Method for predicting residual life of gearbox bearing of wind turbine generator, medium and electronic equipment

Abstract

The invention discloses a wind turbine generator gearbox bearing residual life prediction method, a medium and electronic equipment. The method comprises the following steps: acquiring vibration signal data in a wind turbine generator gearbox bearing; preprocessing the vibration signal data, and adding a residual life percentage label to obtain preprocessed data; inputting the preprocessed data into a life prediction model for residual life prediction to obtain a residual life prediction result; and smoothing the residual life prediction result to obtain a final residual life prediction result. According to the wind turbine generator gearbox bearing residual life prediction method, the medium and the electronic equipment provided by the invention, life prediction is realized by preprocessing the vibration signal data, making the preprocessed data and sending the preprocessed data into the life prediction model, the vibration signal data is convenient to obtain, and residual life prediction is accurate.

Description

Technical Field

[0001] The invention belongs to the technical field of life prediction of key components of a wind turbine generator set, and in particular relates to a method, medium and electronic equipment for predicting the remaining life of a gearbox bearing of a wind turbine generator set. Background Art

[0002] Wind turbines are a type of power generation equipment that converts wind energy into electrical energy and are the core of wind power generation. The operating status and component life of wind turbines have a significant impact on their power generation efficiency and stable operation of the equipment. Therefore, effective collection and analysis of the operating data of key components of wind turbines, identification of the remaining service life of key components of wind turbines, and predictive maintenance are of great significance to improving the operating efficiency of wind turbines.

[0003] In the prior art, for example, a method for predicting the remaining life of a wind turbine gearbox is disclosed in CN117521519A, which predicts the remaining life of a wind turbine gearbox by constructing a gearbox data set for deep learning. However, in the prior art, there are problems such as complex data preprocessing and too many input data features. Summary of the invention

[0004] The purpose of the present invention is to provide a method, medium and electronic equipment for predicting the remaining life of a wind turbine gearbox bearing, so as to solve the problems in the prior art of complex data preprocessing and more input data features.

[0005] To this end, the present invention provides a method, medium and electronic equipment for predicting the remaining life of a wind turbine gearbox bearing, comprising the following steps:

[0006] Obtain vibration signal data in the gearbox bearing of the wind turbine;

[0007] Preprocess the vibration signal data, add a remaining life percentage label, and obtain preprocessed data;

[0008] Inputting the preprocessed data into the life prediction model to perform remaining life prediction and obtain remaining life prediction results;

[0009] The remaining life prediction results are smoothed to obtain the final remaining life prediction results.

[0010] Preferably, the vibration signal data at least includes vibration signal data in the horizontal and vertical directions, and is a complete life cycle.

[0011] Preferably, a remaining life percentage label is added to the vibration signal data.

[0012] Preferably, preprocessing the vibration signal data includes: ignoring the remaining life percentage label added to the vibration signal data, and performing fast Fourier transform on the vibration signal data in the horizontal and vertical directions separately.

[0013] Preferably, the adding of the remaining life percentage label includes: setting the remaining life percentage of the first group of data to 1, setting the remaining life percentage of the last group of data to 0, and gradually decreasing the data in between.

[0014] Preferably, the life prediction model is a bidirectional long short-term memory network-SE module model, which consists of five parts, namely, a data input layer, a bidirectional long short-term memory network model layer, an SE module layer, a fully connected layer and a result output layer.

[0015] Preferably, the preprocessed data is input into a life prediction model, with horizontal and vertical vibration signal data as model input and life percentage labels as prediction output targets.

[0016] Preferably, smoothing the remaining life prediction result includes taking the first value of the prediction result as the starting value of the exponentially weighted moving average, setting the range of the smoothed result to between [0, 1], and taking the smoothed result as the final output.

[0017] A computer-readable storage medium stores a computer program for realizing wind turbine fault prediction based on a neural network, wherein the computer program enables a computer to execute the wind turbine gearbox bearing remaining life prediction method based on a neural network.

[0018] An electronic device comprises at least one processor; and a memory connected to the processor in communication; wherein the memory stores instructions executable by the processor, the instructions are executed by the at least one processor, and the at least one processor is capable of executing the neural network-based method for predicting the remaining life of a wind turbine gearbox bearing.

[0019] Beneficial effects:

[0020] 1. The present invention provides a method, medium and electronic equipment for predicting the remaining life of a wind turbine gearbox bearing. The method preprocesses vibration signal data, produces preprocessed data, and sends the data to a life prediction model to realize life prediction. The vibration signal data is easy to obtain and the remaining life prediction is accurate.

[0021] 2. In the prior art, a large amount of input data with different features is usually required for model training. Data acquisition is difficult, the amount of feature data is large, it is difficult to use in practice, and the steps for preprocessing the data are complicated. In the present invention, the horizontal and vertical vibration signal data in the gearbox bearing of a wind turbine with a complete life cycle are used as the input features of the life prediction model. The number of features is small, and data acquisition is convenient. The vibration signal data is preprocessed by fast Fourier transform to simplify the data preprocessing steps. At the same time, the bidirectional long short-term memory network-SE module model is used as the life prediction model to ensure the accuracy of the remaining life prediction of the wind turbine gearbox bearing. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0023] Figure 1 The present invention is a flowchart of a method for predicting the remaining life of a wind turbine gearbox bearing. DETAILED DESCRIPTION

[0024] The content of the present invention can be more easily understood by selecting the following detailed description of the preferred implementation method of the present invention and the embodiments included. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those of ordinary skill in the art to which the present invention belongs. When there is a conflict, the definition in this specification shall prevail.

[0025] Embodiment 1:

[0026] like Figure 1 As shown, this embodiment provides a method, medium and electronic device for predicting the remaining life of a wind turbine gearbox bearing, including the following steps:

[0027] S1, obtaining vibration signal data in the gearbox bearing of the wind turbine;

[0028] The vibration signal data at least includes vibration signal data in the horizontal and vertical directions, and is a complete life cycle.

[0029] The vibration signal data in the horizontal and vertical directions are sorted and merged, and arranged in order from the size of the remaining life to the smallest, and the remaining life percentage label is added to the vibration signal data. Among them, adding the remaining life percentage label includes: the remaining life percentage of the first group of data is 1, the remaining life percentage of the last group of data is 0, and the data in the middle gradually decreases in sequence, and the life percentage data is expressed in the form of corresponding decimals.

[0030] S2, preprocessing the vibration signal data, adding a remaining life percentage label, and obtaining preprocessed data;

[0031] The preprocessing of the vibration signal data includes: ignoring the remaining life percentage label added to the vibration signal data in S1, and performing fast Fourier transform on the vibration signal data in the horizontal and vertical directions respectively.

[0032] The pre-processed vibration signal data is sorted and merged, arranged in order from the size of the remaining life to the smallest, and the remaining life percentage label is added. Among them, adding the remaining life percentage label includes: the remaining life percentage of the first group of data is 1, the remaining life percentage of the last group of data is 0, and the data in the middle gradually decreases in sequence, and the life percentage data is expressed in the form of corresponding decimals.

[0033] S3, inputting the preprocessed data into the life prediction model to perform remaining life prediction, and obtaining a remaining life prediction result;

[0034] The life prediction model is a bidirectional long short-term memory network-SE module model. The bidirectional long short-term memory network-SE module model consists of five parts, namely the data input layer, the bidirectional long short-term memory network model, the SE module, the fully connected layer and the result output layer. After the input data enters the data input layer, it is processed by the bidirectional long short-term memory network. Then, the SE module further distributes the weights of the data processed by the bidirectional long short-term memory network to enhance the data features. Finally, the final result is obtained through the processing of the fully connected layer and output from the result output layer.

[0035] The preprocessed data is divided into a training set and a test set. The training set is used to train the life prediction model, and the test set is used to test the life prediction. The training set is input into the life prediction model, with the horizontal and vertical vibration signal data as the model input, and the life percentage label as the prediction output target for training to obtain the life training model. The test set is input into the trained life prediction model, with the horizontal and vertical vibration signal data as the model input, and the remaining life prediction result is obtained.

[0036] S4. Smoothing the remaining life prediction result to obtain the final remaining life prediction result.

[0037] The smoothing process of the remaining life prediction result includes taking the first value of the prediction result as the starting value of the exponentially weighted moving average, setting the range of the smoothed result to between [0,1], and taking the smoothed result as the final output.

[0038] Embodiment 2:

[0039] A computer-readable storage medium stores a computer program for realizing wind turbine fault prediction based on a neural network, wherein the computer program enables a computer to execute the wind turbine gearbox bearing remaining life prediction method based on a neural network.

[0040] Embodiment 3:

[0041] An electronic device comprises at least one processor; and a memory connected to the processor in communication; wherein the memory stores instructions executable by the processor, the instructions are executed by the at least one processor, and the at least one processor is capable of executing the neural network-based method for predicting the remaining life of a wind turbine gearbox bearing.

[0042] The embodiments described above are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for predicting the remaining life of a wind turbine gearbox bearing, characterized in that: The steps include: Obtain vibration signal data in the gearbox bearing of the wind turbine; Preprocess the vibration signal data, add a remaining life percentage label, and obtain preprocessed data; Inputting the preprocessed data into the life prediction model to perform remaining life prediction and obtain remaining life prediction results; The remaining life prediction results are smoothed to obtain the final remaining life prediction results.

2. A method for predicting the remaining life of a wind turbine gearbox bearing according to claim 1, characterized in that: The vibration signal data at least includes vibration signal data in the horizontal and vertical directions and is a complete life cycle.

3. A method for predicting the remaining life of a wind turbine gearbox bearing according to claim 2, characterized in that: A remaining life percentage label is added to the vibration signal data.

4. A method for predicting the remaining life of a wind turbine gearbox bearing according to claim 3, characterized in that: The preprocessing of the vibration signal data includes: ignoring the remaining life percentage label added in the vibration signal data, and performing fast Fourier transform on the vibration signal data in the horizontal and vertical directions respectively.

5. A method for predicting the remaining life of a wind turbine gearbox bearing according to claim 4, characterized in that: The adding of the remaining life percentage label includes: setting the remaining life percentage of the first group of data to 1, setting the remaining life percentage of the last group of data to 0, and gradually decreasing the data in between.

6. A method for predicting the remaining life of a wind turbine gearbox bearing according to claim 1, characterized in that: The life prediction model is a bidirectional long short-term memory network-SE module model, which consists of five parts, namely, a data input layer, a bidirectional long short-term memory network model layer, an SE module layer, a fully connected layer and a result output layer.

7. A method for predicting the remaining life of a wind turbine gearbox bearing according to claim 6, characterized in that: The preprocessed data is input into the life prediction model, with the horizontal and vertical vibration signal data as the model input and the life percentage label as the prediction output target.

8. A method for predicting the remaining life of a wind turbine gearbox bearing according to claim 1, characterized in that: The smoothing process of the remaining life prediction result includes taking the first value of the prediction result as the starting value of the exponentially weighted moving average, setting the range of the smoothed result to between [0,1], and taking the smoothed result as the final output.

9. A computer-readable storage medium, characterized in that: It stores a computer program for realizing wind turbine fault prediction based on a neural network, and the computer program enables a computer to execute the method for predicting the remaining life of a wind turbine gearbox bearing based on a neural network as described in any one of claims 1-8.

10. An electronic device, characterized in that: It comprises at least one processor; and a memory connected to the processor in communication; wherein the memory stores instructions executable by the processor, the instructions are executed by the at least one processor, and the at least one processor is capable of executing the wind turbine gearbox bearing remaining life prediction method based on a neural network as described in any one of claims 1-8.

Images