Wind turbine main bearing fault early warning method, device, equipment and storage medium

Abstract

The present disclosure relates to a main bearing fault warning method, device, equipment and storage medium for a wind turbine. The main bearing temperature influencing parameters of the target wind turbine are obtained; the main bearing temperature of the target wind turbine is predicted based on the main bearing temperature influencing parameters, and the main bearing temperature prediction residual of the target wind turbine is calculated; the main bearing temperature prediction residual is estimated by kernel density, and the probability density curve of the main bearing temperature prediction residual is determined; and the main bearing of the target wind turbine is warned of faults according to the probability density curve and a preset probability confidence interval. This fault warning method is convenient and reliable, and can perform timely and effective status monitoring and fault warning on the main bearing, ensuring the safe, reliable and efficient operation of the unit.

Description

Technical Field

[0001] The present disclosure relates to the technical field of wind turbine monitoring, and in particular to a method, device, equipment and storage medium for early warning of main bearing failure of a wind turbine. Background Art

[0002] The main bearing is one of the essential key components of the transmission chain system of a wind turbine. Its performance directly affects the performance and reliability of the transmission chain and even the entire machine.

[0003] Since wind turbines are mostly installed in remote areas with harsh operating environments such as mountains, deserts and seas, they are often operated in variable speed and load conditions, which can easily lead to failures of key components such as the main bearings, resulting in downtime. Therefore, proposing a reliable and efficient method for early warning of main bearing failures of wind turbines is a technical problem that needs to be solved urgently. Summary of the invention

[0004] In order to solve the above technical problem or at least partially solve the above technical problem, the present disclosure provides a main bearing fault early warning method, device, equipment and storage medium for a wind turbine generator set.

[0005] In a first aspect, the present disclosure provides a main bearing fault early warning method for a wind turbine generator set, the method comprising:

[0006] Obtain the main bearing temperature influencing parameters of the target wind turbine;

[0007] Predict the main bearing temperature of the target wind turbine based on the main bearing temperature influencing parameters, and calculate the prediction residual of the main bearing temperature of the target wind turbine;

[0008] Perform kernel density estimation on the main bearing temperature prediction residuals to determine the probability density curve of the main bearing temperature prediction residuals;

[0009] According to the probability density curve and the preset probability confidence interval, fault warning is issued for the main bearing of the target wind turbine.

[0010] In a second aspect, the present disclosure provides a main bearing fault early warning device for a wind turbine generator set, the device comprising:

[0011] A main bearing temperature influencing parameter acquisition module is used to acquire the main bearing temperature influencing parameters of the target wind turbine;

[0012] A main bearing temperature prediction residual calculation module is used to predict the main bearing temperature of a target wind turbine generator set based on the main bearing temperature influencing parameters, and calculate the main bearing temperature prediction residual of the target wind turbine generator set;

[0013] A probability density curve determination module is used to perform kernel density estimation on the main bearing temperature prediction residual and determine the probability density curve of the main bearing temperature prediction residual;

[0014] The main bearing fault warning module is used to provide fault warning for the main bearing of the target wind turbine according to the probability density curve and the preset probability confidence interval.

[0015] In a third aspect, the embodiments of the present disclosure further provide a main bearing fault warning device for a wind turbine generator set, the device comprising:

[0016] one or more processors;

[0017] a storage device for storing one or more programs,

[0018] When one or more programs are executed by one or more processors, the one or more processors implement the method provided in the first aspect.

[0019] In a fourth aspect, an embodiment of the present disclosure further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method provided in the first aspect is implemented.

[0020] Compared with the prior art, the technical solution provided by the embodiments of the present disclosure has the following advantages:

[0021] The disclosed embodiment of a wind turbine main bearing fault warning method, device and storage medium obtains the main bearing temperature influencing parameters of the target wind turbine; predicts the main bearing temperature of the target wind turbine based on the main bearing temperature influencing parameters, and calculates the main bearing temperature prediction residual of the target wind turbine; performs kernel density estimation on the main bearing temperature prediction residual, and determines the probability density curve of the main bearing temperature prediction residual; and performs fault warning on the main bearing of the target wind turbine according to the probability density curve and the preset probability confidence interval. Through the above method, it is possible to perform a warning analysis on the main bearing fault of the wind turbine based on the actual main bearing temperature influencing parameters in the wind turbine, and in turn based on the main bearing temperature prediction method, the kernel density estimation method and the preset probability confidence interval. This fault warning method is convenient and reliable, and can perform timely and effective status monitoring and fault warning on the main bearing, ultimately ensuring the safe, reliable and efficient operation of the unit. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0024] Figure 1 A schematic diagram of a flow chart of a main bearing fault early warning method for a wind turbine set provided in an embodiment of the present disclosure;

[0025] Figure 2 A result diagram of a main bearing temperature prediction provided by an embodiment of the present disclosure;

[0026] Figure 3 A diagram showing an abnormal operating state of a main bearing of a wind turbine provided by an embodiment of the present disclosure;

[0027] Figure 4 A graph showing a change point detection result of an abnormal operation state of a main bearing of a wind turbine set provided by an embodiment of the present disclosure;

[0028] Figure 5 A schematic diagram of a flow chart of a main bearing fault early warning method for a wind turbine set provided in an embodiment of the present disclosure;

[0029] Figure 6 A logical schematic diagram of a main bearing fault early warning method for a wind turbine generator set provided by an embodiment of the present disclosure;

[0030] Figure 7 A schematic structural diagram of a main bearing fault warning device for a wind turbine set provided by an embodiment of the present disclosure;

[0031] Figure 8 A schematic structural diagram of a main bearing fault warning device for a wind turbine set provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0032] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0033] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0034] In the prior art, a physical model-based method or a data-driven method is generally used to perform main bearing fault early warning.

[0035] For the physical model-based method, as wind turbines tend to be larger and more complex, it is difficult to perform physical modeling, making the physical model method difficult to implement.

[0036] In addition, data-driven methods generally use vibration signals, oil, sound signals, and data acquisition and monitoring and control (SCADA) system data to warn of main bearing faults. Among them, the fault warning method based on vibration signals, oil, and sound signals requires additional hardware equipment such as acquisition, storage, and transmission, which is difficult to promote on a large scale; among them, the fault warning method based on SCADA data relies on thresholds to detect whether the main bearing state is abnormal. This method has time uncertainty, and too large or too small thresholds will lead to missed reports and false reports. Therefore, the reliability of this main bearing fault warning method is poor.

[0037] In order to solve the above problems, the embodiments of the present disclosure provide a main bearing fault early warning method, device and storage medium for a wind turbine set, which can ensure safe, reliable and efficient operation of the set.

[0038] Next, first combine Figures 1 to 6 A main bearing fault early warning method for a wind turbine set provided in an embodiment of the present disclosure is described.

[0039] Figure 1 A schematic flow chart of a main bearing fault warning method for a wind turbine set provided by an embodiment of the present disclosure is shown.

[0040] In the disclosed embodiment, Figure 1 The main bearing fault warning method of the wind turbine set shown can be executed by the main bearing fault warning device of the wind turbine set. The main bearing fault warning device of the wind turbine set can be an electronic device or a server. The electronic device can include but is not limited to fixed terminals such as smart phones, laptops, desktop computers, etc. The server can be a cloud server or a server cluster and other devices with storage and computing functions. The embodiments of the present disclosure are explained in detail with the electronic device as the execution subject.

[0041] like Figure 1 As shown, the main bearing fault early warning method of the wind turbine generator set may include the following steps.

[0042] S110, obtaining main bearing temperature influencing parameters of the target wind turbine generator set.

[0043] In actual application, during the operation of the target wind turbine set, the data acquisition system of the target wind turbine set performs real-time data acquisition to obtain the main bearing temperature influencing parameters, and sends the main bearing temperature influencing parameters to the electronic equipment. The electronic equipment can then issue a main bearing fault warning based on the main bearing temperature influencing parameters.

[0044] In the disclosed embodiment, the target wind turbine set may be any wind turbine set that needs to be pre-warned of a main bearing failure. The wind turbine set may be installed in remote areas with harsh operating environments such as mountains, deserts, and seas.

[0045] In the embodiment of the present disclosure, the main bearing temperature influencing parameter refers to data that has an influencing relationship with the main bearing temperature of the wind turbine generator set.

[0046] Optionally, the temperature influencing parameters may include main shaft speed, generator speed, external cabin temperature, hub internal temperature, cabin internal temperature, etc.

[0047] S120, predicting the main bearing temperature of the target wind turbine generator set based on the main bearing temperature influencing parameters, and calculating the main bearing temperature prediction residual of the target wind turbine generator set.

[0048] In practical applications, the electronic device can use the main bearing temperature influencing parameters as analysis factors of the main bearing temperature analysis method to obtain the main bearing predicted temperature of the target wind turbine, and then calculate the main bearing temperature residual based on the main bearing predicted temperature and the main bearing actual temperature.

[0049] In the disclosed embodiment, the main bearing temperature residual refers to the temperature difference between the predicted main bearing temperature and the actual main bearing temperature.

[0050] Specifically, a pre-trained main bearing temperature prediction model can be used to predict the main bearing temperature for the main bearing temperature influencing parameters, and the main bearing predicted temperature can be output; then, the temperature difference between the main bearing predicted temperature and the main bearing actual temperature is calculated as the main bearing temperature residual.

[0051] in, Figure 2 is a result diagram of the predicted temperature of the main bearing. Figure 2 The horizontal axis is time, and the vertical axis is the main bearing temperature, that is, the predicted temperature of the main bearing.

[0052] In order to improve the accuracy of main bearing temperature prediction, before performing main bearing temperature prediction, abnormal data of main bearing temperature influencing parameters can be eliminated to obtain a processed healthy data set; further, the Spearman correlation coefficient correlation analysis method is used to perform correlation analysis on the healthy data set, and temperature influencing parameters whose influence on the main bearing temperature exceeds a preset degree threshold are selected, and the temperature influencing parameters exceeding the preset degree threshold are normalized, so that the normalized temperature influencing parameters are further input into the main bearing temperature prediction model.

[0053] S130, performing kernel density estimation on the main bearing temperature prediction residual to determine a probability density curve of the main bearing temperature prediction residual.

[0054] In practical applications, the electronic device can estimate the probability density distribution of the main bearing temperature prediction residual based on the kernel density estimation method, that is, obtain the probability density curve.

[0055] In the disclosed embodiment, the probability density curve of the main bearing temperature prediction residual may be understood as the probability density distribution curve of the main bearing temperature parameters.

[0056] S140. According to the probability density curve and the preset probability confidence interval, a fault warning is issued for the main bearing of the target wind turbine generator set.

[0057] In actual application, the electronic device can obtain a preset probability confidence interval, use the preset probability confidence interval to identify abnormal temperature residuals from the probability density curve of the main bearing temperature prediction residual, and generate main bearing fault warning information based on the abnormal temperature residual, thereby realizing fault warning for the main bearing of the target wind turbine.

[0058] In the embodiment of the present disclosure, the preset probability confidence interval refers to the temperature residual range used to identify abnormal temperature residuals. Optionally, the preset probability confidence interval may be a fixed interval set according to manual experience. The preset probability confidence interval may be 99.7%.

[0059] In the embodiment of the present disclosure, optionally, S140 may specifically include the following steps:

[0060] If there is an abnormal temperature residual of the main bearing that is not within the preset probability confidence interval on the probability density curve, fault warning information of the target wind turbine generator set is generated.

[0061] It can be understood that if all the main bearing temperature residuals on the probability density curve are within the preset probability confidence interval, it means that the main bearing temperature residual distribution is relatively concentrated and there are no abnormal temperature residuals, thereby determining that the main bearings of the target wind turbine are operating normally; if there are main bearing abnormal temperature residuals on the probability density curve that are not within the preset probability confidence interval, it means that there are abnormal temperature residuals on the probability density curve, thereby determining that the main bearings of the target wind turbine are operating abnormally and generating abnormal information.

[0062] In order to understand the determination method of the abnormal temperature residual of the main bearing, Figure 3 The figure shows the identification result of abnormal operation state of the main bearing of a wind turbine. Figure 3 As shown, during the period from 2020 / 7 / 20 to 2020 / 8 / 26, if there is an abnormal temperature residual of the main bearing that is not in the preset probability confidence interval on the temperature residual density curve, fault warning information of the target wind turbine is generated. Figure 3 The residual in and the main bearing temperature residual on the temperature residual density curve are the same concept.

[0063] The disclosed embodiment provides a main bearing fault warning method for a wind turbine set, which obtains the main bearing temperature influencing parameters of the target wind turbine set; predicts the main bearing temperature of the target wind turbine set based on the main bearing temperature influencing parameters, and calculates the main bearing temperature prediction residual of the target wind turbine set; performs kernel density estimation on the main bearing temperature prediction residual, and determines the probability density curve of the main bearing temperature prediction residual; and performs fault warning on the main bearing of the target wind turbine set according to the probability density curve and the preset probability confidence interval. Through the above method, it is possible to perform a warning analysis on the main bearing fault of the wind turbine set based on the actual main bearing temperature influencing parameters in the wind turbine set, and in turn based on the main bearing temperature prediction method, the kernel density estimation method, and the preset probability confidence interval. This fault warning method is convenient and reliable, and can perform timely and effective status monitoring and fault warning on the main bearing, ultimately ensuring the safe, reliable, and efficient operation of the unit.

[0064] In order to improve the accuracy and reliability of the main bearing fault warning, after determining the abnormal temperature residual of the main bearing, the method may further include the following steps:

[0065] The change point detection analysis of the abnormal temperature residual of the main bearing is carried out to determine the main bearing degradation time corresponding to the abnormal temperature residual of the main bearing.

[0066] Specifically, first, the electronic device can select the main bearing temperature residual about one month before and after the time point when the prediction residual is not in the preset probability confidence interval, and then use the binary segmentation change point detection method to perform change point detection analysis on the main bearing abnormal temperature residual, and accurately locate the deterioration time point of the main bearing state, that is, locate the main bearing degradation time, thereby realizing automatic identification of the main bearing state degradation point.

[0067] Among them, the main bearing deterioration time refers to the specific operating time when the main bearing fails.

[0068] In order to facilitate the understanding of the change point detection analysis results, Figure 4 The figure shows the result of the change point detection of the abnormal operation state of the main bearing of a wind turbine. Figure 4 As shown in the figure, there are four degradation time points on the change point detection result diagram, and these four degradation time points may cause main bearing failure. Figure 4 The residual in and the main bearing temperature residual on the temperature residual density curve are the same concept, the alarm threshold and the preset probability confidence interval are the same concept, and the change points 1-4 refer to four degradation time points.

[0069] Therefore, in the disclosed embodiment, by performing change point detection and analysis on the abnormal temperature residual of the main bearing, the degradation point of the main bearing state can be automatically identified, which facilitates the tracing of the cause of the main bearing degradation and provides a theoretical basis and data support for the post-cause analysis of the main bearing failure.

[0070] In another embodiment of the present disclosure, in order to ensure that the main bearing temperature prediction model can be used to determine the main bearing predicted temperature, the main bearing temperature prediction model may also be trained.

[0071] Figure 5 A schematic flow chart of another main bearing fault warning method for a wind turbine set provided by an embodiment of the present disclosure is shown.

[0072] like Figure 5 As shown, the main bearing fault early warning method of the wind turbine generator set includes the following steps.

[0073] S510, obtaining a sample main bearing temperature influencing parameter and a sample main bearing temperature.

[0074] In the embodiment of the present disclosure, optionally, the “obtaining sample main bearing temperature influencing parameters” in S510 may include the following steps:

[0075] S5101, obtaining the original data set of the wind turbine generator set;

[0076] S5102, removing abnormal data from the original data set of the wind turbine generator set to obtain a healthy data set of the wind turbine generator set;

[0077] S5103. Perform correlation analysis on the wind turbine health data set and sample main bearing temperatures to obtain candidate main bearing temperature influencing parameters;

[0078] S5104. Based on the control relationship between the components in the reference wind turbine generator set, select a sample main bearing temperature influencing parameter from the candidate main bearing temperature influencing parameters.

[0079] Among them, the sample main bearing temperature influencing parameters are parameters that obviously affect the main bearing temperature.

[0080] The sample main bearing temperature refers to the historical temperature of the main bearing of the reference wind turbine.

[0081] The original data set of the wind turbine generator set refers to the original operating data obtained from the reference wind turbine generator set. Optionally, the original data set of the wind turbine generator set may include hundreds of types of parameters or thousands of types of parameters.

[0082] The wind turbine health data set may be a distributed and concentrated data set. Optionally, the wind turbine health data set may include dozens of types of parameters.

[0083] Among them, the candidate main bearing temperature influencing parameters refer to the influencing parameters that have a close influence on the spindle temperature.

[0084] The reference wind turbine generator set and the target wind turbine generator set are in the same or similar environment and have the same or similar operating conditions.

[0085] The control relationship of each component in the reference wind turbine refers to the control principle of the wind turbine.

[0086] Specifically, for S5102, a density-based clustering algorithm (such as a DBSCAN clustering algorithm) may be used to cluster the original data set of the wind turbine generator set, and abnormal data may be removed from the original data set of the wind turbine generator set according to the clustering result.

[0087] Optionally, the abnormal data in the original data set of the wind turbine generator set may be outlier noise point data generated by fluctuations in operating conditions, abnormalities in acquisition or communication equipment, etc., or accumulation point data generated by wind curtailment, power rationing, reduced capacity operation, etc. For example, the normal shutdown, idling or startup state data of the wind turbine generator set with a wind speed less than 3m / s or greater than 25m / s in the original data set of the wind turbine generator set is removed, or the abnormal shutdown state data of the wind turbine generator set with a wind speed greater than 3m / s and a power less than or equal to 0 is removed.

[0088] Specifically, for S5103, the Spearman correlation coefficient correlation analysis method can be used to mine the key influencing parameters of the main bearing temperature and construct a modeling data set for the main bearing temperature prediction model, that is, to select candidate main bearing temperature influencing parameters from the wind turbine health data set.

[0089] Alternatively, the Spearman correlation coefficient calculation formula can be expressed as follows:

[0090]

[0091] Among them, r s is the Spearman correlation coefficient, n is the number of wind turbine health data sets, d i is the grade difference, i is the index of the data in the wind turbine health data set. It is generally believed that if 1≥|r S |≥0.8, it means that the data in the wind turbine health data set is strongly correlated with the main bearing temperature; if 0.8>|r s |≥0.5, it means that the data in the wind turbine health data set is moderately correlated with the main bearing temperature; if 0.5>|r s |≥0.3, it means that the data in the wind turbine health data set is correlated with the low temperature of the main bearing; if |r s |<0.3, it means that the data in the wind turbine health dataset is weakly correlated with the main bearing temperature.

[0092] Specifically, for S5104, based on the Spearman correlation coefficient correlation analysis results and combined with the control relationship of each component in the wind turbine control principle, the state parameters closely related to the main bearing temperature are extracted, that is, the sample main bearing temperature influencing parameters are selected from the candidate main bearing temperature influencing parameters.

[0093] Optionally, the sample main bearing temperature influencing parameters include 12 types of data, namely: selected cabin outer temperature, hub inner temperature, cabin inner temperature, generator non-drive end bearing temperature, main shaft speed, generator active power, gearbox oil pool temperature, generator drive end bearing temperature, generator stator coil temperature, wind speed, generator speed, and wind direction.

[0094] Therefore, in the disclosed embodiment, abnormal data elimination and correlation analysis are performed on the original data set of the wind turbine set in turn, and combined with the control principle of the wind turbine set, sample main bearing temperature influencing parameters that have a significant impact on the main bearing temperature are determined.

[0095] S520, training a preset network using sample temperature influencing parameters and sample main bearing temperature to obtain a main bearing temperature prediction model.

[0096] In the embodiment of the present disclosure, optionally, S520 may specifically include the following steps:

[0097] S5201, using the self-attention layer of the preset network, calculating the attention weight value of the temperature influencing parameter of the sample main bearing;

[0098] S5202, performing weighted summation on the attention weight value and the sample main bearing temperature influencing parameter to obtain a weighted main bearing temperature influencing parameter;

[0099] S5203, inputting the weighted main bearing temperature influencing parameters into the gated cyclic unit network of the preset network to obtain the predicted main bearing temperature;

[0100] S5204. Iteratively adjust the preset network using the predicted main bearing temperature and the sample main bearing temperature to obtain an adjusted main bearing temperature prediction model.

[0101] Among them, the self-attention layer introduces the self-attention mechanism, through which the weight values ​​corresponding to the temperature influencing parameters of the sample main bearing are calculated.

[0102] The weight value can represent the importance of each feature included in the sample temperature influencing parameter.

[0103] Among them, the gated recurrent unit network can extract and fuse the features of the weighted main bearing temperature influencing parameters and output the predicted main bearing temperature.

[0104] Furthermore, after obtaining the predicted main bearing temperature, the predicted main bearing temperature and the sample main bearing temperature can be used to calculate the loss function of the preset network. If the loss function is greater than the preset loss function threshold, the network structure of the self-attention layer and / or the gated recurrent unit network in the preset network is adjusted, and the loss function is recalculated until the recalculated loss function is less than or equal to the preset loss function threshold. It is determined that the main bearing temperature prediction model has been adjusted, and the adjusted main bearing temperature prediction model is obtained.

[0105] In order to reduce the difficulty of model training, the sample main bearing temperature influencing parameters can be normalized before model training to eliminate the dimensional influence of the sample main bearing temperature influencing parameters, and the normalized sample main bearing temperature influencing parameters can be further used for model training.

[0106] Optionally, the normalization method can be implemented as follows:

[0107]

[0108] Among them, x i is the temperature influencing parameter of any sample main bearing, x min is the minimum value of the sample main bearing temperature influence parameter, x max is the maximum value of the sample main bearing temperature influence parameters, x norm It is the normalized processing result of the temperature influencing parameters of the sample main bearing.

[0109] Therefore, in the embodiment of the present disclosure, the main bearing temperature prediction model can improve the temperature prediction accuracy by introducing the self-attention layer and the gated recurrent unit network, thereby ensuring the model accuracy.

[0110] S530, inputting the main bearing temperature influencing parameters into a pre-trained main bearing temperature prediction model to obtain the main bearing predicted temperature of the target wind turbine generator set.

[0111] Based on the training process of the main bearing temperature prediction model, it can be known that S530 can specifically include the following steps: using the self-attention layer of the main bearing temperature prediction model to calculate the attention weight value of the main bearing temperature influencing parameter; performing weighted summation of the attention weight value and the main bearing temperature influencing parameter to obtain the weighted main bearing temperature influencing parameter; inputting the weighted main bearing temperature influencing parameter into the gated recurrent unit network of the main bearing temperature prediction model to obtain the main bearing predicted temperature.

[0112] Therefore, in the main bearing temperature prediction process, the introduction of self-attention layer and gated recurrent unit network can accurately predict the main bearing temperature.

[0113] S540: Calculate the main bearing temperature residual of the target wind turbine generator set by using the predicted main bearing temperature and the actual main bearing temperature of the target wind turbine generator set.

[0114] S550, performing kernel density estimation on the main bearing temperature prediction residual to determine a probability density curve of the main bearing temperature prediction residual.

[0115] S560: According to the probability density curve and the preset probability confidence interval, a fault warning is issued for the main bearing of the target wind turbine generator set.

[0116] Among them, S540~S560 are similar to S120~S140, and will not be described in detail here.

[0117] In order to understand the construction process of the main bearing temperature prediction model and the main bearing fault warning process as a whole, Figure 6 A logical schematic diagram of a main bearing fault early warning method for a wind turbine set provided by an embodiment of the present disclosure is shown.

[0118] like Figure 6 As shown, the main bearing fault early warning method of the wind turbine generator set includes the following steps:

[0119] S610: Obtain offline historical data.

[0120] The offline historical data may include the original data set of the wind turbine and the sample main bearing temperature.

[0121] S620: Perform data cleaning, feature selection and normalization processing on offline historical data.

[0122] Specifically, the abnormal data of the original data set of the wind turbine can be removed first to obtain the original data set of the wind turbine and complete the data cleaning process; then, the correlation analysis of the original data set of the wind turbine and the sample main bearing temperature is performed to obtain the candidate main bearing temperature influencing parameters, and further based on the control relationship of each component in the reference wind turbine, the sample main bearing temperature influencing parameters are selected from the candidate main bearing temperature influencing parameters to complete the data cleaning and feature selection process. Furthermore, before the model training, the sample main bearing temperature influencing parameters can be normalized to eliminate the dimensional influence of the sample main bearing temperature influencing parameters.

[0123] S630: training a main bearing temperature prediction model.

[0124] Specifically, first, the self-attention layer of the preset network is used to calculate the attention weight value of the sample main bearing temperature influencing parameter; the attention weight value and the sample main bearing temperature influencing parameter are weightedly summed to obtain the weighted main bearing temperature influencing parameter; the weighted main bearing temperature influencing parameter is input into the gated recurrent unit network of the preset network to obtain the predicted main bearing temperature; the preset network is iteratively adjusted using the predicted main bearing temperature and the sample main bearing temperature to obtain the adjusted main bearing temperature prediction model.

[0125] S640: Obtain online monitoring data.

[0126] The online monitoring data may include the main bearing temperature influencing parameters of the target wind turbine generator set.

[0127] S650: normalize the online monitoring data.

[0128] S660: Use a main bearing temperature prediction model to predict the main bearing temperature to obtain a predicted main bearing temperature.

[0129] Specifically, the main bearing temperature influencing parameters are input into a pre-trained main bearing temperature prediction model to obtain the main bearing predicted temperature of the target wind turbine set.

[0130] S670, temperature residual calculation.

[0131] Specifically, the predicted temperature value of the main bearing and the actual temperature of the main bearing of the target wind turbine generator set are used to calculate the temperature residual of the main bearing of the target wind turbine generator set.

[0132] S680: Determine whether the threshold is exceeded.

[0133] Specifically, a kernel density estimation is performed on the main bearing temperature residual to determine the probability density curve of the main bearing temperature prediction residual, and a threshold value exceeding judgment is performed based on the probability density curve and a preset probability confidence interval.

[0134] S690: Issue warning information and perform change point detection and analysis.

[0135] Specifically, first, if there is an abnormal temperature residual of the main bearing that is not in the preset probability confidence interval on the probability density curve, fault warning information of the target wind turbine is generated; then, a change point detection analysis is performed on the abnormal temperature residual of the main bearing to determine the main bearing degradation time corresponding to the abnormal temperature residual of the main bearing.

[0136] The embodiment of the present disclosure also provides a main bearing failure warning device for a wind turbine set for implementing the main bearing failure warning method of the wind turbine set. Figure 7 In the embodiment of the present disclosure, the main bearing fault warning device of the wind turbine set can be an electronic device or a server. The electronic device can include a mobile terminal, a tablet computer and other devices with communication functions. The server can be a cloud server or a server cluster and other devices with storage and computing functions.

[0137] Figure 7 A schematic structural diagram of a main bearing fault warning device for a wind turbine set provided in an embodiment of the present disclosure is shown.

[0138] like Figure 7 As shown, the main bearing fault warning device 700 of the wind turbine generator set may include:

[0139] A main bearing temperature influencing parameter acquisition module 710 is used to acquire the main bearing temperature influencing parameters of the target wind turbine generator set;

[0140] A main bearing temperature prediction residual calculation module 720 is used to predict the main bearing temperature of a target wind turbine generator set based on the main bearing temperature influencing parameters, and calculate the main bearing temperature prediction residual of the target wind turbine generator set;

[0141] A probability density curve determination module 730 is used to perform kernel density estimation on the main bearing temperature prediction residual to determine a probability density curve of the main bearing temperature prediction residual;

[0142] The main bearing fault warning module 740 is used to provide fault warning for the main bearing of the target wind turbine generator set according to the probability density curve and the preset probability confidence interval.

[0143] The disclosed embodiment provides a main bearing fault warning device for a wind turbine set, which obtains the main bearing temperature influencing parameters of the target wind turbine set; predicts the main bearing temperature of the target wind turbine set based on the main bearing temperature influencing parameters, and calculates the main bearing temperature prediction residual of the target wind turbine set; performs kernel density estimation on the main bearing temperature prediction residual, and determines the probability density curve of the main bearing temperature prediction residual; and performs fault warning on the main bearing of the target wind turbine set according to the probability density curve and the preset probability confidence interval. Through the above method, it is possible to perform a warning analysis on the main bearing fault of the wind turbine set based on the actual main bearing temperature influencing parameters in the wind turbine set, and in turn based on the main bearing temperature prediction method, the kernel density estimation method and the preset probability confidence interval. This fault warning method is convenient and reliable, and can perform timely and effective status monitoring and fault warning on the main bearing, ultimately ensuring the safe, reliable and efficient operation of the unit.

[0144] In some embodiments, the main bearing temperature prediction residual calculation module 720 includes:

[0145] A main bearing predicted temperature determination unit is used to input the main bearing temperature influencing parameters into a pre-trained main bearing temperature prediction model to obtain the main bearing predicted temperature of the target wind turbine;

[0146] The main bearing temperature residual calculation unit is used to calculate the main bearing temperature residual of the target wind turbine set by using the main bearing predicted temperature and the actual main bearing temperature of the target wind turbine set.

[0147] In some embodiments, the apparatus further comprises:

[0148] A sample acquisition module, used to acquire a sample main bearing temperature influencing parameter and a sample main bearing temperature;

[0149] The model training module is used to train a preset network using sample main bearing temperature influencing parameters and sample main bearing temperatures to obtain a main bearing temperature prediction model.

[0150] In some embodiments, the sample acquisition module includes:

[0151] An acquisition unit, used for acquiring an original data set of a wind turbine generator set;

[0152] A removal unit is used to remove abnormal data from the original data set of the wind turbine generator set to obtain a healthy data set of the wind turbine generator set;

[0153] A correlation analysis unit is used to perform correlation analysis on the wind turbine health data set and the sample main bearing temperature to obtain candidate main bearing temperature influencing parameters;

[0154] The sample selection unit is used to select sample main bearing temperature influencing parameters from candidate main bearing temperature influencing parameters based on the control relationship of each component in the reference wind turbine.

[0155] In some embodiments, the model training module includes:

[0156] An attention weight value calculation unit, used to calculate the attention weight value of the sample main bearing temperature influencing parameter by using the self-attention layer of the preset network;

[0157] A weighting unit is used to perform a weighted summation of the attention weight value and the sample main bearing temperature influencing parameter to obtain a weighted main bearing temperature influencing parameter;

[0158] A main bearing temperature prediction unit, used for inputting the weighted main bearing temperature influencing parameters into a gated loop unit network of a preset network to obtain a predicted main bearing temperature;

[0159] The model training unit is used to iteratively adjust the preset network using the predicted main bearing temperature and the sample main bearing temperature to obtain an adjusted main bearing temperature prediction model.

[0160] In some embodiments, the probability density curve determination module 730 is specifically used to generate fault warning information of the target wind turbine if there is a main bearing abnormal temperature residual that is not within a preset probability confidence interval on the probability density curve.

[0161] In some embodiments, the apparatus further comprises:

[0162] The change point detection and analysis module is used to perform change point detection and analysis on the abnormal temperature residual of the main bearing and determine the main bearing degradation time corresponding to the abnormal temperature residual of the main bearing.

[0163] It should be noted that Figure 7 The main bearing fault warning device 700 of the wind turbine generator set shown can be implemented Figures 1 to 6 The various steps in the method embodiment shown in the figure are implemented Figures 1 to 6 The various processes and effects in the method or system embodiments shown are not described in detail here.

[0164] Figure 8 A schematic structural diagram of a main bearing fault warning device for a wind turbine set provided by an embodiment of the present disclosure is shown.

[0165] like Figure 8 As shown, the main bearing fault warning device of the wind turbine generator set may include a processor 801 and a memory 802 storing computer program instructions.

[0166] Specifically, the processor 801 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0167] The memory 802 may include a large capacity memory for information or instructions. By way of example and not limitation, the memory 802 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a tape, or a universal serial bus (USB) drive or a combination of two or more of these. Where appropriate, the memory 802 may include a removable or non-removable (or fixed) medium. Where appropriate, the memory 802 may be inside or outside the integrated gateway device. In a particular embodiment, the memory 802 is a non-volatile solid-state memory. In a particular embodiment, the memory 802 includes a read-only memory (ROM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (Electrically Erasable Programmable ROM, EEPROM), an electrically rewritable ROM (EAROM) or a flash memory, or a combination of two or more of these.

[0168] The processor 801 reads and executes the computer program instructions stored in the memory 802 to perform the steps of the main bearing fault early warning method for the wind turbine set provided in the embodiment of the present disclosure.

[0169] In one example, the main bearing fault warning device of the wind turbine generator set may also include a transceiver 803 and a bus 804. Figure 8 As shown, the processor 801, the memory 802 and the transceiver 803 are connected via a bus 804 and communicate with each other.

[0170] The bus 804 includes hardware, software, or both. For example, but not limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industrial Standard Architecture (ISA) bus, an InfiniBand interconnect, a Low Pin Count (LPC) bus, a Memory bus, a MicroChannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local Bus (VLB) bus, or other suitable buses or a combination of two or more of these. Where appropriate, the bus 804 may include one or more buses. Although embodiments of the present application describe and illustrate a particular bus, the present application contemplates any suitable bus or interconnect.

[0171] The following is an embodiment of a computer-readable storage medium provided in an embodiment of the present disclosure. The computer-readable storage medium and the main bearing fault warning method for a wind turbine set in the above-mentioned embodiments belong to the same inventive concept. For details not described in detail in the embodiment of the computer-readable storage medium, reference can be made to the embodiment of the main bearing fault warning method for a wind turbine set in the above-mentioned embodiments.

[0172] This embodiment provides a storage medium containing computer executable instructions. When the computer executable instructions are executed by a computer processor, they are used to execute a main bearing fault early warning method for a wind turbine generator set. The method includes:

[0173] Obtain the main bearing temperature influencing parameters of the target wind turbine;

[0174] Predict the main bearing temperature of the target wind turbine based on the main bearing temperature influencing parameters, and calculate the prediction residual of the main bearing temperature of the target wind turbine;

[0175] Perform kernel density estimation on the main bearing temperature prediction residuals to determine the probability density curve of the main bearing temperature prediction residuals;

[0176] According to the probability density curve and the preset probability confidence interval, fault warning is issued for the main bearing of the target wind turbine.

[0177] Of course, the storage medium containing computer executable instructions provided in the embodiment of the present disclosure is not limited to the above method operation, and the computer executable instructions can also execute the relevant operations in the main bearing fault early warning method of the wind turbine provided in any embodiment of the present disclosure. Through the above description of the implementation method, the technicians in the relevant field can clearly understand that the present disclosure can be implemented with the help of software and necessary general hardware, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present disclosure can be essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, a read-only memory (ROM), a random access memory (RAM), a flash memory (FLASH), a hard disk or an optical disk, etc., including a number of instructions to enable a computer cloud platform (which can be a personal computer, a server, or a network cloud platform, etc.) to execute the main bearing fault early warning method of the wind turbine provided in each embodiment of the present disclosure.

[0178] Note that the above are only preferred embodiments of the present disclosure and the technical principles used. Those skilled in the art will understand that the present disclosure is not limited to the specific embodiments herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present disclosure. Therefore, although the present disclosure is described in more detail through the above embodiments, the present disclosure is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present disclosure, and the scope of the present disclosure is determined by the scope of the appended claims.

Claims

1. A main bearing fault early warning method for a wind turbine generator set, characterized in that: include: Obtain the main bearing temperature influencing parameters of the target wind turbine; Predicting the main bearing temperature of the target wind turbine generator set based on the main bearing temperature influencing parameter, and calculating the main bearing temperature prediction residual of the target wind turbine generator set; Performing kernel density estimation on the main bearing temperature prediction residual to determine a probability density curve of the main bearing temperature prediction residual; According to the probability density curve and the preset probability confidence interval, a fault warning is provided for the main bearing of the target wind turbine; The step of predicting the main bearing temperature of the target wind turbine generator set based on the main bearing temperature influencing parameter and calculating the main bearing temperature prediction residual of the target wind turbine generator set includes: Inputting the main bearing temperature influencing parameter into a pre-trained main bearing temperature prediction model to obtain the main bearing predicted temperature of the target wind turbine set; Calculating a main bearing temperature residual of the target wind turbine generator set by using the predicted main bearing temperature and the actual main bearing temperature of the target wind turbine generator set; Also includes: Obtaining the sample main bearing temperature influencing parameters and the sample main bearing temperature; Using the sample main bearing temperature influencing parameters and the sample main bearing temperature to train a preset network to obtain the main bearing temperature prediction model; The method of using the sample main bearing temperature influencing parameters and the sample main bearing temperature to train a preset network to obtain the main bearing temperature prediction model includes: Utilizing the self-attention layer of the preset network, calculating the attention weight value of the temperature influencing parameter of the sample main bearing; Performing a weighted summation on the attention weight value and the sample main bearing temperature influencing parameter to obtain a weighted main bearing temperature influencing parameter; Inputting the weighted main bearing temperature influencing parameter into the gated recurrent unit network of the preset network to obtain the predicted main bearing temperature; The preset network is iteratively adjusted using the predicted main bearing temperature and the sample main bearing temperature to obtain an adjusted main bearing temperature prediction model.

2. The method according to claim 1, characterized in that: The obtaining of the sample main bearing temperature influencing parameters comprises: Obtain the original data set of wind turbines; Eliminating abnormal data from the original data set of the wind turbine generator set to obtain a healthy data set of the wind turbine generator set; Performing a correlation analysis on the wind turbine health data set and the sample main bearing temperature to obtain candidate main bearing temperature influencing parameters; The sample main bearing temperature influencing parameter is selected from the candidate main bearing temperature influencing parameters based on the control relationship of each component in the reference wind turbine.

3. The method according to claim 1, characterized in that The step of providing a fault warning for a main bearing of the target wind turbine generator set according to the probability density curve and a preset probability confidence interval includes: If there is a main bearing abnormal temperature residual that is not within the preset probability confidence interval on the probability density curve, fault warning information of the target wind turbine generator set is generated.

4. The method according to claim 3, characterized in that Also includes: A change point detection analysis is performed on the abnormal temperature residual of the main bearing to determine the main bearing degradation time corresponding to the abnormal temperature residual of the main bearing.

5. A main bearing fault warning device for a wind turbine generator set, characterized in that: include: A main bearing temperature influencing parameter acquisition module is used to acquire the main bearing temperature influencing parameters of the target wind turbine; A main bearing temperature prediction residual calculation module, used to predict the main bearing temperature of the target wind turbine generator set based on the main bearing temperature influencing parameters, and calculate the main bearing temperature prediction residual of the target wind turbine generator set; A probability density curve determination module is used to perform kernel density estimation on the main bearing temperature prediction residual to determine a probability density curve of the main bearing temperature prediction residual; A main bearing fault warning module, used for providing a fault warning for the main bearing of the target wind turbine according to the probability density curve and a preset probability confidence interval; The main bearing temperature prediction residual calculation module includes: A main bearing predicted temperature determination unit, used for inputting the main bearing temperature influencing parameter into a pre-trained main bearing temperature prediction model to obtain the main bearing predicted temperature of the target wind turbine set; A main bearing temperature residual calculation unit, used to calculate the main bearing temperature residual of the target wind turbine group by using the predicted main bearing temperature and the actual main bearing temperature of the target wind turbine group; The main bearing fault early warning device of the wind turbine generator set further comprises: A sample acquisition module, used to acquire a sample main bearing temperature influencing parameter and a sample main bearing temperature; A model training module, used to train a preset network using the sample main bearing temperature influencing parameters and the sample main bearing temperature to obtain the main bearing temperature prediction model; The model training module includes: An attention weight value calculation unit, used to calculate the attention weight value of the sample main bearing temperature influencing parameter by using the self-attention layer of the preset network; A weighting unit, used for performing a weighted summation of the attention weight value and the sample main bearing temperature influencing parameter to obtain a weighted main bearing temperature influencing parameter; A main bearing temperature prediction unit, used for inputting the weighted main bearing temperature influencing parameter into the gated loop unit network of the preset network to obtain a predicted main bearing temperature; The model training unit is used to iteratively adjust the preset network using the predicted main bearing temperature and the sample main bearing temperature to obtain an adjusted main bearing temperature prediction model.

6. A main bearing fault warning device for a wind turbine generator set, characterized in that: include: processor; A memory for storing executable instructions; The processor is used to read the executable instructions from the memory and execute the executable instructions to implement the method according to any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the processor implements the method according to any one of claims 1 to 4.

Images