Wind turbine transmission chain fault monitoring method based on similarity measurement and trend perception network

Abstract

The present invention provides a method for monitoring wind turbine transmission chain faults based on similarity metrics and trend-aware networks, comprising the following steps: acquiring data, integrating a trend-aware algorithm and an attention mechanism, and embedding the data into a long-short-term memory network to form a trend-aware BiLSTM model with bidirectionally arranged network units; integrating a sequence similarity metric and a spatial distribution similarity metric to screen highly similar units based on the data, establishing a single-unit prediction model for each of the highly similar units based on the trend-aware BiLSTM model to form an integrated prediction model; obtaining prediction residuals of monitoring variables using the integrated prediction model, obtaining monitoring indicators using KL divergence, and implementing fault monitoring. The method of the present invention can address the problem of insufficient accuracy of existing methods and effectively improve the accuracy and reliability of wind turbine transmission chain fault monitoring.

Description

Technical Field

[0001] The present invention relates to the field of wind turbine fault monitoring, and in particular to a wind turbine transmission chain fault monitoring method based on similarity measurement and trend perception network. Background Art

[0002] Wind turbines are complex hybrid electromechanical systems designed to convert wind energy into electricity. Because they typically operate in harsh environments, they are highly susceptible to failure. Frequent failures not only reduce operational reliability but also increase operating and maintenance costs. Therefore, it is crucial to develop fault monitoring methods that can adapt to complex conditions and accurately detect abnormal wind turbine operating conditions.

[0003] According to the different analysis methods used, monitoring methods can be divided into three categories: (1) knowledge-based, (2) signal processing-based, and (3) data-driven. Knowledge-based methods rely on expert knowledge or experience, are often subjective and have limited accessibility, and are therefore difficult to promote. Signal processing-based methods usually have reliable monitoring results, but different monitoring signals require different signal analysis techniques, which places high technical requirements on the users. Data-driven methods estimate the values of state variables by learning and inferring a large amount of data, and then judge whether the operating status of unit components or subsystems is normal by comparing actual values and estimated values. This type of method has greater flexibility and practicality. The present invention belongs to a type of data-driven method.

[0004] The current data-driven methods have the following shortcomings: (1) The time correlation of variables is ignored when constructing the state estimation model. The current state of many variables (such as temperature, pressure, and speed) is usually highly correlated with the previous state and evolves over time; (2) Most methods only study a single unit, and due to the limitation of poor data completeness, the generalization ability of the model is insufficient; (3) The existing wind turbine similarity measurement method relies solely on statistical features to determine the similarity between operating conditions, ignoring the time characteristics of state variables. In addition, evaluating the similarity of operating conditions of different wind turbines by comparing only a specific variable makes it difficult to reflect the overall operating conditions and status of the unit. The above three problems lead to the poor reliability and low accuracy of the current wind turbine transmission chain fault monitoring methods.

[0005] In view of this, the present invention is proposed. Summary of the Invention

[0006] In view of this, the present invention proposes a wind turbine transmission chain fault monitoring method based on similarity measurement and trend perception network. This method can solve the problem of insufficient accuracy of existing methods and effectively improve the accuracy and reliability of wind turbine transmission chain fault monitoring.

[0007] Specifically, the present invention is achieved through the following technical solutions:

[0008] The present invention provides a wind turbine transmission chain fault monitoring method based on similarity measurement and trend perception network, comprising the following steps:

[0009] Acquire data, integrate trend perception algorithms and attention mechanisms, and embed them into a long short-term memory network to form a trend perception BiLSTM model with a bidirectional arrangement of network units;

[0010] By integrating the sequence similarity metric and the spatial distribution similarity metric, highly similar units are screened out based on the data, and a single unit prediction model is established for each of the highly similar units based on the trend-aware BiLSTM model to form an integrated prediction model;

[0011] The prediction residual of the monitoring variable is obtained through the integrated prediction model, and the monitoring index is obtained through the KL divergence to realize fault monitoring.

[0012] In addition to providing a wind turbine transmission chain fault monitoring method based on similarity measurement and trend perception network, the present invention also provides a fault monitoring system, including:

[0013] Perception model building module: used to acquire data, integrate the trend perception algorithm and attention mechanism, and embed it into the long short-term memory network to form a trend perception BiLSTM model with a bidirectional arrangement of network units;

[0014] Prediction module: used to integrate sequence similarity measurement and spatial distribution similarity measurement, screen out highly similar units based on the data, establish a single unit prediction model for each of the highly similar units based on the trend-aware BiLSTM model, and form an integrated prediction model;

[0015] Monitoring module: used to obtain the prediction residual of the monitoring variable through the integrated prediction model, obtain the monitoring index through KL divergence, and realize fault monitoring.

[0016] In summary, the solution of the present invention has the following beneficial effects:

[0017] (1) The present invention designs an interval quantile method for removing outliers in SCADA raw data, which overcomes the problem that existing data preprocessing methods are difficult to adapt to the frequently fluctuating operating conditions of wind turbines and cannot effectively remove outliers. The output is a normal data set that can be used for subsequent unit similarity analysis and normal behavior modeling.

[0018] (2) By combining sequence similarity measurement with multidimensional spatial distribution similarity measurement, the present invention can comprehensively compare the similarity between units from multiple perspectives such as operating conditions and unit status. This overcomes the problem of existing methods using few parameter variables and difficulty in measuring the similarity of nonlinear high-dimensional data, and can effectively find similar wind turbines.

[0019] (3) By combining the trend perception algorithm with the attention mechanism, the problem that conventional machine learning models cannot effectively extract trend features in time series data is solved, and the model's hidden feature extraction capability is effectively enhanced; the BiLSTM network is used to effectively mine the spatiotemporal features in multivariate input data, thereby improving the model's prediction accuracy.

[0020] (4) Compared with the traditional single-unit model, the present invention adopts a model integration framework to merge the models of multiple similar units into an overall integrated prediction model, further improving the prediction accuracy and stability; finally, the KL divergence is used to transform the residual sequence to construct a new fault monitoring indicator, which effectively improves the reliability of the monitoring indicator. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Various other advantages and benefits will become apparent to those skilled in the art by reading the detailed description of the preferred embodiment below. The accompanying drawings are only for the purpose of illustrating the preferred embodiment and are not to be considered as limiting the present invention. The same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0022] Figure 1 is a flow chart of wind turbine transmission chain fault monitoring method of the present invention;

[0023] Figure 2 Schematic diagram of the trend perception module structure of the method of the present invention;

[0024] Figure 3 is a schematic diagram of the method of the present invention;

[0025] Figure 4 Schematic diagram of the MSBiLSTM network model structure of the method of the present invention;

[0026] Figure 5 It is a practical case demonstration of the method of the present invention;

[0027] Figure 6A flowchart of a computer device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numerals in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0029] The terms used in the present disclosure are for the purpose of describing particular embodiments only and are not intended to limit the present disclosure. The singular forms "a", "an", and "the" used in the present disclosure and the appended claims are intended to include plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0030] It should be understood that although the terms first, second, third, etc. may be used in the present disclosure to describe various information, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of the present disclosure, the first information may also be referred to as the second information, and similarly, the second information may also be referred to as the first information. Depending on the context, the word "if" as used herein may be interpreted as "at the time" or "when" or "in response to determining",

[0031] Example

[0032] like Figure 1 As shown, a wind turbine transmission chain fault monitoring method based on similarity measurement and trend perception network is provided, comprising the following steps:

[0033] Acquire data, integrate trend perception algorithms and attention mechanisms, and embed them into a long short-term memory network to form a trend perception BiLSTM model with a bidirectional arrangement of network units;

[0034] By integrating the sequence similarity metric and the spatial distribution similarity metric, highly similar units are screened out based on the data, and a single unit prediction model is established for each of the highly similar units based on the trend-aware BiLSTM model to form an integrated prediction model;

[0035] The prediction residual of the monitoring variable is obtained through the integrated prediction model, and the monitoring index is obtained through the KL divergence to realize fault monitoring.

[0036] In this method, the original data set is first collected, and outliers are removed using the interval quantile method. The outlier-cleaned data set X is then normalized using the Max-Min method to align the dimensions of all parameters. Next, a trend-aware deep network model structure is designed to lay the foundation for building a normal behavior model. The trend-aware algorithm is combined with the attention mechanism to form a trend-aware module, which is embedded in a BiLSTM unit. These BiLSTM units are then connected to form an MSBiLSTM network model. Simultaneously, the similarity between the target wind turbine and other turbines in the wind farm is calculated to identify turbines with highly similar operating characteristics. The similarity weights corresponding to each parameter are determined, and the data sequence similarity between the target unit and other units is calculated. The multidimensional spatial distribution similarity between the target unit and other units' data is then calculated. The sequence similarity and spatial distribution similarity are then integrated to calculate the total similarity. Next, for the selected similar units, a single unit normal behavior model is established based on the MSBiLSTM network model structure. Multiple single unit models are further integrated into a single prediction model. Finally, based on this integrated model, the target parameter variables are predicted, resulting in a residual sequence. The KL divergence is calculated as a fault monitoring indicator. A 3-sigma algorithm is used to calculate the threshold of the fault monitoring indicator. Whether the monitoring indicator crosses the threshold determines whether an anomaly exists.

[0037] The present invention proposes a wind turbine transmission chain fault monitoring method based on similarity measurement and trend perception network. It combines the time series similarity measurement with the multidimensional data spatial distribution similarity measurement, and proposes a similarity measurement method that considers sequence similarity and spatial distribution similarity to screen similar wind turbines. In addition, the trend perception mechanism is combined with the long short-term memory neural network to model the normal behavior of the wind turbine transmission chain.

[0038] The steps of the wind turbine transmission chain fault monitoring method based on similarity measurement and trend perception network during use are as follows:

[0039] Step (1): The environmental parameters, state parameters and operating parameters of each wind turbine are collected through the Supervisory Control And Data Acquisition (SCADA) system of the wind farm, and the original data set X = (x1, x2, ..., x k ); k represents the number of parameter variables; due to the influence of various factors such as environmental interference, sensor abnormalities, and communication failures, there is a small amount of abnormal data in the original data set; the present invention uses the interval quantile method to remove abnormal data. First, the data set is divided into multiple intervals according to wind speed, and then the quantile method is used to remove outliers for each parameter variable in a single interval. The calculation process is as follows:

[0040] Arrange the parameter variable data in descending order of value and divide them into four equal parts according to the total number of samples. The dividing lines are marked as Q1, Q2, and Q3, where Q1 represents the quantile value of the lower 25%; Q2 represents the median value, that is, the value at the 50% position; Q3 represents the quantile value of the upper 25%; the interquartile range (IQR) represents the degree of concentration of the value distribution and is calculated as follows:

[0041] IQR=Q3-Q1 (1)

[0042] Further calculate the numerical upper bound B h and numerical lower bound B l :

[0043]

[0044] Among them, α is the normal interval coefficient. The value is greater than B h or less than B l The samples are identified as abnormal data and removed;

[0045] The data set X that has been cleaned of outliers is normalized using the Max-min method, and the parameter variable x i Normalize to the range of [0,1] to make the dimensions of each parameter consistent and obtain the normalized data set X'; the calculation process is:

[0046]

[0047] Among them, e represents the parameter variable x i A sample in ; e' represents the normalized sample value; min(x i ) represents the parameter variable x i The minimum value of max(x i ) represents the parameter variable x i The maximum value of .

[0048] Step (2): Design a trend-aware deep network model structure to lay the foundation for establishing a normal behavior model.

[0049] 2.1) The trend perception algorithm is combined with the attention mechanism to form a trend perception module; this module mainly includes a dynamic trend perception part, an attention part, and a fusion part. Dynamic perception is used to extract local spatiotemporal relationships; attention is used to assign different weights to the extracted local spatiotemporal features; and the fusion part is used to fuse local and global information to obtain the hidden information most relevant to the current local area. The operation process of the trend perception module can be expressed as follows:

[0050]

[0051] q=Norm(S'next ) (5)

[0052]

[0053] Among them, S prev and S next are the previous and next local spatiotemporal hidden states respectively; S' next It's S next The predicted value; Norm is the normalization function; q is the value after normalization; Softmax is the activation function; is the output of BiLSTM; is the learnable parameter of this module; Memory information output by the previous network unit; is the feature vector at the current time point; D h is the hidden state dimension.

[0054] The structure of the trend perception module is as follows Figure 2 As shown in the figure, the memory information provided by the previous network unit is used as the input of the Norm&Drop layer. The output information of this layer enters the Softmax layer for feature extraction to obtain the hidden layer features, which are then input into the memory gating unit. At the same time, the trend variable of the current data is also sent to the memory gating unit to update the memorized trend information.

[0055] 2.2) The trend perception module is embedded in the Bidirectional Long Short-Term Memory (BiLSTM) neural network unit. The trend perception module replaces the linear layer in the BiLSTM to achieve efficient extraction of spatiotemporal information. The processing of the network unit is described as follows:

[0056]

[0057] in, represents the output variable of the first-layer trend perception network; σ represents the activation function; Represents the current time input data; is the hidden state at the previous time step; is the hidden state at the current time; Memory information output by the previous network unit; Represents the first layer learning parameters; represents the first-level trend-aware correlation bias;

[0058]

[0059] in, represents the output variable of the second layer trend perception network; Represents the second layer learning parameters; represents the second-layer trend perception related bias; [·‖·] represents the parallel connection operation; other parameters are consistent with formula (7);

[0060]

[0061] Tanh is the hyperbolic tangent activation function; is a high-order trend perception parameter; are high-order basis vectors.

[0062] The structure of the trend-aware LSTM module is as follows: Figure 3 As shown in the figure, first, the memory information of the previous unit is passed to the LSTM unit structure for memorization and discarding, and the output valid information enters the normalization layer and the weighted part respectively; the normalized information is sent to the attention mechanism and assigned weights according to the importance of the feature information; the current data is input into the attention mechanism and the trend perception module respectively for learning new information and trends; finally, the output information of the LSTM unit, attention mechanism, and trend perception module is fused and output through the weighted part.

[0063] 2.3) Connect the trend-sensing BiLSTM units to form an MSBiLSTM deep network model. Each network unit processes information for k time steps, which reduces the number of iterations while fully considering local spatiotemporal information. The model function can be expressed as follows:

[0064]

[0065] in, represents the i-th local spatiotemporal hidden state; Represents model learning parameters; represents the i-1th local spatiotemporal hidden state; It represents the i-th global state in the l-th layer; Indicates the input signal of this module; Represents a spatiotemporal embedding vector; using the extracted local state Update global information MSBiLSTM is a function of the deep learning model.

[0066] The structural diagram of the specific model is as follows Figure 4As shown in the figure, the network is mainly composed of three layers. The top layer connects the trend-aware LSTM units in forward order, and the information output by each unit enters the next unit and the weighted part respectively; the middle layer is composed of multiple weighted parts arranged in forward order, each weighted part receives information from the forward unit and the reverse unit, and the output information is passed to the next weighted part; the bottom layer connects the trend-aware LSTM units in reverse order, and similarly, the information output by each unit enters the next unit and the weighted part respectively; the current data of the external input enters the beginning of the forward layer and the beginning of the reverse layer respectively, so that the network model can extract the forward features and reverse features of the data at the same time, which can effectively improve the prediction accuracy of the model.

[0067] Step (3): Calculate the similarity between the target wind turbine and other turbines in the wind farm, and find turbines with highly similar operating characteristics.

[0068] 3.1) Calculate the mutual information MI between the target parameter and the candidate parameter i , find the important parameter variables. Calculate the similarity weight corresponding to the parameter using formula (13);

[0069]

[0070] Where k represents the number of input parameters; w i Represents the weight corresponding to parameter i.

[0071] 3.2) Calculate the data sequence similarity between the target unit and the candidate unit using formula (14); first construct the distance matrix Then calculate the cumulative distance matrix L:

[0072]

[0073] L(s,t) represents the cumulative distance between state s and time point t; s and t represent the current index positions of the two time series respectively; dis(x is ,x jt ) represents the distance between the sth element of time series 1 and the tth element of time series 2; min represents the operation of taking the minimum value;

[0074] The sequence similarity between two variables is defined as D DTW (x i ,x j )=L(m,n), where m represents the variable of the target unit and n represents the variable of the candidate unit.

[0075] 3.3) Calculate the multi-dimensional spatial distribution similarity between the target unit and the candidate unit data; X c and Y cRepresenting the center of the data set of the target unit and the candidate unit respectively, they are calculated by formula (15).

[0076]

[0077] Where x and y represent the mean values of parameter variables x and y respectively.

[0078] Based on the center of the dataset, calculate the covariance matrix ε:

[0079]

[0080] The multidimensional space distribution similarity is calculated as follows:

[0081]

[0082] ε -1 represents the inverse covariance matrix;

[0083] Finally, the sequence similarity and spatial distribution similarity are combined to calculate the total similarity S:

[0084]

[0085] w i represents the weight corresponding to the parameter variable i; D DTW(i) represents the sequence similarity of parameter variable i; D MD(i) Indicates the spatial distribution similarity of parameter variable i.

[0086] Step (4): For the similar units selected, based on the deep network model structure, the data of each unit is used to train the corresponding single unit normal behavior model, and then based on the model integration framework, multiple single unit models are integrated into one model.

[0087] 4.1) Based on the integrated model, the target parameter variable is predicted, and the difference between the predicted value and the actual value is calculated to obtain the residual sequence.

[0088] 4.2) Taking the residual sequence of the normal stage as the benchmark, the KL divergence between the real-time residual and the benchmark residual is calculated as the indicator of fault monitoring.

[0089] 4.3) Based on the fault monitoring indicator, a 3-sigma algorithm is used to calculate a threshold value of the fault monitoring indicator; during the monitoring process, if the monitoring indicator exceeds the threshold value, it is determined to be abnormal. If the monitoring indicator does not exceed the threshold value, it is determined to be normal.

[0090] In order to verify the effectiveness of the fault monitoring method of the present invention, a specific actual case is provided. Figure 5As shown in the figure, using SCADA data from a wind farm in Inner Mongolia, the wind turbine in the case had a gearbox oil temperature failure; at the 2000 point position in the figure, the SCADA system detected the abnormality and the unit subsequently stopped operating. Figure 5 (a) is the prediction residual obtained by predicting the gearbox oil temperature using the method of the present invention; the horizontal line in the figure is the alarm threshold calculated using the 3-sigma method; Figure 5 (b) shows the monitoring indicator calculated using KL divergence. The horizontal line represents the alarm threshold, and the vertical line represents the time when the fault was detected, which is approximately 16 hours earlier than the SCADA system alarm time. The figure shows that the KL divergence indicator can clearly reflect the abnormal state of the unit, indicating that the fault monitoring method of the present invention can effectively improve fault monitoring results.

[0091] In addition, in addition to providing a wind turbine transmission chain fault monitoring method based on similarity measurement and trend perception network, the present invention also provides a fault monitoring system, including:

[0092] Perception model building module: used to acquire data, integrate the trend perception algorithm and attention mechanism, and embed it into the long short-term memory network to form a trend perception BiLSTM model with a bidirectional arrangement of network units;

[0093] Prediction module: used to integrate sequence similarity measurement and spatial distribution similarity measurement, screen out highly similar units based on the data, establish a single unit prediction model for each of the highly similar units based on the trend-aware BiLSTM model, and form an integrated prediction model;

[0094] Monitoring module: used to obtain the prediction residual of the monitoring variable through the integrated prediction model, obtain the monitoring index through KL divergence, and realize fault monitoring.

[0095] In specific implementation, the above modules can be implemented as independent entities, or can be arbitrarily combined and implemented as the same or several entities. The specific implementation of the above units can be found in the previous method embodiments and will not be repeated here.

[0096] Figure 6 This is a schematic diagram of the structure of a computer device disclosed in the present invention. Figure 6 As shown, the computer device 400 includes at least a memory 402 and a processor 401; the memory 402 is connected to the processor via a communication bus 403, and is used to store computer instructions executable by the processor 401, and the processor 401 is used to read computer instructions from the memory 402 to implement the steps of the method described in any of the above embodiments.

[0097] For the above-mentioned device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial description of the method embodiments. The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the disclosed solution. A person of ordinary skill in the art can understand and implement it without paying any creative work.

[0098] Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media, and storage devices, including, for example, semiconductor memory devices (e.g., EPROM, EEPROM, and flash memory devices), magnetic disks (e.g., internal or removable), magneto-optical disks, and CD ROM and DVD-ROM disks. The processor and memory can be supplemented by, or incorporated in, special purpose logic circuitry.

[0099] Finally, it should be noted that although this specification contains many specific implementation details, these should not be interpreted as limiting the scope of any invention or the scope of what is claimed, but are primarily intended to describe the features of specific embodiments of a particular invention. Certain features described in multiple embodiments within this specification may also be implemented in combination in a single embodiment. On the other hand, various features described in a single embodiment may also be implemented separately in multiple embodiments or in any suitable sub-combination. In addition, although features may function in certain combinations as described above and may even be initially claimed as such, one or more features from a claimed combination may in some cases be removed from the combination, and a claimed combination may refer to a sub-combination or a variation of a sub-combination.

[0100] Similarly, although operations are depicted in a particular order in the accompanying drawings, this should not be understood as requiring that these operations be performed in the particular order shown or performed sequentially, or that all illustrated operations be performed to achieve the desired results. In some cases, multitasking and parallel processing may be advantageous. In addition, the separation of various system modules and components in the above-described embodiments should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product, or packaged into multiple software products.

[0101] Thus, specific embodiments of the subject matter have been described. Other embodiments are within the scope of the following claims. In some cases, the actions recited in the claims can be performed in a different order and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the particular order shown or sequential sequence to achieve the desired results. In some implementations, multitasking and parallel processing may be advantageous.

[0102] The above description is only a preferred embodiment of the present disclosure and is not intended to limit the present disclosure. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present disclosure should be included in the scope of protection of the present disclosure.

Claims

1. A wind turbine transmission chain fault monitoring method based on similarity measurement and trend perception network, characterized in that: The steps include: Acquire data, integrate the trend perception algorithm and attention mechanism to form a trend perception module, and embed it into the long short-term memory network to form a trend perception BiLSTM model with a bidirectional arrangement of network units; By integrating the sequence similarity metric and the spatial distribution similarity metric, highly similar units are screened out based on the data, and a single unit prediction model is established for each of the highly similar units based on the trend-aware BiLSTM model to form an integrated prediction model; The prediction residual of the monitoring variable is obtained by the integrated prediction model, and the monitoring index is obtained by KL divergence to realize fault monitoring; The trend perception module is embedded in the long short-term memory neural network unit, and the processing process is described as follows: （7） in, represents the output variable of the first layer trend perception network; represents the activation function; Represents the current time input data; is the hidden state at the previous time step; is the hidden state at the current time; Memory information output by the previous network unit; Represents the first layer learning parameters; represents the first-level trend-aware correlation bias; （8） in, represents the output variable of the second layer trend perception network; Represents the second layer learning parameters; represents the second-level trend-aware correlation bias; Indicates parallel connection operation; other parameters are consistent with formula (7); （9） （10） is the hyperbolic tangent activation function; is a high-order trend perception parameter; are high-order basis vectors.

2. The wind turbine transmission chain fault monitoring method according to claim 1, characterized in that: The acquired data were pre-processed to remove outliers using the interval quantile method and then normalized using the Max-min method.

3. The wind turbine transmission chain fault monitoring method according to claim 2, characterized in that: The acquired data is pre-processed to remove outliers using the interval quantile method, and the method of normalizing using the Max-Min method includes the following steps: Collect the environmental parameters, status parameters and operating parameters of each wind turbine to obtain the original data set of the turbine ; k represents the number of parameter variables; Arrange the parameter variable data in descending order of value and divide them into four equal parts according to the total number of samples. The dividing position is recorded as ,in Indicates the 25% percentile value below; Indicates the median value, that is, the value at the 50% position; Represents the 25% quantile value above; the interquartile range IQR represents the concentration of the numerical distribution and is calculated as follows: （1） Further calculation of the numerical upper bound and numerical lower bound : （2） in, is the normal interval coefficient, the value is greater than or less than The samples are identified as abnormal data and removed; The parameter variable Normalize to the range of [0, 1] to make the dimensions of each parameter consistent and obtain a normalized data set ; The calculation process is: （3） in, Represents parameter variables A sample of Represents the normalized sample value; Represents parameter variables The minimum value of Represents parameter variables The maximum value of .

4. The wind turbine transmission chain fault monitoring method according to any one of claims 1 to 2, characterized in that: The trend perception algorithm and attention mechanism are integrated to form the trend perception module, and its operation process is as follows: （4） （5） （6） in, and are the previous and next local spatiotemporal hidden states, respectively; yes The predicted value; Norm is the normalization function; q is the value after normalization; Softmax is the activation function; is the output of BiLSTM; is the learnable parameter of this module; Memory information output by the previous network unit; is the feature vector at the current time point; is the hidden state dimension.

5. The wind turbine transmission chain fault monitoring method according to claim 1, characterized in that: The function of the trend-aware BiLSTM model with bidirectional arrangement of the network units is expressed as follows: （11） （12） in, represents the i-th local spatiotemporal hidden state; Represents model learning parameters; represents the i-1th local spatiotemporal hidden state; It represents the i-th global state in the l-th layer; Indicates the input signal of this module; Represents a spatiotemporal embedding vector; using the extracted local state Update global information ;MSBiLSTM is the function of the model.

6. The wind turbine transmission chain fault monitoring method according to any one of claims 1-2, characterized in that: The method for screening out highly similar units includes the following steps: Calculate the mutual information between the target parameter and the candidate parameter , find the important parameter variables and calculate the similarity weight corresponding to the parameters through formula (13); （13） Where k represents the number of input parameters; Represents the weight corresponding to parameter i; The similarity of the data sequences of the target unit and the candidate units is calculated by formula (14). First, the distance matrix is constructed. , and then calculate the cumulative distance matrix L: （14） L(s,t) represents the cumulative distance between state s and time point t; s and t represent the current index positions of the two time series respectively; Represents the distance between the sth element of time series 1 and the tth element of time series 2; min represents the operation of taking the minimum value; The sequence similarity between two variables is defined as , m represents the variable of the target unit, n represents the variable of the candidate unit; Calculate the multi-dimensional spatial distribution similarity between the target unit and the candidate unit data; The centers of the data sets representing the target unit and the candidate unit are calculated using formula (15): （15） in Represent the average values of parameter variables x and y respectively; Based on the center of the dataset, calculate the covariance matrix : （16） The multidimensional space distribution similarity is calculated as follows: （17） represents the inverse covariance matrix; Finally, the sequence similarity and spatial distribution similarity are combined to calculate the total similarity : （18） Represents the weight corresponding to the parameter variable i; represents the sequence similarity of parameter variable i; Indicates the spatial distribution similarity of parameter variable i.

7. A fault monitoring system using the wind turbine transmission chain fault monitoring method according to any one of claims 1 to 6, characterized in that: include: Perception model building module: used to acquire data, integrate the trend perception algorithm and attention mechanism, and embed it into the long short-term memory network to form a trend perception BiLSTM model with a bidirectional arrangement of network units; Prediction module: used to integrate sequence similarity measurement and spatial distribution similarity measurement, screen out highly similar units based on the data, establish a single unit prediction model for each of the highly similar units based on the trend-aware BiLSTM model, and form an integrated prediction model; Monitoring module: used to obtain the prediction residual of the monitoring variable through the integrated prediction model, obtain the monitoring index through KL divergence, and realize fault monitoring.

8. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the wind turbine transmission chain fault monitoring method according to any one of claims 1 to 6 are executed.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the steps of the wind turbine transmission chain fault monitoring method according to any one of claims 1 to 6 are implemented.

Images