Intelligent identification method and system for input torque of wind turbine generator and storage medium

Abstract

The invention relates to the technical field of wind turbine generator running state monitoring, in particular to a wind turbine generator input torque intelligent identification method and system and a storage medium, and the method comprises the steps: S1, constructing a training data set of state data of a wind turbine generator; s2, constructing an intelligent recognition model based on a radial basis function neural network; s3, inputting the training data set into the intelligent identification model to obtain a trained input torque intelligent identification model; s4, constructing a real-time data set of the state data of the wind turbine generator; and S5, inputting the real-time data set into the trained intelligent identification model to obtain an intelligent identification result of the input torque. According to the intelligent identification method for the input torque of the wind turbine generator, long-term identification and monitoring of the input torque of the wind turbine generator are achieved through an intelligent algorithm, strain gauges do not need to be installed and maintained, and the cost of long-term monitoring of the input torque of the wind turbine generator is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of wind turbine operating status monitoring, and in particular to a method, system and storage medium for intelligently identifying input torque of a wind turbine. Background Art

[0002] The acquisition of wind power input torque is crucial for monitoring the operating status and evaluating the health status of wind turbines. Long-term real-time and accurate measurement and monitoring of wind turbine input torque can support fault diagnosis, status assessment, life prediction and optimize maintenance plans of wind turbine transmission chains. In addition, torque data can also be used to optimize wind turbine control strategies and improve overall wind energy utilization efficiency. In the wind energy industry, torque monitoring is one of the necessary conditions for realizing intelligent operation and maintenance of wind turbines.

[0003] For the measurement, monitoring and collection of wind turbine input torque, referring to IEC61400-13 wind turbine main shaft torque test procedures and related standards, the existing technology generally involves attaching strain gauges on the input main shaft to collect strain, transmitting strain signals through wired or wireless connections, and calculating the corresponding input torque under different strains according to theoretical calculation formulas. However, this type of solution has many shortcomings:

[0004] 1) The wired connection uses a conductive ring to transmit the strain signal to the collection equipment through the conductive ring. With the large-scale wind turbines and the super-large main shaft diameter, the installation and matching conditions of the conductive ring are more stringent and are greatly affected by the actual working conditions. This method is not easy to achieve long-term and stable collection of input torque, and there are problems such as large measurement errors and low reliability;

[0005] 2) Although the wireless method can solve the difficulties faced by using conductive rings, it is difficult to achieve long-term stable and reliable power supply. It requires manual regular replacement of batteries, which may cause signal interference and loss. At the same time, regular replacement of batteries increases the cost of input torque monitoring. In particular, the offshore operation and maintenance time window is greatly affected by environmental factors, making regular replacement difficult.

[0006] 3) In addition, during the 20-25 years of design service life of the wind turbine, the accuracy and reliability of long-term input torque acquisition are difficult to guarantee due to the strain gauge installation method. Summary of the invention

[0007] The present invention provides a method for intelligently identifying the input torque of a wind turbine set, which relies on an intelligent algorithm to achieve long-term identification and monitoring of the input torque of the wind turbine set, while reducing the cost of long-term monitoring of the wind power input torque.

[0008] This application provides the following technical solutions:

[0009] A method for intelligently identifying input torque of a wind turbine generator system comprises the following steps:

[0010] S1. Constructing a training data set of status data of wind turbines;

[0011] S2, construct an intelligent recognition model based on radial basis function neural network;

[0012] S3, inputting the training data set into the intelligent recognition model to obtain a trained input torque intelligent recognition model;

[0013] S4, constructing a real-time data set of status data of the wind turbine generator set;

[0014] S5. Input the real-time data set into the trained intelligent recognition model to obtain the intelligent recognition result of the input torque.

[0015] Further, the S1 includes:

[0016] S11, collecting status data of the wind turbine generator set, wherein the status data includes the main shaft speed n_II i , radial vibration acceleration of the first gear ring a_II 1i , radial vibration acceleration of secondary gear ring a_II 2i 、Radial vibration acceleration of the third gear ring a_II 3i , High-speed output shaft radial vibration acceleration a_II 4i and principal axis strain ε;

[0017] S12. Calculate input torque T_I based on principal axis strain ε i ;

[0018] S13, radial vibration acceleration of the primary gear ring (a_II 1i ), radial vibration acceleration of the secondary gear ring (a_II 2i ), radial vibration acceleration of the third gear ring (a_II 3i ), high-speed output shaft radial vibration acceleration (a_II 4i ), calculate the effective value with an interval of 1s;

[0019]

[0020] n is the number of measurement points per second, which is equivalent to the sampling rate of the vibration sensor;

[0021] S14, spindle speed n_II i 、Input torque T_I i And the effective value of vibration acceleration rms_I ji Perform data cleaning;

[0022] S15, input the speed after data cleaning (N_I i ), effective value of vibration acceleration (RMS_Iji ) and input torque (T1_I i ) Establish a mapping relationship:

[0023] T1_I i =f(N_I i ,RMS_I ji )

[0024] Where: RMS_I ji =(RMS_I 1i ,RMS_I 2i ,RMS_I 3i ,RMS_I 4i ...).

[0025] Further, in step S12, for a solid spindle, the input torque T_I is calculated as follows: i :

[0026]

[0027] Where D is the principal axis diameter, E is the principal axis material elastic modulus, ε is the principal axis strain, μ is the Poisson's ratio, and N is the number of bridge arms;

[0028] For a hollow spindle, calculate the input torque T_I as follows: i :

[0029]

[0030] Where D o The outer diameter of the spindle, D i is the inner diameter of the spindle, E is the elastic modulus of the spindle material, ε is the spindle strain, μ is the Poisson's ratio, and N is the number of bridge arms.

[0031] Further, the data cleaning described in S14 includes:

[0032] S141. Clean the spindle speed data to remove abnormal points, noise points, and interference points in the speed signal. Use the median smoothing method as follows:

[0033]

[0034] Where m is the window length, Med{n_I i-v ,…,n_I i ,…,n_I i+v} is the median value of the speed sequence in the corresponding monitoring window;

[0035] S142, the effective value rms_I of the vibration acceleration calculated for each channel jiData cleaning is performed to remove abnormal points, noise points, and interference points in the effective value of vibration acceleration. The median smoothing method is used as follows:

[0036]

[0037] Where m is the window length, Med{rms_I ji-v ,…,rms_I ji ,…,rms_I ji+v} is the median of the corresponding vibration acceleration effective value sequence;

[0038] S143, input torque T_I after calculating the main shaft strain i Perform data cleaning to remove abnormal points, noise points, and interference points in the input torque signal, and use the median smoothing method as follows:

[0039]

[0040] Where m is the window length, Med{T_I i-v ,…,T_I i ,…,T_I i+v} is the median of the corresponding input torque sequence.

[0041] Furthermore, in step S2, the output of the radial basis function neural network is:

[0042]

[0043] where y j is the network output, w ij is the neuron weight, c i is the cluster center, σ is the variance of the basis function, and h is the number of nodes in the hidden layer.

[0044] Further, the step of constructing the real-time data set in S4 includes:

[0045] S41, online monitoring system speed sensor monitors the spindle speed n_II i , collect the radial vibration acceleration of the first gear ring (a_II 1i ), radial vibration acceleration of the secondary gear ring (a_II 2i ), radial vibration acceleration of the third gear ring (a_II 3i ), high-speed output shaft radial vibration acceleration (a_II 4i );

[0046] S42, the collected vibration acceleration time domain signal (a_II ji ), calculate the effective value with an interval of 1s;

[0047]

[0048] S43, spindle speed n_II i And the effective value rms_II after vibration acceleration calculation ji Perform data cleaning.

[0049] Further, the S43 includes:

[0050] S431, spindle speed n_II i Perform data cleaning to remove abnormal points, noise points, and interference points in the speed signal, and use the median smoothing method as follows:

[0051]

[0052] Where m is the window length, Med{n_I i-v ,…,n_I i ,…,n_I i+v} is the median value of the speed sequence in the corresponding monitoring window;

[0053] S432, effective value rms_II after calculating the vibration acceleration of each channel ji Data cleaning is performed to remove abnormal points, noise points, and interference points in the effective value of vibration acceleration. The median smoothing method is used as follows:

[0054]

[0055] Where m is the window length, Med{rms_II ji-v ,…,rms_II ji ,…,rms_II ji+v} is the median of the corresponding vibration acceleration effective value sequence.

[0056] The principles and advantages of the present invention are:

[0057] Since radial basis function (RBF) and radial basis function network have the characteristics of "local mapping", simple structure, relatively concise training process, high computational efficiency, and RBF network performs well in approximating arbitrary nonlinear functions and has fast learning speed, radial basis function neural network is used to model, learn, and train the recognition model to obtain the invisible nonlinear mapping relationship and realize the recognition of input torque.

[0058] 1. First, use the online monitoring system (CMS) to collect the input spindle speed, vibration acceleration at different gear rings, and short-term spindle strain. Then, smooth and filter the collected signals, remove abnormal values, and calculate the effective value of vibration acceleration.

[0059] 2. Secondly, an input torque identification model based on radial basis function is established. The input is the rotation speed and the effective value of vibration acceleration, and the output is the corresponding torque. The network is solved and trained.

[0060] 3. For the newly monitored vibration acceleration and spindle speed, after smoothing and noise reduction, the effective value of the vibration acceleration is calculated, and the calculated effective value of the vibration acceleration and input speed at different measuring points are substituted into the identification model to perform real-time identification of the input torque.

[0061] The trained input torque identification network relies on intelligent algorithms to achieve long-term identification and monitoring of the input torque of wind turbines. The input parameters only include the main shaft speed, vibration acceleration of different measuring points of the gearbox, etc. The input data of the main shaft stress is not used. Therefore, there is no need to install strain gauges to collect the main shaft strain, which solves the technical difficulties in the long-term input torque monitoring and collection of wind turbines. The collection of status data can be completed based on the online monitoring system, and there is no need for long-term maintenance of the strain gauges, which reduces the cost of long-term monitoring.

[0062] On the other hand, the present invention provides a wind turbine input torque intelligent identification system, including a data acquisition module, a data cleaning module and an intelligent identification module, wherein the data acquisition module is used to collect the status data of the wind turbine, and the status data includes the main shaft speed n_II i , radial vibration acceleration of the first gear ring a_II 1i , radial vibration acceleration of secondary gear ring a_II 2i 、Radial vibration acceleration of the third gear ring a_II 3i , High-speed output shaft radial vibration acceleration a_II 4i and principal axis strain ε;

[0063] The data processing module is used to process the collected wind turbine status data, including an input torque calibration submodule and a data cleaning submodule. The input torque calibration submodule is used to calculate the input torque T_I according to the main shaft strain ε. i , used for calibration of intelligent recognition model, and the acceleration processing submodule is used for spindle speed n_II i , radial vibration acceleration of the first gear ring a_II 1i , radial vibration acceleration of secondary gear ring a_II 2i 、Radial vibration acceleration of the third gear ring a_II 3i , High-speed output shaft radial vibration acceleration a_II 4i Perform data cleaning to obtain the input data set;

[0064] The intelligent recognition module stores the trained wind turbine input torque intelligent recognition model, and the input speed (N_I i), effective value of vibration acceleration (RMS_I ji ) and input torque (T1_I i ) establishes a mapping relationship and generates an input torque intelligent recognition result based on the input data set.

[0065] On the other hand, the present invention provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for intelligently identifying the input torque of a wind turbine generator set as described above is implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 This is a flow chart of a method for intelligently identifying input torque of a wind turbine generator set according to the present invention;

[0067] Figure 2 The original data of the spindle speed in Example 1;

[0068] Figure 3 The original data of the effective value of vibration acceleration in Example 1;

[0069] Figure 4 Inputting original torque data of the main shaft of the first embodiment;

[0070] Figure 5 The spindle speed data after data cleaning in Example 1;

[0071] Figure 6 The vibration acceleration effective value data after data cleaning in Example 1;

[0072] Figure 7 The spindle input torque data after data cleaning of the first embodiment;

[0073] Figure 8 The figure is a comparison between the input torque identification value and the measured value of the first embodiment. DETAILED DESCRIPTION

[0074] The present invention is based on the wind turbine status data collected by the online monitoring system, uses the radial basis function network to design an intelligent recognition model, realizes the long-term recognition and monitoring of the wind turbine input torque, effectively solves the technical difficulties in the long-term input torque monitoring and collection of the wind turbine, and reduces the cost of long-term monitoring of wind power input torque.

[0075] The following is further described in detail through specific implementation methods:

[0076] Embodiment 1

[0077] A method for intelligently identifying the input torque of a wind turbine generator system, such as Figure 1 As shown, the following steps are included:

[0078] S1. Constructing a training data set of wind turbine status data, including the following sub-steps:

[0079] S11, collect the training status data of the wind turbine, and collect the status data of the wind turbine through the online monitoring system, including the main shaft speed n_II i , radial vibration acceleration of the first gear ring a_II 1i , radial vibration acceleration of secondary gear ring a_II 2i 、Radial vibration acceleration of the third gear ring a_II 3i , High-speed output shaft radial vibration acceleration a_II 4i and the principal axis strain ε.

[0080] At the same time, the online monitoring system adheres to the IEC61400-13 wind turbine main shaft torque test procedures and related standards, and sticks strain gauges at a 45° inclination to the main shaft axis to measure the strain at the main shaft.

[0081] Figure 2 Displays the raw data of the spindle speed. Figure 3 The raw data of radial vibration acceleration of each level of gear ring are displayed.

[0082] S12. Calculate input torque T_I based on principal axis strain ε i , used for calibration of intelligent recognition models;

[0083] For a solid spindle, calculate as follows:

[0084]

[0085] Where D is the principal axis diameter, E is the principal axis material elastic modulus, ε is the principal axis strain, μ is the Poisson's ratio, and N is the number of bridge arms;

[0086] For hollow spindles, calculate as follows:

[0087]

[0088] Where D o The outer diameter of the spindle, D i is the inner diameter of the spindle, E is the elastic modulus of the spindle material, ε is the spindle strain, μ is the Poisson's ratio, and N is the number of bridge arms.

[0089] The original data of the spindle input torque is as follows: Figure 4 shown.

[0090] S13, radial vibration acceleration of the primary gear ring (a_II 1i ), radial vibration acceleration of the secondary gear ring (a_II 2i ), radial vibration acceleration of the third gear ring (a_II 3i), high-speed output shaft radial vibration acceleration (a_II 4i ), calculate the effective value with an interval of 1s;

[0091]

[0092] n is the number of measurement points per second, which is equivalent to the sampling rate of the vibration sensor;

[0093] S14, spindle speed n_II i 、Input torque T_I i And the effective value of vibration acceleration rms_I ji Perform data cleaning, including:

[0094] S141. Clean the spindle speed data to remove abnormal points, noise points, and interference points in the speed signal. Use the median smoothing method as follows:

[0095]

[0096] Where m is the window length, Med{n_I i-v ,…,n_I i ,…,n_I i+v} is the median value of the speed sequence in the corresponding monitoring window;

[0097] S142, the effective value rms_I of the vibration acceleration calculated for each channel ji Data cleaning is performed to remove abnormal points, noise points, and interference points in the effective value of vibration acceleration. The median smoothing method is used as follows:

[0098]

[0099] Where m is the window length, Med{rms_I ji-v ,…,rms_I ji ,…,rms_I ji+v} is the median of the corresponding vibration acceleration effective value sequence;

[0100] S143, input torque T_I after calculating the main shaft strain i Perform data cleaning to remove abnormal points, noise points, and interference points in the input torque signal, and use the median smoothing method as follows:

[0101]

[0102] Where m is the window length, Med{T_I i-v ,…,T_I i ,…,T_I i+v} is the median value of the corresponding input torque sequence

[0103] S15, input the speed after data cleaning (N_I i ), effective value of vibration acceleration (RMS_I ji ) and input torque (T1_I i ) Establish a mapping relationship:

[0104] T1_I i =f(N_I i ,RMS_I ji )

[0105] Where: RMS_I ji =(RMS_I 1i ,RMS_I 2i ,RMS_I 3i ,RMS_I 4i ...).

[0106] Figure 5 The spindle speed data after the data cleaning step is shown. Figure 6 The radial vibration acceleration data of each gear ring after the data cleaning step is shown. Figure 7 The input torque data after the data cleaning step is shown.

[0107] S2, construct an intelligent recognition model based on radial basis function neural network;

[0108] The Radial Basis Function Network (RBFN) is a special type of feedforward neural network, mainly used for regression and classification tasks. Unlike the traditional multilayer perceptron (MLP), RBFN uses radial basis function as the activation function of the hidden layer.

[0109] In this embodiment, the output of the radial basis function neural network is:

[0110]

[0111] where y j is the network output, w ij is the neuron weight, c i is the cluster center, σ is the variance of the basis function, and h is the number of nodes in the hidden layer.

[0112] S3, inputting the training data set into the intelligent recognition model to obtain a trained input torque intelligent recognition model;

[0113] Input the training data set into the intelligent recognition model for solving and training, including:

[0114] Use a clustering algorithm to determine the location of the center point;

[0115] The variance is determined based on the average distance between the center points;

[0116] Calculate the hidden layer output;

[0117] Use the least squares method or other optimization algorithm to determine the weight vector and bias term.

[0118] The recognition results generated by the trained input torque intelligent recognition model are evaluated with the actual measured values. The results are as follows: Figure 8 As shown in the figure, it can be seen that the recognition value is very close to the measured value, and the intelligent recognition method has a higher accuracy.

[0119] S4. Construct a real-time data set of status data of wind turbines, including:

[0120] S41, online monitoring system speed sensor monitors the spindle speed n_II i , collect the radial vibration acceleration of the first gear ring (a_II 1i ), radial vibration acceleration of the secondary gear ring (a_II 2i ), radial vibration acceleration of the third gear ring (a_II 3i ), high-speed output shaft radial vibration acceleration (a_II 4i );

[0121] S42, the collected vibration acceleration time domain signal (a_II ji ), calculate the effective value with an interval of 1s;

[0122]

[0123] S43, spindle speed n_II i And the effective value rms_II after vibration acceleration calculation ji Perform data cleaning, including:

[0124] S431, spindle speed n_II i Perform data cleaning to remove abnormal points, noise points, and interference points in the speed signal, and use the median smoothing method as follows:

[0125]

[0126] Where m is the window length, Med{n_I i-ν ,…,n_I i ,…,n_I i+v} is the median value of the speed sequence in the corresponding monitoring window;

[0127] S432, effective value rms_II after calculating the vibration acceleration of each channel jiData cleaning is performed to remove abnormal points, noise points, and interference points in the effective value of vibration acceleration. The median smoothing method is used as follows:

[0128]

[0129] Where m is the window length, Med{rms_II ji-v ,…,rms_II ji ,…,rms_II ji+v} is the median of the corresponding vibration acceleration effective value sequence

[0130] S5. Input the real-time data set into the trained intelligent recognition model to obtain the intelligent recognition result of the input torque.

[0131] Since there is no need to install strain gauges in actual use, nor is there any need for long-term maintenance of the strain gauges, the cost of long-term monitoring of wind power torque is reduced.

[0132] Embodiment 2

[0133] This embodiment is a wind turbine input torque intelligent identification system, including a data acquisition module, a data cleaning module and an intelligent identification module. The data acquisition module is used to collect the status data of the wind turbine, and the status data includes the main shaft speed n_II i , radial vibration acceleration of the first gear ring a_II 1i , radial vibration acceleration of secondary gear ring a_II 2i 、Radial vibration acceleration of the third gear ring a_II 3i , High-speed output shaft radial vibration acceleration a_II 4i and principal axis strain ε;

[0134] The data processing module is used to process the collected wind turbine status data, including an input torque calibration submodule and a data cleaning submodule. The input torque calibration submodule is used to calculate the input torque T_I according to the main shaft strain ε. i , used for calibration of intelligent recognition model, and the acceleration processing submodule is used for spindle speed n_II i , radial vibration acceleration of the first gear ring a_II 1i , radial vibration acceleration of secondary gear ring a_II 2i 、Radial vibration acceleration of the third gear ring a_II 3i , High-speed output shaft radial vibration acceleration a_II 4i Perform data cleaning to obtain the input data set;

[0135] The intelligent recognition module stores the trained wind turbine input torque intelligent recognition model, and the input speed (N_I i ), effective value of vibration acceleration (RMS_Iji ) and input torque (T1_I i ) establishes a mapping relationship and generates an input torque intelligent recognition result based on the input data set.

[0136] Embodiment 3

[0137] This embodiment is a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the above-mentioned method for intelligently identifying the input torque of a wind turbine is implemented. In some possible implementations, various aspects of the present invention may also be implemented in the form of a program product, which includes a program code, and when the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps of various exemplary implementations described in the above-mentioned embodiment section of this specification.

[0138] The above are only embodiments of the present invention. The invention is not limited to the field involved in this implementation case. The common knowledge such as the known specific structure and characteristics in the scheme is not described in detail here. It should be pointed out that for those skilled in the art, several deformations and improvements can be made without departing from the structure of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A method for intelligently identifying input torque of a wind turbine generator set, characterized in that: The steps include: S1. Constructing a training data set of status data of wind turbines; S2, construct an intelligent recognition model based on radial basis function neural network; S3, inputting the training data set into the intelligent recognition model to obtain a trained input torque intelligent recognition model; S4, constructing a real-time data set of status data of the wind turbine generator set; S5. Input the real-time data set into the trained intelligent recognition model to obtain the intelligent recognition result of the input torque.

2. The method for intelligently identifying input torque of a wind turbine generator set according to claim 1, characterized in that: The S1 includes: S11, collecting status data of the wind turbine generator set, wherein the status data includes the main shaft speed n_II i , radial vibration acceleration of the first gear ring a_II 1i , radial vibration acceleration of secondary gear ring a_II 2i 、Radial vibration acceleration of the third gear ring a_II 3i , High-speed output shaft radial vibration acceleration a_II 4i and principal axis strain ε; S12. Calculate input torque T_I based on principal axis strain ε i ,used for calibration of intelligent recognition models; S13, radial vibration acceleration of the primary gear ring (a_II 1i ), radial vibration acceleration of the secondary gear ring (a_II 2i ), radial vibration acceleration of the third gear ring (a_II 3i ), high-speed output shaft radial vibration acceleration (a_II 4i ), calculate the effective value with an interval of 1s; n is the number of measurement points per second, which is equal to the sampling rate of the vibration sensor; S14, spindle speed n_II i 、Input torque T_I i And the effective value of vibration acceleration rms_I ji Perform data cleaning; S15, input the speed after data cleaning (N_I i ), effective value of vibration acceleration (RMS_I jj ) and input torque (T1_I i ) Establish a mapping relationship: T1_I i =f(N_I i ,RMS_I ji ) Where: RMS_I ji =(RMS_I 1i , RMS_I 2i , RMS_I 3i , RMS_I 4i ...).

3. The method for intelligently identifying input torque of a wind turbine generator set according to claim 2, characterized in that: In step S12, for a solid spindle, the input torque T_I is calculated as follows: i : Where D is the principal axis diameter, E is the principal axis material elastic modulus, ε is the principal axis strain, μ is the Poisson's ratio, and N is the number of bridge arms; For a hollow spindle, calculate the input torque T_I as follows: i : Where D o The outer diameter of the spindle, D i is the inner diameter of the spindle, E is the elastic modulus of the spindle material, ε is the spindle strain, μ is the Poisson's ratio, and N is the number of bridge arms.

4. The method for intelligently identifying input torque of a wind turbine generator set according to claim 3, characterized in that: The data cleaning described in S14 includes: S141. Clean the spindle speed data to remove abnormal points, noise points, and interference points in the speed signal. Use the median smoothing method as follows: Where m is the window length, Med{n_I i-v ,…,n_I i ,…,n_I i+v } is the median value of the speed sequence in the corresponding monitoring window; S142, the effective value rms_I of the vibration acceleration calculated for each channel ji Data cleaning is performed to remove abnormal points, noise points, and interference points in the effective value of vibration acceleration. The median smoothing method is used as follows: Where m is the window length, Med{rms_I ji-v ,…,rms_I ji ,…,rms_I ji+v } is the median of the corresponding vibration acceleration effective value sequence; S143, input torque T_I after calculating the main shaft strain i Perform data cleaning to remove abnormal points, noise points, and interference points in the input torque signal, and use the median smoothing method as follows: Where m is the window length, Med{T_I i-v ,…,T_I i ,…,T_I i+v } is the median of the corresponding input torque sequence.

5. The method for intelligently identifying input torque of a wind turbine generator set according to claim 4, characterized in that: In step S2, the output of the radial basis function neural network is: where y j is the network output, w ij is the neuron weight, c i is the cluster center, σ is the variance of the basis function, and h is the number of nodes in the hidden layer.

6. A method for intelligently identifying input torque of a wind turbine generator set according to claim 5, characterized in that: The steps of constructing the real-time data set in S4 include: S41, online monitoring system speed sensor monitors the spindle speed n_II i , collect the radial vibration acceleration of the first gear ring (a_II 1i ), radial vibration acceleration of the secondary gear ring (a_II 2i ), radial vibration acceleration of the third gear ring (a_II 3i ), high-speed output shaft radial vibration acceleration (a_II 4i ); S42, the collected vibration acceleration time domain signal (a_II ji ), calculate the effective value with an interval of 1s; S43, spindle speed n_II i And the effective value of vibration acceleration rms_II ji Perform data cleaning.

7. A method for intelligently identifying input torque of a wind turbine generator set according to claim 6, characterized in that: The S43 includes: S431, spindle speed n_II i Perform data cleaning to remove abnormal points, noise points, and interference points in the speed signal, and use the median smoothing method as follows: Where m is the window length, Med{n_I i-v ,…,n_I i ,…,n_I i+v } is the median value of the speed sequence in the corresponding monitoring window; S432, effective value rms_II after calculating the vibration acceleration of each channel ji Data cleaning is performed to remove abnormal points, noise points, and interference points in the effective value of vibration acceleration. The median smoothing method is used as follows: Where m is the window length, Med{rms_II ji-v ,…,rms_II ji ,…,rms_II ji+v } is the median of the corresponding vibration acceleration effective value sequence.

8. An intelligent identification system for wind turbine input torque, characterized in that: It includes a data acquisition module, a data cleaning module and an intelligent identification module. The data acquisition module is used to collect the status data of the wind turbine set, and the status data includes the main shaft speed n_II i , radial vibration acceleration of the first gear ring a_II 1i , radial vibration acceleration of secondary gear ring a_II 2i 、Radial vibration acceleration of the third gear ring a_II 3i , High-speed output shaft radial vibration acceleration a_II 4i and principal axis strain ε; The data processing module is used to process the collected wind turbine status data, including an input torque calibration submodule and a data cleaning submodule. The input torque calibration submodule is used to calculate the input torque T_I according to the main shaft strain ε. i , used for calibration of intelligent recognition model, and the acceleration processing submodule is used for spindle speed n_II i , radial vibration acceleration of the first gear ring a_II 1i , radial vibration acceleration of secondary gear ring a_II 2i 、Radial vibration acceleration of the third gear ring a_II 3i , High-speed output shaft radial vibration acceleration a_II 4i Perform data cleaning to obtain the input data set; The intelligent recognition module stores the trained wind turbine input torque intelligent recognition model, and the input speed (N_I i ), effective value of vibration acceleration (RNS_I ji ) and input torque (T1_I i ) establishes a mapping relationship and generates an input torque intelligent recognition result based on the input data set.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for intelligently identifying the input torque of a wind turbine set according to any one of claims 1 to 7 is implemented.

Images