Bearing vibration detection method and system for wind power gearbox

Non-contact bearing detection is carried out through pulsed laser rangefinder, and spectrum analysis is used to realize online vibration monitoring and fault warning of wind power gearbox bearings, solving the problem of low detection efficiency and accuracy, and improving the operation and maintenance efficiency of wind power units.

CN118329448BActive Publication Date: 2024-11-19江苏广大鑫盛精密智造有限公司
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
CN202410741048.1
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-06-11
Publication Date
2024-11-19
Estimated Expiration
2044-06-11

AI Technical Summary

Technical Problem

The vibration detection efficiency and accuracy of wind power gearbox bearings are low, resulting in challenges in the operation and maintenance of wind power units.

Method used

The pulsed laser rangefinder is used for non-contact detection, and the bearing distance is calculated by transmitting and receiving laser pulses, the distance timing is generated, and spectrum analysis is used to achieve bearing vibration detection and fault warning.

Benefits of technology

It realizes the online full-time vibration detection of wind power gearbox bearings, improves detection efficiency and accuracy, and supports real-time fault monitoring and early warning of wind power units.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118329448B_ABST
    Figure CN118329448B_ABST
Patent Text Reader

Abstract

The present invention discloses a bearing vibration detection method and system for a wind turbine gearbox, which belongs to the field of intelligent detection, wherein the method includes: sending a first pulse laser beam to a target bearing through a pulse laser rangefinder, and receiving a reflected laser beam of the target bearing, recorded as a second pulse laser beam; reading the first duration data from the timer sending the first pulse laser beam to the receiving head receiving the second pulse laser beam; generating a target distance value time series for the target bearing based on the first duration data and a predetermined pulse laser beam speed; converting the processed target distance value time series into a target distance spectrum diagram; analyzing the multi-dimensional features of the target distance spectrum diagram to obtain a target vibration detection result for the target bearing. The present application solves the technical problem of low efficiency and accuracy of wind turbine gearbox bearing vibration detection in the prior art, and achieves the technical effect of realizing online full-time vibration detection of wind turbine gearbox bearings and improving detection efficiency and accuracy.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of intelligent detection, and in particular to a bearing vibration detection method and system for a wind power gearbox. Background Art

[0002] With the rapid growth of wind power installed capacity, the operation and maintenance of wind turbines are facing new challenges. Among them, wind turbine gearboxes are one of the key parts in the unit that is most prone to failure. During long-term operation, wind turbine gearboxes will produce problems such as gear meshing and bearing wear, which directly affect the safe operation of wind turbines. Traditional wind turbine gearbox status monitoring mostly uses contact sensors for detection and offline analysis, which is limited in installation location, resulting in low detection efficiency and accuracy. Summary of the invention

[0003] The present application aims to solve the technical problems of low efficiency and accuracy in wind turbine gearbox bearing vibration detection in the prior art by providing a bearing vibration detection method and system for a wind turbine gearbox.

[0004] In view of the above problems, the present application provides a bearing vibration detection method and system for a wind power gearbox.

[0005] The first aspect disclosed in the present application provides a bearing vibration detection method for a wind turbine gearbox, the method comprising: emitting a first pulse laser beam to a target bearing through a transmitter head of a pulse laser rangefinder; receiving a reflected laser beam of the target bearing to the first pulse laser beam through a receiver head of the pulse laser rangefinder, recorded as a second pulse laser beam; reading first duration data of a timer, the first duration data referring to the duration from when the transmitter emits the first pulse laser beam to when the receiver head receives the second pulse laser beam; calculating a first distance value based on the first duration data and a predetermined pulse laser beam speed to generate a target distance value timing sequence for the target bearing; converting the target distance value timing sequence processed by a data conditioner into a target distance spectrum diagram by using the principle of spectrum analysis; and analyzing the multi-dimensional features of the target distance spectrum diagram to obtain a target vibration detection result for the target bearing.

[0006] Another aspect disclosed in the present application provides a bearing vibration detection system for a wind turbine gearbox, the system comprising: a pulse laser beam emitting module, for emitting a first pulse laser beam to a target bearing through a transmitting head of a pulsed laser rangefinder; a reflected laser beam receiving module, for receiving a reflected laser beam of the target bearing to the first pulse laser beam through a receiving head of the pulsed laser rangefinder, recorded as a second pulse laser beam; a duration data reading module, for reading first duration data of a timer, the first duration data referring to the duration from when the transmitter emits the first pulse laser beam to when the receiving head receives the second pulse laser beam; a target distance value timing module, for calculating a first distance value based on the first duration data and a predetermined pulse laser beam speed to generate a target distance value timing of a target bearing; a target distance spectrum diagram module, for converting the target distance value timing processed by a data conditioner into a target distance spectrum diagram using the principle of spectrum analysis; and a vibration detection result module, for analyzing the multi-dimensional features of the target distance spectrum diagram to obtain a target vibration detection result of the target bearing.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] A controlled pulse laser rangefinder is used to periodically send laser pulses to the target bearing and receive the light signal reflected by the bearing. The real-time distance between the bearing and the rangefinder is calculated by recording the emission and reception time of the laser pulses. The real-time distance data of the bearing is collected, and a target distance value time series is generated, including bearing vibration information, to avoid interference with the operation of the unit. The target distance value time series is processed using spectrum analysis to obtain a target vibration spectrum diagram of the bearing. Different types of bearing faults are detected based on the spectrum characteristics, and a technical solution for online monitoring and fault warning of wind turbine gearbox bearing vibration is realized, which solves the technical problem of low efficiency and accuracy of wind turbine gearbox bearing vibration detection in the prior art, and achieves the technical effect of realizing online full-time vibration detection of wind turbine gearbox bearings and improving detection efficiency and accuracy.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 A flow chart of a bearing vibration detection method for a wind power gearbox is provided for an embodiment of the present application;

[0011] Figure 2 A flow chart of obtaining a target vibration detection result in a bearing vibration detection method for a wind power gearbox is provided for an embodiment of the present application;

[0012] Figure 3 A structural schematic diagram of a bearing vibration detection system for a wind turbine gearbox is provided for an embodiment of the present application.

[0013] Explanation of the reference numerals: pulse laser beam emitting module 11 , reflected laser beam receiving module 12 , duration data reading module 13 , target distance value timing module 14 , target distance spectrum diagram module 15 , vibration detection result module 16 . DETAILED DESCRIPTION

[0014] The overall idea of ​​the technical solution provided by this application is as follows:

[0015] The embodiments of the present application provide a bearing vibration detection method and system for a wind turbine gearbox. First, a detection connection is established between a pulse laser rangefinder and a wind turbine gearbox bearing to achieve non-contact detection of the target bearing. Then, the transmitting and receiving system of the laser rangefinder is used to periodically send laser pulses to the bearing and receive the reflected signal. Based on the time-of-flight principle of laser, the distance between the bearing and the rangefinder is calculated in real time by recording the time difference between the emission and reception. The micro-displacement information of the bearing is collected to obtain the time series data of the target distance change. Next, the obtained bearing distance time series is subjected to spectral analysis to analyze the characteristics of the distance time series at different frequencies. According to the bearing spectrum characteristics, the multi-dimensional features reflecting the working state of the bearing are extracted from the spectrum diagram to obtain the target vibration detection result. Thereby, online bearing status monitoring and fault warning are achieved.

[0016] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically described below in conjunction with the drawings in the specification. Example 1

[0017] like Figure 1 As shown, an embodiment of the present application provides a bearing vibration detection method for a wind turbine gearbox, which is applied to a bearing vibration detection system for a wind turbine gearbox. The system is communicatively connected to a pulsed laser rangefinder, and the pulsed laser rangefinder is fixed at a predetermined position above a target bearing of the wind turbine gearbox.

[0018] Specifically, an embodiment of the present application discloses a bearing vibration detection method for a wind turbine gearbox, which is applied to a bearing vibration detection system for a wind turbine gearbox. The system is communicatively connected to a pulsed laser rangefinder, wherein the pulsed laser rangefinder is a measuring device that can emit a short-time pulsed laser and measure the flight time of the pulsed light beam, and realizes data interaction connection with the bearing vibration detection system through a standard digital communication interface.

[0019] In order to realize the vibration detection of the target bearing in the wind turbine gearbox, the pulsed laser rangefinder is fixed at a predetermined position at a certain height above the target bearing. The position should ensure that the transmitter head of the rangefinder can irradiate the bearing surface and the receiver head can receive the light beam reflected by the bearing.

[0020] Through the communication connection with the pulse laser rangefinder and the determined layout of the rangefinder, long-distance, non-contact vibration detection of the target bearing of the wind turbine gearbox can be achieved, which effectively solves the problem that the internal space of the wind turbine gearbox is limited and the traditional contact vibration sensor cannot be arranged.

[0021] Bearing vibration detection methods include:

[0022] Sending a first pulse laser beam to the target bearing through the transmitting head of the pulse laser rangefinder;

[0023] In the implementation of the present application, after the pulsed laser rangefinder is fixed at a predetermined position above the target bearing, the bearing vibration detection process begins. The pulsed laser rangefinder includes a transmitter head and a receiver head. The laser transmitter head of the pulsed laser rangefinder emits a first pulsed laser beam to the target bearing below at a certain emission frequency, and the pulsed laser beam irradiates the surface position of the target bearing. The first pulsed laser beam refers to a laser pulse signal.

[0024] receiving, by means of a receiving head of the pulsed laser rangefinder, a laser beam reflected by the target bearing from the first pulsed laser beam, which is recorded as a second pulsed laser beam;

[0025] In the embodiment of the present application, after the first pulse laser beam emitted by the transmitter of the pulse laser rangefinder irradiates the target bearing surface, the target bearing surface will be reflected, and the reflected part of the first pulse beam enters the receiving head of the pulse laser rangefinder. The receiving head of the pulse laser rangefinder receives the part of the first pulse laser beam reflected back from the bearing surface and records it as the second pulse laser beam, providing a basis for monitoring the micro displacement and vibration of the bearing.

[0026] Reading first duration data of the timer, where the first duration data refers to the duration from when the transmitter emits the first pulse laser beam to when the receiver receives the second pulse laser beam;

[0027] In the embodiment of the present application, after the transmitter emits the first pulse laser beam and the receiver receives the second pulse laser beam, the time interval between the two is measured, that is, the flight time of the first pulse laser beam. To this end, a high-precision timer is provided in the pulse laser rangefinder to record the emission and reception time data of the pulse laser beam. The time data recorded in the timer is read to obtain the accurate time interval between the emission of the first pulse laser beam and the reception of the second pulse laser beam, and the time interval is defined as the first duration data, which provides a basis for calculating the bearing displacement and performing vibration detection.

[0028] A first distance value is calculated based on the first duration data and a predetermined pulse laser beam speed to generate a target distance value time sequence for the target bearing;

[0029] In the embodiment of the present application, after obtaining the first duration data, the propagation distance of the first pulse laser beam is calculated, that is, the instantaneous distance value between the bearing surface and the pulse laser rangefinder. The first distance value is calculated by multiplying the first duration data by the predetermined pulse laser beam speed and dividing by 2. In the process of continuous multiple pulse laser beam ranging, each first distance value is recorded to obtain the time series of the distance value of the target bearing surface relative to the pulse laser rangefinder over time, that is, the target distance value time series, which reflects the micro displacement information of the bearing surface during the movement process, and is the key data for realizing bearing vibration detection.

[0030] The target distance value time series after being processed by the data conditioner is converted into a target distance spectrum diagram by using the spectrum analysis principle;

[0031] In an embodiment of the present application, after obtaining the target distance value time series of the target bearing, first, the target distance value time series is preprocessed, including filtering, smoothing and other means to remove noise to improve data quality. Then, a threshold judgment condition is set to scan the target distance value time series conversion, and abnormal data points that obviously deviate from the normal range are eliminated to obtain the processed target distance value time series, thereby ensuring that the data of the target distance value time series is true and reliable. Subsequently, the processed distance time series signal is subjected to a fast Fourier transform to complete the conversion of the time domain distance signal to the frequency domain distance spectrum, and obtain a distance fluctuation spectrum diagram of the target bearing within a certain frequency range, that is, a target distance spectrum diagram, which is used for subsequent bearing fault detection and diagnosis.

[0032] The multi-dimensional features of the target distance spectrum diagram are analyzed to obtain the target vibration detection result of the target bearing.

[0033] In the embodiment of the present application, after obtaining the distance spectrum of the target bearing, the multi-dimensional features of the target distance spectrum are extracted, including the frequency, amplitude, phase and other parameters of each main component of the spectrum, which reflect the vibration response characteristics of the bearing at different frequencies during motion. Among them, the enhancement of components in certain specific frequency ranges is often related to the typical failure mode of the bearing.

[0034] Subsequently, matching analysis is performed based on the bearing fault spectrum feature data stored in the experience knowledge base and the extracted target spectrum multi-dimensional features. When the target spectrum graph has characteristic frequencies related to the fault, it can be detected that the bearing has the corresponding type of fault, such as inner and outer ring faults, rolling element faults, etc., and finally the target vibration detection result of the target bearing is obtained.

[0035] Furthermore, the embodiment of the present application also includes:

[0036] Before analyzing the multi-dimensional features of the target distance spectrum graph, the target distance spectrum graph corresponding to the preset duration value determined by analyzing the average preheating duration of the wind turbine gearbox is deleted.

[0037] In an embodiment of the present application, during the bearing vibration detection process, the wind turbine gearbox needs to go through a preheating process from startup to stable operation. During this process, the target distance spectrum diagram detected by the pulsed laser rangefinder is the preheating state of the wind turbine gearbox. To this end, the average preheating time of this model of wind turbine gearbox is determined through multiple test analyses and determined as a preset time value. Then, after analyzing the multi-dimensional features of the target distance spectrum diagram, the target distance spectrum diagram obtained during the period corresponding to the preset time value is first deleted, effectively removing the unstable spectrum results that may appear in the target bearing during the preheating stage, ensuring the reliability of the distance spectrum data entering the spectrum analysis stage, and improving the reliability and accuracy of bearing vibration detection.

[0038] Furthermore, the embodiment of the present application also includes:

[0039] After the first duration data of the reading timer is read, a detection loss function is obtained, and the expression of the detection loss function is as follows:

[0040] ;

[0041] in, refers to the detection loss function, It refers to the actual first duration data when the pulse laser rangefinder emits a pulse laser beam to the target bearing for the i-th time, refers to the first duration data recorded by the timer when the pulse laser rangefinder emits a pulse laser beam to the target bearing for the i-th time, n refers to the number of times the pulse laser rangefinder detects the distance to the target bearing, is the feedback adjustment coefficient, and , refers to the timing loss adjustment coefficient for emitting the first pulse laser beam, is the timing loss adjustment coefficient for receiving the second pulse laser beam, refers to the timing accuracy loss adjustment coefficient of the timer;

[0042] A loss analysis is performed on the first duration data according to the detection loss function, and the first duration data is corrected according to the analysis result.

[0043] In a preferred embodiment, after reading the first duration data of the timer, the accuracy of the duration data is evaluated by constructing a detection loss function. First, the expression for constructing the detection loss function is: , is the detection loss function, It refers to the actual first duration data when the pulse laser rangefinder emits the pulse beam for the i-th time. is the first duration data recorded by the timer during the i-th distance measurement, n is the number of distance measurements, is the feedback adjustment coefficient, and , is the timing loss adjustment factor for emitting the first pulse laser beam, is the timing loss adjustment factor for receiving the second pulse laser beam, Refers to the adjustment coefficient of the timer's timing accuracy loss.

[0044] First duration data recorded in the completed read timer Then, substitute it into the loss function Calculate and obtain the loss function analysis result, and then use the corresponding method to correct the first duration data to further improve the accuracy of vibration detection. For example, when the error mainly comes from the transmitter, adjust Parameters, perform transmitter calibration; if the error mainly comes from the receiving head, adjust Parameters, receive head calibration, etc.

[0045] Furthermore, the embodiment of the present application also includes:

[0046] Extracting the target distance value in the target distance value time series by the data conditioner;

[0047] An abnormal support function is introduced to determine whether the target distance value is abnormal data. The abnormal support function is expressed as follows:

[0048] ;

[0049] in, Refers to the adjacent distance value of the target distance value The distance value to the target The first abnormal support of refers to the distance data adjacent to the target distance value in the target distance value time series, k refers to the amplitude of the abnormal support function, and , is the attenuation factor, and ;

[0050] If the first abnormal support reaches a predetermined threshold, the target distance value is determined to be abnormal data, and the target distance value is removed from the target distance value time series.

[0051] In a preferred embodiment, before converting the target distance value time series into a target distance spectrum, first, each target distance value in the target distance value time series is scanned by a data conditioner, and each target distance value in the time series is detected and extracted, that is, the instantaneous distance data of the bearing at each sampling time point. The data conditioner is a signal processing device used to purify and improve data quality. Secondly, in order to realize outlier detection, an abnormal support function is introduced , judge whether each target distance value is normal to eliminate abnormal values; for the target distance value , calculate its adjacent distance value through the abnormal support function The first abnormal support of , realizes abnormal detection and screening of target distance values. Distance value to adjacent If there is a significant difference, the function value will be smaller; if Distance value to adjacent The closer, the larger the function value. Where k refers to the amplitude of the anomaly support function, which is set according to the accuracy requirement, and k∈[0,1], e is the attenuation factor, and e≥0;

[0052] After obtaining the first abnormal support of each target distance value, the first abnormal support of each target distance value is compared with the predetermined threshold value according to the preset first abnormal support. If the first abnormal support of a certain target distance value is lower than the predetermined threshold, the distance value is determined to be an abnormal value. Correspondingly, the target distance value determined to be abnormal is removed from the obtained complete target distance value sequence, ensuring that the target distance value sequence entering the subsequent spectrum analysis is within the normal range, thereby improving the reliability of the result analysis.

[0053] Further, such as Figure 2 As shown, the embodiment of the present application also includes:

[0054] Respectively extracting a first distance value corresponding to a first time and a second distance value corresponding to a second time in the target distance spectrum graph, wherein the first time and the second time have a sequential order;

[0055] generating a target vibration curve based on a calculated first distance difference between the first distance value and the second distance value;

[0056] The target vibration detection result is obtained by analyzing the target distance spectrum diagram and the target vibration curve.

[0057] In a preferred embodiment, the target distance spectrum diagram contains the distance value information of the target bearing at different sampling times. After obtaining the target distance spectrum diagram, the distance parameter corresponding to a specific moment is extracted from the target distance spectrum diagram. First, a first time is selected, and the first distance value corresponding to the moment is retrieved from the target distance spectrum diagram. At the same time, a second time after the first time is selected, and the corresponding second distance value is also extracted from the target distance spectrum diagram. Then, the difference between the first distance value and the second distance value is calculated to obtain the first distance difference, and the distance values ​​at different moments in the spectrum diagram are continued to be selected to calculate more distance differences. These time-sequential distance differences are connected to form a distance change curve of the target bearing, that is, a target vibration curve, which reflects the change in the distance of the bearing during the vibration process.

[0058] Subsequently, the obtained target distance spectrum diagram and target vibration curve are combined for comprehensive analysis to obtain the target vibration detection result of the target bearing. Among them, the target distance spectrum diagram reflects the information of different frequency components, and the target vibration curve reflects the distance change in the time domain. For example, the spectrum diagram shows that a certain characteristic frequency is abnormal, and the continuous oscillation of the vibration curve can further verify the vibration problem at this frequency. Finally, through the combined analysis of the target distance spectrum diagram and the target vibration curve, the fault type and severity of the bearing are detected, the target vibration detection result is obtained, and the accuracy of bearing vibration detection is improved.

[0059] Furthermore, the embodiment of the present application also includes:

[0060] Collecting a target spectrum feature set of the target distance spectrum graph based on a predetermined feature indicator set, wherein the predetermined feature indicator set includes a center frequency, a frequency bandwidth, a peak frequency, and a peak amplitude;

[0061] Obtaining a target slope of the target vibration curve;

[0062] The target vibration detection result is determined according to the target frequency spectrum feature set and the target slope.

[0063] In a preferred embodiment, after obtaining the target distance spectrum graph, it is necessary to extract characteristic parameters in the spectrum graph. To this end, a set of indicators for describing the spectrum characteristics is pre-set, that is, a predetermined characteristic indicator set, including center frequency, frequency bandwidth, peak frequency and peak amplitude. According to the predetermined characteristic indicator set, the corresponding characteristic values ​​are detected and extracted from the target distance spectrum graph to form a target spectrum characteristic set, which comprehensively reflects the response characteristics of the bearing at different frequencies. Subsequently, the changing trend of the target vibration curve is calculated, the slope characteristic parameters of the target vibration curve are obtained, and the representative slope values ​​in different time periods are recorded to form the target slope of the target vibration curve. Afterwards, the target distance spectrum graph and the target vibration curve are used as the target vibration detection result.

[0064] Furthermore, the embodiment of the present application also includes:

[0065] Construct a fault prediction model based on the principle of neural network;

[0066] Analyzing the data in the target vibration detection result by using the fault prediction model to obtain a target predicted fault;

[0067] Predictive maintenance is performed on the target predicted fault.

[0068] In a preferred embodiment, first, characteristic data in the bearing vibration detection results are collected, including the spectral characteristics of the target distance spectrum diagram, such as the center frequency, frequency bandwidth, etc., and the slope characteristics of the target vibration curve, to form the input layer of the fault prediction model, and the output layer is the specific fault content. The network model is trained using an annotated data set, wherein the annotated data set contains the vibration signals and corresponding labels under various states of the bearing. Back propagation is used to continuously update the network weights and thresholds so that the output of the fault prediction model is close to the actual label. When the model error converges, the model parameters are saved to obtain a fault prediction model.

[0069] Subsequently, the spectral features of the target distance spectrum graph and the slope features of the target vibration curve are input into the trained fault prediction model. The model analyzes the data and obtains the corresponding model output, i.e., the prediction result of the bearing state, to predict the target fault. Afterwards, the severity and scope of the fault are evaluated according to the type, location, and probability of the target predicted fault, so as to determine the urgency of maintenance and plan the maintenance time. A suitable maintenance plan is formulated, including replacing or repairing the faulty parts, adjusting relevant parameters, etc., to guide maintenance personnel to accurately find the fault location and replace the faulty parts to achieve predictive maintenance.

[0070] In summary, the bearing vibration detection method for a wind turbine gearbox provided in the embodiment of the present application has the following technical effects:

[0071] The first pulse laser beam is sent to the target bearing through the transmitter of the pulse laser rangefinder; the reflected laser beam of the target bearing to the first pulse laser beam is received by the receiver of the pulse laser rangefinder, which is recorded as the second pulse laser beam, providing a basis for obtaining the first duration data and the target distance value sequence, and realizing non-contact detection. The first duration data of the timer is read, and the first duration data refers to the time from the transmitter sending the first pulse laser beam to the receiver receiving the second pulse laser beam; the first distance value is calculated based on the first duration data and the predetermined pulse laser beam speed to generate the target distance value sequence of the target bearing, and the distance change time series caused by bearing vibration is generated. The target distance value sequence processed by the data conditioner is converted into a target distance spectrum diagram using the spectrum analysis principle to provide support for obtaining the target vibration detection result. The multi-dimensional characteristics of the target distance spectrum diagram are analyzed to obtain the target vibration detection result of the target bearing, realize the online monitoring and fault prediction of the wind turbine gearbox bearing vibration, and improve the detection efficiency and accuracy. Example 2

[0072] Based on the same inventive concept as the bearing vibration detection method for a wind turbine gearbox in the aforementioned embodiment, Figure 3 As shown, an embodiment of the present application provides a bearing vibration detection system for a wind turbine gearbox, the system is communicatively connected with a pulsed laser rangefinder, and the pulsed laser rangefinder is fixed at a predetermined position above a target bearing of the wind turbine gearbox, and the system includes:

[0073] A pulse laser beam emitting module 11, used for emitting a first pulse laser beam to the target bearing through a emitting head of the pulse laser rangefinder;

[0074] A reflected laser beam receiving module 12 is used to receive the reflected laser beam of the target bearing to the first pulse laser beam through the receiving head of the pulse laser rangefinder, which is recorded as a second pulse laser beam;

[0075] The duration data reading module 13 is used to read the first duration data of the timer, where the first duration data refers to the duration from when the transmitter emits the first pulse laser beam to when the receiver receives the second pulse laser beam;

[0076] A target distance value timing module 14 is used to calculate a first distance value based on the first duration data and a predetermined pulse laser beam speed to generate a target distance value timing of the target bearing;

[0077] A target distance spectrum diagram module 15 is used to convert the target distance value time series processed by the data conditioner into a target distance spectrum diagram by using the spectrum analysis principle;

[0078] The vibration detection result module 16 is used to analyze the multi-dimensional features of the target distance spectrum diagram to obtain the target vibration detection result of the target bearing.

[0079] Further, the vibration detection result module 16 includes the following execution steps:

[0080] Before analyzing the multi-dimensional features of the target distance spectrum graph, the target distance spectrum graph corresponding to the preset duration value determined by analyzing the average preheating duration of the wind turbine gearbox is deleted.

[0081] Furthermore, the duration data reading module 13 includes the following execution steps:

[0082] Get the detection loss function, the expression of the detection loss function is as follows:

[0083] ;

[0084] in, refers to the detection loss function, It refers to the actual first duration data when the pulse laser rangefinder emits a pulse laser beam to the target bearing for the i-th time, refers to the first duration data recorded by the timer when the pulse laser rangefinder emits a pulse laser beam to the target bearing for the i-th time, n refers to the number of times the pulse laser rangefinder detects the distance to the target bearing, is the feedback adjustment coefficient, and , refers to the timing loss adjustment coefficient for emitting the first pulse laser beam, is the timing loss adjustment coefficient for receiving the second pulse laser beam, refers to the timing accuracy loss adjustment coefficient of the timer;

[0085] A loss analysis is performed on the first duration data according to the detection loss function, and the first duration data is corrected according to the analysis result.

[0086] Furthermore, the target distance spectrum diagram module 15 includes the following execution steps:

[0087] Extracting the target distance value in the target distance value time series by the data conditioner;

[0088] An abnormal support function is introduced to determine whether the target distance value is abnormal data. The abnormal support function is expressed as follows:

[0089] ;

[0090] in, Refers to the adjacent distance value of the target distance value The distance value to the target The first abnormal support of refers to the distance data adjacent to the target distance value in the target distance value time series, k refers to the amplitude of the abnormal support function, and , is the attenuation factor, and ;

[0091] If the first abnormal support reaches a predetermined threshold, the target distance value is determined to be abnormal data, and the target distance value is removed from the target distance value time series.

[0092] Furthermore, the vibration detection result module 16 further includes the following execution steps:

[0093] Respectively extracting a first distance value corresponding to a first time and a second distance value corresponding to a second time in the target distance spectrum graph, wherein the first time and the second time have a sequential order;

[0094] generating a target vibration curve based on a calculated first distance difference between the first distance value and the second distance value;

[0095] The target vibration detection result is obtained by analyzing the target distance spectrum diagram and the target vibration curve.

[0096] Furthermore, the vibration detection result module 16 further includes the following execution steps:

[0097] Collecting a target spectrum feature set of the target distance spectrum graph based on a predetermined feature indicator set, wherein the predetermined feature indicator set includes a center frequency, a frequency bandwidth, a peak frequency, and a peak amplitude;

[0098] Obtaining a target slope of the target vibration curve;

[0099] The target vibration detection result is determined according to the target frequency spectrum feature set and the target slope.

[0100] Furthermore, the vibration detection result module 16 further includes the following execution steps:

[0101] Construct a fault prediction model based on the principle of neural network;

[0102] Analyzing the data in the target vibration detection result by using the fault prediction model to obtain a target predicted fault;

[0103] Predictive maintenance is performed on the target predicted fault.

[0104] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.

[0105] Furthermore, the first or second mentioned above may not only represent an order relationship, but may also represent a specific concept, and / or refer to multiple elements that can be selected individually or in full. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A bearing vibration detection method for a wind turbine gearbox, characterized in that: The method is applied to a bearing vibration detection system for a wind turbine gearbox, the system is communicatively connected with a pulsed laser rangefinder, and the pulsed laser rangefinder is fixed at a predetermined position above a target bearing of the wind turbine gearbox, and the method comprises: Sending a first pulse laser beam to the target bearing through the transmitting head of the pulse laser rangefinder; receiving, by means of a receiving head of the pulsed laser rangefinder, a laser beam reflected by the target bearing from the first pulsed laser beam, which is recorded as a second pulsed laser beam; Reading first duration data of the timer, where the first duration data refers to the duration from when the transmitter emits the first pulse laser beam to when the receiver receives the second pulse laser beam; A first distance value is calculated based on the first duration data and a predetermined pulse laser beam speed to generate a target distance value time sequence for the target bearing; The target distance value time series processed by the data conditioner is converted into a target distance spectrum diagram by using the spectrum analysis principle; Analyzing the multi-dimensional features of the target distance spectrum diagram to obtain a target vibration detection result of the target bearing; Wherein, before analyzing the multi-dimensional features of the target distance spectrum graph, the target distance spectrum graph corresponding to the preset duration value determined by analyzing the average preheating duration of the wind turbine gearbox is deleted; Wherein, after reading the first duration data of the timer, the method further includes: Get the detection loss function, the expression of the detection loss function is as follows: in, refers to the detection loss function, It refers to the actual first duration data when the pulse laser rangefinder emits a pulse laser beam to the target bearing for the i-th time, refers to the first duration data recorded by the timer when the pulse laser rangefinder emits a pulse laser beam to the target bearing for the i-th time, n refers to the number of times the pulse laser rangefinder detects the distance to the target bearing, is the feedback adjustment coefficient, and , refers to the timing loss adjustment coefficient for emitting the first pulse laser beam, is the timing loss adjustment coefficient for receiving the second pulse laser beam, refers to the timing accuracy loss adjustment coefficient of the timer; performing loss analysis on the first duration data according to the detection loss function, and correcting the first duration data according to the analysis result; Wherein, before converting the target distance value time series processed by the data conditioner into a target distance spectrum diagram by using the spectrum analysis principle, it includes: Extracting the target distance value in the target distance value time series by the data conditioner; An abnormal support function is introduced to determine whether the target distance value is abnormal data. The abnormal support function is expressed as follows: in, Refers to the adjacent distance value of the target distance value The distance value to the target The first abnormal support of refers to the distance data adjacent to the target distance value in the target distance value time series, k refers to the amplitude of the abnormal support function, and , is the attenuation factor, and If the first abnormal support reaches a predetermined threshold, the target distance value is determined to be abnormal data, and the target distance value is removed from the target distance value time series; The step of analyzing the multi-dimensional features of the target distance spectrum diagram to obtain the target vibration detection result of the target bearing includes: Respectively extracting a first distance value corresponding to a first time and a second distance value corresponding to a second time in the target distance spectrum graph, wherein the first time and the second time have a sequential order; generating a target vibration curve based on a calculated first distance difference between the first distance value and the second distance value; Obtaining the target vibration detection result by analyzing the target distance spectrum diagram and the target vibration curve; Wherein, obtaining the target vibration detection result by analyzing the target distance spectrum diagram and the target vibration curve includes: Collecting a target spectrum feature set of the target distance spectrum graph based on a predetermined feature indicator set, wherein the predetermined feature indicator set includes a center frequency, a frequency bandwidth, a peak frequency, and a peak amplitude; Acquiring a target slope of the target vibration curve; The target vibration detection result is determined according to the target frequency spectrum feature set and the target slope.

2. The method according to claim 1, characterized in that: The method further comprises: Construct a fault prediction model based on the principle of neural network; Analyzing the data in the target vibration detection result by using the fault prediction model to obtain a target predicted fault; Predictive maintenance is performed on the target predicted fault.

3. Bearing vibration detection system for wind turbine gearbox, characterized in that: The method for detecting bearing vibration of a wind turbine gearbox according to any one of claims 1 to 2 is used for implementing the method, wherein the system is communicatively connected with a pulsed laser rangefinder, and the pulsed laser rangefinder is fixed at a predetermined position above a target bearing of the wind turbine gearbox, and the system comprises: A pulse laser beam emitting module, the pulse laser beam emitting module is used to emit a first pulse laser beam to the target bearing through the emitting head of the pulse laser rangefinder; A reflected laser beam receiving module, wherein the reflected laser beam receiving module is used to receive a reflected laser beam of the target bearing to the first pulsed laser beam through a receiving head of the pulsed laser rangefinder, which is recorded as a second pulsed laser beam; A duration data reading module, the duration data reading module is used to read the first duration data of the timer, the first duration data refers to the duration from the transmitter emitting the first pulse laser beam to the receiver receiving the second pulse laser beam; A target distance value timing module, the target distance value timing module is used to calculate a first distance value based on the first duration data and a predetermined pulse laser beam speed to generate a target distance value timing of the target bearing; A target distance spectrum diagram module, the target distance spectrum diagram module is used to convert the target distance value time series processed by the data conditioner into a target distance spectrum diagram by using the spectrum analysis principle; A vibration detection result module is used to analyze the multi-dimensional features of the target distance spectrum diagram to obtain a target vibration detection result of the target bearing.

Citation Information

Patent Citations

  • System, apparatus and method of determining condition of a bearing

    CN114026564A

  • Method and system for monitoring planet-level running state of wind power gear box

    CN114279704A

  • Vibration measurement device and vibration measurement method

    CN114838803A

  • Bearing fault intelligent detection system and method based on inspection and vibration signals

    CN115165362A

  • Rolling bearing fault diagnosis method based on event camera

    CN115452373A