Wind turbine generator performance analysis method, device and equipment and storage medium

Abstract

The invention discloses a wind turbine generator performance analysis method and device, equipment and a storage medium, and relates to the technical field of computers, and the method comprises the steps: predicting each input sample in a target wind turbine generator data set through a preset prediction model, so as to obtain a plurality of power prediction values corresponding to each input sample; determining a plurality of power actual values corresponding to each input sample according to the target wind turbine generator data set, and calculating a plurality of target residual values corresponding to each input sample according to the plurality of power actual values and the plurality of power predicted values; determining a target replacement part, and determining a plurality of first residual values corresponding to the target replacement part before replacement and a plurality of second residual values corresponding to the target replacement part after replacement from the plurality of target residual values; and based on the plurality of first residual values, the plurality of second residual values, the target residual value and the plurality of actual power values, calculating the performance reduction contribution degree of the wind turbine generator corresponding to the target replacement part. Therefore, accurate quantification of the aging contribution degree of the wind turbine generator components can be realized.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method, apparatus, equipment, and storage medium for analyzing the performance of wind turbine generators. Background Technology

[0002] With the rapid growth of installed wind power capacity, more and more aging wind farms are reaching the end of their service life and will soon face the issues of replacing major components and increasing power generation capacity. However, most concerns about wind turbine aging focus on reliability and failure rate, rather than performance degradation. Therefore, it is necessary to analyze wind power performance from the massive database of monitoring data.

[0003] Currently, traditional power curve analysis methods have significant errors when wind speed fluctuates, making it impossible to accurately quantify wind turbine performance degradation. Some technologies require additional sensors, increasing deployment costs, while others can only diagnose single component failures, failing to quantify the impact of coordinated aging of multiple components. Furthermore, when the cumulative performance loss of the unit reaches a certain threshold, existing technologies struggle to accurately pinpoint the main degradation sources, leading to high gearbox replacement rates and excessively high costs per maintenance. Summary of the Invention

[0004] In view of this, the purpose of this invention is to provide a wind turbine performance analysis method that can accurately quantify the contribution of wind turbine component aging by quantifying the performance deviation of the wind turbine before and after component replacement. The specific solution is as follows:

[0005] In a first aspect, this application discloses a method for analyzing the performance of wind turbine generators, including:

[0006] The preset prediction model is used to predict each input sample in the target wind turbine dataset to obtain several power prediction values ​​corresponding to each input sample.

[0007] Based on the target wind turbine dataset, determine several actual power values ​​corresponding to each input sample, and calculate several target residual values ​​corresponding to each input sample based on the several actual power values ​​and the several predicted power values.

[0008] Identify the target replacement component, and determine from the plurality of target residual values ​​a plurality of first residual values ​​corresponding to the target replacement component before replacement and a plurality of second residual values ​​corresponding to the target replacement component after replacement;

[0009] The contribution of the wind turbine performance degradation corresponding to the target replacement component is calculated based on the plurality of first residual values, the plurality of second residual values, the target residual value, and the plurality of actual power values.

[0010] Optionally, before predicting each input sample in the target wind turbine dataset using a preset prediction model to obtain several predicted power values ​​corresponding to each input sample, the method further includes:

[0011] The annual operation dataset of the target wind turbine is obtained through a pre-set data acquisition system;

[0012] Target power data that is less than a preset rated power threshold are selected from the annual operating dataset;

[0013] The current air density is determined based on standard air density, standard temperature, and actual ambient temperature. The wind speed data in the annual operational dataset is then normalized using the current air density and the standard air density to obtain the target wind speed data.

[0014] A target wind turbine dataset is generated based on the target power data, the target wind speed data, and the annual operation dataset.

[0015] Optionally, before predicting each input sample in the target wind turbine dataset using a preset prediction model to obtain several predicted power values ​​corresponding to each input sample, the method further includes:

[0016] Several input vectors are constructed based on the target wind turbine dataset corresponding to the wind turbine, and the kernel function values ​​between the several input vectors are calculated through the target kernel function;

[0017] Accordingly, the step of predicting each input sample in the target wind turbine dataset using a preset prediction model to obtain several predicted power values ​​corresponding to each input sample includes:

[0018] The prediction model uses the kernel function value to predict the power prediction value corresponding to each input sample in the target wind turbine dataset.

[0019] Optionally, the step of determining several actual power values ​​corresponding to each input sample based on the target wind turbine dataset, and calculating several target residual values ​​corresponding to each input sample based on the several actual power values ​​and the several predicted power values, includes:

[0020] Select several actual power values ​​corresponding to each input sample from the target wind turbine dataset;

[0021] Determine the target actual power value and the target predicted power value corresponding to each input sample among the plurality of actual power values ​​and the plurality of predicted power values;

[0022] The difference between the actual value of the target power and the predicted value of the target power is calculated to obtain several target residual values ​​corresponding to each input sample.

[0023] Optionally, determining the target replacement component and determining, from the plurality of target residual values, a plurality of first residual values ​​corresponding to the target replacement component before replacement and a plurality of second residual values ​​corresponding to the target replacement component after replacement, includes:

[0024] Identify the target replacement component and determine the component replacement time of the target replacement component, and split each input sample based on the component replacement time to determine a first input sample before the component replacement time and a second input sample after the component replacement time;

[0025] From the plurality of target residual values, determine a plurality of first residual values ​​corresponding to the first input sample and a plurality of second residual values ​​corresponding to the second input sample.

[0026] Optionally, the step of calculating the contribution of the wind turbine performance degradation corresponding to the target replacement component based on the plurality of first residual values, the plurality of second residual values, the target residual value, and the plurality of actual power values ​​includes:

[0027] From the plurality of actual power values, determine a plurality of first actual power values ​​corresponding to the plurality of first residual values, and from the plurality of actual power values, determine a plurality of second actual power values ​​corresponding to the plurality of second residual values;

[0028] Calculate the ratio between each residual value in the plurality of first residual values ​​and the corresponding actual power value in the plurality of first actual power values ​​to obtain a plurality of first ratios, and calculate the average value of the plurality of first ratios to use the obtained average value as the first deviation index;

[0029] Calculate the ratio between each residual value in the plurality of second residual values ​​and the corresponding actual power value in the plurality of second actual power values ​​to obtain a plurality of second ratios, and calculate the average value of the plurality of second ratios to use the obtained average value as the second deviation index;

[0030] Calculate the ratio between each residual value in the plurality of target residual values ​​and the corresponding actual power value in the plurality of actual power values ​​to obtain a plurality of third ratios, and calculate the average value of the plurality of third ratios to use the obtained average value as the third deviation index;

[0031] The contribution of the wind turbine performance degradation corresponding to the target replacement component is calculated based on the first deviation index, the second deviation index, and the third deviation index.

[0032] Optionally, calculating the contribution of the wind turbine performance degradation corresponding to the target replacement component based on the first deviation index, the second deviation index, and the third deviation index includes:

[0033] Calculate the difference between the first deviation index and the second deviation index to obtain the target difference;

[0034] The absolute value of the target difference is calculated as the ratio of the third ratio, and the obtained ratio is used as the contribution of the wind turbine performance degradation corresponding to the target replacement component.

[0035] Secondly, this application discloses a wind turbine performance analysis device, comprising:

[0036] The power prediction module is used to predict each input sample in the target wind turbine dataset using a preset prediction model, so as to obtain several power prediction values ​​corresponding to each input sample.

[0037] The residual calculation module is used to determine several actual power values ​​corresponding to each input sample based on the target wind turbine dataset, and to calculate several target residual values ​​corresponding to each input sample based on the several actual power values ​​and the several predicted power values.

[0038] The data classification module is used to determine the target replacement part, and to determine, from the plurality of target residual values, a plurality of first residual values ​​corresponding to the target replacement part before replacement and a plurality of second residual values ​​corresponding to the target replacement part after replacement;

[0039] The contribution determination module is used to calculate the contribution of the wind turbine performance degradation corresponding to the target replacement component based on the plurality of first residual values, the plurality of second residual values, the target residual value, and the plurality of actual power values.

[0040] Thirdly, this application discloses an electronic device, including:

[0041] Memory, used to store computer programs;

[0042] A processor is used to execute the computer program to implement the wind turbine performance analysis method described above.

[0043] Fourthly, this application discloses a computer-readable storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the wind turbine performance analysis method as described above.

[0044] Therefore, the method of this application can predict each input sample in the target wind turbine dataset using a preset prediction model to obtain several predicted power values ​​corresponding to each input sample; determine several actual power values ​​corresponding to each input sample based on the target wind turbine dataset, and calculate several target residual values ​​corresponding to each input sample based on the several actual power values ​​and the several predicted power values; determine the target replacement component, and determine several first residual values ​​corresponding to the target replacement component before replacement and several second residual values ​​corresponding to the target replacement component after replacement from the several target residual values; calculate the contribution of the target replacement component to the wind turbine performance degradation based on the several first residual values, the several second residual values, the target residual values, and the several actual power values. In this way, the aging contribution of multiple components of the wind turbine can be accurately quantified. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0046] Figure 1 This is a flowchart of a wind turbine performance analysis method disclosed in this application;

[0047] Figure 2 This is a schematic diagram of the structure of a wind turbine performance analysis device disclosed in this application;

[0048] Figure 3 This is a structural diagram of an electronic device disclosed in this application. Detailed Implementation

[0049] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0050] Currently, traditional power curve analysis methods have significant errors when wind speed fluctuates, making it impossible to accurately quantify wind turbine performance degradation. Some technologies require additional sensors, increasing deployment costs, while others can only diagnose single component failures, failing to quantify the impact of coordinated aging of multiple components. Furthermore, when the cumulative performance loss of the unit reaches a certain threshold, existing technologies struggle to accurately pinpoint the main degradation sources, leading to high gearbox replacement rates and excessively high costs per maintenance.

[0051] To overcome the aforementioned technical problems, this application discloses a wind turbine performance analysis method, which can accurately quantify the contribution of wind turbine component aging by quantifying the performance deviation of the wind turbine before and after component replacement.

[0052] See Figure 1 As shown in the figure, an embodiment of the present invention discloses a method for wind turbine performance analysis, including:

[0053] Step S11: Predict each input sample in the target wind turbine dataset using a preset prediction model to obtain several power prediction values ​​corresponding to each input sample.

[0054] In this embodiment, the power prediction of the input samples is performed using the constructed prediction model to obtain the predicted power value corresponding to each input sample. However, it should be noted that before performing power prediction, it is necessary to first construct a target wind turbine dataset corresponding to the wind turbine. Specifically, the annual operation dataset of the target wind turbine needs to be obtained through a preset data acquisition system. This dataset needs to be obtained through a SCADA (Supervisory Control and Data Acquisition) system. Then, the data in the annual operation dataset needs to be filtered to select target power data that are lower than a preset rated power threshold. It should be noted that the annual operation dataset is the dataset corresponding to the base year of the wind turbine, and the base year data refers to the various operating data recorded by the SCADA system during the first full year of operation of the wind turbine, including information such as wind speed, wind direction, power output, and rotational speed under different operating conditions, which can reflect the performance of the wind turbine in its initial state. The data of the first full year of operation is chosen as the base because the wind turbine is relatively new at this time, and its performance has not been significantly affected by aging, so it can serve as a benchmark for subsequent performance comparison and degradation analysis. This effectively ensures data quality and avoids inaccurate performance analysis due to poor data quality.

[0055] Furthermore, the wind speed data in the annual operational dataset needs to be normalized. The current air density needs to be determined based on standard air density, standard temperature, and actual ambient temperature. The expression for the current air density is as follows:

[0056] ;

[0057] in, Given the current air density, With a standard air density of 1.225 kg / m³, T ref The standard temperature is 288.15 K, and T is the actual ambient temperature. The units for both the actual ambient temperature and the standard temperature are Kelvin (T).

[0058] Furthermore, the wind speed data in the annual operational dataset needs to be normalized using the current air density and the standard air density to obtain the target wind speed data. The expression for normalizing the wind speed data is as follows:

[0059] ;

[0060] in, The wind speed data is after normalization. For wind speed data in the annual operational dataset, Given the current air density, This is the standard air density.

[0061] Ultimately, a target wind turbine dataset needs to be generated based on the target power data, target wind speed data, and annual operation dataset.

[0062] The next step requires constructing several input vectors based on the target wind turbine dataset. Then, the kernel function values ​​between these input vectors are calculated using the target kernel function. The constructed input vectors are as follows: Where X is the input vector, The wind speed data is after normalization. This represents the turbulence intensity, which is the ratio of the standard deviation of wind speed to the average wind speed. The pitch angle is the propeller angle. The wind turbine rotor speed is the speed of the wind turbine unit. This refers to the generator speed of the wind turbine.

[0063] In the actual embodiment, three operating conditions are illustrated: low temperature in winter, high temperature in summer, and the edge of a typhoon. The input vector X for the three operating conditions is respectively... The following data is included: The three operating conditions are shown in Table 1: Table 1 ;

[0064] After the input vectors are constructed, the baseline year data needs to be trained using a Gaussian kernel support vector regression model to obtain the kernel function values ​​between the input vectors. The kernel function is defined as follows:

[0065] ;

[0066] Wherein K(X) i X j ) is the input vector X i With X j The kernel function values ​​between is the parameter of the Gaussian kernel function, which controls the width of the kernel function. The larger the value, the more sensitive the model is to local features. In this embodiment, the value is 0.5. exp() is the exponential function with the natural constant e as the base. ||.|| is the norm of the vector, which represents the Euclidean distance between two input vectors in the expression.

[0067] After obtaining the kernel function values ​​for each input sample, the prediction model can be used to predict the power of each input sample in the target wind turbine dataset based on these kernel function values. It should be noted that the prediction model is a pre-trained support vector regression model, and the corresponding regression function is:

[0068] ;

[0069] in, and For Lagrange multipliers, Let be the Lagrange multiplier for the i-th sample, used in the optimization process of the support vector regression model. Let K(X) be the dual variable of the Lagrange multiplier for the i-th sample, N be the total number of samples, and K(X) be the dual variable of the Lagrange multiplier for the i-th sample. i (X) is a vector X i The corresponding kernel function value is X, which is the support vector in the target wind turbine dataset, and b is the bias term obtained through training.

[0070] It should be noted that the constraint conditions of the Lagrange multipliers satisfy... as well as And C=1.0.

[0071] It should be noted that the support vector regression model determines the Lagrange multipliers by optimizing the following problem, which is as follows:

[0072] ;

[0073] Where ε = 0.01.

[0074] If the following data are used in the embodiments, three sets of data are given as examples:

[0075] Among them, the Lagrange multipliers are: = 0.05, = 0.01; = 0.03, [[ID=1s]]= 0.025; = 0.04, s= 0.08; and 0 < < C, C = 0.1; Among them, the bias term b = 50; Among them, the Gaussian kernel parameter: ; [[ID=s1]]

[0076] The following uses the trained parameters and combines them with the input vector X test (the working condition to be predicted) to demonstrate the calculation process of the regression function.

[0077] Step 1: Calculate the Gaussian kernel function:

[0078] ;

[0079] Among them , ||X i -X j || 2 is the square of the Euclidean distance between the input vector X i and X j .

[0080] Step 2: Substitute the support vectors and multipliers and expand the regression function:

[0081] Among them, the regression function is: [[ID=s3]]

[0082] .

[0083] Step 3: Input the test vector X test , and calculate the predicted value:

[0084] Assume that the test working condition is the stable working condition in autumn, and the input vector: X test = [9.00, 0.14, 16, 1850], then it is necessary to calculate the kernel function values of X text and 3 support vectors, and then substitute the multiplier difference and the bias term;

[0085] First, it is necessary to calculate K(X text , X1) (the support vector for low temperature in winter), and its vector difference is: X test −X1 = [9.00 - 8.30, 0.14 - 0.12, 16 - 15, 1850 - 1800] = [0.70, 0.02, 1, 50];

[0086] It should be noted that there seems to be some incorrect or unclear notations in the original text (such as "1s" in the translation), but I have translated it as accurately as possible according to the rules.Its squared Euclidean distance is:

[0087] =0.70 2 +0.02 2 +1 2 +50 2 =2501.49;

[0088] Its kernel function value is:

[0089] =exp(−0.001×2501.4904)≈0.08.

[0090] Secondly, it is necessary to calculate K(X) text X2) (Summer high temperature support vectors), their vector difference is: X test −X2=[9.00−7.26, 0.14−0.18, 16−14, 1850−1750]=[1.74, −0.04, 2, 100];

[0091] Its squared Euclidean distance is:

[0092] =1.74 2 +(−0.04) 2 +2 2 +100 2 =10007.02;

[0093] Its kernel function value is:

[0094] =exp(−0.001×10007.0292)=4.5×10 −5 .

[0095] Then, it is necessary to calculate K(X) text， X3) (Typhoon edge support vectors), where the vector difference is: X test −X3=[9.00−11.89, 0.14−0.25, 16−20, 1850−2200]=[−2.89, −0.11, −4, −350];

[0096] The square of the Euclidean distance is:

[0097] =(−2.89) 2 +(−0.11) 2 +(−4) 2 +(−350) 2 =122524.36;

[0098] The kernel function value is:

[0099] =exp(−0.001×122524.3642)≈1.5×10−53.

[0100] Finally, we need to substitute the values ​​into the regression function to calculate the predicted power:

[0101] =0.040; =0.025; =0.032;

[0102] Substituting the kernel function value, multiplier difference, and bias term b=100, we can obtain:

[0103] =0.04×0.0818+0.025×4.5×10 −5 +0.032×1.5×10 −53 +100≈100KW.

[0104] Step S12: Determine several actual power values ​​corresponding to each input sample based on the target wind turbine dataset, and calculate several target residual values ​​corresponding to each input sample based on the several actual power values ​​and the several predicted power values.

[0105] In this embodiment, it is necessary to determine the actual power value corresponding to each input sample based on the wind turbine dataset. Specifically, it is necessary to select several actual power values ​​corresponding to each input sample from the target wind turbine dataset. Since the target wind turbine dataset contains the actual power values ​​corresponding to each input sample, the corresponding actual power values ​​can be directly determined from the target wind turbine dataset based on the input samples.

[0106] Furthermore, it is necessary to determine the target actual power value and target predicted power value corresponding to each input sample from among several actual power values ​​and several predicted power values. For example, if the current input samples are X1, X2, and X3, then it is necessary to determine the target actual power values ​​Y1, Y2, Y3 and the target predicted power values ​​f(X1), f(X2), f(X3) corresponding to X1, X2, and X3 respectively from the determined several actual power values ​​and several predicted power values. Further, it is necessary to calculate the difference between the target actual power value and the target predicted power value to obtain several target residual values ​​corresponding to each input sample, and the expression for the residual value is as follows:

[0107] ;

[0108] Wherein, R(X) i ) is the input sample X i The corresponding residual value, Y iFor X i The corresponding actual value of the target power, f(X) i ) is X i The corresponding target power prediction value.

[0109] Step S13: Determine the target replacement component, and determine from the plurality of target residual values ​​a plurality of first residual values ​​corresponding to the target replacement component before replacement and a plurality of second residual values ​​corresponding to the target replacement component after replacement.

[0110] In this embodiment, it is necessary to determine the residual values ​​before and after the replacement of the target component. Specifically, it is necessary to determine the target component to be replaced and its replacement time, so that each input sample can be split based on the replacement time to determine the first input sample before the replacement time and the second input sample after the replacement time. From a number of target residual values, a number of first residual values ​​corresponding to the first input sample and a number of second residual values ​​corresponding to the second input sample are determined. For example, if the target component to be replaced is a gearbox, and the replacement time of the gearbox is A month B day of the base year, then A month B day of the base year is needed as the replacement time of the target component to be replaced. The samples before A month B day of the base year are used as the first input sample, and the samples after A month B day of the base year are used as the second input sample. Then, from the calculated number of target residual values, a number of first residual values ​​corresponding to the input samples and a number of second residual values ​​corresponding to the second input samples are determined respectively.

[0111] Step S14: Calculate the contribution of the wind turbine performance degradation corresponding to the target replacement component based on the plurality of first residual values, the plurality of second residual values, the target residual value, and the plurality of actual power values.

[0112] In this embodiment, it is necessary to calculate the contribution of the wind turbine performance degradation corresponding to the target replacement component based on several first residual values, several second residual values, the target residual value, and several actual power values. Specifically, it is necessary to calculate the annual performance deviation index before the replacement component. That is, the first deviation index corresponding to the first input sample; the annual performance deviation index after component replacement. That is, the second deviation index corresponding to the second input sample; the total annual performance deviation index of the wind turbine. This refers to the third deviation index corresponding to all input samples. Specifically, it is necessary to determine several first actual power values ​​corresponding to several first residual values ​​from several actual power values, and several second actual power values ​​corresponding to several second residual values ​​from several actual power values. Taking each input sample in the first input sample as an example, it is necessary to calculate the ratio between each residual value in the several first residual values ​​and the corresponding actual power value in the several first actual power values ​​to obtain several first ratios, and calculate the average of the several first ratios to use the average as the first deviation index. Similarly, it is necessary to calculate the ratio between each residual value in the several second residual values ​​and the corresponding actual power value in the several second actual power values ​​to obtain several second ratios, and calculate the average of the several second ratios to use the average as the second deviation index. Furthermore, it is necessary to calculate the ratio between each residual value in the several target residual values ​​and the corresponding actual power value in the several actual power values ​​to obtain several third ratios, and calculate the average of the several third ratios to use the average as the third deviation index. It should be noted that the expression corresponding to the deviation index is as follows:

[0113] ;

[0114] in, N is the deviation index, used to quantify the degree of performance degradation of wind turbines in year t. t Let R(X) be the total number of samples in year t. k ) is the sample X k The residual, Y k For sample X k The actual output.

[0115] Taking a real-world example, if the actual output and predicted output of the wind turbine are as shown in Table 2, then the deviation index corresponding to the sample in Table 1 can be calculated using the data in Table 2. Table 2 is shown below: Table 2 ;

[0116] The residuals for sample 1 are R(X1) = 850 − 860 = −10, for sample 2 R(X2) = 1100 − 1090 = 10, for sample 3 R(X3) = 1400 − 1380 = 20, for sample 4 R(X4) = 1600 − 1620 = −20, and for sample 5 R(X5) = 1800 − 1790 = 10. The deviation index corresponding to the above data is 1.066%, and the calculation process is as follows:

[0117] ;

[0118] Finally, the contribution of the target replaced component to the wind turbine's performance degradation can be calculated based on the first, second, and third deviation indices. Taking a wind turbine with a replaced gearbox as an example, and considering the annual performance deviation indices before and after the replacement as follows: =2.5% and =1.8%, the total annual performance deviation index is =4.0%, and the aging contribution C of component k k The calculation formula is as follows:

[0119] ;

[0120] Therefore, the contribution of a component to the aging of the wind turbine can be calculated before and after replacement, enabling precise quantification of the aging contribution of multiple components in the wind turbine. It should be noted that the annual energy production (AEP) loss can also be verified using the IEC (International Electrotechnical Commission) compartmentalization method.

[0121] In this embodiment, a preset prediction model can be used to predict each input sample in the target wind turbine dataset to obtain several predicted power values ​​corresponding to each input sample. Based on the target wind turbine dataset, several actual power values ​​corresponding to each input sample are determined, and several target residual values ​​corresponding to each input sample are calculated based on the several actual power values ​​and the several predicted power values. A target replacement component is identified, and several first residual values ​​corresponding to the target replacement component before replacement and several second residual values ​​corresponding to the target replacement component after replacement are determined from the several target residual values. Based on the several first residual values, the several second residual values, the target residual values, and the several actual power values, the contribution of the target replacement component to the wind turbine performance degradation is calculated. In this way, the aging contribution of multiple components of the wind turbine can be accurately quantified.

[0122] See Figure 2 As shown in the figure, an embodiment of the present invention discloses a wind turbine performance analysis device, comprising:

[0123] The power prediction module 11 is used to predict each input sample in the target wind turbine dataset through a preset prediction model, so as to obtain several power prediction values ​​corresponding to each input sample.

[0124] The residual calculation module 12 is used to determine several actual power values ​​corresponding to each input sample based on the target wind turbine dataset, and to calculate several target residual values ​​corresponding to each input sample based on the several actual power values ​​and the several predicted power values.

[0125] Data classification module 13 is used to determine the target replacement part, and to determine a number of first residual values ​​corresponding to the target replacement part before replacement and a number of second residual values ​​corresponding to the target replacement part after replacement from the number of target residual values.

[0126] The contribution determination module 14 is used to calculate the contribution of the wind turbine performance degradation corresponding to the target replacement component based on the plurality of first residual values, the plurality of second residual values, the target residual value and the plurality of actual power values.

[0127] In this embodiment, a preset prediction model can be used to predict each input sample in the target wind turbine dataset to obtain several predicted power values ​​corresponding to each input sample. Based on the target wind turbine dataset, several actual power values ​​corresponding to each input sample are determined, and several target residual values ​​corresponding to each input sample are calculated based on the several actual power values ​​and the several predicted power values. A target replacement component is identified, and several first residual values ​​corresponding to the target replacement component before replacement and several second residual values ​​corresponding to the target replacement component after replacement are determined from the several target residual values. Based on the several first residual values, the several second residual values, the target residual values, and the several actual power values, the contribution of the target replacement component to the wind turbine performance degradation is calculated. In this way, the aging contribution of multiple components of the wind turbine can be accurately quantified.

[0128] In some embodiments, the wind turbine performance analysis device may further include:

[0129] The dataset acquisition unit is used to acquire the annual operation dataset of the target wind turbine through a preset data acquisition system.

[0130] The data filtering unit is used to filter out target power data that is less than a preset rated power threshold from the annual operating dataset;

[0131] The data processing unit is used to determine the current air density based on standard air density, standard temperature, and actual ambient temperature, and to normalize the wind speed data in the annual operational dataset using the current air density and the standard air density to obtain the target wind speed data.

[0132] The dataset generation unit is used to generate a target wind turbine dataset based on the target power data, the target wind speed data, and the annual operation dataset.

[0133] In some embodiments, the wind turbine performance analysis device may further include:

[0134] The kernel function value calculation unit is used to construct several input vectors based on the target wind turbine dataset corresponding to the wind turbine, and to calculate the kernel function value between the several input vectors through the target kernel function.

[0135] In some embodiments, the power prediction module 11 may specifically include:

[0136] The power prediction unit is used to predict each input sample in the target wind turbine dataset based on the kernel function value using a prediction model, so as to obtain the power prediction value corresponding to each input sample.

[0137] In some embodiments, the residual calculation module 12 may specifically include:

[0138] The first power data determination unit is used to filter out several actual power values ​​corresponding to each input sample from the target wind turbine dataset.

[0139] The second power data determination unit is used to determine the target actual power value and the target power prediction value corresponding to each input sample among the plurality of actual power values ​​and the plurality of predicted power values;

[0140] The residual calculation unit is used to calculate the difference between the actual value of the target power and the predicted value of the target power, so as to obtain several target residual values ​​corresponding to each input sample.

[0141] In some embodiments, the data classification module 13 may specifically include:

[0142] A data splitting unit is used to determine the target replacement component and the component replacement time of the target replacement component, so as to split each input sample based on the component replacement time to determine a first input sample before the component replacement time and a second input sample after the component replacement time;

[0143] The residual determination unit is used to determine, from the plurality of target residual values, a plurality of first residual values ​​corresponding to the first input sample and a plurality of second residual values ​​corresponding to the second input sample.

[0144] In some embodiments, the contribution determination module 14 may specifically include:

[0145] The power data determination submodule is used to determine a plurality of first power actual values ​​corresponding to the plurality of first residual values ​​from the plurality of power actual values, and to determine a plurality of second power actual values ​​corresponding to the plurality of second residual values ​​from the plurality of power actual values;

[0146] The first deviation index calculation submodule is used to calculate the ratio between each residual value in the plurality of first residual values ​​and the corresponding actual power value in the plurality of first actual power values ​​to obtain a plurality of first ratios, and to calculate the average value of the plurality of first ratios, so as to use the obtained average value as the first deviation index.

[0147] The second deviation index calculation submodule is used to calculate the ratio between each residual value in the plurality of second residual values ​​and the corresponding actual power value in the plurality of second actual power values ​​to obtain a plurality of second ratios, and to calculate the average value of the plurality of second ratios to use the obtained average value as the second deviation index.

[0148] The third deviation index calculation submodule is used to calculate the ratio between each residual value in the plurality of target residual values ​​and the corresponding actual power value in the plurality of actual power values ​​to obtain a plurality of third ratios, and to calculate the average value of the plurality of third ratios, so as to use the obtained average value as the third deviation index.

[0149] The component contribution calculation submodule is used to calculate the contribution of the wind turbine performance degradation corresponding to the target replacement component based on the first deviation index, the second deviation index, and the third deviation index.

[0150] In some embodiments, the component contribution calculation submodule may specifically include:

[0151] The difference calculation unit is used to calculate the difference between the first deviation index and the second deviation index to obtain the target difference.

[0152] The contribution calculation unit is used to calculate the ratio of the absolute value of the target difference to the third ratio, so as to use the obtained ratio as the contribution of the wind turbine performance decline corresponding to the target replacement component.

[0153] Furthermore, embodiments of this application also disclose an electronic device, Figure 3 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content of the diagram should not be construed as limiting the scope of this application.

[0154] Figure 3This is a schematic diagram of the structure of an electronic device 20 provided in an embodiment of this application. Specifically, the electronic device 20 may include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 stores a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the wind turbine performance analysis method disclosed in any of the foregoing embodiments. Alternatively, the electronic device 20 in this embodiment may specifically be a computer.

[0155] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows can be any communication protocol applicable to the technical solution of this application, and is not specifically limited here; the input / output interface 25 is used to acquire external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.

[0156] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or optical disk, etc. The resources stored thereon can include operating system 221, computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0157] The operating system 221 is used to manage and control the various hardware devices on the electronic device 20 and the computer program 222, which may be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of performing the wind turbine performance analysis method executed by the electronic device 20 as disclosed in any of the foregoing embodiments, the computer program 222 may further include computer programs capable of performing other specific tasks.

[0158] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned wind turbine performance analysis method. Specific steps of this method can be found in the corresponding content disclosed in the foregoing embodiments, and will not be repeated here.

[0159] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0160] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0161] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0162] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0163] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for analyzing the performance of wind turbine generators, characterized in that, include: The preset prediction model is used to predict each input sample in the target wind turbine dataset to obtain several power prediction values ​​corresponding to each input sample. Based on the target wind turbine dataset, determine several actual power values ​​corresponding to each input sample, and calculate several target residual values ​​corresponding to each input sample based on the several actual power values ​​and the several predicted power values. Identify the target replacement component, and determine from the plurality of target residual values ​​a plurality of first residual values ​​corresponding to the target replacement component before replacement and a plurality of second residual values ​​corresponding to the target replacement component after replacement; The contribution of the wind turbine performance degradation corresponding to the target replacement component is calculated based on the first residual value, the second residual value, the target residual value, and the actual power value.

2. The wind turbine performance analysis method according to claim 1, characterized in that, Before the step of predicting each input sample in the target wind turbine dataset using a preset prediction model to obtain several predicted power values ​​corresponding to each input sample, the method further includes: The annual operation dataset of the target wind turbine is obtained through a pre-set data acquisition system; Target power data that is less than a preset rated power threshold are selected from the annual operating dataset; The current air density is determined based on standard air density, standard temperature, and actual ambient temperature. The wind speed data in the annual operational dataset is then normalized using the current air density and the standard air density to obtain the target wind speed data. A target wind turbine dataset is generated based on the target power data, the target wind speed data, and the annual operation dataset.

3. The wind turbine performance analysis method according to claim 1, characterized in that, Before the step of predicting each input sample in the target wind turbine dataset using a preset prediction model to obtain several predicted power values ​​corresponding to each input sample, the method further includes: Several input vectors are constructed based on the target wind turbine dataset corresponding to the wind turbine, and the kernel function values ​​between the several input vectors are calculated through the target kernel function; Accordingly, the step of predicting each input sample in the target wind turbine dataset using a preset prediction model to obtain several predicted power values ​​corresponding to each input sample includes: The prediction model uses the kernel function value to predict the power prediction value corresponding to each input sample in the target wind turbine dataset.

4. The wind turbine performance analysis method according to claim 1, characterized in that, The step of determining several actual power values ​​corresponding to each input sample based on the target wind turbine dataset, and calculating several target residual values ​​corresponding to each input sample based on the several actual power values ​​and the several predicted power values, includes: Select several actual power values ​​corresponding to each input sample from the target wind turbine dataset; Determine the target actual power value and the target predicted power value corresponding to each input sample among the plurality of actual power values ​​and the plurality of predicted power values; The difference between the actual value of the target power and the predicted value of the target power is calculated to obtain several target residual values ​​corresponding to each input sample.

5. The wind turbine performance analysis method according to claim 1, characterized in that, The step of determining the target replacement component and determining, from the plurality of target residual values, a plurality of first residual values ​​corresponding to the target replacement component before replacement and a plurality of second residual values ​​corresponding to the target replacement component after replacement, includes: Identify the target replacement component and determine the component replacement time of the target replacement component, and split each input sample based on the component replacement time to determine a first input sample before the component replacement time and a second input sample after the component replacement time; From the plurality of target residual values, determine a plurality of first residual values ​​corresponding to the first input sample and a plurality of second residual values ​​corresponding to the second input sample.

6. The wind turbine performance analysis method according to any one of claims 1 to 5, characterized in that, The calculation of the contribution of the wind turbine performance degradation corresponding to the target replacement component based on the plurality of first residual values, the plurality of second residual values, the target residual value, and the plurality of actual power values ​​includes: From the plurality of actual power values, determine a plurality of first actual power values ​​corresponding to the plurality of first residual values, and from the plurality of actual power values, determine a plurality of second actual power values ​​corresponding to the plurality of second residual values; Calculate the ratio between each residual value in the plurality of first residual values ​​and the corresponding actual power value in the plurality of first actual power values ​​to obtain a plurality of first ratios, and calculate the average value of the plurality of first ratios to use the obtained average value as the first deviation index; Calculate the ratio between each residual value in the plurality of second residual values ​​and the corresponding actual power value in the plurality of second actual power values ​​to obtain a plurality of second ratios, and calculate the average value of the plurality of second ratios to use the obtained average value as the second deviation index; Calculate the ratio between each residual value in the plurality of target residual values ​​and the corresponding actual power value in the plurality of actual power values ​​to obtain a plurality of third ratios, and calculate the average value of the plurality of third ratios to use the obtained average value as the third deviation index; The contribution of the wind turbine performance degradation corresponding to the target replacement component is calculated based on the first deviation index, the second deviation index, and the third deviation index.

7. The wind turbine performance analysis method according to claim 6, characterized in that, The calculation of the contribution of the target replacement component to the performance degradation of the wind turbine is based on the first deviation index, the second deviation index, and the third deviation index, including: Calculate the difference between the first deviation index and the second deviation index to obtain the target difference; The absolute value of the target difference is calculated as the ratio of the third ratio, and the obtained ratio is used as the contribution of the wind turbine performance degradation corresponding to the target replacement component.

8. A wind turbine performance analysis device, characterized in that, include: The power prediction module is used to predict each input sample in the target wind turbine dataset using a preset prediction model, so as to obtain several power prediction values ​​corresponding to each input sample. The residual calculation module is used to determine several actual power values ​​corresponding to each input sample based on the target wind turbine dataset, and to calculate several target residual values ​​corresponding to each input sample based on the several actual power values ​​and the several predicted power values. The data classification module is used to determine the target replacement part, and to determine, from the plurality of target residual values, a plurality of first residual values ​​corresponding to the target replacement part before replacement and a plurality of second residual values ​​corresponding to the target replacement part after replacement; The contribution determination module is used to calculate the contribution of the wind turbine performance degradation corresponding to the target replacement component based on the plurality of first residual values, the plurality of second residual values, the target residual value, and the plurality of actual power values.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor for executing the computer program to implement the wind turbine performance analysis method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Used to store a computer program, wherein the computer program, when executed by a processor, implements the wind turbine performance analysis method as described in any one of claims 1 to 7.

Images