14, the wind condition prediction unit 113 corrects the global wind condition profile function. For example, the wind resource prediction unit 113 replaces the ground surface roughness and zero surface displacement included in the global wind resource profile function with the ground surface roughness and zero surface displacement included in the local wind resource profile function. Note that the ground surface roughness and zero surface displacement are parameters determined by the terrain on the windward side of the prediction target point.
In step S15, the wind condition prediction unit 113 predicts wind conditions using the global wind condition profile function corrected in step S14.

≪Characteristics of the wind forecast device≫
In the global meteorological analysis, the area to be analyzed is divided into meshes (grids) of several hundred meters to several kilometers, and data such as wind speed and wind direction can be obtained on the intersections and center points of the grids. The prediction target point is not limited to the intersection point or center point of the grid. In addition, in the vicinity of the prediction target point, there may be changes in terrain that cannot be resolved with a grid of several hundred meters to several kilometers, and there are changes in the altitude of the ground surface. Therefore, prediction of wind conditions based on only the global wind condition data 130 has a large error.

The wind condition prediction device 100 corrects the global wind condition profile function using the local wind condition profile function to predict the wind condition at the prediction target point. For example, the wind resource prediction device 100 corrects by replacing some of the parameters included in the global wind resource profile function with parameters of the local wind resource profile function that are determined by topography.
Even if it is short-term local wind resource data, the parameters of the local wind resource profile function reflect the topography, so annual wind resource can be predicted with high accuracy.

≪Modification: Correction 1 considering the appearance frequency of wind direction and wind speed≫
In the embodiment described above, the parameters included in the global wind profile function are replaced with the parameters included in the local wind profile function. Instead of simply replacing the parameters, the parameters may be calculated based on the appearance frequency of the wind direction and wind speed.

FIG. 4 is a flowchart of wind condition prediction processing according to a modification of the present embodiment.
Step S21 is the same as step S11 (see FIG. 3).
In step S22, the wind condition prediction unit 113 calculates the friction velocity of the global wind condition profile function based on the appearance frequency for each wind direction. Specifically, the wind condition prediction unit 113 weights the friction velocity for each wind direction of the local wind condition profile function by the appearance frequency for each wind direction obtained from the global wind condition data 130, and obtains the friction velocity for the global wind condition profile function. . In other words, the friction velocity of the global wind profile is calculated by Equation (3), where Σ is the sum of the wind directions.

Friction velocity of global wind profile = Σ (Frequency of appearance for each wind direction obtained from global wind
× Friction velocity for each wind direction of the local wind profile function) (3)

Steps S23 to S26 are the same as steps S12 to S15, respectively. However, the friction velocity of the corrected global wind condition profile function in step S25 is the friction velocity calculated in step S22.

In the embodiment described above, the friction velocity is not corrected. On the other hand, in the modified example, the friction velocity according to the topography is used, and the prediction accuracy is improved.

≪Modification: Correction 2 considering the appearance frequency of wind direction and wind speed≫
In the wind condition prediction processing shown in FIG. Correction may be made based on the appearance frequency of wind speed for each wind direction obtained from the global wind condition data 130 .

FIG. 5 is a rose map 220 showing the appearance frequency distribution of wind direction and wind speed according to a modification of the present embodiment. In the rose map 220, the wind direction is separated by 30 degrees, and the wind speed is separated by 5 m/s from the darker to the lighter hatched areas. The rose map 220 shows that the frequency of winds in the direction of 90° to 180° is high, and the frequency of wind speeds of 10 to 20 m/s is high among them.

FIG. 6 is a flowchart of wind condition prediction processing according to a modification of the present embodiment. Steps S31 to S33 and S35 are the same as S11 to S13 and S15 shown in FIG. 3, respectively.
In step S34, the wind condition prediction unit 113 corrects the global wind condition profile function. The wind resource prediction unit 113 replaces the ground surface roughness and zero surface displacement included in the global wind resource profile function with the ground surface roughness and zero surface displacement included in the local wind resource profile function. Furthermore, the wind resource prediction unit 113 calculates the friction speed by weighting the friction speed for each wind speed included in the local wind resource profile function by the appearance frequency of the wind speed obtained from the global wind resource data, and uses this friction speed to calculate the global wind resource. Compensate by replacing the friction velocity included in the profile function. In other words, the friction velocity of the global wind profile is calculated by equation (4), where Σ is the sum of the wind velocity. Note that this correction is performed for each wind direction in correspondence with the repeated processing for each wind direction in steps S33 and S34.

Friction velocity of global wind profile by wind direction = Σ (Frequency of occurrence of wind speed by wind direction obtained from global wind data
× Friction velocity for each wind speed in each wind direction of the local wind profile function) (4)

In the embodiment described above, the friction velocity is not corrected. On the other hand, in the modified example, the friction velocity according to the topography is used, and the prediction accuracy is improved.

≪Modification: Correction considering temperature≫
In the above embodiments and modifications, the wind condition prediction unit 113 corrects the global wind condition profile function for each wind direction, but it may be corrected for each temperature (temperature class) in addition to the wind direction.

FIG. 7 is a flow chart of wind condition prediction processing according to a modification of the present embodiment.
In step S41, the wind condition prediction unit 113 divides the global wind condition data 130 and the local wind condition data 140 by wind direction and temperature. The air temperature is divided by the temperature of the ground surface, or by the altitude distribution near the ground surface.
In step S42, the wind condition prediction unit 113 starts a process of repeating steps S43 and S44 for each wind direction and temperature divided in step S41.

In step S<b>43 , the function acquisition unit 112 determines wind condition profile functions of the global wind condition data 130 and the local wind condition data 140 . For example, the function acquisition unit 112 selects a logarithmic function as the function and uses it as a parameter to calculate the proportional coefficient, the frictional velocity, the ground surface roughness, and the zero surface displacement.
In step S44, the wind condition prediction unit 113 corrects the global wind condition profile function. For example, the wind resource prediction unit 113 replaces the ground surface roughness and zero surface displacement included in the global wind resource profile function with the ground surface roughness and zero surface displacement included in the local wind resource profile function.

In the embodiment described above, the parameters are corrected for each wind direction. On the other hand, in the modified example, the parameters are corrected for each wind direction and temperature, which improves the accuracy of prediction.

<<Other Modifications>>
Although several embodiments of the present invention have been described above, these embodiments are merely examples and do not limit the technical scope of the present invention. The present invention can take various other embodiments, and various modifications such as omissions and substitutions can be made without departing from the gist of the present invention. These embodiments and modifications thereof are included in the scope and gist of the invention described in this specification and the like, and are included in the scope of the invention described in the claims and equivalents thereof.

100 wind condition prediction device 112 function acquisition unit 113 wind condition prediction unit 130 global wind condition data 140 local wind condition data

Claims (7)

a function acquisition unit that acquires a local wind condition function that is a function that indicates a local wind condition at a prediction target point and a global wind condition function that is a function that indicates a global wind condition;
a wind resource prediction unit that corrects the global wind resource function using parameters included in the local wind resource function, and predicts the wind resource at the prediction target point using the corrected global wind resource function. A wind condition prediction device characterized by: The function acquisition unit
Obtained by deriving the local wind condition function that indicates the distribution of wind speed in the direction of altitude from the local wind condition data of the prediction target point based on a wind condition physical model near the ground surface where the wind speed increases from the ground to the sky. The wind condition prediction device according to claim 1, characterized in that: The local wind condition data is an actual measured value of the wind condition measured at the prediction target point during a first period,
The global wind condition function is a function derived based on the wind condition physical model from global wind condition data including the prediction target point in a second period longer than the first period. The wind condition prediction device according to claim 2. The local wind resource function is derived from numerical analysis using topographical data in the vicinity of the prediction target point and the global wind resource in the first period as inputs,
The global wind condition function is a function derived based on the wind condition physical model from global wind condition data including the prediction target point in a second period longer than the first period. The wind condition prediction device according to claim 2. The wind forecasting unit
correcting the global wind function using the parameters included in the local wind function changed according to the frequency of appearance of the wind direction or wind speed at the prediction target point, and using the corrected global wind function, The wind condition prediction device according to claim 1, which predicts wind conditions at a prediction target point. The wind resource prediction device according to claim 1, wherein the parameter of the local wind resource function used for correcting the global wind resource function is a parameter determined by terrain near the prediction target point. The wind forecast device
a step of acquiring a local wind resource function that indicates a local wind resource at a prediction target point and a global wind resource function that indicates a global wind resource;
correcting the global wind condition function using parameters included in the local wind condition function, and predicting the wind condition at the prediction target point using the corrected global wind condition function. and the wind forecast method.

Images