CN114184375A - Intelligent diagnosis method for common faults of gear box - Google Patents

Intelligent diagnosis method for common faults of gear box Download PDF

Info

Publication number
CN114184375A
CN114184375A CN202111559562.6A CN202111559562A CN114184375A CN 114184375 A CN114184375 A CN 114184375A CN 202111559562 A CN202111559562 A CN 202111559562A CN 114184375 A CN114184375 A CN 114184375A
Authority
CN
China
Prior art keywords
fault
analysis
gearbox
vibration
steps
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111559562.6A
Other languages
Chinese (zh)
Inventor
杜志强
付杰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Boming Chuang Neng Tianjin Technology Co ltd
Original Assignee
Boming Chuang Neng Tianjin Technology Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Boming Chuang Neng Tianjin Technology Co ltd filed Critical Boming Chuang Neng Tianjin Technology Co ltd
Priority to CN202111559562.6A priority Critical patent/CN114184375A/en
Publication of CN114184375A publication Critical patent/CN114184375A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00—Testing of machine parts
    • G01M13/02—Gearings; Transmission mechanisms
    • G01M13/021—Gearings
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01D—MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
    • G01D21/00—Measuring or testing not otherwise provided for
    • G01D21/02—Measuring two or more variables by means not covered by a single other subclass
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00—Testing of machine parts
    • G01M13/02—Gearings; Transmission mechanisms
    • G01M13/028—Acoustic or vibration analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/02—Preprocessing
    • G06F2218/04—Denoising
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/08—Feature extraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G06F2218/12—Classification; Matching

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Acoustics & Sound (AREA)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

The invention relates to the technical field of gearbox fault diagnosis, and discloses an intelligent diagnosis method for common faults of a gearbox. According to the invention, by carrying out oil product detection and analysis on lubricating oil, the risk of the fault of the gearbox possibly caused by the lubricating oil problem can be found in time, whether the gearbox part is abraded or not is determined by detecting specific metal elements, so that a detector is guided to carry out subsequent detection in a targeted manner, the fault position and fault level can be analyzed by vibration detection, the exact position, fault type and fault degree of the fault of the gearbox can be intuitively determined by an endoscope, so that the detector is guided to specifically analyze the fault reason according to fault information, and the problem is solved from the source.

Description

Intelligent diagnosis method for common faults of gear box
Technical Field
The invention relates to the technical field of gearbox fault diagnosis, in particular to an intelligent diagnosis method for common faults of a gearbox.
Background
In recent years, great progress has been made in wind power generation, but maintenance strategies for wind power projects require more initiative than conventional power generation systems such as coal, natural gas, etc. due to the relatively high operating and maintenance costs. Therefore, there is a need to reduce maintenance costs of wind turbines through condition monitoring, diagnostics, prognostics, and health management. While most of the gearbox failures are gear failures. Also, maintenance operations typically associated with gearbox failure are quite complex, and their disassembly, transportation, and repair costs are also quite high. Therefore, in order to ensure the normal operation of the wind turbine generator, the study on the fault of the gearbox of the wind turbine generator is of great importance, and the fault of the fan is frequently generated along with the continuous increase of the installed capacity of the fan. The slight fault needs to carry out maintenance detection on the fan; major failures require shutdown for maintenance, which not only causes serious economic loss, but also causes a series of potential safety hazards. The gearbox is statistically used as an important transmission component of the fan, and the failure rate of the component is very high. Therefore, in order to ensure the safe and stable operation of the wind turbine and the durable and effective power generation, the fault diagnosis work of the gearbox is very necessary. However, although the conventional methods such as spectrum analysis can detect a single and simple fault, they cannot be satisfied with the identification of complex faults.
For this reason, we propose a common fault intelligent diagnosis method for the gearbox.
Disclosure of Invention
The invention mainly solves the technical problems in the prior art and provides an intelligent diagnosis method for common faults of a gearbox.
In order to achieve the purpose, the invention adopts the following technical scheme that the method for intelligently diagnosing the common faults of the gearbox comprises the following steps,
s1, detecting the traditional content;
s2, mounting vibration sensors in a fixed-point dispersing manner;
and S3, extracting vibration signal sampling points, analyzing and calculating.
Preferably, the conventional content testing in S1 includes oil product testing analysis, which includes: performing physical and chemical index analysis and spectral analysis on the lubricating oil; the physical and chemical index analysis comprises kinematic viscosity analysis, total acid value analysis, moisture content analysis and particle counting analysis; when the physical and chemical index analysis result does not meet the standard, judging that the wind driven generator gearbox has a fault risk and needing to purify or replace lubricating oil; the spectral analysis is used for analyzing the content of specific metal elements in the lubricating oil, the specific metal elements comprise metal elements in an additive and metal elements contained in the gearbox of the wind driven generator, and when the metal elements in the additive are lower than an index value, the condition that the gearbox of the wind driven generator is in a failure risk is judged, and the additive needs to be added or the lubricating oil needs to be replaced properly; and when the metal element contained in the wind driven generator gear box is detected to exceed the standard, judging that the wind driven generator gear box is abraded and possibly fails.
Preferably, the fixed-point distributed mounting vibration sensor in S2 is specifically: the method comprises the steps of selecting a plurality of effective characteristics of a gearbox, setting sensors respectively, collecting signals transmitted by the set sensors in real time, receiving and classifying vibration signals of the sensors at the same time, dividing according to vibration intensity, and additionally summarizing according to vibration distribution during division.
Preferably, the vibration signal sampling point extraction and analysis calculation in S3 specifically includes: the method comprises the steps of carrying out three-layer decomposition analysis on collected vibration signals by adopting a wavelet packet analysis method, carrying out empirical mode decomposition on the vibration signals after the wavelet packet analysis, extracting a first component of the signals, carrying out characteristic value extraction on the extracted first signal component to be used as a characteristic vector used in fault diagnosis, obtaining a characteristic vector sample of historical fault data of the gearbox, training the characteristic vector sample by using a support vector machine, using a group with the highest classification accuracy as a parameter used in the fault diagnosis, obtaining real-time operation data of the gearbox, obtaining the characteristic vector, classifying the characteristic vector by using the support vector machine, and outputting a diagnosis result.
Preferably, the method further performs Fourier transform and normalization on each original vibration signal to obtain a spectrum signal corresponding to each original vibration signal, forms training data from all spectrum signals, performs unsupervised training on a plurality of denoising autoencoders by the training data, stacks hidden layers of the denoising autoencoders after training, adds a logistic regression layer to form a stacked denoising autoencoder, performs supervised training optimization on the stacked denoising autoencoder by using a quantum particle swarm optimization method to obtain an optimized stacked denoising autoencoder, performs fault diagnosis by using the optimized stacked denoising autoencoder, and initializes parameters of the stacked denoising autoencoder by using the optimal parameters of each denoising autoencoder obtained in the unsupervised training process, then, updating the weight of the stacking denoising self-encoder by using a random gradient descent method, and performing fault diagnosis through the optimized stacking denoising self-encoder, wherein the fault diagnosis comprises the following steps: acquiring a target vibration signal of a wind driven generator gearbox to be diagnosed, and performing Fourier transform and normalization processing on the target vibration signal to obtain a target frequency spectrum signal; and extracting fault characteristic signals by the stacking denoising autoencoder, and identifying the fault characteristic signals by a least square support vector machine to obtain fault types.
Preferably, the wavelet packet three-layer decomposition analysis method comprises the following steps:
establishing a wavelet packet change three-layer decomposition relational expression; selecting a corresponding threshold according to the soft threshold function to carry out threshold quantization processing on the high-frequency coefficient under each decomposition scale; and performing one-dimensional wavelet reconstruction on the decomposed lowest layer low-frequency coefficient and high-frequency coefficient.
Preferably, the vibration sensor mounting points are respectively distributed at the axial position and the radial position of the low-speed end, the middle end and the high-speed end of the fan gearbox body.
Preferably, after the vibration signal is collected, the endoscope is used for detecting the position with large vibration inside the gearbox, so that the fault point is detected and confirmed, the type and the degree of the fault are determined, and the follow-up examination can be directly finished after the fault point is confirmed to be an obvious simple problem.
Preferably, the characteristic values are obtained from a time domain and a frequency domain respectively, the time domain characteristic value includes a peak index, a kurtosis index, a skewness index and a margin index, the frequency domain characteristic value includes a frequency center of gravity, a frequency standard deviation and a root mean square frequency, and the wavelet packet change three-layer decomposition relation is expressed as:
S＝AAA3+DAA3+ADA3+DDA3+AAD3+DAD3+ADD3+DDD3；
where A is the low frequency portion of the signal and D is the high frequency portion of the signal.
Preferably, the reason integration analysis and the actual reinforcement detection in S4 are specifically: and integrating the obtained conclusions, carrying out actual detection on the gear box if a plurality of different conclusions are obtained so as to obtain the most accurate diagnosis result, repairing the gear box according to the conclusions if the same conclusion is obtained, wherein the actual reinforced detection comprises gear box endoscopy and bearing disassembly inspection, and the most intuitive accurate diagnosis is carried out through the gear tooth surface and bearing damage characteristics.
Advantageous effects
The invention provides an intelligent diagnosis method for common faults of a gearbox. The method has the following beneficial effects:
(1) according to the intelligent diagnosis method for common faults of the gear box, the risk that the gear box possibly fails due to the lubricating oil problem can be found in time by performing oil product detection and analysis on the lubricating oil, and whether the gear box part is abraded or not is determined by detecting specific metal elements, so that a detector is guided to perform targeted subsequent detection.
(2) According to the intelligent diagnosis method for the common faults of the gear box, the positions and the fault levels of the faults can be analyzed through vibration detection, and the exact positions, the types and the fault degrees of the faults of the gear box can be visually determined through an endoscope, so that a detector is guided to specifically analyze the fault reasons according to fault information, and the problems are solved from the source.
(3) According to the intelligent diagnosis method for common faults of the gearbox, the noise of the vibration signals is reduced by a method of combining wavelet packet decomposition and empirical mode decomposition. Firstly, vibration sensors are arranged on the low-speed shaft, the intermediate shaft and the high-speed shaft of the gearbox in the axial direction and the radial direction, and vibration signals generated when the gearbox runs are collected through the vibration sensors. And secondly, carrying out wavelet packet decomposition and reconstruction on the acquired vibration signal line. After the signal is reconstructed, the signal is processed again by using empirical mode decomposition. The method can effectively eliminate the interference of noise on the signal.
(4) According to the intelligent diagnosis method for common faults of the gearbox, the characteristic values extracted from the vibration signals after two times of processing can represent the running state of equipment better, and the characteristic values are provided for a support vector machine to perform sample training, so that the accuracy of classification results can be improved. The method can diagnose the fault type of the gear box, can determine the fault position, and has the advantages of accurate diagnosis result, clear result display and the like.
(5) The method for intelligently diagnosing common faults of the gear box comprises the steps that a stacking denoising self-encoder is constructed through collected gear box fault signals, signals used in the construction process of the stacking denoising self-encoder come from the gear box, fault features in the gear box signals can be effectively extracted through the stacking denoising self-encoder after optimization, the extracted fault features contain high-dimensional information of original vibration signals, the fault types can be effectively distinguished when the feature signals are input into a least square support vector machine, powerful basis is provided for finding the positions of the faults of the gear box and maintaining, and stable and reliable operation of equipment is guaranteed.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is a flow chart of a conventional detection method according to the present invention;
fig. 3 is a flow chart of vibration signal sampling point extraction and analysis calculation according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Example (b): the intelligent diagnosis method for common faults of the gearbox comprises the following steps,
s1, detecting the traditional content;
s2, mounting vibration sensors in a fixed-point dispersing manner;
and S3, extracting vibration signal sampling points, analyzing and calculating.
The traditional content detection in the S1 comprises oil product detection and analysis, and the oil product detection and analysis comprises the following steps: performing physical and chemical index analysis and spectral analysis on the lubricating oil; the physical and chemical index analysis comprises kinematic viscosity analysis, total acid value analysis, moisture content analysis and particle counting analysis; when the physical and chemical index analysis result does not meet the standard, judging that the wind driven generator gearbox has a fault risk and needing to purify or replace lubricating oil; the spectral analysis is used for analyzing the content of specific metal elements in the lubricating oil, the specific metal elements comprise metal elements in an additive and metal elements contained in the gearbox of the wind driven generator, and when the metal elements in the additive are lower than an index value, the condition that the gearbox of the wind driven generator is in a failure risk is judged, and the additive needs to be added or the lubricating oil needs to be replaced properly; when the metal elements contained in the wind driven generator gear box exceed the standard, the wind driven generator gear box is judged to be worn and possibly broken down, and the specific steps of dispersedly installing the vibration sensors at fixed points in the step S2 are as follows: select a plurality of effective characteristic of gear box and set up the sensor respectively, gather in real time the signal of the sensor transmission that sets up to the vibration signal to each sensor is received and is categorised simultaneously, divides according to the vibration intensity, and additionally sums up according to vibration distribution ground when dividing, vibration signal sampling point extraction and analysis calculation in S3 specifically are: carrying out three-layer decomposition analysis on the acquired vibration signals by adopting a wavelet packet analysis method, carrying out empirical mode decomposition on the vibration signals after the wavelet packet analysis, extracting a first component of the signals, carrying out characteristic value extraction on the extracted first signal component, obtaining a characteristic vector sample of historical fault data of the gearbox as a characteristic vector used in fault diagnosis, training the characteristic vector sample by using a support vector machine, obtaining real-time operation data of the gearbox by using a group with the highest classification accuracy as a parameter used in the subsequent fault diagnosis, obtaining the characteristic vector, classifying the characteristic vector by using the support vector machine, and outputting a diagnosis result, wherein on the basis, Fourier transform and normalization processing are carried out on each original vibration signal to obtain a frequency spectrum signal corresponding to each original vibration signal, the method comprises the following steps of forming training data by all spectrum signals, carrying out unsupervised training on a plurality of denoising autocoders by the training data, stacking hidden layers of the denoising autocoders after the training is finished, adding a logistic regression layer to form a stacked denoising autocoder, carrying out supervised training optimization on the stacked denoising autocoders by adopting a quantum particle swarm optimization method to obtain optimized stacked denoising autocoders, carrying out fault diagnosis by the optimized stacked denoising autocoders, initializing parameters of the stacked denoising autocoders by using optimal parameters of the denoising autocoders obtained in the unsupervised training process, updating weights of the stacked denoising autocoders by using a random gradient descent method, and carrying out fault diagnosis by the optimized stacked denoising autocoders, wherein the method comprises the following steps: acquiring a target vibration signal of a wind driven generator gearbox to be diagnosed, and performing Fourier transform and normalization processing on the target vibration signal to obtain a target frequency spectrum signal; extracting a fault characteristic signal by the stacking denoising autoencoder, and identifying the fault characteristic signal by a least square support vector machine to obtain a fault type, wherein the wavelet packet three-layer decomposition analysis method comprises the following steps:
establishing a wavelet packet change three-layer decomposition relational expression; selecting a corresponding threshold according to the soft threshold function to carry out threshold quantization processing on the high-frequency coefficient under each decomposition scale; the method comprises the following steps of performing one-dimensional wavelet reconstruction on a low-frequency coefficient and a high-frequency coefficient of the bottom layer after decomposition, wherein mounting points of vibration sensors are respectively distributed at axial and radial positions of a low-speed end, a middle end and a high-speed end of a gearbox body of a fan, after vibration signals are collected, detecting a large vibration position in a gearbox through an endoscope, detecting and confirming a fault point, determining the type and the degree of the fault, and directly finishing subsequent inspection after confirming that the simple problem is obvious, wherein characteristic values are respectively obtained from a time domain and a frequency domain, time domain characteristic values comprise a peak value index, a kurtosis index, a skewness index and a margin index, frequency domain characteristic values comprise a frequency center of gravity, a frequency standard deviation and a root-mean-square frequency, and wavelet packet change three-layer decomposition relational expressions are as follows:
S＝AAA3+DAA3+ADA3+DDA3+AAD3+DAD3+ADD3+DDD3；
where A is the low frequency portion of the signal and D is the high frequency portion of the signal.
The reason integration analysis and the actual strengthening detection in S4 are specifically: and integrating the obtained conclusions, carrying out actual detection on the gear box if a plurality of different conclusions are obtained so as to obtain the most accurate diagnosis result, repairing the gear box according to the conclusions if the same conclusion is obtained, wherein the actual reinforced detection comprises gear box endoscopy and bearing disassembly inspection, and the most intuitive accurate diagnosis is carried out through the gear tooth surface and bearing damage characteristics.
By carrying out oil product detection and analysis on lubricating oil, the risk that the gear box possibly fails due to the lubricating oil problem can be found in time, whether the gear box part is abraded or not is determined through specific metal element detection, subsequent detection is guided to detection personnel pertinently, the position and the fault level of the fault can be analyzed through vibration detection, the exact position of the gear box fault, the type and the fault degree of the fault can be visually determined through an endoscope, the detection personnel is guided to specifically analyze the fault reason according to fault information, the problem is solved from the source, and the method of combining wavelet packet decomposition and empirical mode decomposition is utilized to reduce the noise of a vibration signal. Firstly, vibration sensors are arranged on the low-speed shaft, the intermediate shaft and the high-speed shaft of the gearbox in the axial direction and the radial direction, and vibration signals generated when the gearbox runs are collected through the vibration sensors. And secondly, carrying out wavelet packet decomposition and reconstruction on the acquired vibration signal line. After the signal is reconstructed, the signal is processed again by using empirical mode decomposition. The method can effectively eliminate the interference of noise on signals, the characteristic value extracted from the vibration signals after two times of processing can better represent the running state of the equipment, and the characteristic value is provided for a support vector machine to carry out sample training, so that the accuracy of the classification result can be improved. The method can diagnose the fault type of the gearbox, can determine the fault position, has the advantages of accurate diagnosis result, clear result display and the like, constructs the stacking denoising autoencoder through the collected gearbox fault signals, because the signals used in the construction process come from the gearbox, the stacking denoising autoencoder can effectively extract the fault characteristics in the gearbox signals after optimization, the extracted fault characteristics contain high-dimensional information of original vibration signals, the fault types can be effectively distinguished by inputting the characteristic signals into a least square support vector machine, powerful basis is provided for finding the fault position of the gearbox and maintaining, and stable and reliable operation of equipment is ensured
The foregoing shows and describes the general principles and broad features of the present invention and advantages thereof. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims (10)

1. The intelligent diagnosis method for common faults of the gear box is characterized in that: comprises the following steps of (a) carrying out,
s1, detecting the traditional content;
s2, mounting vibration sensors in a fixed-point dispersing manner;
s3, extracting, analyzing and calculating vibration signal sampling points;
and S4, integrating and analyzing reasons and actually strengthening detection.
2. The intelligent diagnosis method for common faults of gearboxes according to claim 1, wherein the method comprises the following steps: the traditional content detection in the S1 comprises oil product detection and analysis, and the oil product detection and analysis comprises the following steps: performing physical and chemical index analysis and spectral analysis on the lubricating oil; the physical and chemical index analysis comprises kinematic viscosity analysis, total acid value analysis, moisture content analysis and particle counting analysis; when the physical and chemical index analysis result does not meet the standard, judging that the wind driven generator gearbox has a fault risk and needing to purify or replace lubricating oil; the spectral analysis is used for analyzing the content of specific metal elements in the lubricating oil, the specific metal elements comprise metal elements in an additive and metal elements contained in the gearbox of the wind driven generator, and when the metal elements in the additive are lower than an index value, the condition that the gearbox of the wind driven generator is in a failure risk is judged, and the additive needs to be added or the lubricating oil needs to be replaced properly; and when the metal element contained in the wind driven generator gear box is detected to exceed the standard, judging that the wind driven generator gear box is abraded and possibly fails.
3. The intelligent diagnosis method for common faults of gearboxes according to claim 1, wherein the method comprises the following steps: the fixed-point scattered installation vibration sensor in the step S2 specifically includes: the method comprises the steps of selecting a plurality of effective characteristics of a gearbox, setting sensors respectively, collecting signals transmitted by the set sensors in real time, receiving and classifying vibration signals of the sensors at the same time, dividing according to vibration intensity, and additionally summarizing according to vibration distribution during division.
4. The intelligent diagnosis method for common faults of gearboxes according to claim 1, wherein the method comprises the following steps: the vibration signal sampling point extraction and analysis calculation in the step S3 specifically includes: the method comprises the steps of carrying out three-layer decomposition analysis on collected vibration signals by adopting a wavelet packet analysis method, carrying out empirical mode decomposition on the vibration signals after the wavelet packet analysis, extracting a first component of the signals, carrying out characteristic value extraction on the extracted first signal component to be used as a characteristic vector used in fault diagnosis, obtaining a characteristic vector sample of historical fault data of the gearbox, training the characteristic vector sample by using a support vector machine, using a group with the highest classification accuracy as a parameter used in the fault diagnosis, obtaining real-time operation data of the gearbox, obtaining the characteristic vector, classifying the characteristic vector by using the support vector machine, and outputting a diagnosis result.
5. The intelligent diagnosis method for common faults of gearboxes according to claim 4, wherein the method comprises the following steps: on the basis, the method additionally carries out Fourier transform and normalization processing on each original vibration signal to obtain a spectrum signal corresponding to each original vibration signal, training data is formed by all spectrum signals, unsupervised training is carried out on a plurality of denoising autoencoders by the training data, hidden layers of the denoising autoencoders after training are stacked together, a logistic regression layer is added to form a stacked denoising autoencoder, supervised training optimization is carried out on the stacked denoising autoencoder by adopting a quantum particle swarm optimization method to obtain an optimized stacked denoising autoencoder, fault diagnosis is carried out by the optimized stacked denoising autoencoder, parameters of the stacked denoising autoencoder are initialized by utilizing the optimal parameters of each denoising autoencoder obtained in the unsupervised training process, and then the weight of the stacked denoising autoencoder is updated by utilizing a random gradient descent method, fault diagnosis is carried out through the optimized stacked denoising self-encoder, and the fault diagnosis method comprises the following steps: acquiring a target vibration signal of a wind driven generator gearbox to be diagnosed, and performing Fourier transform and normalization processing on the target vibration signal to obtain a target frequency spectrum signal; and extracting fault characteristic signals by the stacking denoising autoencoder, and identifying the fault characteristic signals by a least square support vector machine to obtain fault types.
6. The intelligent diagnosis method for common faults of gearboxes according to claim 4, wherein the method comprises the following steps: the wavelet packet three-layer decomposition analysis method comprises the following steps:
establishing a wavelet packet change three-layer decomposition relational expression; selecting a corresponding threshold according to the soft threshold function to carry out threshold quantization processing on the high-frequency coefficient under each decomposition scale; and performing one-dimensional wavelet reconstruction on the decomposed lowest layer low-frequency coefficient and high-frequency coefficient.
7. The intelligent diagnosis method for common faults of gearboxes according to claim 3, wherein the method comprises the following steps: the mounting points of the vibration sensors are respectively distributed at the axial and radial positions of the low-speed end, the middle end and the high-speed end of the gearbox body of the fan.
8. The intelligent diagnosis method for common faults of gearboxes according to claim 4, wherein the method comprises the following steps: according to the method, after the vibration signals are collected, the positions with obvious or judged fault characteristics of the internal vibration abnormal characteristics of the gear box are detected through the endoscope, so that the fault points are inspected and confirmed, the type and the degree of the fault are determined, the problem of obvious fault/failure characteristics is confirmed, and the follow-up inspection is directly finished after an effective solution or measure is made.
9. The intelligent diagnosis method for common faults of gearboxes according to claim 4, wherein the method comprises the following steps: the characteristic values are respectively obtained from a time domain and a frequency domain, the time domain characteristic values comprise a peak value index, a kurtosis index, a skewness index and a margin index, the frequency domain characteristic values comprise a frequency center of gravity, a frequency standard deviation and a root mean square frequency, and the wavelet packet change three-layer decomposition relation is expressed as follows:
S＝AAA3+DAA3+ADA3+DDA3+AAD3+DAD3+ADD3+DDD3；
where A is the low frequency portion of the signal and D is the high frequency portion of the signal.
10. The intelligent diagnosis method for common faults of gearboxes according to claim 1, wherein the method comprises the following steps: the reason integration analysis and the actual strengthening detection in S4 are specifically: and integrating the obtained conclusions, carrying out actual detection on the gear box if a plurality of different conclusions are obtained so as to obtain the most accurate diagnosis result, repairing the gear box according to the conclusions if the same conclusion is obtained, wherein the actual reinforced detection comprises gear box endoscopy and bearing disassembly inspection, and the most intuitive accurate diagnosis is carried out through the gear tooth surface and bearing damage characteristics.
CN202111559562.6A 2021-12-20 2021-12-20 Intelligent diagnosis method for common faults of gear box Pending CN114184375A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111559562.6A CN114184375A (en) 2021-12-20 2021-12-20 Intelligent diagnosis method for common faults of gear box

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111559562.6A CN114184375A (en) 2021-12-20 2021-12-20 Intelligent diagnosis method for common faults of gear box

Publications (1)

Publication Number Publication Date
CN114184375A true CN114184375A (en) 2022-03-15

Family

ID=80605640

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111559562.6A Pending CN114184375A (en) 2021-12-20 2021-12-20 Intelligent diagnosis method for common faults of gear box

Country Status (1)

Country Link
CN (1) CN114184375A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118051860A (en) * 2024-04-15 2024-05-17 西安瓦力机电科技有限公司 Intelligent high-precision gear reducer monitoring and diagnosing system

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118051860A (en) * 2024-04-15 2024-05-17 西安瓦力机电科技有限公司 Intelligent high-precision gear reducer monitoring and diagnosing system

Similar Documents

Publication Publication Date Title
Saufi et al. Challenges and opportunities of deep learning models for machinery fault detection and diagnosis: A review
Wang et al. An integrated fault diagnosis and prognosis approach for predictive maintenance of wind turbine bearing with limited samples
Vamsi et al. Comparison of condition monitoring techniques in assessing fault severity for a wind turbine gearbox under non-stationary loading
CN106197996A (en) Offshore crane Fault Diagnosis of Gear Case device and method based on multivariate data
CN113469060A (en) Multi-sensor fusion convolution neural network aeroengine bearing fault diagnosis method
CN112179691B (en) Mechanical equipment running state abnormity detection system and method based on counterstudy strategy
Li et al. Bearing fault feature selection method based on weighted multidimensional feature fusion
CN110823576B (en) Mechanical anomaly detection method based on generation of countermeasure network
CN107797537A (en) A kind of prognostic and health management method applied to automatic production line
Barbieri et al. Analysis of automotive gearbox faults using vibration signal
CN112393906B (en) Method for diagnosing, classifying and evaluating health of weak signal fault of bogie bearing of metro vehicle
Alkhadafe et al. Condition monitoring of helical gears using automated selection of features and sensors
CN107229269A (en) A kind of wind-driven generator wheel-box method for diagnosing faults of depth belief network
CN110398362B (en) Robot RV reducer fault diagnosis and positioning method
Cao et al. Deterioration state diagnosis and wear evolution evaluation of planetary gearbox using vibration and wear debris analysis
CN114184375A (en) Intelligent diagnosis method for common faults of gear box
Mollazade et al. An intelligent combined method based on power spectral density, decision trees and fuzzy logic for hydraulic pumps fault diagnosis
CN115165326A (en) Fan fault diagnosis method through mechanical transmission chain lubricating oil (grease) impurity analysis
CN109580218B (en) Blower gear box state identification method based on likelihood learning machine
CN117370879A (en) Real-time online fault diagnosis method and system for wind turbine gearbox
CN115326393A (en) Wind turbine generator bearing pair fault diagnosis method based on temperature information
CN107727392B (en) State index evaluation and optimization method based on signal detection and ROC analysis
CN113447267B (en) Vibration signal analysis-based complete gearbox state evaluation method and system
CN114417514A (en) Wind turbine generator gearbox fault diagnosis system and method, computer equipment and storage medium
Mollazade et al. An intelligent model based on data mining and fuzzy logic for fault diagnosis of external gear hydraulic pumps

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination