Wind farm wind direction numerical calculation method and system based on wind direction correlation

Abstract

The present invention discloses a method and system for numerically calculating the wind direction of a wind farm based on wind direction correlation. The method includes: obtaining the wind direction data of the wind turbines in the wind farm, mapping the wind direction data of each wind turbine onto a two-dimensional unit circle to generate vector calculation data of the wind direction and the included angle; based on the vector calculation data of the wind direction and the included angle, according to the correlation between the wind turbines, when a preset part of the wind turbines in the wind farm is abnormal, calculating the wind direction value of the abnormal part of the wind turbines. The present invention proposes to utilize the extremely strong correlation between the wind direction data of the wind turbines, train a wind direction numerical model using the wind direction operation data of the wind turbines, and determine whether the wind direction measurement results between the wind turbines meet the correlation or there are measurement abnormalities, and establish various neural network models of the wind direction numerical model. By making full use of the extremely strong correlation between the wind direction data of the wind turbines to establish a wind direction numerical model, the measurement verification of the wind speed and wind direction sensors of the wind turbines can be realized at low cost.

Description

Technical Field

[0001] The present invention relates to the technical field of wind farm measurement, and particularly relates to a method and system for numerically calculating the wind direction of a wind farm based on wind direction correlation. Background Art

[0002] At present, both the newly installed capacity and the cumulative installed capacity of wind power in China rank first in the world. The wind power industry has shifted from large-scale development to lean development, and grid parity has become the main trend in the current wind power industry. Based on this, on the premise of not increasing costs, improving the wind energy conversion rate of equipment through algorithm optimization has become the primary task of industry experts' research and development.

[0003] Affected by many factors such as terrain, climate, and the surrounding environment, wind power generation has strong intermittency, volatility, and uncertainty, which pose hazards to the power generation planning and economic dispatching of the power system and increase the operation and maintenance costs of wind turbines. Therefore, low-cost operation of wind power requires accurate and efficient data analysis algorithms to find the sources of abnormal operation and eliminate them in a timely manner. A wind measurement system intelligent linkage that can realize on-line calibration and numerical simulation of the wind speed and wind direction sensors and measurement results of wind turbines has important application value.

[0004] At the present stage, the wind power industry realizes the comparative measurement between the original wind measurement sensors of the unit and the additionally installed wind measurement sensors by installing a wind measurement tower or lidar, etc., obtains the deviation results between the actual wind speed and wind direction of the wind turbine and the measured wind speed and wind direction, and then realizes the calibration of the wind measurement system sensors of a single unit.

[0005] In traditional research, due to the inability to effectively overcome the "circumferential characteristic" of the wind direction, that is, in the wind direction representation method, the problem that 0° and 359° are numerically discontinuous, researchers only analyze and discuss the trend of the wind direction time series of a single unit. For example, when the wind direction changes from 359° to 1°, such singular points can only satisfy the wind direction change trend by adding the modulus 360°, that is, 359° changes to 361°, and the correlation analysis of the wind direction data between wind turbines cannot be realized. For example, when the wind direction changes 2° clockwise from 359°, according to the "circumferential characteristic", the result is 1°, but it is extremely inconvenient for the machine learning calculation process. Such singular points can only satisfy the wind direction change trend by adding the modulus 360°. There is a risk of being unable to realize the numerical simulation of the wind measurement system and failing to utilize the operation data of wind turbines and the data correlation between units. Summary of the Invention

[0006] Therefore, the method and system for numerically calculating the wind direction of a wind farm based on wind direction correlation provided by the present invention overcome the defects in the prior art that the numerical simulation of the wind measurement system cannot be realized and the operation data of wind turbines and the data correlation between units cannot be utilized.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] In a first aspect, an embodiment of the present invention provides a method for numerically calculating the wind direction of a wind farm based on wind direction correlation, including:

[0009] Obtain the wind direction data of the wind farm units, map the wind direction data of each unit to a two-dimensional unit circle, and generate vector calculation data of the wind direction and the included angle;

[0010] Based on the vector calculation data of the wind direction and the included angle, according to the correlation between wind turbines, when a preset part of the units in the wind farm is abnormal, calculate the wind direction value of the abnormal part of the units.

[0011] Optionally, the calculating the wind direction value of the abnormal part of the units based on the vector calculation data of the wind direction and the included angle, according to the correlation between wind turbines, when a preset part of the units in the wind farm is abnormal, includes:

[0012] According to the correlation of the wind direction data between the wind farm units, use the operation data of the anemometry system of the existing wind turbines in the wind farm to judge whether the wind direction measurement results between the units meet the correlation and whether there are abnormalities;

[0013] Establish a recurrent neural network model in the horizontal direction of the unit wind direction. According to the accurate anemometry system data model and numerical simulation between wind turbines, when the anemometry system of a preset part of the units in the wind farm is abnormal, simulate and calculate the wind direction data of the abnormal units through the wind direction data of the remaining units, and calculate the wind direction value of the abnormal units.

[0014] Optionally, map the wind direction data of each unit to a two-dimensional unit circle through the following formula:

[0015]

[0016] Among them, taking a group of m units as the object, the wind direction vectors are respectively denoted as Taking the mth unit as the target unit, θ is the wind direction angle of each unit at each moment, α1 is the first deviation angle, α2 is the second deviation angle, and after mapping, it is transformed into a unit vector with a length of 1 as (sinθ, cosθ).

[0017] Optionally, calculate the wind direction vector through the following formula:

[0018]

[0019] Among them, is the wind direction vector.

[0020] Optionally, the recurrent neural network model established in the horizontal direction of the unit wind direction is a recurrent neural network model with a single hidden layer.

[0021] In a second aspect, an embodiment of the present invention provides a numerical calculation system for wind farm wind directions based on wind direction correlation, including:

[0022] A data acquisition module, configured to acquire wind direction data of wind farm units, map the wind direction data of each unit onto a two-dimensional unit circle, and generate vector calculation data of wind direction and included angle;

[0023] A wind direction numerical calculation module, configured to calculate the wind direction numerical values of abnormal units based on the vector calculation data of wind direction and included angle and according to the correlation between wind turbines when a preset part of the units in the wind farm are abnormal.

[0024] In a third aspect, an embodiment of the present invention provides a terminal, including: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method for calculating the wind direction numerical value of a wind farm based on wind direction correlation according to the first aspect of the embodiment of the present invention.

[0025] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the method for calculating the wind direction numerical value of a wind farm based on wind direction correlation according to the first aspect of the embodiment of the present invention.

[0026] The technical solution of the present invention has the following advantages:

[0027] 1. The wind farm wind direction numerical modeling system of the present invention based on wind direction correlation and recurrent neural network algorithm, according to the extremely strong correlation of wind direction data between wind turbines, uses the wind direction operation data of wind turbines to train the wind direction numerical model and judge whether the wind direction measurement results between the units meet the correlation or there are measurement abnormalities, and establishes a variety of neural network models for the wind direction numerical model. By making full use of the extremely strong correlation of wind direction data between the units to establish the wind direction numerical model, there is no need to additionally install wind measurement sensors, and the measurement verification of the wind speed and wind direction sensors of the wind turbines can be realized at low cost.

[0028] 2. The present invention completely overcomes the "circumference characteristic" of wind direction data, that is, the problem that the numerical value is discontinuous when the traditional wind direction representation method crosses the 0° and 360° boundaries. After mapping the unit wind direction data onto a two-dimensional unit circle, the included angle vector is calculated, realizing the superposition conversion calculation of the wind direction vector and the included angle vector, thereby ensuring the continuity of the wind direction data. According to the accurate wind measurement system data model and numerical simulation between wind turbines, when the wind measurement systems of some units in the wind farm are damaged, the wind direction data of the abnormal units can be simulated and calculated through the wind direction data of the remaining units, and the yaw accuracy requirements can be met. Description of the Drawings

[0029] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0030] Figure 1 A flowchart of a specific example of a method for calculating the wind direction value of a wind farm based on wind direction correlation provided by an embodiment of the present invention;

[0031] Figure 2 A schematic diagram of unit vectors and their included angles at different angles of a specific example of a method for calculating the wind direction value of a wind farm based on wind direction correlation provided by an embodiment of the present invention;

[0032] Figure 3 A diagram of an RNN neural network wind direction correlation model of a specific example of a method for calculating the wind direction value of a wind farm based on wind direction correlation provided by an embodiment of the present invention;

[0033] Figure 4 A flowchart of training an RNN neural network vector wind direction correlation model of a specific example of a method for calculating the wind direction value of a wind farm based on wind direction correlation provided by an embodiment of the present invention;

[0034] Figure 5 A diagram of the module composition of a system for calculating the wind direction value of a wind farm based on wind direction correlation provided by an embodiment of the present invention;

[0035] Figure 6 A diagram of the composition of a specific example of a terminal provided by an embodiment of the present invention. Detailed Embodiments

[0036] The following will clearly and completely describe the technical solutions of the present invention in conjunction with the drawings. Obviously, the described embodiments are some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

[0037] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, the terms "first", "second", "third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0038] In the description of the present invention, it should be noted that unless otherwise clearly specified and limited, the terms "installed", "connected", "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can also be the communication inside two elements. It can be a wireless connection or a wired connection. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0039] In addition, the technical features involved in different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0040] Embodiment 1

[0041] A wind farm wind direction numerical calculation method based on wind direction correlation provided by an embodiment of the present invention, as Figure 1 shown, includes the following steps:

[0042] Step S1: Obtain the wind direction data of the wind farm units, map the wind direction data of each unit to the two-dimensional unit circle, and generate vector calculation data of the wind direction and the included angle.

[0043] In a specific embodiment, as Figure 2 shown, the one-dimensional representation of the wind direction data from 0° to 359° is mapped to a two-dimensional unit vector of (sinθ, cosθ), realizing the continuity of the wind direction data expression, and proposing the included angle and the included angle vector of two vectors. The included angle and the vector of the included angle are calculated by the following formula:

[0044]

[0045] Among them, taking the fleet of m units as the object, the wind direction vectors are respectively denoted as Taking the mth unit as the target unit, θ is the wind direction angle of each unit at each moment, α1 is the first deviation angle, α2 is the second deviation angle, and after mapping, it is transformed into a unit vector with a length of 1 as (sinθ, cosθ).

[0046] The wind direction vector becomes the wind direction vector after being rotated by an angle. According to the trigonometric function sum - of - angles formula, the formula is as follows:

[0047]

[0048] Through the above formula method, it is extremely convenient for machine learning to calculate the direction vector and direction difference, and in the process of applying to the calculation of the recurrent neural network model, the operation is always completed in the form of an angle.

[0049] Step S2: Based on the vector calculation data of the wind direction and the included angle, according to the correlation between wind turbines, when there are abnormalities in a preset part of the wind turbines in the wind farm, calculate the wind direction values of the abnormal part of the wind turbines.

[0050] In the embodiment of the present invention, based on the vector calculation data of the wind direction and the included angle, according to the correlation between wind turbines, when there are abnormalities in a preset part of the wind turbines in the wind farm, calculating the wind direction values of the abnormal part of the wind turbines includes: According to the correlation of the wind direction data between the wind turbines in the wind farm, using the operation data of the anemometry system of the existing wind turbines in the wind farm, determining whether the wind direction measurement results between the wind turbines meet the correlation and whether there are abnormalities. Establish a recurrent neural network model in the horizontal direction of the wind direction of the wind turbines. According to the accurate anemometry system data model and numerical simulation between the wind turbines, when there are abnormalities in the anemometry system of a preset part of the wind turbines in the wind farm, simulate and calculate the wind direction data of the abnormal wind turbines through the wind direction data of the remaining wind turbines, and calculate the wind direction values of the abnormal wind turbines.

[0051] In practice, the Recurrent Neural Network (RNN) is a classic feedback neural network, which increases the connection between the hidden layer nodes at different times and solves the defect that the traditional neural network cannot model time. The most basic RNN model consists of an input layer X, a hidden layer H, and an output layer Y. The output of the hidden layer node at the previous moment is also passed to the hidden layer at the next moment.

[0052] As the training time in the model progresses, the RNN has the problem of "gradient disappearance" where the ability to perceive early data information decreases. The Long - Short Term Memory neural network model (LSTM) makes up for the "gradient disappearance" problem of the RNN model and effectively utilizes the "long - distance" data information. Different from the natural language analysis problem with a large time series span and strong correlation, the wind direction is in a state of frequent fluctuation and "reversal". Considering factors such as the frequent fluctuation of wind direction data, small relevant time series span, large number of nodes in the multi - layer neural network, complex calculation, etc., therefore, the present invention selects a recurrent neural network model with a single hidden layer.

[0053] In the embodiments of the present invention, taking a fleet of m units as the object, the wind direction vectors are respectively denoted as Taking the mth unit as the target unit, the output wind direction vector is denoted as Initial offset angle vector Adopt an RNN neural network structure with a single hidden layer. The number of nodes in the input layer X is m - 1, the number of nodes in the output layer Y is 1, and the time series expansion structure is as follows Figure 3 As shown.

[0054] Forward propagation of the RNN model:

[0055]

[0056] In the formula, is the input of the ith input layer node at the tth moment, is the input of the jth hidden layer node at the tth moment, U ij is the transfer parameter from the ith input layer node to the jth hidden layer node, is the output of the jth hidden layer node at the tth moment, W j is the time series transfer parameter of the jth hidden layer node, y (t) is the output of the output layer node at the tth moment, V j is the transfer parameter from the jth hidden layer node to the output layer node, L (t) is the loss function at the tth moment.

[0057] The embodiments of the present invention adopt a grid search method to optimize the number of hidden layer nodes n and the learning rate η. After comparison, it is determined that the number of nodes in the hidden layer H is 7, and the learning rate η = 0.105. Only for example, not limited to this, the corresponding number of nodes is selected according to the actual situation in practical applications.

[0058] In a specific embodiment, as Figure 4 shown, the training flow chart of the RNN neural network vector wind direction correlation model. The working principle of the entire system:

[0059] Collect and organize the operation data of the wind turbine anemometry system, and select data segments with a specified average wind speed. Generally, the greater the average wind speed, the more stable the full-field airflow flow field, and the relatively smaller the fluctuations in wind speed and wind direction.

[0060] Preprocess the wind direction and wind speed data within the corresponding data segments respectively. Among them, the wind direction angle θ of each unit at each moment, after mapping, is transformed into a unit vector of length 1 as (sinθ, cosθ); the wind speed value v of each unit at each moment is directly substituted into the subsequent calculation.

[0061] First, it is the wind direction data model. Initialize the weight coefficient Weights, deviation angle α (sinα, cosα), number of training times, learning rate, and minimum error Loss_min in the recurrent neural network. The initialization parameters are not restricted here and should be selected according to the actual situation. Convolve the unit vector (sinθ, cosθ) of the wind direction data θ of all the units in the field with the weight coefficient Weights and then convert it into a unit vector with a length of 1. Perform angular superposition with the initialized deviation angle α (sinα, cosα), and then calculate the error angle with the actual wind direction data of the target unit. Adjust the weight coefficient Weights and the deviation angle α (sinα, cosα) again according to the error and the learning rate, and iterate repeatedly until the error obtained in a certain training is greater than the error obtained in the previous training, or the number of training times reaches the limit, thereby ending the model training and obtaining the weight coefficient Weights and the deviation angle α (sinα, cosα) of the recurrent neural network.

[0062] In the embodiment of the present invention, a wind farm wind direction numerical calculation method based on wind direction correlation is provided, which overcomes the "circumference characteristic" of wind direction data, that is, in the wind direction representation method, the problem that 0° and 359° are numerically discontinuous. After mapping the wind direction data of each unit to the two-dimensional unit circle, calculate the included angle vector, and successfully realize the superposition conversion calculation of the wind direction vector and the included angle vector, thereby ensuring the continuity of the wind direction data. Establish an accurate wind measurement system data model and numerical simulation between wind turbines, and realize that when the wind measurement system of some units in the wind farm is damaged, the wind direction data of the abnormal unit can be simulated and calculated through the wind direction data of the remaining units to meet the yaw accuracy requirements.

[0063] Embodiment 2

[0064] The embodiment of the present invention provides a wind farm wind direction numerical calculation system based on wind direction correlation, as Figure 5 shown, including:

[0065] The data acquisition module 1 is used to acquire the wind direction data of the wind farm units, map the wind direction data of each unit to the two-dimensional unit circle, and generate vector calculation data of the wind direction and the included angle; this module executes the method described in step S1 of Embodiment 1 and will not be elaborated here.

[0066] The wind direction numerical calculation module 2 is used to calculate the wind direction numerical values of the abnormal part of the units based on the vector calculation data of the wind direction and the included angle and according to the correlation between the wind turbines when a preset part of the units in the wind farm are abnormal; this module executes the method described in step S2 of Embodiment 1 and will not be elaborated here.

[0067] An embodiment of the present invention provides a wind farm wind direction numerical calculation system based on wind direction correlation, which overcomes the "circular characteristic" of wind direction data, that is, the problem that 0° and 359° are numerically discontinuous in wind direction representation. After mapping the wind direction data of each unit to a two-dimensional unit circle, the included angle vector is calculated, and the superposition conversion calculation of the wind direction vector and the included angle vector is successfully realized, thereby ensuring the continuity of the wind direction data. An accurate wind measurement system data model and numerical simulation between wind turbines are established, and when the wind measurement systems of some units in the wind farm are damaged, the wind direction data of abnormal units can be simulated and calculated through the wind direction data of the remaining units to meet the yaw accuracy requirements.

[0068] Embodiment 3

[0069] An embodiment of the present invention provides a terminal, as Figure 6 shown, including: at least one processor 401, such as a CPU (Central Processing Unit), at least one communication interface 403, a memory 404, and at least one communication bus 402. Among them, the communication bus 402 is used to realize the connection communication between these components. Among them, the communication interface 403 may include a display screen and a keyboard. Optionally, the communication interface 403 may further include a standard wired interface and a wireless interface. The memory 404 may be a high-speed RAM memory (Random Access Memory), or a non-volatile memory, such as at least one disk memory. Optionally, the memory 404 may further be at least one storage device located far from the aforementioned processor 401. Among them, the processor 401 may execute the wind farm wind direction numerical calculation method based on wind direction correlation in Embodiment 1. A set of program codes are stored in the memory 404, and the processor 401 calls the program codes stored in the memory 404 to execute the wind farm wind direction numerical calculation method based on wind direction correlation in Embodiment 1. Among them, the communication bus 402 may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The communication bus 402 may be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, Figure 6It is represented by only one line in the figure, but it does not mean that there is only one bus or one type of bus. Among them, the memory 404 may include volatile memory (English: volatile memory), such as random access memory (English: random-access memory, abbreviation: RAM); the memory may also include non-volatile memory (English: non-volatile memory), such as flash memory (English: flash memory), hard disk (English: hard disk drive, abbreviation: HDD) or solid-state drive (English: solid-state drive, abbreviation: SSD); the memory 404 may also include a combination of the above types of memories. Among them, the processor 401 may be a central processing unit (English: central processing unit, abbreviation: CPU), a network processor (English: network processor, abbreviation: NP) or a combination of CPU and NP.

[0070] Among them, the memory 404 may include volatile memory (English: volatile memory), such as random access memory (English: random-access memory, abbreviation: RAM); the memory may also include non-volatile memory (English: non-volatile memory), such as flash memory (English: flash memory), hard disk (English: hard disk drive, abbreviation: HDD) or solid-state drive (English: solid-state drive, abbreviation: SSD); the memory 404 may also include a combination of the above types of memories.

[0071] Among them, the processor 401 may be a central processing unit (English: central processing unit, abbreviation: CPU), a network processor (English: network processor, abbreviation: NP) or a combination of CPU and NP.

[0072] Among them, the processor 401 may further include a hardware chip. The above-mentioned hardware chip may be an application-specific integrated circuit (ASIC), a programmable logic device (PLD), or a combination thereof. The above-mentioned PLD may be a complex programmable logic device (CPLD), a field-programmable gate array (FPGA), a generic array logic (GAL), or any combination thereof.

[0073] Optionally, the memory 404 is further configured to store program instructions. The processor 401 may invoke the program instructions to implement the wind farm wind direction numerical calculation method based on wind direction correlation in Embodiment 1 of the present application.

[0074] The embodiment of the present invention further provides a computer-readable storage medium, on which computer-executable instructions are stored, and the computer-executable instructions can execute the wind farm wind direction numerical calculation method based on wind direction correlation in Embodiment 1. Among them, the storage medium may be a magnetic disk, an optical disc, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk drive (HDD), or a solid-state drive (SSD), etc.; the storage medium may also include a combination of the above-mentioned types of memories.

[0075] Obviously, the above embodiments are merely examples for clear illustration and are not limitations on the implementation manners. For those of ordinary skill in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation manners here. And the obvious changes or modifications derived therefrom are still within the protection scope of the present invention.

Claims

1. A numerical calculation method for wind directions in a wind farm based on wind direction correlation, characterized in that, Including: Obtain the wind direction data of the wind farm units, map the wind direction data of each unit onto a two-dimensional unit circle, and generate vector calculation data of the wind direction and included angle; The process of mapping wind direction data in the range of angular measure from 0° to 360° to a two-dimensional unit circle is essentially the conversion of the wind direction angle in the polar coordinate system to a unit vector in the Cartesian coordinate system. Specifically, for each wind turbine, the wind direction angle θ corresponds to the coordinates (sinθ, cosθ) on the unit circle, forming a standardized vector. The vector calculation data includes the unit vectors corresponding to each wind turbine and the angles between the vectors. Among them, the unit vector represents the wind direction, and the angle between the vectors represents the wind direction difference. Wind direction vectors and and the included angle vector between them can complete vector calculations through the following formula; Based on the vector calculation data of the wind direction and included angle, according to the correlation between wind turbines, when a preset part of the units in the wind farm is abnormal, calculate the wind direction value of the abnormal part of the units; Map the wind direction data of each unit onto a two-dimensional unit circle through the following formula: Among them, θ dif is the included angle itself, is the included angle vector. Taking the fleet of m units as the object, the wind direction vectors are respectively denoted as Taking the m-th unit as the target unit, θ is the wind direction angle of each unit at each moment, α1 is the first deviation angle, α2 is the second deviation angle. After mapping, it is transformed into a unit vector with a length of 1 as (sinθ, cosθ); Calculate the wind direction vector through the following formula: Among them, is the wind direction vector.

2. The method for numerically calculating the wind direction of a wind farm based on wind direction correlation according to claim 1, wherein The above-mentioned vector calculation data based on the wind direction and included angle, according to the correlation between wind turbines, when a preset part of the units in the wind farm is abnormal, calculate the wind direction value of the abnormal part of the units, including: According to the correlation of the wind direction data between the wind farm units, use the operation data of the anemometry system of the existing wind turbines in the wind farm to judge whether the wind direction measurement results between the units meet the correlation and whether there is an abnormality; Establish a recurrent neural network model for the horizontal wind direction of the units. According to the accurate anemometry system data model and numerical simulation between the wind turbines, when the anemometry system of a preset part of the units in the wind farm is abnormal, simulate and calculate the wind direction data of the abnormal units through the wind direction data of the remaining units, and calculate the wind direction value of the abnormal units.

3. The wind farm wind direction numerical calculation method based on wind direction correlation according to claim 2, characterized in that The recurrent neural network model established for the horizontal wind direction of the units is a recurrent neural network model with a single hidden layer.

4. A system for implementing the numerical calculation method of the wind direction of a wind farm based on wind direction correlation as described in any one of claims 1-3, characterized in that, Including: A data acquisition module, configured to obtain the wind direction data of the wind farm units, map the wind direction data of each unit onto a two-dimensional unit circle, and generate vector calculation data of the wind direction and included angle; A wind direction value calculation module, configured to calculate the wind direction value of the abnormal part of the units based on the vector calculation data of the wind direction and included angle, according to the correlation between wind turbines, when a preset part of the units in the wind farm is abnormal.

5. A terminal, characterized in that, Including: At least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the method for calculating the wind direction value of a wind farm based on wind direction correlation according to any one of claims 1-3.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, and the computer instructions are used to cause the computer to execute the method for calculating the wind direction value of a wind farm based on wind direction correlation according to any one of claims 1-3.

Images