Generator main bearing lubricating method and device of wind generating set

Abstract

The invention relates to a generator main bearing lubricating method and device of a wind generating set. The generator main bearing lubricating method comprises the steps of determining a generator main bearing temperature characteristic variable based on operation data of a wind generating set; establishing a plurality of generator main bearing state prediction models based on the generator main bearing temperature characteristic variables, and predicting the wear condition of a front bearing of the generator by using the plurality of generator main bearing state prediction models; based on the prediction results of the plurality of generator main bearing state prediction models, determining the wear state of the generator front bearing; and determining a lubricating strategy based on the wear state of the front bearing of the generator, and lubricating the front bearing of the generator according to the lubricating strategy. According to the invention, the abnormity of the main bearing of the generator can be accurately and reliably identified in advance, the corresponding maintenance strategy is executed, and the maintenance cost is reduced.

Description

Technical Field

[0001] The present disclosure relates to the technical field of wind power generation, and more specifically, to a method and device for lubricating a main bearing of a wind power generator set. Background Art

[0002] The axial cross-sectional structure of the permanent magnet direct drive outer rotor wind turbine in the prior art is as follows: Figure 8 As shown, it mainly includes rotor poles 801, stator coils 802, stator brackets 803, rotor front brackets 804, stator cores 805, generator rotor connection flanges 806, rear bearings 807, bearing and rotor connection bolts 808, stator and bearing connection flanges 809, stator connection flanges 810, rotor front brackets, rotor front end covers 812, etc. The permanent magnet synchronous generator is one of the main components of a direct-drive wind turbine generator set. The impeller is connected to the generator through the generator rotor connection flange 806, and the generator is connected to the fixed axis of the nacelle through the stator connection flange 810. When the impeller rotates under the action of the wind, it drives the generator rotor to rotate, and the generator stator coils 802 cut the magnetic field generated by the rotor poles 801 on the rotor, and the generator generates electrical energy, which is then transmitted and converted through cables and converters and delivered to the power grid to realize the power generation function of the wind turbine generator set. Among them, the main bearing includes a front bearing (not shown in the figure) and a rear bearing 807.

[0003] The greasing methods of the main bearings of the generators of wind turbines include manual greasing and automatic greasing. Manual greasing refers to, for example, greasing according to the provisions of the maintenance and overhaul manual during inspections (for example, half-year or annual overhauls) or special greasing when abnormalities are found in the main bearings of the generators, usually using quantitative greasing guns and quantitative grease bags. In addition, the main bearings of the generators of wind turbines can be equipped with automatic lubrication systems. The automatic lubrication system can add lubricating grease to the main bearings of the generators to reduce the friction and heat of the main bearings of the generators during the operation of the generators, so as to make the power generation process smoother. However, the existing automatic lubrication systems usually adopt a lubrication strategy of lubricating grease with fixed quantity, that is, the interval time and amount of lubricating grease are fixed. Although this lubrication strategy realizes automatic lubrication, it does not take into account the actual operation of the wind turbines, and there is a problem that the amount of lubricating grease injected is not equal to the amount required, resulting in the main bearings of the generators being unable to operate smoothly due to insufficient lubrication, or even leading to failures or waste due to excessive amount of lubricating grease injected. Summary of the invention

[0004] In order to solve the above problems, the present disclosure proposes a method and device for lubricating a main bearing of a wind turbine generator set, a computing system, and a computer-readable storage medium.

[0005] According to one aspect of the present disclosure, a method for lubricating a generator main bearing of a wind turbine generator set is provided, the method comprising: determining a temperature characteristic variable of a generator main bearing based on operating data of the wind turbine generator set; establishing a plurality of generator main bearing state prediction models based on the generator main bearing temperature characteristic variable, and using the plurality of generator main bearing state prediction models to predict the wear of a generator front bearing; determining a generator front bearing wear state based on prediction results of the plurality of generator main bearing state prediction models; determining a lubrication strategy based on the generator front bearing wear state, and lubricating the generator front bearing according to the lubrication strategy.

[0006] Optionally, the operating data of the wind turbine generator set includes generator rotation speed, ambient temperature, nacelle temperature, and front bearing temperature and rear bearing temperature of a generator main bearing.

[0007] Optionally, the main bearing includes a front bearing sleeved on the front end of the generator main shaft and a rear bearing sleeved on the middle part of the generator main shaft.

[0008] Optionally, the step of predicting the wear condition of the generator front bearing includes: in response to the difference between the generator main bearing temperature characteristic variable of the rear bearing and the corresponding generator main bearing temperature characteristic variable of the front bearing exceeding a predetermined threshold, determining that the generator front bearing is in an abnormal wear state.

[0009] Optionally, the wind turbine generator set is an outer rotor inner stator type permanent magnet direct drive wind turbine generator set.

[0010] Optionally, the generator main bearing temperature characteristic variables include the temperature rise of the front bearing temperature relative to the cabin temperature, the temperature rise of the front bearing temperature relative to the ambient temperature, the temperature rise of the rear bearing temperature relative to the cabin temperature, the temperature rise of the rear bearing temperature relative to the ambient temperature, the difference between the temperature rise of the front bearing temperature relative to the cabin temperature and the temperature rise of the rear bearing temperature relative to the cabin temperature, the difference between the temperature rise of the front bearing temperature relative to the ambient temperature and the temperature rise of the rear bearing temperature relative to the ambient temperature, the average value of the front bearing temperature, the average value of the rear bearing temperature, the ratio of the average front bearing temperature to the average rear bearing temperature, the ratio of the front bearing temperature to the generator speed, and the difference between the average front bearing temperature and the average rear bearing temperature.

[0011] Optionally, the multiple generator main bearing state prediction models include two or more of the following models: a prediction model based on multivariable fusion analysis, a prediction model based on time series trend decomposition, a prediction model based on a Gaussian mixture model, a prediction model based on a support vector machine, and a prediction model based on deep learning.

[0012] Optionally, the step of predicting the wear condition of the generator front bearing includes: comparing the temperature characteristic variables of the generator main bearing with corresponding thresholds respectively; in response to at least one of the temperature characteristic variables of the generator main bearing exceeding the corresponding threshold, determining that the generator front bearing is in an abnormal wear state.

[0013] Optionally, the step of predicting the wear condition of the generator front bearing includes: determining a trend factor based on the temperature characteristic variable of the generator main bearing; and determining that the generator front bearing is in an abnormal wear state in response to the trend factor being greater than a specific threshold.

[0014] Optionally, the step of predicting the wear condition of the generator front bearing includes: in response to the front bearing temperature presenting a binormal distribution, determining that the generator front bearing is in an abnormal wear state.

[0015] Optionally, the step of determining the wear state of the generator front bearing based on the prediction results of the multiple generator main bearing state prediction models includes: in response to the number of prediction results indicating that the generator front bearing is in an abnormal wear state being greater than zero, determining the wear level of the generator front bearing based on the number of prediction results indicating that the generator front bearing is in an abnormal wear state; in response to the number of prediction results indicating that the generator front bearing is in an abnormal wear state being zero, determining that the generator front bearing is in a normal wear state.

[0016] Optionally, the step of predicting the wear condition of the generator front bearing includes: in response to at least one of the generator main bearing temperature characteristic variables exceeding a corresponding first threshold, determining that the wear level of the generator front bearing is a first level; in response to at least one of the generator main bearing temperature characteristic variables exceeding a corresponding second threshold, determining that the wear level of the generator front bearing is a second level, wherein the second threshold is higher than the first threshold, and the degree of wear of the generator front bearing indicated by the second level is higher than the degree of wear of the generator front bearing indicated by the first level.

[0017] Optionally, the step of determining the lubrication strategy based on the wear state of the generator main bearing includes: based on the wear level of the generator front bearing, determining the lubrication parameters corresponding to the wear level, the lubrication parameters including at least one of a single greasing time, a lubrication pump flow rate and a greasing time interval.

[0018] According to another aspect of the present disclosure, a generator main bearing lubrication device of a wind turbine generator set is provided, the generator main bearing lubrication device comprising: a characteristic variable determination unit, which determines the generator main bearing temperature characteristic variable based on the operation data of the wind turbine generator set; a prediction model unit, which establishes multiple generator main bearing state prediction models based on the generator main bearing temperature characteristic variable, and uses the multiple generator main bearing state prediction models to predict the wear of the generator front bearing; a wear state determination unit, which determines the generator front bearing wear state based on the prediction results of the multiple generator main bearing state prediction models; and a lubrication control unit, which determines a lubrication strategy based on the generator front bearing wear state, and lubricates the generator front bearing according to the lubrication strategy.

[0019] Optionally, the operating data of the wind turbine generator set includes generator rotation speed, ambient temperature, nacelle temperature, and front bearing temperature and rear bearing temperature of a generator main bearing.

[0020] Optionally, the main bearing includes a front bearing sleeved on the front end of the generator main shaft and a rear bearing sleeved on the middle part of the generator main shaft.

[0021] Optionally, the prediction model unit is configured to determine that the generator front bearing is in an abnormal wear state in response to a difference between a generator main bearing temperature characteristic variable of the rear bearing and a corresponding generator main bearing temperature characteristic variable of the front bearing exceeding a predetermined threshold.

[0022] Optionally, the wind turbine generator set is an outer rotor inner stator type permanent magnet direct drive wind turbine generator set.

[0023] Optionally, the generator main bearing temperature characteristic variables include the temperature rise of the front bearing temperature relative to the cabin temperature, the temperature rise of the front bearing temperature relative to the ambient temperature, the temperature rise of the rear bearing temperature relative to the cabin temperature, the temperature rise of the rear bearing temperature relative to the ambient temperature, the difference between the temperature rise of the front bearing temperature relative to the cabin temperature and the temperature rise of the rear bearing temperature relative to the cabin temperature, the difference between the temperature rise of the front bearing temperature relative to the ambient temperature and the temperature rise of the rear bearing temperature relative to the ambient temperature, the average value of the front bearing temperature, the average value of the rear bearing temperature, the ratio of the average front bearing temperature to the average rear bearing temperature, the ratio of the front bearing temperature to the generator speed, and the difference between the average front bearing temperature and the average rear bearing temperature.

[0024] Optionally, the multiple generator main bearing state prediction models include two or more of the following models: a prediction model based on multivariable fusion analysis, a prediction model based on time series trend decomposition, a prediction model based on a Gaussian mixture model, a prediction model based on a support vector machine, and a prediction model based on deep learning.

[0025] Optionally, the prediction model unit is configured to: compare the generator main bearing temperature characteristic variables with corresponding thresholds respectively; and determine that the generator front bearing is in an abnormal wear state in response to at least one of the generator main bearing temperature characteristic variables exceeding the corresponding threshold.

[0026] Optionally, the prediction model unit is configured to: determine a trend factor based on the temperature characteristic variable of the generator main bearing; and determine that the generator front bearing is in an abnormal wear state in response to the trend factor being greater than a specific threshold.

[0027] Optionally, the prediction model unit is configured to: in response to the front bearing temperature presenting a binormal distribution, determine that the generator front bearing is in an abnormal wear state.

[0028] Optionally, the wear status determination unit is configured to: in response to the number of prediction results indicating that the front bearing of the generator is in an abnormal wear state being greater than zero, determine the wear level of the front bearing of the generator based on the number of prediction results indicating that the front bearing of the generator is in an abnormal wear state; in response to the number of prediction results indicating that the front bearing of the generator is in an abnormal wear state being zero, determine that the front bearing of the generator is in a normal wear state.

[0029] For the prediction model based on multivariate fusion analysis, the prediction model unit is configured to: in response to at least one of the generator main bearing temperature characteristic variables exceeding the corresponding first threshold, determine that the wear level of the generator front bearing is a first level; in response to at least one of the generator main bearing temperature characteristic variables exceeding the corresponding second threshold, determine that the wear level of the generator front bearing is a second level, wherein the second threshold is higher than the first threshold, and the degree of wear of the generator front bearing indicated by the second level is higher than the degree of wear of the generator front bearing indicated by the first level.

[0030] Optionally, the lubrication control unit is configured to determine lubrication parameters corresponding to the wear level based on the wear level of the generator front bearing, and the lubrication parameters include at least one of a single greasing time, a lubrication pump flow rate and a greasing time interval.

[0031] According to another aspect of the present disclosure, a computing system is provided that includes at least one computing device and at least one storage device storing instructions, wherein the instructions, when executed by the at least one computing device, prompt the at least one computing device to execute the generator main bearing lubrication method of the wind turbine generator set as described above.

[0032] According to another aspect of the present disclosure, a computer-readable storage medium storing instructions is provided, wherein when the instructions are executed by at least one computing device, the at least one computing device is prompted to execute the above-mentioned method for lubricating a main bearing of a generator of a wind turbine generator set.

[0033] By adopting the present disclosure, it is possible to accurately, reliably and in advance identify abnormalities in the main bearing of the generator, execute corresponding maintenance strategies, reduce maintenance costs, and provide valuable data support for business decisions; without having to invest large costs, it is possible to achieve coverage of a wider range of units and identification of hidden risks, thereby reducing the cost and investment of manual inspections. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The above and / or other objects and advantages of the present disclosure will become more apparent through the following description of embodiments in conjunction with the accompanying drawings, in which:

[0035] Figure 1 is a flow chart showing a method for lubricating a main bearing of a wind turbine generator according to an exemplary embodiment of the present disclosure;

[0036] Figure 2 is a graph showing the decomposition of the time series trend;

[0037] Figure 3 is a graph showing a binormal distribution of the front bearing temperature;

[0038] Figure 4 is a flow chart illustrating a method of determining a lubrication strategy according to an exemplary embodiment of the present disclosure;

[0039] Figure 5 is a block diagram showing a generator main bearing lubrication device of a wind turbine generator set according to an exemplary embodiment of the present disclosure;

[0040] Figure 6 is a block diagram illustrating a computing system including at least one computing device and at least one storage device storing instructions according to an exemplary embodiment of the present disclosure;

[0041] Figure 7 is a schematic diagram showing the structure of a wind turbine generator according to an exemplary embodiment of the present disclosure;

[0042] Figure 8 The figure is a schematic diagram showing the structure of a wind turbine generator in the prior art. DETAILED DESCRIPTION

[0043] Below, in conjunction with the accompanying drawings, a description of a specific embodiment is provided to help the reader obtain a comprehensive understanding of the method, device and / or system described herein. However, after understanding the disclosure of the present application, various changes, modifications and equivalents of the method, device and / or system described herein will be clear. For example, the order of operations described herein is only an example and is not limited to those orders set forth herein, but can be changed as will be clear after understanding the disclosure of the present application, except for operations that must occur in a specific order. In addition, for greater clarity and simplicity, the description of features known in the art may be omitted.

[0044] Figure 7 Schematic diagram showing the structure of a wind turbine according to an exemplary embodiment of the present disclosure.

[0045] The outer rotor type permanent magnet direct drive wind turbine generator comprises a stator, a rotor, a fixed shaft 701, a rotating shaft 702 and a bearing assembly. The rotating shaft 702 is sleeved outside the fixed shaft 701 through the bearing assembly and can rotate relative to the fixed shaft 701. The bearing assembly comprises a front bearing 703 and a rear bearing 704, wherein the front bearing 703 is arranged between the front end of the rotating shaft and the front end of the fixed shaft 701; the rear bearing 704 is arranged between the rear end of the rotating shaft 702 and the fixed shaft 1. The rear end of the fixed shaft 701 is connected to an external base (not shown). The stator is mounted on the fixed shaft 701 through a stator bracket 705, one end of the stator bracket 705 is connected to the fixed shaft 701, and the other end supports the iron core winding 706. The rotor is supported by a rotor bracket 707 and mounted on the rotating shaft 702 through the rotor bracket, one end of the rotor bracket 707 is connected to the rotating shaft 702, and the other end is connected to the rotor yoke 719.

[0046] Factors that affect the operation of the generator main bearing include ambient temperature, generator speed, operating time, lubrication, heat dissipation, its own level, fault status, etc. Considering these factors and selecting one or more appropriate characteristic indicators can more accurately evaluate and detect the operation of the main bearing. In the units with damaged main bearings, the front bearings are more likely to be damaged, while the rear bearings are relatively stable and rarely damaged. Therefore, the rear bearings can be used as a comparative reference for changes in the operation of the front bearings.

[0047] According to the exemplary embodiment of the present invention, the lubrication method for the main bearing of a wind turbine generator set can be based on, for example, SCADA data, and can diagnose the wear of the main bearing of the generator through big data and artificial intelligence technologies such as multivariate fusion analysis of specific working conditions, statistical analysis, time series analysis, machine learning, deep learning, etc., and classify the severity of the wear. Different maintenance and greasing strategies can be implemented for main bearings with different levels of abnormal wear (including executing different lubrication times, different lubrication times, different greasing amounts, special maintenance, etc. and the optimal maintenance strategies corresponding to the severity levels) to achieve improvement and life extension for the aggravated wear of the main bearing of the generator.

[0048] SCADA data may include relevant measurement point data such as wind speed, power, ambient temperature, bearing temperature, etc., for example, as shown in Table 1 below.

[0049] Table 1 SCADA data measurement points

[0050]

[0051] In the example, the time period of one year + 10 days before the fault determination time can be collected for the faulty unit (for example, for a unit determined to be abnormal on December 23, 2022, the 7S data or 10min data of the unit from December 13, 2021 to December 23, 2022) can be collected to verify the amount of 7S data and 10min data stored in the big data platform for the abnormal unit during this period. Usually, the amount of unit data obtained is higher than 80% of the theoretical data amount. The rules for data acquisition include: when the 10-min data volume of the unit is higher than 80% of the theoretical data volume, the 10-min data of the unit is acquired; when the 10-min data volume of the unit is lower than 80% of the theoretical data volume and the 7S data volume of the unit is higher than 80% of the theoretical data volume, the 7S data of the unit is acquired; when the data volume of the 10-min data and 7S data of the unit are both lower than 80%, the data of a larger type is acquired, for example, the 7S data includes 365 days of data and the 10-min data includes 200 days of data, in which case the 7S data is acquired as the data of the unit; when the data volumes are roughly the same, the 10-min data is used preferentially.

[0052] In addition, after obtaining the unit data, the data can be preprocessed (including, for example, processing of missing values, processing of duplicate values, elimination of noise and data that does not conform to business logic, unit maintenance process data, etc.). In addition, the processing involving feature engineering mainly includes: availability assessment, used to determine the effectiveness of features; feature cleaning, processing the whole or part of the features that do not meet the requirements; feature screening and extraction: used for feature analysis and extraction based on mechanisms and algorithms to reduce the number of features; derived features and reconstruction, used to reconstruct the original features and increase the derived feature indicators.

[0053] Figure 1 Detailed description is a flow chart showing a method for lubricating a main bearing of a wind turbine generator according to an exemplary embodiment of the present disclosure. Figure 2 is a graph showing the decomposition of time series trends. Figure 3 is a graph showing a binormal distribution of the front bearing temperature.

[0054] like Figure 1As shown, in step S101, based on the operating data of the wind turbine generator set, the temperature characteristic variable of the main bearing of the generator is determined. In the example, the main bearing includes a front bearing sleeved at the front end of the generator main shaft and a rear bearing sleeved at the middle of the generator main shaft. In the example, the wind turbine generator set is an outer rotor inner stator type permanent magnet direct drive wind turbine generator set. In the example, the operating data of the wind turbine generator set includes the generator speed, the ambient temperature, the cabin temperature, and the front bearing temperature and the rear bearing temperature of the generator main bearing. In the example, the generator main bearing temperature characteristic variables include the temperature rise of the front bearing temperature relative to the cabin temperature, the temperature rise of the front bearing temperature relative to the ambient temperature, the temperature rise of the rear bearing temperature relative to the cabin temperature, the temperature rise of the rear bearing temperature relative to the ambient temperature, the difference between the temperature rise of the front bearing temperature relative to the cabin temperature and the temperature rise of the rear bearing temperature relative to the cabin temperature, the difference between the temperature rise of the front bearing temperature relative to the ambient temperature and the temperature rise of the rear bearing temperature relative to the ambient temperature, the average front bearing temperature, the average rear bearing temperature, the ratio of the average front bearing temperature to the average rear bearing temperature, the ratio of the front bearing temperature to the generator speed, and the difference between the average front bearing temperature and the average rear bearing temperature.

[0055] For example, the front bearing temperature at two or more points is obtained and the obtained front bearing temperature is averaged to obtain the front bearing temperature average value, the rear bearing temperature at two or more points is obtained and the obtained rear bearing temperature is averaged to obtain the rear bearing temperature average value, and the front and rear bearing temperature difference (i.e., the difference between the front bearing temperature average value and the rear bearing temperature average value) is determined based on the front bearing temperature average value and the rear bearing temperature average value. In addition, the ambient temperature and the cabin temperature are variables that are strongly correlated with the main bearing temperature. The main bearing ambient temperature rise can be calculated to offset the influence of the ambient temperature on the main bearing temperature, and is used for the unified comparative analysis of historical data and current data, current data and other unit data, etc. In addition, in order to evaluate the influence of the generator speed on the main bearing temperature, the relative temperature of the generator main bearing temperature can be calculated, that is, the main bearing temperature corresponding to one rotation of the generator. The relative temperature of the main bearing at different speeds is different. The cabin temperature is relatively stable and close to the generator. This temperature is used as one of the fault judgment conditions. The cabin temperature rise can reflect the operation of the generator main bearing to a certain extent. The front bearing temperature rise includes the increase of the front bearing temperature relative to the cabin temperature or the increase of the front bearing temperature relative to the ambient temperature, and the rear bearing temperature rise includes the increase of the rear bearing temperature relative to the cabin temperature or the increase of the rear bearing temperature relative to the ambient temperature. The difference between the front bearing temperature rise and the rear bearing temperature rise may include the difference between the increase of the front bearing temperature relative to the cabin temperature and the increase of the rear bearing temperature relative to the cabin temperature or the difference between the increase of the front bearing temperature relative to the ambient temperature and the increase of the rear bearing temperature relative to the ambient temperature. In another example, the generator main bearing temperature characteristic variable also includes the upper quartile of the front bearing temperature, the relative temperature of the main bearing (the ratio of the generator front bearing temperature to the generator speed), and the peak ratio of the front and rear bearing temperatures (the front bearing temperature is a multiple of the rear bearing temperature, the larger the ratio, the higher the front bearing temperature is than the rear bearing temperature; when the ratio is positive, the front bearing temperature and the rear bearing temperature are both above zero or below zero; when the ratio is negative, it means that the front bearing temperature and the rear bearing temperature are above zero and below zero respectively).

[0056] In the process of early warning based on the actual data of the unit, it is not easy to objectively, accurately and in advance discover the initial abnormality of the unit based on the abnormality of a single indicator. Therefore, it is necessary to perform fusion analysis through multiple variables. The generator main bearing lubrication method of the wind turbine generator set according to the exemplary embodiment of the present disclosure takes into account the various generator main bearing temperature characteristic variables as described above, and can accurately determine the wear condition of the generator main bearing according to the temperature of the bearing. In step S102, based on the generator main bearing temperature characteristic variables, multiple generator main bearing state prediction models are established, and the wear condition of the generator front bearing is predicted using multiple generator main bearing state prediction models. In the example, in response to the difference between the generator main bearing temperature characteristic variable of the rear bearing and the corresponding generator main bearing temperature characteristic variable of the front bearing exceeding a predetermined threshold, it can be determined that the front bearing is in an abnormal wear state. For example, in response to the difference between the temperature rise of the front bearing relative to the cabin temperature and the temperature rise of the rear bearing relative to the cabin temperature exceeding a predetermined threshold, or the difference between the temperature rise of the front bearing relative to the ambient temperature and the temperature rise of the rear bearing relative to the ambient temperature exceeding a predetermined threshold, or the difference between the ratio of the front bearing temperature to the generator speed and the ratio of the rear bearing temperature to the generator speed exceeding a predetermined threshold, it can be determined that the front bearing is in an abnormal wear state. In the example, the multiple generator main bearing state prediction models include two or more of the following models: a prediction model based on multivariate fusion analysis, a prediction model based on time series trend decomposition, a prediction model based on a Gaussian mixture model, a prediction model based on a support vector machine, and a prediction model based on deep learning. In the multivariate fusion analysis of specific working conditions, correlation analysis is performed based on conventional characteristics of the unit, main bearing related characteristics, and feature derivative reconstruction. Compared with the analysis of a single feature, the reliability, accuracy, and timeliness of the main bearing warning results can be improved. In the time series trend decomposition analysis, through the SCADA data analysis of the unit, it is found that there is a strong correlation between the main bearing temperature and the ambient temperature, and the main bearing temperature data has a strong seasonal change. Through the time series trend decomposition analysis, trend factors, seasonal factors, and random factors can be decomposed, and fault diagnosis and prediction can be achieved based on trend factors. Prediction models based on machine learning may include Gaussian mixture model clustering (GMM) classification models and support vector machine (SVM) classification models (support vector machine is a type of supervised machine learning model that can be used for classification and regression. It aims to find an optimal hyperplane in multidimensional space that can divide all samples into two categories. This plane makes the distance between the closest points in the two categories as large as possible. The points on the boundary of the distance are called support vectors, and the segmented hyperplane is located in the middle of the distance. For example, the data before the abnormality is found is marked as 0 as the normal state, and the data after that is marked as abnormal data and abnormal state as 1.Parameter tuning: nonlinear kernel function - radial basis function kernel, maps low-dimensional vectors to high-dimensional vectors, based on the grid search method, different transformations and combinations of gamma and cost parameters, balance overfitting and underfitting, and output the optimal model parameters among the candidate parameters. Model fitting: fit the model using the optimal parameters. Evaluation cross-validation: predict the test set, and output the confusion matrix (horizontal: real; vertical: predicted) and effect visualization ROC, evaluate the model effect), cluster and classify the main bearing temperature and related measurement point data in different dimensions through clustering and classification models, and accurately identify the characteristics of the gradual change process of abnormal main bearing wear.

[0057] In the prediction model based on deep learning, using the backend engine of TensorFlow and the Keras deep learning framework, deep learning models (networks) can be trained and developed to learn the characteristic laws between abnormal units and normal units based on SCADA data, map abnormal features and normal features to the deep model network, and finally, use the trained deep learning model to generate predictions for test data and new data, iteratively optimize the model so that it can be gradually improved, so that it can accurately identify abnormal patterns in complex working conditions. For example, with regard to the Keras workflow, you can first define the data, input tensors and target tensors, and then configure the learning process by defining the model (model structure / layer, neuron, activation function), compiling the model (loss, optimizer, monitoring indicators), and calling FIT iterative optimization. The deep learning model will learn the SCADA data characteristics between normal units and abnormal units. Finally, use the deep learning model to generate predictions for test data and new data, verify whether these predictions match the actual situation, and continuously iterate and optimize the model. Regarding data preparation, after storing the tensor-converted unit sensor data in a two-dimensional or three-dimensional tensor format of (30, 12342, 6) and performing other operations such as standardization, a generator function can be written using the preprocessed data to generate the most recent batch data and future target data. A generator function is written for the unit's multi-point data, which mainly generates a list (samples, targets), where samples is a set of input data and targets is the corresponding target data. Three generators are instantiated using the abstract generator function: one for training (train_gen), one for validation (halogen), and one for testing (test_gen). Each generator will look at a different time segment of the original data (custom t), the training generator looks at the first 0-t ten thousand time steps, the validation generator looks at the next t+1 to t+n data, and the test generator looks at the rest. Regarding defining, compiling, and training models, you can establish the first loop baseline, use loop dropout (a floating point number that specifies the drop rate of the input units of the layer) to combat overfitting, train and evaluate the dropout regularized model, stack the loop network, and use the model development technology route of deep learning networks (including reverse training and evaluation of LSTM) to train and test the model.

[0058] Through statistical analysis, multivariate fusion analysis of specific working conditions, time series analysis, machine learning and deep learning related algorithms, the abnormal state of the main bearing can be accurately diagnosed and the severity of the main bearing abnormality can be graded, thereby implementing maintenance and greasing strategies for different levels, achieving accurate early warning and maintenance of the abnormal state of the main bearing and extending the life of the main bearing.

[0059] In the example, for the prediction model based on multivariate fusion analysis, the step of predicting the wear of the front bearing of the generator includes: comparing the temperature characteristic variables of the main bearing of the generator with the corresponding thresholds respectively; in response to at least one of the temperature characteristic variables of the main bearing of the generator exceeding the corresponding threshold, determining that the front bearing of the generator is in an abnormal wear state. For example, the temperature rise of the front bearing and / or the temperature rise of the rear bearing can be compared with the corresponding threshold (for example, 27°C), and when the temperature rise of the front bearing and / or the temperature rise of the rear bearing exceeds the corresponding threshold, it is determined that the front bearing of the generator is in an abnormal wear state. For example, the difference between the temperature rise of the front bearing and the temperature rise of the rear bearing can be compared with the corresponding threshold (for example, 12°C), and when the difference between the temperature rise of the front bearing and the temperature rise of the rear bearing exceeds the corresponding threshold, it is determined that the front bearing of the generator is in an abnormal wear state. For example, the ratio of the average temperature of the front bearing and the average temperature of the rear bearing can be compared with the corresponding threshold (for example, 1.5), and when the ratio of the average temperature of the front bearing and the average temperature of the rear bearing exceeds the corresponding threshold, it is determined that the front bearing of the generator is in an abnormal wear state. For example, the difference between the average temperature of the front bearing and the average temperature of the rear bearing may be compared with a corresponding threshold value (e.g., 12°C). When the difference between the average temperature of the front bearing and the average temperature of the rear bearing exceeds the corresponding threshold value, it is determined that the front bearing of the generator is in an abnormal wear state. For example, the specific operating temperature of the front bearing and / or the specific operating temperature of the rear bearing may be compared with a corresponding threshold value (e.g., 45°C). When the specific operating temperature of the front bearing and / or the specific operating temperature of the rear bearing exceeds the corresponding threshold value, it is determined that the front bearing of the generator is in an abnormal wear state. In addition, in the example, the above conditions may also be considered in combination to determine whether the front bearing of the generator is in an abnormal wear state. In addition, different thresholds may be set to divide the wear levels of abnormal wear states, for example, higher thresholds correspond to higher wear levels. In the example, for the prediction model based on multivariate fusion analysis, the step of predicting the wear condition of the generator front bearing includes: in response to at least one of the generator main bearing temperature characteristic variables exceeding the corresponding first threshold, determining that the wear level of the generator front bearing is a first level; in response to at least one of the generator main bearing temperature characteristic variables exceeding the corresponding second threshold, determining that the wear level of the generator front bearing is a second level, wherein the second threshold is higher than the first threshold, and the degree of wear of the generator front bearing indicated by the second level is higher than the degree of wear of the generator front bearing indicated by the first level.

[0060] In addition, through the curves of the mean, maximum value, quantile value, second-order central moment (measures the fluctuation of random variable data around the mean, the larger the variance, the greater the volatility), third-order central moment (measures the degree to which the random density function of bearing-related indicator data is skewed to the left or right, characterizes the data distribution characteristics, and is calculated based on the mean, variance and skewness of the data. The result can be positive, negative or zero. When the result is zero, it indicates that the data distribution is symmetrical; when the result is positive, it indicates that the data distribution is skewed to the right; when the result is negative, it indicates that the data distribution is skewed to the left), and fourth-order central moment of the generator main bearing temperature characteristic variable, it is possible to determine whether the generator main bearing is in an abnormal wear state and the wear level of the abnormal wear state.

[0061] In the example, for a prediction model based on time series trend decomposition, the steps of predicting the wear condition of the generator front bearing include: determining a trend factor based on the temperature characteristic variable of the generator main bearing; and determining that the generator front bearing is in an abnormal wear state in response to the trend factor being greater than a specific threshold.

[0062] For example, a trend decomposition multiplication model or a trend decomposition addition model can be used to decompose the characteristic variable of the generator main bearing temperature, extract the trend factor in the data, and trigger an early warning when the extracted trend factor is greater than the set threshold. Figure 2 As shown, the four curves on the left are the time series decomposition curves obtained by the STL method, which are the original data curve, the periodic component curve, the trend component curve and the remainder curve from top to bottom; the four curves on the right are the time series decomposition curves obtained by the decomposition method, which are the original data curve, the trend component curve, the periodic component curve and the random component curve from top to bottom. Figure 2 The data shown are abnormal unit data. It can be seen from the separated trend component curve that the relevant indicator data show an obvious upward trend.

[0063] In the example, for the prediction model based on the Gaussian mixture model, the step of predicting the wear of the generator front bearing includes: in response to the front bearing temperature presenting a binormal distribution, determining that the generator front bearing is in an abnormal wear state. The Gaussian mixture model is a practical application of the expectation-maximization (EM) algorithm. When performing cluster analysis, all samples are considered to come from several normal populations with different parameter controls. For example, in the gradual wear process of the main bearing, the parameter distribution problem under specific working conditions at different periods can be regarded as a binormal distribution mixed probability model. Figure 3 In the probability density curve shown, after the wear abnormality, the generator main bearing temperature related indicators (for example, the front bearing temperature) show a bimodal change. In the clustering model, the data after the wear abnormality is significantly different from the data before the model abnormality. The algorithm can accurately predict that the generator main bearing is in an abnormal wear state.

[0064] In step S103, the wear state of the generator front bearing is determined based on the prediction results of multiple generator main bearing state prediction models. In the example, in response to the number of prediction results indicating that the generator front bearing is in an abnormal wear state being greater than zero, the wear level of the generator front bearing is determined based on the number of prediction results indicating that the generator front bearing is in an abnormal wear state; in response to the number of prediction results indicating that the generator front bearing is in an abnormal wear state being zero, it is determined that the generator front bearing is in a normal wear state. In step S104, a lubrication strategy is determined based on the wear state of the generator front bearing, and the generator front bearing is lubricated according to the lubrication strategy. In the example, based on the wear level of the generator front bearing, a lubrication parameter corresponding to the wear level is determined, and the lubrication parameter includes at least one of a single greasing time, a lubrication pump flow rate, and a greasing time interval. For steps S103 and S104, please refer to the following for Figure 4 Detailed description.

[0065] The generator main bearing lubrication method of a wind turbine generator set according to the exemplary embodiment of the present disclosure can diagnose the wear condition of the main bearing, determine the wear level, and adjust and execute different greasing strategies and maintenance strategies according to the wear level, so that the greasing strategy is more in line with the current operating state of the bearing, improve the operating condition of the bearing, and achieve the effect of extending the service life and reducing operation and maintenance costs.

[0066] Figure 4 is a flow chart illustrating a method of determining a lubrication strategy according to an exemplary embodiment of the present disclosure.

[0067] like Figure 4 As shown, in step S401, the prediction results of multiple generator main bearing state prediction models are obtained. For example, after obtaining the SCADA predetermined measurement point data of several units, the amount of original data and the amount of pre-processed data are judged, and after determining that the amount of data meets the predetermined conditions, the data processing, feature engineering, sub-condition (sequence) division and processing steps are entered; the pre-processed data of a single unit or two or more units are input into the trained prediction model (for example, a multivariate fusion model for a specific condition, a Gaussian mixture clustering model, a support vector machine classification model, a deep learning model), and the corresponding prediction results are obtained from each model.

[0068] In step S402, it is determined whether the number of prediction results indicating that the generator front bearing is in an abnormal wear state is greater than zero. If it is determined in step S402 that the number of prediction results indicating that the generator front bearing is in an abnormal wear state is not greater than zero (that is, the number of prediction results indicating that the generator front bearing is in an abnormal wear state is zero), then in step S403, it is determined that the generator front bearing is in a normal wear state. If it is determined in step S402 that the number of prediction results indicating that the generator front bearing is in an abnormal wear state is greater than zero, then in step S404, based on the number of prediction results indicating that the generator front bearing is in an abnormal wear state, the wear level of the generator front bearing is determined. For example, taking the case including four types of prediction models as an example, when the prediction results of the four prediction models all indicate that the front bearing of the generator is not in an abnormal wear state, it can be determined that the front bearing of the generator is in a normal wear state and the current lubrication strategy is maintained; when the prediction result of one of the four prediction models indicates that the front bearing of the generator is in an abnormal wear state, the wear level of the front bearing of the generator can be determined as level 1; when the prediction results of two of the four prediction models indicate that the front bearing of the generator is in an abnormal wear state, the wear level of the front bearing of the generator can be determined as level 2; when the prediction results of three of the four prediction models indicate that the front bearing of the generator is in an abnormal wear state, the wear level of the front bearing of the generator can be determined as level 3; when the prediction results of the four prediction models all indicate that the front bearing of the generator is in an abnormal wear state, the wear level of the front bearing of the generator can be determined as level 4.

[0069] In step S405, based on the wear level of the generator front bearing, lubrication parameters corresponding to the wear level are determined, and the generator front bearing is lubricated according to the determined lubrication parameters. In the example, the lubrication parameters include at least one of a single greasing time, a lubrication pump flow rate, and a greasing time interval. Continuing the example given above for steps S402 to S404, when the wear level of the generator front bearing is determined to be level 1, compared with the current lubrication strategy, at least one of the single greasing time, the lubrication pump flow rate, and the greasing time interval will be adjusted, for example, increasing the single greasing time, increasing the lubrication pump flow rate, or reducing the greasing time interval; similarly, as the wear level increases, the lubrication parameters will be adjusted to increase the lubrication amount per unit time of the generator front bearing. In another example, online automatic lubrication is performed or a manual lubrication reminder is issued according to whether the unit is equipped with an online automatic lubrication device. In the case of automatic lubrication, the automatic lubrication process is controlled according to the determined number of lubrications, lubrication duration, lubrication amount, or other special maintenance strategies. In the case of manual lubrication, the manual lubrication reminder may include recommended data for a single greasing time, lubrication pump flow rate, and greasing time interval. In addition, although the above examples are described by taking four prediction models and four wear levels as an example, those skilled in the art should understand that the examples of the present disclosure may include more or fewer prediction models and wear levels.

[0070] By adopting the method for determining the lubrication strategy according to the exemplary embodiment of the present disclosure, it is possible to implement a main bearing wear warning, accurately identify the abnormal wear state of the main bearing and quantify the severity level of the abnormal wear state (for example, output the main bearing abnormality level of "normal", "attention", "warning", "serious", etc.), and execute different optimal maintenance strategies corresponding to the wear level, such as different lubrication times, lubrication durations, grease addition amounts, special maintenance, etc., according to the main bearing wear level.

[0071] Figure 5 A block diagram of a generator main bearing lubrication device of a wind turbine generator set according to an exemplary embodiment of the present disclosure is shown.

[0072] like Figure 5As shown, a generator main bearing lubrication device 500 of a wind turbine generator set according to an exemplary embodiment of the present disclosure includes: a characteristic variable determination unit 501, which determines the characteristic variable of the generator main bearing temperature based on the operating data of the wind turbine generator set; a prediction model unit 502, which establishes multiple generator main bearing state prediction models based on the generator main bearing temperature characteristic variable, and uses multiple generator main bearing state prediction models to predict the wear of the generator front bearing; a wear state determination unit 503, which determines the wear state of the generator front bearing based on the prediction results of multiple generator main bearing state prediction models; a lubrication control unit 504, which determines the lubrication strategy based on the wear state of the generator front bearing, and lubricates the generator front bearing according to the lubrication strategy.

[0073] In the example, the main bearing includes a front bearing sleeved at the front end of the generator main shaft and a rear bearing sleeved at the middle of the generator main shaft. In the example, the prediction model unit 502 is configured to: in response to the difference between the generator main bearing temperature characteristic variable of the rear bearing and the corresponding generator main bearing temperature characteristic variable of the front bearing exceeding a predetermined threshold, determine that the front bearing is in an abnormal wear state. In addition, the wind turbine generator set is an outer rotor inner stator type permanent magnet direct drive wind turbine generator set.

[0074] In the example, the operation data of the wind turbine generator set includes the generator speed, the ambient temperature, the cabin temperature, and the front bearing temperature and the rear bearing temperature of the generator main bearing. In the example, the generator main bearing temperature characteristic variables include the temperature rise of the front bearing temperature relative to the cabin temperature, the temperature rise of the front bearing temperature relative to the ambient temperature, the temperature rise of the rear bearing temperature relative to the cabin temperature, the temperature rise of the rear bearing temperature relative to the ambient temperature, the difference between the temperature rise of the front bearing temperature relative to the cabin temperature and the temperature rise of the rear bearing temperature relative to the cabin temperature, the difference between the temperature rise of the front bearing temperature relative to the ambient temperature and the temperature rise of the rear bearing temperature relative to the ambient temperature, the average value of the front bearing temperature, the average value of the rear bearing temperature, the ratio of the front bearing temperature to the average value of the rear bearing temperature, the ratio of the front bearing temperature to the generator speed, and the difference between the average value of the front bearing temperature and the average value of the rear bearing temperature. In the example, the multiple generator main bearing state prediction models include two or more of the following models: a prediction model based on multivariate fusion analysis, a prediction model based on time series trend decomposition, a prediction model based on a Gaussian mixture model, a prediction model based on a support vector machine, and a prediction model based on deep learning.

[0075] In the example, for the prediction model based on multivariate fusion analysis, the prediction model unit 502 is configured to: compare the generator main bearing temperature characteristic variables with the corresponding thresholds respectively; in response to at least one of the generator main bearing temperature characteristic variables exceeding the corresponding threshold, determine that the generator front bearing is in an abnormal wear state. In the example, for the prediction model based on multivariate fusion analysis, the prediction model unit 502 is configured to: in response to at least one of the generator main bearing temperature characteristic variables exceeding the corresponding first threshold, determine that the wear level of the generator front bearing is the first level; in response to at least one of the generator main bearing temperature characteristic variables exceeding the corresponding second threshold, determine that the wear level of the generator front bearing is the second level, wherein the second threshold is higher than the first threshold, and the wear degree of the generator front bearing indicated by the second level is higher than the wear degree of the generator front bearing indicated by the first level. In the example, for the prediction model based on time series trend decomposition, the prediction model unit 502 is configured to: determine the trend factor based on the generator main bearing temperature characteristic variables; in response to the trend factor being greater than a specific threshold, determine that the generator front bearing is in an abnormal wear state. In an example, for a prediction model based on a Gaussian mixture model, the prediction model unit 502 is configured to: in response to the front bearing temperature presenting a binormal distribution, determine that the generator front bearing is in an abnormal wear state.

[0076] In the example, the wear state determination unit 503 is configured to: in response to the number of prediction results indicating that the generator front bearing is in an abnormal wear state being greater than zero, determine the wear level of the generator front bearing based on the number of prediction results indicating that the generator front bearing is in an abnormal wear state; in response to the number of prediction results indicating that the generator front bearing is in an abnormal wear state being zero, determine that the generator front bearing is in a normal wear state.

[0077] In the example, the lubrication control unit 504 is configured to determine lubrication parameters corresponding to the wear level based on the wear level of the generator front bearing, the lubrication parameters including at least one of a single greasing time, a lubrication pump flow rate and a greasing time interval.

[0078] Combination of the above Figures 1 to 4 The specific operations shown are respectively Figure 5 The corresponding units in the generator main bearing lubrication device 500 of the wind turbine generator set shown are executed, and the specific operation details will not be repeated here. The generator main bearing lubrication device of the wind turbine generator set according to the exemplary embodiment of the present disclosure can realize the diagnosis of the wear condition of the main bearing, determine the wear level, and adjust and execute different greasing strategies and maintenance strategies according to the wear level, so that the greasing strategy is more in line with the current operating state of the bearing, improve the operating condition of the bearing, and achieve the effect of extending the life and reducing the operation and maintenance cost.

[0079] Figure 6 is a block diagram illustrating a computing system including at least one computing device and at least one storage device storing instructions according to an exemplary embodiment of the present disclosure.

[0080] like Figure 6 As shown, a computing system 600 provided according to an exemplary embodiment of the present invention includes a computing device 601 and a storage device 602, wherein the storage device 602 stores computer executable instructions. When the computer executable instructions are executed by the computing device 601, the generator main bearing lubrication method of the wind turbine generator set described in any of the aforementioned embodiments is executed.

[0081] The computing device 601 is deployed in a server or client, and may also be deployed on a node device in a distributed network environment. In addition, the computing device 601 may be a PC computer, a tablet device, a personal digital assistant, a smart phone, a web application or other device capable of executing the above-mentioned instruction set. Here, the computing device is not necessarily a single computing device, but may also be any device or circuit capable of executing the above-mentioned instruction (or instruction set) individually or jointly. The computing device may also be a part of an integrated control system or a system manager, or may be configured as a portable electronic device interconnected with an interface locally or remotely (e.g., via wireless transmission). In the computing device, the processor includes a central processing unit (CPU), a graphics processing unit (GPU), a programmable logic device, a dedicated processor system, a microcontroller or a microprocessor. As an example and not limitation, the processor also includes an analog processor, a digital processor, a microprocessor, a multi-core processor, a processor array, a network processor, etc.

[0082] According to another aspect of the present disclosure, a computer-readable storage medium storing instructions is provided, and when the instructions are executed by at least one computing device, the at least one computing device is prompted to execute the method for lubricating the main bearing of the generator of the wind turbine generator set described in any of the aforementioned embodiments. The computer-readable storage medium includes magnetic media such as floppy disks and tapes, optical media (including compact disk (CD) ROM and DVD ROM), magneto-optical media such as floppy disks, hardware devices such as ROM, RAM and flash memory designed to store and execute program commands. The instructions may include language codes executable by a computer using an interpreter and machine language codes generated by a compiler.

[0083] By adopting the present disclosure, it is possible to accurately, reliably and in advance identify abnormalities in the main bearing of the generator, execute corresponding maintenance strategies, reduce maintenance costs, and provide valuable data support for business decisions; without having to invest large costs, it is possible to achieve coverage of a wider range of units and identification of hidden risks, thereby reducing the cost and investment of manual inspections.

[0084] The processes, methods or algorithms disclosed herein may be transmitted to or implemented by a processing device, a controller or a computer, which may include any existing programmable electronic control unit or a dedicated electronic control unit. Similarly, the processes, methods or algorithms may be stored in a variety of forms as data and instructions that can be executed by a controller or a computer, including but not limited to information being permanently stored on a non-writable storage medium (such as a ROM device) and information being variably stored on a writable storage medium (such as a floppy disk, a tape, a CD, a RAM device, and other magnetic and optical media). The processes, methods or algorithms may also be implemented in a software executable object. Optionally, the processes, methods or algorithms may be implemented in whole or in part using suitable hardware components (such as an ASIC, an FPGA, a state machine, a controller or other hardware components or devices) or a combination of hardware components, software components and firmware components.

[0085] Although the present disclosure includes specific examples, it will be apparent to those of ordinary skill in the art that various changes in form and detail may be made in these examples without departing from the spirit and scope of the claims and their equivalents. The examples described herein will be considered to be descriptive only and not for limiting purposes. The description of the features or aspects in each example will be considered to be applicable to similar features or aspects in other examples. Suitable results may be obtained if the described techniques are performed in a different order, and / or if the components in the described systems, architectures, devices, or circuits are combined in different ways and / or replaced or supplemented with other components or their equivalents in the described systems, architectures, devices, or circuits. Therefore, the scope of the present disclosure is not limited by specific embodiments, but by the claims and their equivalents, and all variations within the scope of the claims and their equivalents will be interpreted as included in the present disclosure.

Claims

1. A method for lubricating a main bearing of a wind turbine generator set, characterized in that: The method comprises: Based on the operation data of the wind turbine generator set, determine the temperature characteristic variables of the generator main bearing; Based on the generator main bearing temperature characteristic variable, multiple generator main bearing state prediction models are established, and the wear condition of the generator front bearing is predicted using the multiple generator main bearing state prediction models; Determining the wear state of the front bearing of the generator based on the prediction results of the plurality of generator main bearing state prediction models; A lubrication strategy is determined based on the wear state of the generator front bearing, and the generator front bearing is lubricated according to the lubrication strategy.

2. The method for lubricating the main bearing of a wind turbine generator set according to claim 1, characterized in that: The operation data of the wind turbine generator set includes generator rotation speed, ambient temperature, cabin temperature, and front bearing temperature and rear bearing temperature of the generator main bearing.

3. The method for lubricating the main bearing of a wind turbine generator set according to claim 1, characterized in that: The generator main bearing comprises a front bearing sleeved on the front end of the generator main shaft and a rear bearing sleeved on the middle part of the generator main shaft.

4. The method for lubricating the main bearing of a wind turbine generator set according to claim 3, characterized in that: The step of predicting the wear condition of the generator front bearing includes: in response to the difference between the generator main bearing temperature characteristic variable of the rear bearing and the corresponding generator main bearing temperature characteristic variable of the front bearing exceeding a predetermined threshold, determining that the generator front bearing is in an abnormal wear state.

5. The method for lubricating the main bearing of a wind turbine generator set according to claim 1, characterized in that: The wind turbine generator set is an outer rotor inner stator type permanent magnet direct drive wind turbine generator set.

6. The method for lubricating the main bearing of a wind turbine generator set according to claim 1, characterized in that: The multiple generator main bearing state prediction models include two or more of the following models: a prediction model based on multivariate fusion analysis, a prediction model based on time series trend decomposition, a prediction model based on a Gaussian mixture model, a prediction model based on a support vector machine, and a prediction model based on deep learning.

7. The method for lubricating the main bearing of a wind turbine generator set according to claim 1, characterized in that: The step of predicting the wear condition of the generator front bearing includes: comparing the temperature characteristic variables of the generator main bearing with corresponding thresholds respectively; in response to at least one of the temperature characteristic variables of the generator main bearing exceeding the corresponding threshold, determining that the generator front bearing is in an abnormal wear state.

8. The method for lubricating the main bearing of a wind turbine generator set according to claim 1, characterized in that: The step of predicting the wear condition of the generator front bearing includes: determining a trend factor based on the temperature characteristic variable of the generator main bearing; and determining that the generator front bearing is in an abnormal wear state in response to the trend factor being greater than a specific threshold.

9. The method for lubricating the main bearing of a wind turbine generator set according to claim 1, characterized in that: The step of predicting the wear condition of the generator front bearing includes: in response to the generator front bearing temperature presenting a binormal distribution, determining that the generator front bearing is in an abnormal wear state.

10. The method for lubricating the main bearing of a wind turbine generator set according to any one of claims 1 to 9, characterized in that: The step of determining the wear state of the front bearing of the generator based on the prediction results of the multiple generator main bearing state prediction models comprises: In response to the number of prediction results indicating that the generator front bearing is in an abnormal wear state being greater than zero, determining a wear level of the generator front bearing based on the number of prediction results indicating that the generator front bearing is in an abnormal wear state; In response to the number of prediction results indicating that the generator front bearing is in an abnormal wear state being zero, it is determined that the generator front bearing is in a normal wear state.

11. The method for lubricating the main bearing of a wind turbine generator set according to claim 9, characterized in that: The step of determining the lubrication strategy based on the wear state of the generator front bearing comprises: Based on the wear level of the front bearing of the generator, a lubrication parameter corresponding to the wear level is determined, wherein the lubrication parameter includes at least one of a single greasing time, a lubrication pump flow rate, and a greasing time interval.

12. The method for lubricating the main bearing of a wind turbine generator set according to claim 1, characterized in that: The step of predicting the wear condition of the front bearing of the generator comprises: In response to at least one of the generator main bearing temperature characteristic variables exceeding a corresponding first threshold, determining that the wear level of the generator front bearing is a first level; In response to at least one of the generator main bearing temperature characteristic variables exceeding the corresponding second threshold, the wear level of the generator front bearing is determined to be a second level, wherein the second threshold is higher than the first threshold, and the wear level of the generator front bearing indicated by the second level is higher than the wear level of the generator front bearing indicated by the first level.

13. The method for lubricating the main bearing of a wind turbine generator set according to claim 1, characterized in that: The step of determining the lubrication strategy based on the wear state of the generator front bearing comprises: Based on the wear level of the front bearing of the generator, a lubrication parameter corresponding to the wear level is determined, wherein the lubrication parameter includes at least one of a single greasing time, a lubrication pump flow rate, and a greasing time interval.

14. The method for lubricating the main bearing of a wind turbine generator set according to claim 1, characterized in that: The generator main bearing temperature characteristic variables include the temperature rise of the front bearing temperature relative to the cabin temperature, the temperature rise of the front bearing temperature relative to the ambient temperature, the temperature rise of the rear bearing temperature relative to the cabin temperature, the temperature rise of the rear bearing temperature relative to the ambient temperature, the difference between the temperature rise of the front bearing temperature relative to the cabin temperature and the temperature rise of the rear bearing temperature relative to the cabin temperature, the difference between the temperature rise of the front bearing temperature relative to the ambient temperature and the temperature rise of the rear bearing temperature relative to the ambient temperature, the average value of the front bearing temperature, the average value of the rear bearing temperature, the ratio of the average value of the front bearing temperature to the average value of the rear bearing temperature, the ratio of the front bearing temperature to the generator speed, and the difference between the average value of the front bearing temperature and the average value of the rear bearing temperature.

15. A lubrication device for a main bearing of a wind turbine generator, characterized in that: The generator main bearing lubrication device comprises: A characteristic variable determination unit, which determines a characteristic variable of a main bearing temperature of a generator based on operating data of the wind turbine generator set; A prediction model unit, based on the generator main bearing temperature characteristic variable, establishes multiple generator main bearing state prediction models, and uses the multiple generator main bearing state prediction models to predict the wear of the generator front bearing; a wear state determining unit, which determines the wear state of the generator front bearing based on the prediction results of the plurality of generator main bearing state prediction models; A lubrication control unit determines a lubrication strategy based on the wear state of the generator front bearing, and lubricates the generator front bearing according to the lubrication strategy.

16. A computing system comprising at least one computing device and at least one storage device storing instructions, characterized in that: When the instructions are executed by the at least one computing device, the at least one computing device is prompted to execute the method for lubricating a main bearing of a generator of a wind turbine generator set according to any one of claims 1-14.

17. A computer-readable storage medium storing instructions, characterized in that: When the instructions are executed by at least one computing device, the at least one computing device is prompted to execute the method for lubricating a main bearing of a generator of a wind turbine generator set according to any one of claims 1-14.

Images