Wind condition information processing device, wind condition information processing method, and wind condition information processing program

Abstract

To provide a wind condition information processing device that can perform suitable wind condition analysis. [Solution] According to one embodiment, the wind condition information processing device includes a simulation unit that simulates wind conditions in a first region and outputs a first simulation result, which is the simulation result of wind conditions under first conditions at a first location within the first region, and a second simulation result, which is the simulation result of wind conditions under second conditions at a second location within the first region. The device further includes a model construction unit that constructs a prediction model of the wind conditions in the first region based on the first simulation result and the second simulation result. The device further includes a model processing unit that performs information processing on the wind conditions in the first region based on the prediction model of the first region and displays the results of the information processing on a display unit.

Description

[Technical Field]

[0001] Embodiments of the present invention relate to a wind condition information processing device, a wind condition information processing method, and a wind condition information processing program. [Background technology]

[0002] When designing a wind power generation facility that includes multiple wind turbines, it is desirable to evaluate the wind conditions at the installation site in advance. Examples of such wind power generation facilities include wind farms installed on land or at sea. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2023-069910 [Overview of the project] [Problems that the invention aims to solve]

[0004] Wind characteristics are evaluated, for example, by wind condition analysis. In this case, as wind turbine power generation facilities become large-scale, problems arise such as increased computational load and longer computation time for wind condition analysis. Furthermore, when the computational load of wind condition analysis becomes high, it becomes difficult to evaluate wind characteristics with a single wind condition analysis.

[0005] Therefore, embodiments of the present invention provide a wind condition information processing device, a wind condition information processing method, and a wind condition information processing program that are capable of performing suitable wind condition analysis. [Means for solving the problem]

[0006] According to one embodiment, the wind condition information processing device includes a simulation unit that simulates wind conditions in a first region and outputs a first simulation result, which is the simulation result of wind conditions under first conditions at a first location within the first region, and a second simulation result, which is the simulation result of wind conditions under second conditions at a second location within the first region. The device further includes a model construction unit that constructs a prediction model of the wind conditions in the first region based on the first simulation result and the second simulation result. The device further includes a model processing unit that performs information processing on the wind conditions in the first region based on the prediction model of the first region and displays the results of the information processing on a display unit. [Brief explanation of the drawing]

[0007] [Figure 1] This is a block diagram showing the configuration of the wind turbine information processing device according to the first embodiment. [Figure 2] This is a plan view showing an example of a wind condition simulation model according to the first embodiment. [Figure 3] This figure shows an example of the governing equations for the first embodiment. [Figure 4] This is a side view showing an example of a wind turbine according to the first embodiment. [Figure 5] This figure illustrates the method for combining simulation results in the first embodiment. [Figure 6] This is a flowchart showing the flow of the wind turbine information processing method according to the first embodiment. [Modes for carrying out the invention]

[0008] Embodiments of the present invention will now be described with reference to the drawings. In Figures 1 to 6, identical components are denoted by the same reference numerals, and redundant descriptions are omitted.

[0009] (First Embodiment) Figure 1 is a block diagram showing the configuration of the wind turbine information processing device 1 according to the first embodiment. The wind turbine information processing device 1 is an example of a wind condition information processing device.

[0010] The wind turbine information processing device 1 comprises an input unit 11, an information processing unit 12, and a display unit 13. The information processing unit 12 comprises a storage unit 21 and a calculation unit 22. The storage unit 21 comprises a wind condition data storage unit 21a, a land data storage unit 21b, and a set value storage unit 21c. The calculation unit 22 comprises a simulation unit 22a, a data merging unit 22b, a model construction unit 22c, and a model processing unit 22d.

[0011] The wind turbine information processing device 1 is a device that processes information related to wind turbines. The wind turbine information processing device 1 is used, for example, when designing a wind power generation facility that includes multiple wind turbines, to evaluate the wind conditions at the locations where these wind turbines will be installed (wind turbine installation sites) in advance. Examples of wind power generation facilities include wind farms installed on land or at sea. The wind turbine information processing device 1 of this embodiment calculates and outputs the wind conditions at the wind turbine installation sites based on the wind conditions and land conditions of the wind power generation facility.

[0012] Wind conditions include various factors related to wind. Examples of wind conditions include the annual energy production (AEP) obtained from the multiple wind turbines mentioned above, the wind updraft angle of the wind flowing into each turbine, and the wake effect that the leeward turbine receives from the windward turbine. Another example of wind conditions is the extreme wind speed V at the site of the wind power generation facility. ref , V e50 These include an index I indicating the turbulent state at the site, and an index indicating the change in wind speed in the vertical direction during a storm at the site. The wind conditions may also include various constraints related to wind.

[0013] Land conditions include, for example, various constraints related to the land (site) on which the wind power generation facility will be installed. Examples of land conditions include distance constraints related to the distance from buildings, roads, rivers, etc., to the wind turbine, and pollution constraints related to shading and noise caused by the wind turbine. Other examples of land conditions include construction constraints related to the slope and ground conditions on which the wind turbine will be installed, and rights constraints related to land rights, etc. Land conditions may also include various information related to the topography and geography of the land (site) on which the wind power generation facility will be installed.

[0014] The wind turbine information processing device 1 is, for example, a computer such as a PC (personal computer). The wind turbine information processing device 1 is implemented, for example, by installing a computer program for processing information about wind turbines on the computer. This installation may be performed by inserting a recording medium on which the computer program is stored into the computer, or by downloading the computer program from a server to the computer via a network.

[0015] The details of the wind turbine information processing device 1 will be explained below with reference to Figure 1. Other figures will also be referred to as appropriate in this explanation.

[0016] The input unit 11 receives various input operations from the operator and outputs information corresponding to the input operations to the information processing unit 12. The input unit 11 is composed of, for example, input devices such as a mouse or keyboard.

[0017] The information processing unit 12 performs various information processing. The information processing unit 12 comprises a storage unit 21 for storing various data and programs, and an arithmetic unit 22 for performing various calculations. The storage unit 21 is composed of memory and storage such as ROM (Read Only Memory), RAM (Random Access Memory), HDD (Hard Disk Drive), and SSD (Solid State Drive). On the other hand, the arithmetic unit 22 is composed of a processor such as a CPU (Central Processing Unit) or ASIC (Application Specific Integrated Circuit). The information processing unit 12 may also be equipped with a communication device for controlling communication between the wind turbine information processing device 1 and other devices, and a memory port for attaching external memory to the wind turbine information processing device 1.

[0018] The display unit 13 displays various information output from the information processing unit 12 on its screen. The display unit 13 is composed of a display device such as a liquid crystal display. The information from the information processing unit 12 may be displayed locally on the display unit 13 of the wind turbine information processing device 1, or it may be displayed remotely on the display unit of a device other than the wind turbine information processing device 1.

[0019] The wind condition data storage unit 21a stores wind condition data showing the simulation results when the simulation unit 22a performs a wind condition simulation. The land data storage unit 21b stores data related to the land (site) on which the wind power generation equipment will be installed, such as the land conditions mentioned above. The setting value storage unit 21c stores various setting values ​​used by the information processing unit 12 when performing information processing.

[0020] In this embodiment, before the information processing unit 12 performs information processing, the operator inputs the data and settings necessary for information processing in advance through the input unit 11. The data and settings input from the input unit 11 are stored in the land data storage unit 21b, the setting value storage unit 21c, etc., as shown in Figure 1. Note that instead of the operator manually inputting such data and settings into the wind turbine information processing device 1, the wind turbine information processing device 1 may automatically acquire them.

[0021] Next, we will describe the details of the functional blocks such as the simulation unit 22a, the data merging unit 22b, the model construction unit 22c, and the model processing unit 22d. These functional blocks are realized, for example, by the processor in the wind turbine information processing device 1 executing a computer program for processing information related to wind turbines (wind turbine information processing program). This program is an example of a wind condition information processing program.

[0022] [Simulation section 22a] The simulation unit 22a simulates wind conditions in the analysis domain and outputs the simulation results of the wind conditions in the analysis domain. The simulation unit 22a simulates wind conditions in the analysis domain using wind condition analysis software such as MASCOT® or RIAM-COMPACT®. The analysis domain is a computational domain that corresponds to the real-world area (study area) where the placement of wind turbines is being considered. An example of a study area is the area within the site of a wind power generation facility. The simulation unit 22a stores wind condition data showing the simulation results of the wind conditions in the wind condition data storage unit 21a. The wind condition simulation performed by the simulation unit 22a is also called wind condition analysis.

[0023] In this embodiment, the simulation unit 22a simulates wind conditions in the same analysis domain under different conditions. Specifically, the simulation unit 22a simulates wind conditions under low load conditions and wind conditions under high load conditions in the analysis domain. In this case, the simulation unit 22a simulates wind conditions under low load conditions at many locations within the analysis domain, and simulates wind conditions under high load conditions at fewer locations within the analysis domain. The simulation of wind conditions under low load conditions is a simulation under conditions where the computational load of the wind turbine information processing device 1 is low, and the simulation of wind conditions under high load conditions is a simulation under conditions where the computational load of the wind turbine information processing device 1 is high. For example, low load conditions are conditions where the occurrence of a predetermined meteorological phenomenon (e.g., a typhoon) is not considered, and high load conditions are conditions where the occurrence of the predetermined meteorological phenomenon is considered. However, the predetermined meteorological condition may be something other than a typhoon, for example, snowfall. The analysis domain is an example of the first domain, and the low load conditions and high load conditions are examples of the first and second conditions, respectively.

[0024] Figure 5 shows region R as an example of the analysis region. The upper left figure in Figure 5 shows the simulation of wind conditions under low load conditions using many grid points within region R. The lower left figure in Figure 5 shows the simulation of wind conditions under high load conditions using fewer grid points within region R. The grid points in the latter figure include multiple grid points P1. The grid points in the former figure include these multiple grid points P1 and multiple grid points P2. As a result, the grid points in the former figure include many grid points, while the grid points in the latter figure include few grid points. Grid points P1 and P2 are examples of the first location and the first point. Grid point P2 is an example of the second location and the second point. Grid point P1 is an example of the fourth location and the fourth point. Grid point P2 is an example of the fourth location and the fourth point. In this embodiment, the first and second locations include the overlapping grid point P1, but they do not necessarily have to include the overlapping point. Further details of Figure 5 will be described later.

[0025] The simulation unit 22a outputs simulation results for wind conditions under low load conditions at many locations within the analysis domain, and simulation results for wind conditions under high load conditions at fewer locations within the analysis domain. For example, the simulation unit 22a outputs simulation results for wind conditions under low load conditions at grid points P1 and P2 within domain R, and simulation results for wind conditions under high load conditions at grid point P2 within domain R. Hereinafter, the simulation and simulation results for wind conditions under low load conditions will also be referred to as "low load simulation" and "low load simulation results," respectively, and the simulation and simulation results for wind conditions under high load conditions will also be referred to as "high load simulation" and "high load simulation results." The low load simulation results are an example of the first simulation results, and the high load simulation results are an example of the second simulation results.

[0026] Here, we assume a case where low-load and high-load simulations are performed at the same location within the same analysis domain. For example, we assume that low-load and high-load simulations are performed at grid points P1 and P2 within domain R. In this case, the computational load when performing the high-load simulation will be higher than the computational load when performing the low-load simulation. However, the simulation unit 22a of this embodiment performs the low-load simulation at grid points P1 and P2 within domain R, and the high-load simulation at grid point P1 within domain R. This makes it possible to reduce the computational load of the high-load simulation. Instead of performing the high-load simulation at grid points P1 and P2, the simulation unit 22a of this embodiment performs the high-load simulation at grid point P1, performs the low-load simulation at grid points P1 and P2, and combines the results of the high-load and low-load simulations. This makes it possible to obtain the simulation results at grid points P1 and P2 with a low computational load, even though a high-load simulation is performed. Further details of the combination of simulation results will be described later.

[0027] Alternatively, the simulation unit 22a may simulate wind conditions at multiple points on one or more contour lines of the terrain within the analysis domain, instead of simulating wind conditions at multiple grid points within the analysis domain. In this case, low-load simulations are performed at many points on these contour lines, while high-load simulations are performed at fewer points on these contour lines.

[0028] Furthermore, the wind turbine information processing device 1 of this embodiment may be used to optimize the wind turbine arrangement when designing a wind power generation facility equipped with multiple wind turbines. In this case, the simulation unit 22a samples the position coordinates of various wind turbine arrangements of multiple wind turbines, for example, using random numbers (quasi-random numbers). The simulation unit 22a further samples from among the various wind turbine arrangements the wind turbine arrangement that maximizes the total power generated by all wind turbines, using a greedy method or the like. Alternatively, the simulation unit 22a samples from among the various wind turbine arrangements the wind turbine arrangement with the smallest difference from the current wind turbine arrangement. When sampling various wind turbine arrangements, the type of each wind turbine may be changed to various types.

[0029] Figure 2 is a plan view showing an example of the wind condition simulation model 2 of the first embodiment.

[0030] In Figure 2, the wind condition simulation model 2 includes an analysis center 31, a minimum analysis grid range 32, a target region 33, an analysis region 34, an additional region 35, an upstream buffer region 41, a downstream buffer region 42, a lateral buffer region 43, and a lateral buffer region 44. Figure 2 further shows the wind turbine position P and the inflow wind W. The simulation unit 22a of this embodiment simulates wind conditions using the wind condition simulation model 2 shown in Figure 2.

[0031] Figure 2 further shows the X, Y, and Z directions, which are perpendicular to each other. In this embodiment, the +Z direction is upward, the -Z direction is downward, and the XY plane is the horizontal plane. In this embodiment, the -Z direction coincides with the direction of gravity, but it may be inclined from the direction of gravity.

[0032] The wind turbine position P is the location where the wind turbine is placed. The incoming wind W is the wind flowing into the wind turbine from the windward side. The analysis center 31 is the central position of the analysis of the wind condition simulation model 2. The minimum analysis grid range 32 is the region of the smallest constituent unit of the target region 33 of the wind condition simulation. The analysis region 34 is the region that encloses the target region 33 in a ring shape. The additional region 35 is a region provided on the windward side of the analysis region 34 and added to the analysis region 34. Note that the above-mentioned "analysis region" in which the simulation unit 22a simulates wind conditions may be the same as the analysis region 34, or it may be different from the analysis region 34.

[0033] The upstream buffer region 41 is the region adjacent to the additional region 35 on the upwind side of the additional region 35. The downstream buffer region 42 is the region adjacent to the analysis region 34 on the downwind side of the analysis region 34. The lateral buffer regions 43 and 44 are the regions adjacent to the sides of the analysis region 34 and the additional region 35. In Figure 2, the lateral buffer region 43 is located in the +Y direction of the analysis region 34 and the additional region 35, and the lateral buffer region 44 is located in the -Y direction of the analysis region 34 and the additional region 35.

[0034] In this embodiment, the operator pre-specifies the above-mentioned study area using the input unit 11 before the wind condition simulation. The operator further inputs topographic data of the region including this study area from the input unit 11. The operator further inputs wind inflow conditions such as wind direction, wind speed, and turbulence intensity, as well as wind turbine information such as wind turbine shape and number of wind turbines, from the input unit 11. The operator further inputs other wind conditions and land conditions from the input unit 11. The data input by the operator in this way is used in the wind condition simulation. For example, the specification of the study area and the contents of the topographic data are reflected in each region of the wind condition simulation model 2.

[0035] Figure 3 shows an example of the governing equations in the first embodiment.

[0036] The simulation unit 22a of this embodiment generates a wind condition simulation model 2 for calculating predetermined physical quantities based on governing equations. The governing equations of this embodiment are the Navier-Stokes equations, as shown in Figure 3. In the wind condition simulation of this embodiment, the Navier-Stokes equations are adopted as the governing equations that express the physical laws within the mesh model in mathematical equations.

[0037] In Figure 3, ρ represents the fluid density, μ represents the fluid viscosity, and ν represents the fluid kinematic viscosity. In this embodiment, the fluid is air. The Navier-Stokes equations shown in Figure 3 include a time term, a pressure term, an advection term, and a viscosity term, as well as an external force term derived from the external force F from the wind turbine rotor. When the Navier-Stokes equations are expressed as three equations for the X, Y, and Z directions, the external force term is the X component F of the external force F. X , Y component F Y , and Z component F Z This represents a vector quantity expressed as follows:

[0038] Figure 4 is a side view showing an example of the wind turbine 3 of the first embodiment.

[0039] The wind turbine 3 shown in Figure 4 comprises a tower 51, a nacelle 52, a hub 53, and a plurality of blades 54.

[0040] The tower 51 extends in the Z direction from its lower end on the ground side to its upper end on the nacelle 52 side. The nacelle 52 is attached to the upper end of the tower 51 and houses a generator (not shown). The hub 53 is attached to the rotor of the generator. Each blade 54 is attached to the hub 53.

[0041] In Figure 4, when the multiple blades 54 rotate due to wind power, the rotation of these blades 54 is transmitted to the generator via the hub 53 and the rotating shaft. As a result, the generator is driven by wind power and generates an alternating current voltage.

[0042] The explanation of the wind turbine information processing device 1 will now resume, again referring to Figure 1. Other figures will also be referenced as appropriate during this explanation.

[0043] [Data merging section 22b] The data merging unit 22b combines the low-load simulation results and the high-load simulation results in the same analysis domain to generate new simulation results for that analysis domain. However, the low-load simulation results before merging represent the wind conditions under low-load conditions at many locations within the analysis domain, and the high-load simulation results before merging represent the wind conditions under high-load conditions at fewer locations within the analysis domain. On the other hand, the simulation results after merging represent the wind conditions at many locations within the analysis domain. The simulation results after merging reflect the low-load simulation results before merging and the high-load simulation results before merging. Hereafter, the simulation results after merging will also be referred to as "merged simulation results" to distinguish them from the low-load simulation results and high-load simulation results before merging. The merged simulation results are an example of the third simulation result.

[0044] Figure 5 shows how the low-load simulation results shown in the upper left of Figure 5 and the high-load simulation results shown in the lower left of Figure 5 are combined to generate a combined simulation result. The low-load simulation results shown in the upper left of Figure 5 are simulation results of wind conditions under low-load conditions at grid points P1 and P2 in region R, and the high-load simulation results shown in the lower left of Figure 5 are simulation results of wind conditions under high-load conditions at grid point P1 in region R. As a result, the combined simulation result is the simulation result of wind conditions at grid points P1 and P2 in region R. The combined simulation result includes, for example, the simulation results of wind conditions under low-load conditions at grid points P1 and P2 in region R (low-load simulation result) and the simulation results of wind conditions under high-load conditions at grid point P1 in region R (high-load simulation result). The data combining unit 22b performs the simulation result combining process using, for example, the wind condition data stored in the wind condition data storage unit 21a. Grid points P1 and P2 are examples of a third location and a third point. In this embodiment, the first, second, and third locations include a point called grid point P1, but they do not necessarily have to include a point of overlap. Further details of Figure 5 will be described later.

[0045] [Model Construction Section 22c] The model building unit 22c constructs a predictive model of the wind conditions in the analysis domain based on the combined simulation results described above. The model building unit 22c constructs the predictive model using, for example, regression analysis methods such as Gaussian process regression. The model building unit 22c may also construct the predictive model using a neural network, random forest, or support vector regression.

[0046] The wind condition prediction model in this embodiment is a model that uses the position coordinates of a single wind turbine as an explanatory variable. Examples of wind conditions handled by the prediction model include the aforementioned AEP, updraft angle, wake effect, and extreme wind speed V. ref , V e50For example, they are the index I of the turbulent state, the index of the wind speed change during a storm, etc. The wind conditions handled by the prediction model are called wind condition parameters. Note that the model construction unit 22c may construct a prediction model using the wind conditions input from the input unit 11.

[0047] In FIG. 5, the model construction unit 22c constructs a prediction model of the wind conditions in the region R based on the coupling simulation results of the region R. For example, the model construction unit 22c considers the influence of a typhoon and constructs a prediction model for predicting the extreme wind speeds V ref V e50 . In FIG. 5, the model construction unit 22c constructs a prediction model of the wind conditions under high load conditions at various points (coordinates) within the region R including the grid points P1 and P2. Thereby, for example, it becomes possible to predict the wind conditions under high load conditions at any point (coordinate) within the region R. The above "various points" is an example in the sixth case. Further details of FIG. 5 will be described later.

[0048] [Model processing unit 22d] The model processing unit 22d performs information processing regarding the wind conditions in the above analysis region based on the above prediction model and displays the result of the information processing on the display unit 13. For example, the model processing unit 22d predicts the wind conditions in the analysis region based on the prediction model and displays the prediction result of the wind conditions on the display unit 13. Examples of the predicted wind conditions are the extreme wind speeds V ref V e50 . In this case, the prediction results of the extreme wind speeds V ref V e50 are displayed on the display unit 13.

[0049] The model processing unit 22d may perform information processing regarding the wind conditions based on the prediction model constructed by the model construction unit 22c and the land conditions stored in the land data storage unit 21b. For example, the model processing unit 22d may perform information processing regarding the wind conditions based on the land condition of whether the actual region (study region) corresponding to the analysis region is located in a place where it is likely to be affected by a typhoon.

[0050] In Figure 5, the model processing unit 22d performs information processing on the wind conditions of region R based on the prediction model of region R, and displays the results of this information processing on the display unit 13. For example, the predicted wind conditions of region R are displayed as a result of this information processing. The figure on the right in Figure 5 shows an example of the predicted wind conditions of region R. For example, region R shown in the figure on the right in Figure 5 is displayed on the display unit 13. The figure on the right in Figure 5 shows the predicted wind conditions under high load conditions at various points (coordinates) within region R, including grid points P1 and P2.

[0051] Furthermore, the wind turbine information processing device 1 of this embodiment may be used to optimize the wind turbine arrangement when designing a wind power generation facility equipped with multiple wind turbines. In this case, the model processing unit 22d may perform a process to determine candidate wind turbine arrangements in region R as information processing regarding the wind conditions in region R, or it may display the candidate wind turbine arrangements in region R on the display unit 13.

[0052] As described above, the wind turbine information processing device 1 of this embodiment performs a high-load simulation at grid point P1, performs a low-load simulation at grid points P1 and P2, and combines the results of the high-load and low-load simulations, instead of performing a high-load simulation at grid points P1 and P2. This makes it possible to obtain the simulation results (combined simulation results) at grid points P1 and P2 with a low computational load, even though a high-load simulation is performed.

[0053] When performing high-load simulations at grid points P1 and P2, precise calculations that take into account the characteristics of the object being analyzed can be performed. However, this leads to problems such as increased computational load and longer simulation times. Furthermore, as wind power generation facilities become larger in scale, it becomes difficult to perform high-load simulations at grid points P1 and P2 within a single analysis domain R. For example, if the site of a wind power generation facility is represented across multiple analysis domains R, the computational load, computation time, and calculation accuracy of the simulation may deteriorate.

[0054] On the other hand, the wind turbine information processing device 1 of this embodiment combines the high-load simulation results at grid point P1 with the low-load simulation results at grid points P1 and P2 to generate simulation results (combined simulation results) at grid points P1 and P2. This makes it possible to perform precise calculations that take into account the characteristics of the object being analyzed with a low computational load. As a result, it becomes easy to represent the site of the wind power generation facility in a single analysis domain R. In this embodiment, by utilizing machine learning, high-load analysis values ​​are predicted based on the low-load analysis results.

[0055] Thus, according to this embodiment, it is possible to perform wind condition analysis that is suitable in terms of computational load, computation time, and computational accuracy. For example, it is possible to reduce the computational load while maintaining computational accuracy. Furthermore, by reducing the computational load, it is possible to increase the number of analysis trials.

[0056] The wind turbine information processing device 1 of this embodiment may have a structure other than that shown in Figure 1. For example, the calculation unit 22 does not need to include a data merging unit 22b. In this case, the model construction unit 22c constructs a prediction model of the wind conditions in the analysis domain based on the low-load simulation results and high-load simulation results output from the simulation unit 22a. The model construction unit 22c constructs the prediction model using, for example, wind condition data stored in the wind condition data storage unit 21a. In this case, the model construction unit 22c may perform all the processing performed by the data merging unit 22b and the model construction unit 22c, or it may perform only some of the processing performed by the data merging unit 22b and the model construction unit 22c. For example, in this case, the model construction unit 22c may omit the processing of merging the low-load simulation results and the high-load simulation results.

[0057] Next, referring to Figure 1, we will explain an example of the operation of the wind turbine information processing device 1. Other figures will also be referenced as appropriate during this explanation.

[0058] The simulation unit 22a performs, for example, a high-load simulation at coordinates from x1 to x n and outputs wind conditions (e.g., AEP, upwind angle, wake influence, extreme wind speed V ref , V e50 , an index I indicating the state of turbulence, an exponent indicating the change in wind speed in the height direction during a storm, etc.) under high-load conditions. For example, the extreme wind speed at coordinate x i under high-load conditions is represented by y i 1 , and data (x i , y i 1 ) i=1 n is output. Also, the simulation unit 22a performs, for example, a low-load simulation at coordinates from x1 to x m and outputs wind conditions under low-load conditions. For example, the extreme wind speed at coordinate x i under low-load conditions is represented by y i 2 , and data (x j , y j 2 ) j=1 m is output. Here, n represents the number of grid points in the high-load simulation, and m represents the number of grid points in the low-load simulation. Therefore, n corresponds to the number of grid points P1, and m corresponds to the number of grid points P1 and P2. In this embodiment, n < m holds. The data combining unit 22b combines and stores, for example, two pieces of data (x i , y i 1 ) i=1 n and (x j , y j 2 ) j=1 m that are outputs of the simulation unit 22a.

[0059] The model construction unit 22c, for example, uses extreme wind speeds V ref , V e50A prediction model is constructed using two-output Gaussian process regression. In this case, the model construction unit 22c uses ICM (Intrinsic Coregionalization Model) as the kernel method for the two outputs of the Gaussian process regression. The model construction unit 22c uses coordinates as explanatory variables and two extreme wind speeds V obtained from two types of wind condition analysis. ref , V e50 Predict the dependent variable.

[0060] Here, coordinate x i The extreme wind speed under high load conditions in y i 1 Represented by, coordinate x i The extreme wind speed under low load conditions is y i 2 It is represented as follows. Also, machine learning in the model building unit 22c is performed on data (x i ,y i 1 ) i=1 n and data (x j ,y j 2 ) j=1 m Let's assume this is done using the following. In this case, machine learning may be performed directly using the two data outputs of the simulation unit 22a, or the data obtained by combining the two data may be used to combine the two data (x i ,y i 1 ) i=1 n and data (x j ,y j 2 ) j=1 m You can calculate the two data points and then perform machine learning using those two data points.

[0061] In this case, the extreme wind velocity μ under high load conditions at a new (arbitrary) coordinate x'. 1 (x') is expressed by equation (1).

number

Number

Number

Number

Number

Number

[0062] Note that the model construction unit 22c predicts the augmentation coefficient E ref , V e50 and the correction coefficient E tV instead of directly predicting the extreme wind speeds V tI . From the prediction results of the coefficients E tV , E tI , the extreme wind speeds V ref , V e50The coefficient E under high load conditions may be calculated using the method shown in equations (1) to (6). tV , E tI And the coefficient E under low load conditions tV , E tI This predicts that. Furthermore, the model building unit 22c may use a multi-output neural network, random forest, or support vector regression instead of a two-output Gaussian process regression.

[0063] According to this embodiment, by constructing a predictive model based on two sets of data, it is possible to predict high-load simulation results at any coordinate in a short time and with high accuracy. Conventionally, data (x i ,y i 1 ) i=1 n Because the prediction model was built based solely on data (x), the prediction accuracy was low. In this embodiment, the data (x i ,y i 1 ) i=1 n In addition, data (x j ,y j 2 ) j=1 m By using this method, the prediction accuracy of the prediction model is improved. Furthermore, according to this embodiment, even if the number of n is reduced and the number of simulations under high load conditions is reduced compared to conventional methods, a highly accurate prediction model can be obtained, and as a result, the computational load required for simulation can be significantly reduced.

[0064] The following describes in more detail the information processing performed by the wind turbine information processing device 1 with reference to Figures 5 and 6. The reference numerals shown in Figure 1 will also be used as appropriate in this description.

[0065] Figure 5 is a diagram illustrating the method for combining simulation results in the first embodiment.

[0066] Figure 5 shows region R as an example of one analysis region. The upper left figure in Figure 5 shows the simulation of wind conditions under low load conditions at grid points P1 and P2 within region R. The lower left figure in Figure 5 shows the simulation of wind conditions under high load conditions at grid point P1 within region R. As shown in Figure 5, the simulation unit 22a performs low-load simulations at many grid points within region R and high-load simulations at fewer grid points within region R.

[0067] Figure 5 further illustrates how the low-load simulation results shown in the upper left of Figure 5 and the high-load simulation results shown in the lower left of Figure 5 are combined to generate the combined simulation results. The low-load simulation results shown in the upper left of Figure 5 represent the wind conditions under low-load conditions at grid points P1 and P2 within region R, while the high-load simulation results shown in the lower left of Figure 5 represent the wind conditions under high-load conditions at grid point P1 within region R. As a result, the combined simulation results represent the wind conditions at grid points P1 and P2 within region R. The combined simulation results reflect both the low-load and high-load simulation results.

[0068] Figure 6 is a flowchart showing the flow of the wind turbine information processing method in the first embodiment. The wind turbine information processing method shown in Figure 6 is performed by the wind turbine information processing device 1 of this embodiment. For the sake of clarity, the following explanation will use the region R and grid points P1 and P2 shown in Figure 5. This wind turbine information processing method is an example of a wind condition information processing method.

[0069] First, the simulation unit 22a sets various conditions based on information input by the operator from the input unit 11 (step S1). For example, the operator inputs information about the area under consideration for placing wind turbines, and the various information mentioned above that is stored in the memory unit 21.

[0070] Next, the simulation unit 22a performs a low-load simulation (low-load simulation) at grid points P1 and P2 within region R and outputs the low-load simulation results (step S2). As a result, wind condition data showing the low-load simulation results is stored in the wind condition data storage unit 21a. The low-load simulation is performed using the conditions set in step S1.

[0071] Next, the simulation unit 22a performs a high-load calculation simulation (high-load simulation) at grid point P1 within region R and outputs the high-load simulation results (step S3). As a result, wind condition data showing the high-load simulation results is stored in the wind condition data storage unit 21a. The high-load simulation is performed using the conditions set in step S1.

[0072] Next, the data merging unit 22b combines the low-load simulation results for grid points P1 and P2 within region R with the high-load simulation results for grid point P1 within region R to generate the simulation results (merged simulation results) for grid points P1 and P2 within region R (step S4). The merging process of the simulation results is performed, for example, using the wind condition data stored in the wind condition data storage unit 21a.

[0073] Next, the model building unit 22c constructs a predictive model of the wind conditions in region R based on the simulation results (joint simulation results) at grid points P1 and P2 within region R (step S5). For example, the model building unit 22c considers the effects of typhoons and constructs an extreme wind speed V ref , V e50 A predictive model is constructed to forecast the wind conditions under high load conditions at various points (coordinates) within region R, including grid points P1 and P2. In Figure 5, the model construction unit 22c constructs a predictive model for wind conditions under high load conditions at various points (coordinates) within region R.

[0074] Next, the model processing unit 22d performs information processing on the wind conditions of region R based on the prediction model for the wind conditions of region R, and displays the results of the information processing on the display unit 13 (step S6). For example, the model processing unit 22d predicts the wind conditions of region R and displays the prediction results of the wind conditions on the display unit 13. An example of a predicted wind condition is the extreme wind speed V mentioned above. ref , V e50 In this case, the extreme wind speed V ref , V e50 The prediction results are displayed on the display unit 13. For example, the region R shown in the right-hand figure in Figure 5 is displayed on the display unit 13. The right-hand figure in Figure 5 shows the prediction results for wind conditions under high load conditions at various points (coordinates) within region R, including grid points P1 and P2.

[0075] As described above, the wind turbine information processing device 1 of this embodiment performs a high-load simulation at grid point P1, performs a low-load simulation at grid points P1 and P2, and combines the results of the high-load and low-load simulations, instead of performing a high-load simulation at grid points P1 and P2. This makes it possible to obtain the simulation results (combined simulation results) at grid points P1 and P2 with a low computational load, even though a high-load simulation is performed. According to this embodiment, it is possible to perform a wind condition analysis that is suitable in terms of computational load, computation time, and calculation accuracy.

[0076] In this embodiment, as an example of a process for combining simulation results under a first condition and simulation results under a second condition, a process for combining low-load simulation results and high-load simulation results was described. Here, in the process for combining simulation results under a first condition and simulation results under a second condition, only simulation results under two conditions, the first and second conditions, may be combined, or simulation results under three or more conditions, including the first and second conditions, may be combined.

[0077] Furthermore, the contents of this embodiment can be applied to devices other than the wind turbine information processing device 1, which performs information processing related to wind turbines. For example, the contents of this embodiment can be applied to a wind condition information processing device that performs information processing related to wind conditions unrelated to wind turbines. In this case as well, it is possible to perform wind condition analysis that is suitable in terms of computation load, computation time, and calculation accuracy.

[0078] Although several embodiments have been described above, these embodiments are presented only as examples and are not intended to limit the scope of the invention. The novel apparatus, methods, and programs described herein can be implemented in a variety of other forms. Furthermore, various omissions, substitutions, and modifications can be made to the forms of apparatus, methods, and programs described herein, without departing from the spirit of the invention. The appended claims and equivalents are intended to include such forms and modifications that are included in the scope and spirit of the invention. [Explanation of symbols]

[0079] 1: Wind turbine information processing device, 2: Wind condition simulation model, 3: Wind turbine 11: Input unit, 12: Information processing unit, 13: Display unit, 21: Memory unit, 21a: Wind condition data storage unit, 21b: Land data storage unit, 21c: Setting value storage unit, 22: Calculation unit, 22a: Simulation unit, 22b: Data merging unit, 22c: Model construction unit, 22d: Model processing unit, 31: Analysis center, 32: Minimum analysis grid range, 33: Target area, 34: Analysis area, 35: Additional area, 41: Upstream buffer area, 42: Downstream buffer area, 43: Side buffer area, 44: Side buffer area, 51: Tower, 52: Nacelle, 53: Hub, 54: Blade

Claims

1. A simulation unit that simulates wind conditions in a first region and outputs a first simulation result which is the simulation result of wind conditions under first conditions at a first location within the first region, and a second simulation result which is the simulation result of wind conditions under second conditions at a second location within the first region. A model building unit constructs a prediction model of the wind conditions in the first region based on the first simulation results and the second simulation results, A model processing unit that performs information processing on the wind conditions of the first region based on the prediction model of the first region and displays the results of the information processing on a display unit, A wind condition information processing device equipped with the following features.

2. The system further includes a data merging unit that combines the first simulation result and the second simulation result to generate a third simulation result, which is a simulation result of the wind conditions at a third location within the first region. The wind condition information processing device according to claim 1, wherein the model construction unit constructs the prediction model for the first region based on the third simulation results.

3. The first location includes a plurality of first points within the first region, The second location includes a plurality of second points within the first region, The third location includes a plurality of third points within the first region, The first, second, and third locations include multiple grid points within the first region, multiple points on one or more contour lines within the first region, or multiple points sampled from the first region. The wind condition information processing device according to claim 2.

4. The wind condition information processing device according to claim 3, wherein the number of the first location is greater than the number of the second location.

5. The wind condition information processing device according to claim 3, wherein the number of the third location is greater than the number of the second location.

6. The first simulation result is the simulation result of the wind conditions under the first conditions at the fourth and fifth locations within the first region. The second simulation result is the simulation result of the wind conditions under the second condition at the fourth location within the first region. The third simulation result is the simulation result of wind conditions at the fourth and fifth locations within the first region. The wind condition information processing device according to claim 2.

7. The fourth location includes a plurality of fourth points within the first region, The fifth location includes a plurality of fifth points within the first region, The fourth and fifth locations include multiple grid points within the first region, multiple points on one or more contour lines within the first region, or multiple points sampled from the first region. The wind condition information processing device according to claim 6.

8. The wind condition information processing device according to claim 1, wherein when the wind conditions under the first condition and the wind conditions under the second condition are simulated at the same location within the first region, the computational load when simulating the wind conditions under the second condition is higher than the computational load when simulating the wind conditions under the first condition.

9. The first condition is one that does not take into account the occurrence of a specific meteorological phenomenon. The second condition described above is a condition that takes into account the occurrence of the predetermined meteorological phenomenon. The wind condition information processing device according to claim 1.

10. The wind condition information processing device according to claim 1, wherein the simulation unit simulates wind conditions in the first region using wind condition analysis software.

11. The wind condition information processing device according to claim 1, wherein the model construction unit constructs the prediction model for the first region using a regression analysis method.

12. The model construction unit constructs the prediction model for the wind conditions under the second conditions at the sixth location within the first region. The sixth location includes the fourth and fifth locations, The wind condition information processing device according to claim 6.

13. The wind condition information processing device according to claim 1, wherein the model processing unit predicts the wind conditions of the first region based on the prediction model of the first region, and displays the prediction result of the wind conditions of the first region on the display unit.

14. The wind condition information processing apparatus according to claim 1, wherein the model processing unit performs the information processing based on the prediction model of the first region and the land conditions of the first region.

15. The system simulates wind conditions in the first region and outputs a first simulation result, which is the simulation result of wind conditions under first conditions at a first location within the first region, and a second simulation result, which is the simulation result of wind conditions under second conditions at a second location within the first region. Based on the first and second simulation results, a predictive model for the wind conditions in the first region is constructed. Based on the prediction model for the first region, information processing is performed regarding the wind conditions for the first region, and the results of the information processing are displayed on the display unit. A wind condition information processing method that includes the following.

16. The system simulates wind conditions in the first region and outputs a first simulation result, which is the simulation result of wind conditions under first conditions at a first location within the first region, and a second simulation result, which is the simulation result of wind conditions under second conditions at a second location within the first region. Based on the first and second simulation results, a predictive model for the wind conditions in the first region is constructed. Based on the prediction model for the first region, information processing is performed regarding the wind conditions for the first region, and the results of the information processing are displayed on the display unit. A wind condition information processing program that causes a computer to execute a wind condition information processing method that includes the following.

Images