Wind farm theoretical power calculation method and device

Abstract

The present invention discloses a method and device for calculating the theoretical power of a wind farm. The method includes: pairing the wind turbines in the wind farm in pairs to obtain a plurality of wind turbine combinations; for each wind turbine combination, calculating the power correlation coefficient and the wind speed correlation coefficient of the wind turbine combination, and calculating the performance correlation coefficient according to the product of the power correlation coefficient and the wind speed correlation coefficient; according to the serial number of the abnormal wind turbine, determining the wind turbine combination with the largest performance correlation coefficient in the combination formed by the wind turbine paired with the abnormal wind turbine, taking the wind turbine paired with the abnormal wind turbine in the wind turbine combination as the target wind turbine, and using the actual power of the target wind turbine as the base number to calculate the abnormal theoretical power. By applying the embodiments of the present invention, the theoretical power of the abnormal wind turbine can be accurately estimated.

Description

Technical Field

[0001] The present invention relates to the technical field of wind farm management, and more particularly to a method and device for calculating the theoretical power of a wind farm. Background Art

[0002] To provide basic data for improving the acceptance capacity of wind power, the dispatching department evaluates the output level of wind power and the curtailment electricity. For the economic and fair distribution of the grid connection capacity of wind power and the promotion of the healthy and orderly development of wind power, improving the monitoring function of the real-time balancing capacity of the power grid, alleviating the contradictions between power plants and the grid, and scientifically and accurately calculating the theoretical power of the wind farm is of great significance. The theoretical power generation of a wind farm refers to the power that can be generated when all the wind turbines in the field can operate normally under the current wind conditions.

[0003] Currently, the existing technologies for calculating the theoretical power of a wind farm are mainly the template machine method, the extrapolation method of wind measurement data, and the nacelle wind speed method.

[0004] The template machine method stipulates that several wind turbines in the wind farm are template turbines, and the theoretical power of the turbines due to faults, maintenance, or power curtailment is calculated based on the output of the template turbines. The extrapolation method of wind measurement data is to calculate the theoretical power of the turbines due to faults, maintenance, or power curtailment according to the wind measurement data of the wind measurement tower and the wind speed-power curve of the wind turbines. The nacelle wind speed method is similar to the extrapolation method of wind measurement data, and the theoretical power is calculated using the self-measured nacelle wind speed of the corresponding wind turbine. Since the template turbines do not participate in regulation and the number cannot exceed 10% of the total number of wind turbines, when the terrain of the wind farm is complex and the wind turbine models are diverse, the performance differences of each wind turbine are large, and the output of the template turbines cannot accurately reflect the actual power generation of other wind turbines. The template turbines need to be filed with the grid dispatching department and are basically fixed. However, the template turbines also have faults and maintenance situations, which are more unfavorable for calculating the theoretical power. The extrapolation method of wind measurement data fits the wind measurement data of the wind measurement tower and the power curve of the wind turbines to calculate the theoretical power. Generally, there is only one wind measurement tower in a wind farm, and the wind farm has a large overall area. The wind measurement data of the wind measurement tower cannot accurately reflect the wind speed at the position of the turbines due to faults, maintenance, or power curtailment, and there are also deviations in the wind speed-power curve of the wind turbines. The nacelle wind speed method uses the self-measured wind measurement data of the wind turbine and calculates according to the wind speed-power curve of the wind turbine. However, the wind speed-power curve of the wind turbine is only a probability distribution, with large deviations, and the quality of the self-measured wind measurement data of the wind turbine is poor, and there are also equipment damages from time to time, resulting in inaccurate wind speed measurement results.

[0005] Therefore, the existing technologies have the technical problem that the theoretical power of abnormal wind turbines in the wind farm cannot be accurately calculated. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a method and device for calculating the theoretical power of a wind farm to accurately calculate the theoretical power of abnormal wind turbines.

[0007] The present invention solves the above technical problems through the following technical solutions:

[0008] The present invention provides a method for calculating the theoretical power of a wind farm, and the method includes:

[0009] Pair up the wind turbines in the wind farm two by two to obtain a number of wind turbine combinations;

[0010] For each wind turbine combination, calculate the power correlation coefficient and the wind speed correlation coefficient of the wind turbine combination, and calculate the performance correlation coefficient according to the product of the power correlation coefficient and the wind speed correlation coefficient;

[0011] According to the serial number of the abnormal wind turbine, determine the wind turbine combination with the largest performance correlation coefficient in the combination formed by the wind turbine paired with the abnormal wind turbine, use the wind turbine paired with the abnormal wind turbine in the wind turbine combination as the target wind turbine, and use the actual power of the target wind turbine as the base number to calculate the theoretical power of the abnormal wind turbine.

[0012] Optionally, the calculating the power correlation coefficient and the wind speed correlation coefficient of the wind turbine combination includes:

[0013] According to the power data within the set time range before the current moment, calculate the power correlation coefficient and the wind speed correlation coefficient of the wind turbine combination.

[0014] Optionally, the calculation process of the power correlation coefficient includes:

[0015] Using the formula, Calculate the power correlation coefficient of the current moment power generation of the two wind turbines in each combination, where

[0016] ρ is the power correlation coefficient; Cov(X,Y) is the covariance of the power generation of the two wind turbines; Cov(X,Y) = E[(X - μ X )(Y - μ Y )], and X is the power generation of the wind turbine with serial number x at the current moment; is the power generation of the wind turbine with serial number y at the current moment; μ X is the average power generation of the wind turbine with serial number x within the set time; μ Y is the average power generation of the wind turbine with serial number y within the set time; E[] is the expectation; σ X is the standard deviation of the time series power data array corresponding to the power curve of the wind turbine with serial number x in the combination; σ Y is the standard deviation of the time series power data array corresponding to the power curve of the wind turbine with serial number y in the combination.

[0017] Optionally, the calculating the performance correlation coefficient according to the product of the power correlation coefficient and the wind speed correlation coefficient includes:

[0018] Take the Hadamard product of the power correlation coefficient matrix and the wind speed correlation coefficient matrix as the performance correlation coefficient.

[0019] Optionally, determining the fan combination with the largest performance correlation coefficient among the combinations formed by pairing fans with the abnormal fan includes:

[0020] Arrange all fans according to the fan numbers to obtain a horizontal header, and set the horizontal header vertically to obtain a vertical header. Generate an empty table based on the horizontal header and the vertical header, and fill in the corresponding correlation coefficients at the positions of the corresponding fan combinations in the empty table to obtain a performance correlation coefficient table;

[0021] Determine the fan combination with the largest performance correlation coefficient among the combinations formed by pairing fans with the abnormal fan from the performance correlation coefficient table.

[0022] Optionally, using the actual power of the target fan as the base for calculating the theoretical power of the abnormal fan includes:

[0023] Use the actual power of the target fan as the base, and multiply the base by the rated capacity ratio of the abnormal fan to obtain the theoretical power of the abnormal fan.

[0024] Optionally, the method further includes:

[0025] Estimate the overall theoretical power of the wind farm based on the theoretical power of the combination where the abnormal fan is located.

[0026] The present invention also provides a device for calculating the theoretical power of a wind farm, and the device includes:

[0027] A pairing module for pairing the fans in the wind farm in pairs to obtain a number of fan combinations;

[0028] A calculation module for calculating the power correlation coefficient and the wind speed correlation coefficient of each fan combination, and calculating the performance correlation coefficient according to the product of the power correlation coefficient and the wind speed correlation coefficient;

[0029] A determination module for determining the fan combination with the largest performance correlation coefficient among the combinations formed by pairing fans with the abnormal fan according to the serial number of the abnormal fan, taking the fan paired with the abnormal fan in the fan combination as the target fan, and using the actual power of the target fan as the base for calculating the theoretical power of the abnormal fan.

[0030] Optionally, the calculation module is used for:

[0031] Calculate the power correlation coefficient and the wind speed correlation coefficient of the fan combination according to the power data within the set time range before the current moment.

[0032] Optionally, the calculation module is used for:

[0033] Using the formula, calculate the power correlation coefficient of the current power generation of two wind turbines in each combination, where

[0034] ρ is the power correlation coefficient; Cov(X,Y) is the covariance of the power generation of two wind turbines; Cov(X,Y) = E[(X - μ X )(Y - μ Y )], and X is the current power generation of the wind turbine with serial number x; is the current power generation of the wind turbine with serial number y; μ X is the average power generation of the wind turbine with serial number x within the set time; μ Y is the average power generation of the wind turbine with serial number y within the set time; E[] is the expectation; σ X is the standard deviation of the time-series power data array corresponding to the power curve of the wind turbine with serial number x in the combination; σ Y is the standard deviation of the time-series power data array corresponding to the power curve of the wind turbine with serial number y in the combination.

[0035] The present invention has the following advantages compared with the prior art:

[0036] The present invention groups the wind turbines in the wind farm in pairs, lists all combinations, calculates the correlation coefficient of two wind turbines according to the power generation and wind speed of the two wind turbines in all combinations for a period of time, and generates a performance correlation coefficient matrix that is the Hadamard product of the power correlation coefficient matrix and the wind speed correlation coefficient matrix of each wind turbine combination. By analyzing the wind turbine performance correlation coefficient matrix, the wind turbine combination with the closest power generation performance can be found, and then the theoretical power of the abnormal wind turbine can be accurately estimated. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 is a schematic flowchart of a method for calculating the theoretical power of a wind farm provided by an embodiment of the present invention;

[0038] Figure 2 is a schematic diagram of the principle of a method for calculating the theoretical power of a wind farm provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0039] The following is a detailed description of the embodiments of the present invention. The embodiments are implemented on the premise of the technical solution of the present invention, and detailed implementation manners and specific operation processes are given. However, the protection scope of the present invention is not limited to the following embodiments.

[0040] Figure 1 is a schematic flowchart of a method for calculating the theoretical power of a wind farm provided by an embodiment of the present invention; Figure 2 is a schematic diagram of the principle of a method for calculating the theoretical power of a wind farm provided by an embodiment of the present invention, asFigure 1 and Figure 2 As shown in Figure 2 , the wind farm theoretical power calculation method provided by the invention embodiment includes the following steps:

[0041] S101: Pair up the wind turbines in the wind farm to obtain a number of wind turbine combinations.

[0042] Pair up the n wind turbines (wind turbine numbers are #1, #2, #3... #n) in a certain wind farm, then there are a total of m = C n 2 kinds of combination methods.

[0043] Wind turbine combinations:

[0044] [1,2]\[1,3]\[1,4]...[1,n]\[2,3]\[2,4]...[2,n]\[3,4]...[3,n]\...\[n - 1,n]

[0045] S102: For each wind turbine combination, calculate the power correlation coefficient matrix and the wind speed correlation coefficient matrix of the combination, and calculate the performance correlation coefficient according to the product of the power correlation coefficient and the wind speed correlation coefficient.

[0046] The system statistically calculates the power generation power of each wind turbine in real time and draws the power curve of each wind turbine. At the current moment, when the system monitors that the power generation power of a certain wind turbine is abnormal, taking 60 minutes as the set time length as an example, then filter the power generation power of each wind turbine in the time period from 11:00:00 to 12:00:00 to obtain the power curves of each wind turbine.

[0047] Then use the formula, to calculate the power correlation coefficient of the current moment power generation power of the two wind turbines in each combination, where,

[0048] ρ is the power correlation coefficient; Cov(X,Y) is the covariance of the power generation power of the two wind turbines; Cov(X,Y) = E[(X - μ X )(Y - μ Y )], and X is the power generation power of the wind turbine with serial number x at the current moment; is the power generation power of the wind turbine with serial number y at the current moment; μ X is the average power generation power of the wind turbine with serial number x within the set time; μ Y is the average power generation power of the wind turbine with serial number y within the set time; E[] is the expectation function; σ X is the standard deviation of the time series power data array corresponding to the power curve of the wind turbine with serial number x in the combination; σ Y is the standard deviation of the time series power data array corresponding to the power curve of the wind turbine with serial number y in the combination.

[0049] Furthermore, a power correlation coefficient matrix can be constructed based on the power correlation coefficients of each fan combination. Table 1 shows the power correlation coefficient matrix P corresponding to all combinations obtained 功率 .

[0050] Table 1

[0051]

[0052]

[0053] ρ 功率12 is the power correlation coefficient between the fan numbered 1 and the fan numbered 2.

[0054] This matrix is a semi - amplitude matrix, symmetric about the diagonal 1 (the correlation coefficient of the same fan combination is 1), and ρ 功率nm = ρ 功率mn .

[0055] Similarly, according to the above calculation method, the only difference is that all parameters are nacelle wind speed parameters when generating Table 2. Then, the wind speed correlation coefficient matrix W of all combinations as shown in Table 2 is calculated 机舱风速 matrix.

[0056] Table 2

[0057] #1 Fan #2 Fan #3 Fan …… #n Fan #1 Fan 1 <![CDATA[ρ 机舱风速12 > <![CDATA[ρ 机舱风速13 > …… <![CDATA[ρ 机舱风速1n > #2 Fan <![CDATA[ρ 机舱风速21 > 1 <![CDATA[ρ 机舱风速23 > …… <![CDATA[ρ 机舱风速2n > #3 Fan <![CDATA[ρ 机舱风速31 > <![CDATA[ρ 机舱风速32 > 1 …… <![CDATA[ρ 机舱风速3n > …… …… …… …… 1 …… #n Fan <![CDATA[ρ 机舱风速n1 > <![CDATA[ρ 功率n2 > <![CDATA[ρ 机舱风速n3 > …… 1

[0058] ρ 机舱风速12 is the wind speed correlation coefficient based on the nacelle wind speed between the fan numbered 1 and the fan numbered 2.

[0059] S103 (not shown in the figure): Arrange all fans according to the fan numbers to obtain a horizontal header, and set the horizontal header vertically to obtain a vertical header. Generate an empty table according to the horizontal header and the vertical header, and fill in the corresponding correlation coefficients at the positions of the corresponding fan combinations in the empty table to obtain a performance correlation coefficient table.

[0060] Table 3 shows the performance correlation matrix C corresponding to the fan power generation performance 性能 = P 功率 (power correlation matrix) · W 机舱风速 (nacelle wind speed correlation matrix). As shown in Table 3, the first row of Table 3 is the horizontal header, and the first column of Table 3 is the vertical header. The sorting order of the fan numbers in the horizontal header from left to right is the same as the sorting order of the fan numbers in the vertical header from top to bottom.

[0061] Table 3

[0062]

[0063] ρ 性能n1It is the performance correlation coefficient between the fan numbered 1 and the fan numbered n.

[0064] It should be emphasized that the step S103 of the embodiment of the present invention is an optional step. Combining the step S103 with other steps can further improve the query efficiency and be more intuitive.

[0065] S104: According to the serial number of the abnormal fan, determine the fan combination with the largest performance correlation coefficient in the combination formed by the fans paired with the abnormal fan from the performance correlation coefficient table. Take the fan paired with the abnormal fan in this fan combination as the target fan, and use the actual power of the target fan as the base to calculate the theoretical power of the abnormal fan.

[0066] When there are fans with faults, maintenance, or power restrictions, these fans are regarded as abnormal fans. Then, find the maximum value of the performance correlation matrix of the abnormal fans in the performance correlation coefficient table in step S103 to obtain the target fan with the highest power generation performance correlation coefficient with this fan. At this time, the target fan is a fan operating normally. Take the actual power of the target fan at the current moment as the base, and multiply the base by the rated capacity ratio of the fan with faults, maintenance, or power restrictions as the theoretical power of the fan with faults, maintenance, or power restrictions.

[0067] It can be understood that the rated capacity ratio of the abnormal fan is preset.

[0068] Furthermore, the power generation performance correlation coefficient matrix of the fans in all combinations can be calculated in real time and scrolled according to the operating states of each fan, and the correlation coefficient matrix is dynamically refreshed to ensure the real-time performance of the fan performance correlation.

[0069] Since various types of fans from different manufacturers and different models are equipped in the wind farm, there are significant differences in the fan models in the same wind farm, and their power generation performances are different; even for the same model of fans, there will also be differences in power generation performance during the manufacturing process and operation process. Coupled with the randomness of the wind farm and the large area of the wind farm, the power generation states of the same model of fans in different positions in the same wind farm will also have significant differences. Therefore, only from the correlation of the power generation power, it is impossible to determine the closeness of the power generation performance of the fans. For example, even if the power correlation of fans with different capacities is very high, it cannot be determined that their power generation performances are close.

[0070] The power generation performance of the fan is mainly determined by the power curve of the fan, and the power curve consists of two parameters: wind speed and power. It is necessary to add the wind speed parameter to determine the power generation performance of the fan. Therefore, it is also necessary to calculate the correlation coefficient of the nacelle wind speed of the two fans in each combination for a period of time for all two-fan combinations, and generate the nacelle wind speed correlation coefficient matrix of the two fans in each combination.

[0071] Since the correlation coefficient is dimensionless, there is no mathematical relationship between the power correlation coefficient matrix and the nacelle wind speed correlation coefficient matrix. To avoid irrelevant correlations and keep the correlation coefficient between ±1, the Hadamard product of the matrices with corresponding elements multiplied is selected as the third correlation coefficient matrix, which is defined as the wind turbine power generation performance correlation matrix. In the wind turbine power generation performance correlation matrix, the maximum coefficient value indicates that the power generation performances of the two wind turbines in this combination are the closest.

[0072] The present invention can avoid the deviation of the measured wind data of the reference wind turbine, the anemometer tower and the wind turbine itself, as well as the power curve, according to the actual operating conditions of each wind turbine in the wind farm. Using the data closest to the actual operating state, the theoretical power of the wind farm can be calculated. Moreover, since the correlation coefficient matrix is calculated and analyzed in real time under the same operating conditions, the change in the wind turbine performance is also taken into account, and the theoretical power of the wind farm can be accurately calculated.

[0073] Furthermore, the sum of the theoretical powers of all abnormal wind turbines in the wind farm and the actual powers of normal wind turbines can be used to calculate the overall theoretical power of the wind farm.

[0074] Corresponding to the above method embodiments of the present invention, the present invention also provides a device for calculating the theoretical power of a wind farm, and the device includes:

[0075] A pairing module, configured to pair the wind turbines in the wind farm in pairs to obtain a plurality of wind turbine combinations;

[0076] A calculation module, configured to calculate the power correlation coefficient and the wind speed correlation coefficient of each wind turbine combination, and calculate the performance correlation coefficient according to the product of the power correlation coefficient and the wind speed correlation coefficient;

[0077] A determination module, configured to determine, according to the serial number of the abnormal wind turbine, the wind turbine combination with the largest performance correlation coefficient in the combination formed by the wind turbine paired with the abnormal wind turbine, use the wind turbine paired with the abnormal wind turbine in this wind turbine combination as the target wind turbine, and use the actual power of the target wind turbine as the base number to calculate the abnormal theoretical power.

[0078] In a specific implementation manner of the embodiment of the present invention, the calculation module is configured to:

[0079] Calculate the power correlation coefficient and the wind speed correlation coefficient of the wind turbine combination according to the power data within the set duration range before the current moment.

[0080] In a specific implementation manner of the embodiment of the present invention, the calculation module is configured to:

[0081] Use the formula To calculate the power correlation coefficient of the current moment power generation of the two wind turbines in each combination, where

[0082] ρ is the power correlation coefficient; Cov(X,Y) is the covariance of the power generation of two wind turbines; Cov(X,Y) = E[(X - μ X )(Y - μ Y ), where X is the power generation of the wind turbine with serial number x at the current moment; is the power generation of the wind turbine with serial number y at the current moment; μ X is the average power generation of the wind turbine with serial number x within the set time; μ Y is the average power generation of the wind turbine with serial number y within the set time; E[] is the expectation; σ X is the standard deviation of the time-series power data array corresponding to the power curve of the wind turbine with serial number x in the combination; σ Y is the standard deviation of the time-series power data array corresponding to the power curve of the wind turbine with serial number y in the combination.

[0083] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. Based on the theoretical power calculation method for a wind farm, characterized in that, the method includes: Pairing the wind turbines in the wind farm in pairs to obtain a number of wind turbine combinations; For each wind turbine combination, calculating the power correlation coefficient and the wind speed correlation coefficient of the combination, and calculating the performance correlation coefficient according to the product of the power correlation coefficient and the wind speed correlation coefficient; According to the serial number of the abnormal wind turbine, determining the wind turbine combination with the largest performance correlation coefficient among the combinations formed by the wind turbines paired with the abnormal wind turbine, taking the wind turbine paired with the abnormal wind turbine in this combination as the target wind turbine, and using the actual power of the target wind turbine as the base to calculate the theoretical power of the abnormal wind turbine.

2. The theoretical power calculation method for a wind farm according to claim 1, characterized in that, the calculating the power correlation coefficient and the wind speed correlation coefficient of the wind turbine combination includes: Calculating the power correlation coefficient of the wind turbine combination according to the power data within a set time range before the current moment; The calculation process of the power correlation coefficient includes: Using the formula, calculate the power correlation coefficient of the current power generation of the two wind turbines in each combination, where ρ is the power correlation coefficient; Cov(X,Y) is the covariance of the power generation of two wind turbines; Cov(X,Y) = E[(X - μ X )(Y - μ Y ), where X is the power generation of the wind turbine with serial number x at the current moment; Y is the power generation of the wind turbine with serial number y at the current moment; μ X is the average power generation of the wind turbine with serial number x within the set time; μ Y is the average power generation of the wind turbine with serial number y within the set time; E[] is the expectation; σ X is the standard deviation of the time-series power data array corresponding to the power curve of the wind turbine with serial number x in the combination; σ Y is the standard deviation of the time-series power data array corresponding to the power curve of the wind turbine with serial number y in the combination; Calculating the wind speed correlation coefficient of the wind turbine combination according to the wind speed data within a set time range before the current moment; The calculation process of the wind speed correlation coefficient includes: Using the formula, calculate the wind speed correlation coefficient of the current wind speeds of the two fans in each combination, where ρ′ is the wind speed correlation coefficient; Cov(M,N) is the covariance of the wind speeds of two wind turbines; Cov(M,N) = E′[(M - μ m )(N - μ n ), where M is the wind speed of the wind turbine with serial number x at the current moment; N is the wind speed of the wind turbine with serial number y at the current moment; μ m is the average wind speed of the wind turbine with serial number m within the set time; μ n is the average wind speed of the wind turbine with serial number n within the set time; E′[] is the expectation; σ M is the standard deviation of the time series power data array corresponding to the wind speed curve of the wind turbine with serial number m in the combination; σ N is the standard deviation of the time series power data array corresponding to the wind speed curve of the wind turbine with serial number n in the combination.

3. The theoretical power calculation method for a wind farm according to claim 1, characterized in that, the calculating the performance correlation coefficient according to the product of the power correlation coefficient and the wind speed correlation coefficient includes: Taking the Hadamard product of the power correlation coefficient matrix and the wind speed correlation coefficient matrix as the performance correlation coefficient.

4. The theoretical power calculation method for a wind farm according to claim 1, characterized in that, the determining the wind turbine combination with the largest performance correlation coefficient among the combinations formed by the wind turbines paired with the abnormal wind turbine includes: Arranging all the wind turbines in order of the wind turbine serial numbers to obtain a horizontal header, vertically setting the horizontal header to obtain a vertical header, generating an empty table according to the horizontal header and the vertical header, and filling in the corresponding correlation coefficients at the positions of the corresponding wind turbine combinations in the empty table to obtain a performance correlation coefficient table; Determining the wind turbine combination with the largest performance correlation coefficient among the combinations formed by the wind turbines paired with the abnormal wind turbine from the performance correlation coefficient table.

5. The theoretical power calculation method for a wind farm according to claim 1, characterized in that, the using the actual power of the target wind turbine as the base to calculate the theoretical power of the abnormal wind turbine includes: Using the actual power of the target wind turbine as the base, and multiplying the base by the rated capacity ratio of the abnormal wind turbine to obtain the theoretical power of the abnormal wind turbine.

6. The theoretical power calculation method for a wind farm according to claim 1, characterized in that, the method further includes: Estimating the overall theoretical power of the wind farm according to the theoretical power of the combination where the abnormal wind turbine is located.

7. A theoretical power calculation device for a wind farm, characterized in that, the device includes: A pairing module for pairing the wind turbines in the wind farm in pairs to obtain a number of wind turbine combinations; A calculation module for calculating the power correlation coefficient and the wind speed correlation coefficient of each wind turbine combination, and calculating the performance correlation coefficient according to the product of the power correlation coefficient and the wind speed correlation coefficient; A determination module, configured to determine, according to the serial number of an abnormal fan, the fan combination with the largest performance correlation coefficient in the combinations formed by the fans paired with the abnormal fan, use the fan paired with the abnormal fan in the fan combination as the target fan, and use the actual power of the target fan as the base number to calculate the theoretical power of the abnormal fan.

8. The wind farm theoretical power calculation device according to claim 7, wherein, the calculation module is configured to: calculate the power correlation coefficient of the fan combination according to the power data within a set time range before the current moment; specifically, the calculation module is configured to: ρ is the power correlation coefficient; Cov(X,Y) is the covariance of the generated powers of two wind turbines; Cov(X,Y) = E[(X - μ X )(Y - μ Y ), where X is the generated power of the wind turbine with serial number x at the current moment; Y is the generated power of the wind turbine with serial number y at the current moment; μ X is the mean value of the generated power of the wind turbine with serial number x within the set time period; μ Y is the mean value of the generated power of the wind turbine with serial number y within the set time period; E[] is the expectation; σ X is the standard deviation of the time-series power data array corresponding to the power curve of the wind turbine with serial number x in the combination; σ Y is the standard deviation of the time-series power data array corresponding to the power curve of the wind turbine with serial number y in the combination; calculate the wind speed correlation coefficient of the fan combination according to the wind speed data within a set time range before the current moment; the calculation process of the wind speed correlation coefficient includes: Using the formula, calculate the wind speed correlation coefficient of the current wind speeds of the two fans in each combination, where ρ′ is the wind speed correlation coefficient; Cov(M,N) is the covariance of the wind speeds of two wind turbines; Cov(M,N) = E′[(M - μ m )(N - μ n ), where M is the wind speed of the wind turbine with serial number x at the current moment; N is the wind speed of the wind turbine with serial number y at the current moment; μ m is the average wind speed of the wind turbine with serial number m within the set time; μ n is the average wind speed of the wind turbine with serial number n within the set time; E′[] is the expectation; σ M is the standard deviation of the time series power data array corresponding to the wind speed curve of the wind turbine with serial number m in the combination; σ N is the standard deviation of the time series power data array corresponding to the wind speed curve of the wind turbine with serial number n in the combination.

Images