CN115018181A - Online monitoring and service life prediction method for speed reducer gear box of light rail unit - Google Patents
Online monitoring and service life prediction method for speed reducer gear box of light rail unit Download PDFInfo
- Publication number
- CN115018181A CN115018181A CN202210735193.XA CN202210735193A CN115018181A CN 115018181 A CN115018181 A CN 115018181A CN 202210735193 A CN202210735193 A CN 202210735193A CN 115018181 A CN115018181 A CN 115018181A
- Authority
- CN
- China
- Prior art keywords
- factor
- gearbox
- light rail
- speed reducer
- dictionary
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 68
- 239000003638 chemical reducing agent Substances 0.000 title claims abstract description 55
- 238000012544 monitoring process Methods 0.000 title claims abstract description 50
- 230000036541 health Effects 0.000 claims abstract description 47
- 238000004458 analytical method Methods 0.000 claims abstract description 28
- 238000006731 degradation reaction Methods 0.000 claims abstract description 19
- 230000015556 catabolic process Effects 0.000 claims abstract description 17
- 230000001133 acceleration Effects 0.000 claims abstract description 11
- 230000004044 response Effects 0.000 claims abstract description 10
- 238000005516 engineering process Methods 0.000 claims abstract description 9
- 230000004927 fusion Effects 0.000 claims abstract description 9
- 230000005540 biological transmission Effects 0.000 claims abstract description 7
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 238000012423 maintenance Methods 0.000 claims description 18
- 238000000605 extraction Methods 0.000 claims description 15
- 238000001914 filtration Methods 0.000 claims description 15
- 238000000354 decomposition reaction Methods 0.000 claims description 13
- 238000010183 spectrum analysis Methods 0.000 claims description 12
- 230000003595 spectral effect Effects 0.000 claims description 8
- 238000003745 diagnosis Methods 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 7
- 230000006870 function Effects 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 5
- 230000006698 induction Effects 0.000 claims description 5
- 238000003860 storage Methods 0.000 claims description 5
- 238000010276 construction Methods 0.000 claims description 4
- 238000011478 gradient descent method Methods 0.000 claims description 4
- 238000013016 damping Methods 0.000 claims description 3
- 238000012847 principal component analysis method Methods 0.000 claims description 3
- 238000007639 printing Methods 0.000 claims description 3
- 238000010223 real-time analysis Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 230000005236 sound signal Effects 0.000 claims description 3
- 238000001228 spectrum Methods 0.000 claims description 3
- 238000010977 unit operation Methods 0.000 claims description 3
- 230000003190 augmentative effect Effects 0.000 claims description 2
- 230000001052 transient effect Effects 0.000 claims description 2
- 238000010586 diagram Methods 0.000 description 16
- 238000005096 rolling process Methods 0.000 description 13
- 238000013398 bayesian method Methods 0.000 description 3
- 238000009826 distribution Methods 0.000 description 2
- 230000005284 excitation Effects 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000008054 signal transmission Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 210000001367 artery Anatomy 0.000 description 1
- 230000001174 ascending effect Effects 0.000 description 1
- 230000003416 augmentation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
- G06F17/141—Discrete Fourier transforms
- G06F17/142—Fast Fourier transforms, e.g. using a Cooley-Tukey type algorithm
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Tourism & Hospitality (AREA)
- Operations Research (AREA)
- Algebra (AREA)
- General Engineering & Computer Science (AREA)
- General Business, Economics & Management (AREA)
- Software Systems (AREA)
- Databases & Information Systems (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Discrete Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- General Health & Medical Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Primary Health Care (AREA)
- Quality & Reliability (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention discloses an online monitoring and service life prediction method for a speed reducer gearbox of a light rail unit, which comprises the steps of utilizing a wireless WIFI acceleration sensor to online pick up health information such as vibration acceleration, force, temperature, sound and the like of the speed reducer gearbox of the urban light rail unit in real time, and utilizing a wireless transmission technology to transmit data to an operation monitoring center and reconstruct the data; preprocessing the reconstructed data in an operation monitoring center; diagnosing the fault degree and position of the equipment and key parts thereof based on a time domain/frequency domain analysis method; and extracting a fusion health factor time sequence reflecting the states of the equipment and key parts thereof based on the preprocessed data, and performing degradation tracking prediction and residual life prediction by using a quaternion prediction model. The invention can reconstruct the data collected by the remote subway light rail set in real time, monitor and predict on line, has high fault location and identification precision and high prediction response speed, and can be applied to the fields of urban light rail and high-speed rail health monitoring and the like.
Description
Technical Field
The invention relates to the technical field of mechanical equipment prediction and health management, in particular to a method for on-line monitoring and service life prediction of a speed reducer gear box of a light rail unit.
Background
The operation health state prediction and health management of key rotating parts of the urban light rail directly influence the safe and reliable operation of a train set system. The gear and the rolling bearing are used as a 'main artery' and a 'joint' of a key power transmission system of an urban light rail train set, the whole operation performance and the service reliability of system equipment of the train set are directly determined, however, service environments of key rotating parts such as the gear and the rolling part are complex, severe working conditions such as variable load, strong impact, large disturbance, acid-base corrosion and the like are frequent, and once a fault occurs, a series of linkage accidents are caused. Therefore, in order to ensure the safety and stable operation of the light rail unit, the operation health state and the remaining service life of the key rotating parts of the light rail unit need to be monitored and analyzed in real time.
In the era of industrial internet of things and big data, 3 main challenges exist in the research of the urban light rail intelligent operation and maintenance technology at present: (1) the method has the advantages that the number of subway lines in each city is large, the coverage range of a subway unit gear box in each line is wide, the equipment distribution scale is large, the number of equipment measuring points is large, the sampling frequency is high, so that the challenges of data storage capacity are caused due to the fact that a large amount of collected health degradation data needs to be transmitted and analyzed in real time through a terminal, meanwhile, the real-time online state monitoring and predicting precision is not high, and a unit health diagnosis and prediction report and an operation and maintenance decision scheme cannot be provided for a maintenance department in real time online. (2) The subway unit gear box system has a complex internal structure, the distribution characteristic difference of degradation characteristics of the same type of equipment or parts is large, the external working condition environment is variable, random interference factors are multiple, the large-range fluctuation of the operation rotating speed, load and noise and the mutual coupling of internal excitation and external excitation cause the whole life cycle health prediction and intelligent operation and maintenance of the subway unit gear box system and the parts thereof to be frosted on snow. (3) The maintenance after the fact and the operation and maintenance at regular time increase the maintenance labor cost, the maintenance efficiency and level are not high, the maintenance resources of the train are wasted, the real-time performance is not strong, and the like.
Therefore, in order to overcome the challenges and difficulties, the invention provides an online monitoring and service life prediction system for the speed reducer gearbox of the urban light rail unit, which is applied to health monitoring and operation and maintenance of urban light rails and high-speed rails, and has the characteristics of real-time transmission of big data, high fault positioning and identification precision, and high response speed, and is very necessary.
Disclosure of Invention
The invention aims to: the method for on-line monitoring and service life prediction of the reducer gearbox of the light rail unit and the preparation method are provided to solve the defects.
In order to achieve the above purpose, the invention provides the following technical scheme:
a method for on-line monitoring and service life prediction of a gearbox of a light rail unit speed reducer comprises the following steps:
s1: the method comprises the steps that a wireless WIFI acceleration sensor, a force sensor, a microphone and a temperature sensor are utilized to pick up vibration acceleration signals, force signals, sound signals and real-time health information data of temperature of a gearbox of a speed reducer of the light rail unit on line in real time;
s2: performing sparse representation on real-time health monitoring information data of a gearbox of the speed reducer by using a compressed sensing and dictionary learning technology, remotely transmitting sparse representation information by using a wireless transmission network, and reconstructing the real-time health monitoring information data in a health monitoring center;
s3: in a monitoring operation center, carrying out preprocessing analysis on reconstructed unit reducer gearbox monitoring data, wherein the preprocessing analysis comprises outlier removal, signal filtering, signal denoising or signal decomposition processing, and obtaining pollution-free or near-zero pollution data;
s4: based on a time domain analysis method and a frequency domain analysis method, fault diagnosis and fault feature extraction are carried out on equipment and key parts thereof, including gears and bearings, so that fault positions and fault signal sources are identified;
s5: extracting a health factor time sequence of key parts of a gearbox of a unit speed reducer based on a time domain/frequency domain feature extraction method; performing dimensionality reduction and fusion on the time sequence of the time domain/frequency domain health factor by using a principal component analysis method, and extracting a Fused health factor time sequence (FHITS) of a key component of a gearbox of a unit speed reducer; meanwhile, constructing a quaternion prediction model;
s6: inputting a multi-channel health factor time sequence of a key component of a gearbox of the unit speed reducer into a quaternion prediction model, and predicting the residual life of the key component of the gearbox of the unit speed reducer;
s7: printing an output report of the analysis result of the last step, and providing a real-time analysis report for a unit operation supervision department and a maintenance department; and performing predictive maintenance on the unit speed reducer equipment and key parts thereof according to the diagnosis and prediction analysis results.
Preferably, in step S2, the compressed sensing and dictionary learning technique is a compressed sensing and sparse bayesian dictionary learning technique, and includes dictionary construction and algorithm reconstruction, and the specific implementation method thereof is as follows:
a1: constructing an over-complete dictionary atom and initializing a dictionary;
a2: based on an over-complete dictionary, carrying out sparse representation on an original signal;
a3: and reconstructing original monitoring data by using a sparse Bayesian algorithm, analyzing a reconstruction result and an error, and optimizing a model parameter of the sparse Bayesian algorithm.
Preferably, in step S3, the signal filtering includes low-pass filtering, high-pass filtering and band-pass filtering; the signal denoising specifically includes: denoising by using a wavelet soft threshold and sparse regularization; the signal decomposition processing specifically includes: empirical mode decomposition method and variation mode decomposition method.
Preferably, in step S4, the time domain analysis method refers to extracting time domain signal dimension factor features and extracting non-dimension factor features; the dimension factor features comprise: maximum value, minimum value, peak-to-peak value, mean value, variance, standard deviation, mean square value, root mean square value RMS, mean square error MSE, root mean square error RMSE and root mean square amplitude; the non-dimensional factor features include: a peak factor, a pulse factor, a margin factor, a kurtosis factor, a waviness factor, and a kurtosis factor and a skewness factor.
Preferably, in step S4, the frequency domain analysis method is a spectrum analysis method, and includes: the method comprises a signal FFT amplitude spectrum analysis method, a Hilbert envelope spectrum analysis method, a Hilbert marginal spectrum analysis method and a rapid spectrum kurtosis analysis method.
Preferably, in step S5, the time domain/frequency domain feature extraction method, that is, performing time domain feature extraction and frequency domain feature extraction on the reconstructed monitoring signal, where the time domain feature extraction includes: the average value, standard deviation, skewness, kurtosis, peak-to-peak value, root mean square, peak coefficient factor, waveform coefficient factor, pulse coefficient factor, edge coefficient factor and energy factor, wherein the total index is 11 indexes; the frequency domain feature extraction comprises: spectral kurtosis-mean, spectral kurtosis-standard, spectral kurtosis-skewness, spectral kurtosis-kurtosis, and 4 indexes.
Preferably, in step S5, the construction of the quaternion prediction model includes the following main steps:
b1: constructing a target cost function based on a quaternion prediction model;
b2: updating the filter weight of the target cost function by using a gradient descent method;
b3: inputting multi-channel data as input into a quaternion prediction model after weight updating, and predicting the degradation operation trend of a gearbox of a unit speed reducer and key parts thereof in real time;
b4: and obtaining the residual life time of the gearbox of the unit speed reducer and key parts of the gearbox according to the degradation operation trend of the prediction curve and the fault threshold value.
Preferably, in a1 of step S2, the building overcomplete dictionary atom and dictionary initialization are specifically:
in order to prevent the distortion of a reconstructed signal and ensure that the waveform of redundant dictionary atoms is matched with the physical structure of an instantaneous pulse signal, a stepping-pulse dictionary atom is adopted, and the expression is as follows:
d＝η 1 ·a·d imp +η 2 ·d step (1)，
in the formula: a is the step-to-pulse peak ratio, d imp Is a single degree of freedom pulse atom, dstep is a single degree of freedom stepping atom, eta 1 And η 2 The amplitude factor is used for adjusting the amplitude of the stepping-pulse dictionary atoms to be consistent with the amplitude of the low-frequency resonance component;
d imp and d step These two classes of dictionary atoms are defined as:
in the formula: f. of n And the system natural frequency is tau, the system damping coefficient is u, the response time of the step-pulse dictionary is started, and delta t is the interval time from the moment that the bearing ball starts to contact the fault point to the moment that the bearing ball leaves the fault point.
Preferably, in step S5, the constructed quaternion prediction model is a 1/2 th power quaternion prediction model, which is specifically:
wherein, | e * (n)e(n)e * (n)e(n)| 1/2 Is a term raised to the power of 1/2 as the minimum mean,
is a Gaussian correlation entropy-induced penalty term, 1/2 is a scoreThe number order, the weight factor beta is more than or equal to 0 and is used for balancing the minimum mean value 1/2 power term of augmentation and the Gaussian correlation entropy induction punishment term, w i And (n) is the filter weight.
Preferably, a computer readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the steps of the online monitoring and life prediction method for the reducer gearbox of the urban light rail set.
The invention has the beneficial effects that:
the invention relates to a method for on-line monitoring and service life prediction of a gearbox of a light rail unit speed reducer, which is based on a quaternion model, can predict a multichannel health factor sequence constructed by collected multichannel and high-dimensional data in real time at one time, does not need channel-by-channel or dimensionality-by-dimensionality prediction, and has high prediction precision and high prediction response speed. Compared with a remote signal transmission method, the method of the invention does not need a health center to reserve extra-large memory and storage equipment, and can reconstruct the data acquired by the remote subway and light rail set in real time. The method provided by the invention does not need to consider the influence of the physical degradation process, equipment degradation mechanism and complex external working conditions of the subway light rail unit gearbox and key components thereof on the stability of the prediction model, can be applied to the fields of urban light rail and high-speed rail health monitoring and the like, and has good industrial application value.
Drawings
FIG. 1: a method flow diagram of an embodiment of the invention;
FIG. 2: the invention discloses a life-cycle vibration acceleration signal diagram of a rolling bearing;
FIG. 3: the invention discloses an atomic time domain oscillogram of a stepping-pulse dictionary;
FIG. 4: the rolling bearing of the embodiment of the invention comprises an original vibration signal of a rolling bearing and a reconstructed vibration signal diagram based on a sparse Bayesian method;
FIG. 5: the fusion health factor sequence chart of the rolling bearing of the embodiment of the invention;
FIG. 6: the invention discloses a rolling bearing degradation trend prediction result graph.
Detailed Description
The present invention is further described with reference to the following examples, which are intended to be illustrative and illustrative only, and various modifications, additions and substitutions for the specific embodiments described herein may be made by those skilled in the art without departing from the spirit of the invention or exceeding the scope of the claims.
Example 1:
fig. 1 is a flowchart of a method for on-line monitoring and life prediction of a gearbox of a light rail unit speed reducer according to an embodiment of the invention. As shown in fig. 1, an online monitoring and life prediction system for a speed reducer gearbox of an urban light rail unit specifically includes: the system comprises a remote data acquisition and terminal restoration module, a data preprocessing module, an equipment fault diagnosis and positioning identification module, an equipment key part service life prediction module and a decision analysis module. The system platform can transmit big data in real time, has high fault positioning and identification precision and high response speed, can be applied to the fields of urban light rail and high-speed rail health monitoring and operation and maintenance, and has good market popularization value.
As shown in fig. 1, the on-line monitoring and life predicting method for the reducer gearbox of the light rail set based on the on-line monitoring and life predicting system for the reducer gearbox of the urban light rail set comprises the following steps:
s1: the method comprises the steps that a wireless WIFI acceleration sensor (the model is WD-ACWL500), a force sensor, a microphone and a temperature sensor are utilized to online pick up vibration acceleration signals, force signals, sound signals and real-time health information data of temperature of a gearbox of a speed reducer of the light rail unit in real time.
FIG. 2 is a diagram of a life-cycle vibration acceleration signal of a rolling bearing according to the present invention, wherein the sampling frequency is 25.6 KHz. As can be seen from fig. 2, the bearing operation stages are roughly divided into: a normal operation phase, an early failure phase and a failure worsening phase. In the normal operation stage of the bearing, the vibration amplitude is stable and low; in the early failure stage, the vibration amplitude fluctuates slightly and is slightly higher than that in the normal operation stage; and in the fault deterioration stage, the vibration amplitude is greatly increased, and when the vibration amplitude exceeds a fault threshold value in the later stage, the bearing is in a scrapped state.
S2: the real-time health monitoring information data of the speed reducer gearbox is sparsely represented by utilizing a compressed sensing and dictionary learning technology, the sparsely represented information is remotely transmitted by utilizing a wireless transmission network, and the real-time health monitoring information data is reconstructed in a health monitoring center.
The compressed sensing and dictionary learning technology is a compressed sensing and sparse Bayesian dictionary learning technology, and comprises dictionary construction and algorithm reconstruction, and the specific implementation method comprises the following steps:
a1: and constructing overcomplete dictionary atoms and initializing a dictionary. In order to prevent the distortion of the reconstructed signal and ensure that the waveform of the redundant dictionary atom is matched with the physical structure of the transient pulse signal, a step-pulse dictionary atom is adopted, and the expression is as follows:
d＝η 1 ·a·d imp +η 2 ·d step (1)，
in the formula: a is the step-to-pulse peak ratio, d imp Is a single degree of freedom pulse atom, d step Is a single degree of freedom stepping atom, eta 1 And η 2 Is an amplitude factor used for adjusting the amplitude of the atoms of the step-pulse dictionary to be consistent with the amplitude of the low-frequency resonance component;
d imp and d step These two classes of dictionary atoms are defined as:
in the formula: f. of n For the natural frequency of the system, τ is the damping coefficient of the system, u is the response time of the start of the step-pulse dictionary, and Δ t is the contact time, such as the time interval from the start of the contact of the bearing ball to the departure from the fault point [ see document [1 ]]Sawalhi N.,Randall R.B.Vibration response of spalled rolling element bearing:observations,simulations and signal processing techniques to track the spall size.Mech.Syst.Signal Process.,2011,25:846-870.]。
Fig. 3 is a time-domain waveform diagram of step-pulse dictionary atoms, wherein (a) the diagram is a pulse dictionary single atom diagram, (b) the diagram is a step dictionary single atom diagram, (c) the diagram is a step-pulse dictionary single atom diagram, and (d) the diagram is a step-pulse dictionary atom signal diagram without noise. From the finally constructed step-pulse dictionary atomic signal, as shown in fig. 3, it can be seen that the constructed step-pulse dictionary atomic time domain waveform diagram has two peaks, and a peak appears when the bearing ball starts to contact the fault point and a peak appears when the bearing ball leaves the fault point, which is consistent with the actual bearing fault (especially pitting, pit and crack) operation condition.
A2: and based on the over-complete dictionary, performing sparse representation on the original signal. The method specifically comprises the following steps: and performing sparse representation on the real-time health monitoring information data of the rolling bearing based on the over-complete dictionary, and remotely transmitting sparse representation information by using a wireless transmission network.
A3: in the health monitoring center, a sparse Bayesian algorithm is used to reconstruct original monitoring data [ construct document [2] Qing Li, Wei Hu, Erfei ping, Steven y. liang, Multichannel signals recovery base on tunable Q-factor wave transform-dynamic component analysis and sparse Bayesian estimation for rotating models, control, 2018,20(4),263 ], and fig. 4 is a graph of an original vibration signal of a rolling bearing and a reconstructed vibration signal based on the sparse Bayesian method, wherein (a) is a graph of the original vibration signal, and (b) is a graph of a reconstructed vibration signal based on the sparse Bayesian method. As shown in fig. 4, in order to observe the vibration signal prediction effect, only 1024 point data is given to the vibration signal, that is, 0.04 second, it can be seen that the waveform of the vibration signal reconstructed by the sparse bayes method is substantially consistent with that of the original vibration signal, and the effectiveness of the reconstructed sparse bayes method is proved.
And simultaneously, analyzing the reconstruction result and the error, and optimizing the sparse Bayesian algorithm model parameters.
S3: and in the monitoring operation center, preprocessing and analyzing the reconstructed monitoring data of the gearbox of the unit reducer, including outlier removal, signal filtering, signal denoising or signal decomposition processing, and acquiring pollution-free or near-zero pollution data.
The signal filtering specifically includes low-pass filtering, high-pass filtering, and band-pass filtering. Signal denoising, which specifically includes: wavelet soft threshold denoising and sparse regularization denoising. Signal decomposition processing, which specifically includes: empirical mode decomposition method and variation mode decomposition method.
S4: based on a time domain analysis method and a frequency domain analysis method, fault diagnosis and fault feature extraction are carried out on equipment and key parts thereof, including gears and bearings, so that fault positions and fault signal sources are identified.
The time domain analysis method comprises the steps of extracting the dimension factor characteristics of a time domain signal and extracting the non-dimension factor characteristics. Dimensional factor features, including: maximum, minimum, peak-to-peak, mean, variance, standard deviation, mean square, root mean square RMS, mean square MSE, root mean square error RMSE, root mean square amplitude. Non-dimensional factor features, including: a peak factor, a pulse factor, a margin factor, a kurtosis factor, a waviness factor, and a kurtosis factor and a skewness factor.
The frequency domain analysis method is a spectrum analysis method and comprises the following steps: the method comprises a signal FFT amplitude spectrum analysis method, a Hilbert envelope spectrum analysis method, a Hilbert marginal spectrum analysis method and a rapid spectrum kurtosis analysis method.
S5: based on a time domain/frequency domain feature extraction method, a health factor time sequence of key components of a gearbox of a unit speed reducer is extracted. Specifically, the method comprises the following steps: and (2) performing monotonicity sequencing on 15 time/frequency domain characteristic factor time sequences aiming at 11 time domain and 4 frequency domain characteristic factor time sequences of the vibration acceleration data, selecting the first 5 characteristic factor time sequences with larger monotonicity values, and performing dimensionality reduction and characteristic fusion on the first 5 time domain and frequency domain characteristic factor time sequences with larger monotonicity values by using a principal component analysis method to determine a final fusion health factor sequence of the bearing. FIG. 5 is a sequence diagram of the fusion health factor of a rolling bearing according to the present invention. As shown in fig. 5, the fusion health factor sequence fuses the time sequences of the first 5 time-domain and frequency-domain characteristic factors with larger monotonicity values, and the sequences have an overall ascending trend and conform to the random degradation trend of the bearing.
Meanwhile, a quaternion prediction model is constructed, and the method mainly comprises the following steps:
b1: constructing a target cost function based on a quaternion prediction model;
b2: updating the filter weight of the target cost function by using a gradient descent method;
b3: inputting multi-channel data as input into a quaternion prediction model after weight updating, and predicting the operation trend of a gearbox of a unit speed reducer and key parts of the gearbox in real time;
b4: and obtaining the residual life time of the gear box of the unit speed reducer and key parts of the gear box according to the trend of the prediction curve and the fault threshold value.
The constructed quaternion prediction model is an 1/2 th power quaternion prediction model, and specifically comprises the following steps:
in the formula (4, | e * (n)e(n)e * (n)e(n)| 1/2 Is a term raised to the power of 1/2 as the minimum mean,
is a Gaussian correlation entropy induction penalty term, 1/2 is a fractional order, a weight factor beta is more than or equal to 0 and is used for balancing an augmented minimum mean value 1/2 power term and a Gaussian correlation entropy induction penalty term, w i And (n) is the filter weight.
S6: updating the filter weight of the constructed quaternion prediction model by using a gradient descent method, inputting a multichannel health factor time sequence of the key component of the gearbox of the unit speed reducer into the quaternion prediction model, and predicting the future operation degradation trend and the residual life of the key component of the gearbox of the unit speed reducer.
FIG. 6 is a diagram illustrating the prediction result of the degradation trend of the rolling bearing according to the present invention. As can be seen from fig. 6, the bearing degradation trend predicted by the quaternion prediction method is consistent with the original degradation data.
S7: printing an output report of the analysis result of the last step, and providing a real-time analysis report for a unit operation supervision department and a maintenance department; and performing predictive maintenance on the unit speed reducer equipment and key parts thereof according to the diagnosis and prediction analysis results.
Meanwhile, the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the steps of the online monitoring and service life prediction method for the speed reducer gearbox of the urban light rail unit.
The online monitoring and service life prediction system for the speed reducer gear box of the urban light rail unit can transmit big data in real time, has high fault positioning and identification precision and high response speed, can be applied to the fields of urban light rail and high-speed rail health monitoring and operation and maintenance, and has good market popularization value.
The invention relates to a method for on-line monitoring and service life prediction of a gearbox of a light rail unit speed reducer, which is based on a quaternion model, can predict a multichannel health factor sequence constructed by collected multichannel and high-dimensional data in real time at one time, does not need channel-by-channel or dimension-by-dimension prediction, and has high prediction precision and high prediction response speed. Compared with a remote signal transmission method, the method of the invention does not need a health center to reserve extra-large memory and storage equipment, and can reconstruct the data acquired by the remote subway and light rail set in real time. The method provided by the invention does not need to consider the influence of the physical degradation process, equipment degradation mechanism and complex external working conditions of the subway light rail unit gearbox and key components thereof on the stability of the prediction model, can be applied to the fields of urban light rail and high-speed rail health monitoring and the like, and has good industrial application value.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for on-line monitoring and service life prediction of a gearbox of a light rail unit speed reducer is characterized by comprising the following steps:
s1: the method comprises the steps that a wireless WIFI acceleration sensor, a force sensor, a microphone and a temperature sensor are utilized to pick up vibration acceleration signals, force signals, sound signals and temperature real-time health information data of a gearbox of a speed reducer of the light rail unit on line in real time;
s2: performing sparse representation on real-time health monitoring information data of a gearbox of the speed reducer by using a compressed sensing and dictionary learning technology, remotely transmitting sparse representation information by using a wireless transmission network, and reconstructing the real-time health monitoring information data in a health monitoring center;
s3: in a monitoring operation center, carrying out preprocessing analysis on reconstructed unit reducer gearbox monitoring data, wherein the preprocessing analysis comprises outlier removal, signal filtering, signal denoising or signal decomposition processing, and obtaining pollution-free or near-zero pollution data;
s4: based on a time domain analysis method and a frequency domain analysis method, analyzing the fault degradation degree, diagnosing the fault and extracting the fault characteristics of equipment and key parts thereof, including gears and bearings, so as to identify the fault position and source;
s5: extracting a health factor time sequence of key parts of a gearbox of a unit speed reducer based on a time domain/frequency domain feature extraction method; performing dimension reduction and fusion on the time sequence of the time domain/frequency domain health factor by using a principal component analysis method, and extracting a fusion health factor time sequence of key parts of a gearbox of a unit speed reducer; meanwhile, constructing a quaternion prediction model;
s6: inputting a multi-channel health factor time sequence of a key component of a gearbox of the unit speed reducer into a quaternion prediction model, and predicting a degradation track and the residual life of the key component of the gearbox of the unit speed reducer;
s7: printing an output report according to the analysis result of the last step, and providing a real-time analysis report for a unit operation supervision department and a maintenance department; and performing predictive maintenance on the unit speed reducer equipment and key parts thereof according to the diagnosis and prediction analysis results.
2. The on-line monitoring and life predicting method for the gearbox of the light rail unit reducer according to claim 1, wherein in step S2, the compressed sensing and dictionary learning technology is a compressed sensing and sparse bayesian dictionary learning technology, and comprises dictionary construction, sparse representation and algorithm reconstruction, and the specific implementation method is as follows:
a1: constructing an over-complete dictionary atom and initializing a dictionary;
a2: based on an over-complete dictionary, carrying out sparse representation on an original signal;
a3: and reconstructing the original monitoring data by using a sparse Bayesian algorithm, simultaneously comparing a reconstruction result with the original data, analyzing a comparison error, and optimizing the model parameters of the sparse Bayesian algorithm.
3. The method for on-line monitoring and life prediction of the gearbox of the light rail unit reducer according to claim 1, wherein in step S3, the signal filtering specifically comprises low-pass filtering, high-pass filtering and band-pass filtering; the signal denoising specifically includes: denoising by using a wavelet soft threshold and sparse regularization; the signal decomposition processing specifically includes: empirical mode decomposition method, local mean decomposition and variation mode decomposition method.
4. The method for on-line monitoring and life prediction of the gearbox of the light rail unit reducer according to claim 1, wherein in step S4, the time domain analysis method refers to time domain signal dimension factor feature extraction and non-dimension factor feature extraction; the dimension factor features comprise: maximum value, minimum value, peak-to-peak value, mean value, variance, standard deviation, mean square value, root mean square value RMS, mean square error MSE, root mean square error RMSE and root mean square amplitude; the non-dimensional factor features include: a peak factor, a pulse factor, a margin factor, a kurtosis factor, a waviness factor, and a kurtosis factor and a skewness factor.
5. The method for on-line monitoring and life prediction of the gearbox of the light rail unit reducer according to claim 1, wherein in step S4, the frequency domain analysis method is a spectrum analysis method, and comprises the following steps: the method comprises a signal FFT amplitude spectrum analysis method, a Hilbert envelope spectrum analysis method, a Hilbert marginal spectrum analysis method and a rapid spectrum kurtosis analysis method.
6. The on-line monitoring and life predicting method for the gearbox of the light rail unit reducer according to claim 1, wherein in step S5, the time domain/frequency domain feature extraction method is to extract the time domain feature and the frequency domain feature of the reconstructed monitoring signal, and the time domain feature extraction includes: the average value, standard deviation, skewness, kurtosis, peak-to-peak value, root mean square, peak coefficient factor, waveform coefficient factor, pulse coefficient factor, edge coefficient factor and energy factor, wherein the total index is 11 indexes; the frequency domain feature extraction comprises: spectral kurtosis-mean, spectral kurtosis-standard, spectral kurtosis-skewness, spectral kurtosis-kurtosis, and 4 indexes.
7. The on-line monitoring and life prediction method for the gearbox of the light rail unit speed reducer according to claim 1, wherein in step S5, a quaternion prediction model is constructed, and the method mainly comprises the following steps:
b1: constructing a target cost function based on a quaternion prediction model;
b2: updating the filter weight of the target cost function by using a gradient descent method;
b3: inputting multi-channel data as input into a quaternion prediction model after weight updating, and predicting the degradation operation trend of a gearbox of a unit speed reducer and key parts thereof in real time;
b4: and obtaining the residual life time of the gearbox of the unit speed reducer and key parts of the gearbox according to the degradation operation trend of the prediction curve and the fault threshold value.
8. The method for on-line monitoring and life prediction of the gearbox of the light rail unit reducer according to claim 2, wherein in step S2, in a1, the constructing overcomplete dictionary atom and dictionary initialization are specifically:
in order to prevent the distortion of the reconstructed signal and ensure that the waveform of the redundant dictionary atom is matched with the physical structure of the transient pulse signal, a step-pulse dictionary atom is adopted, and the expression is as follows:
d＝η 1 ·a·d imp +η 2 ·d step (1)，
in the formula: a is the step-to-pulse peak ratio, d imp Is a single degree of freedom pulse atom, d step Is a single degree of freedom stepping atom, eta 1 And η 2 Is an amplitude factor used for adjusting the amplitude of the atoms of the step-pulse dictionary to be consistent with the amplitude of the low-frequency resonance component;
d imp and d step These two classes of dictionary atoms are defined as:
in the formula: f. of n And tau is the system natural frequency, u is the system damping coefficient, u is the starting response time of the step-pulse dictionary, and delta t is the interval time from the bearing ball starting to contact the fault point to leave the fault point.
9. The method for on-line monitoring and life prediction of the gearbox of the light rail unit reducer according to claim 1, wherein the quaternion prediction model constructed in the step S5 is a 1/2 th power quaternion prediction model, which is specifically as follows:
wherein, | e * (n)e(n)e * (n)e(n)| 1/2 Is a term raised to the power of 1/2 as the minimum mean,
is a Gaussian correlation entropy induction penalty term, 1/2 is a fractional order, a weight factor beta is more than or equal to 0 and is used for balancing an augmented minimum mean value 1/2 power term and a Gaussian correlation entropy induction penalty term, w i And (n) is the filter weight.
10. A computer-readable storage medium, storing a computer program which, when executed by a processor, causes the processor to carry out the steps of the method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210735193.XA CN115018181B (en) | 2022-06-27 | Online monitoring and life predicting method for light rail unit speed reducer gearbox |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210735193.XA CN115018181B (en) | 2022-06-27 | Online monitoring and life predicting method for light rail unit speed reducer gearbox |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115018181A true CN115018181A (en) | 2022-09-06 |
| CN115018181B CN115018181B (en) | 2024-06-25 |
Family
ID=
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116226646A (en) * | 2023-05-05 | 2023-06-06 | 国家石油天然气管网集团有限公司 | Method, system, equipment and medium for predicting health state and residual life of bearing |
| CN117647392A (en) * | 2024-01-30 | 2024-03-05 | 成都三一能源环保技术有限公司 | Downhole drilling instrument scrapping monitoring and early warning system based on data analysis |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108280424A (en) * | 2018-01-22 | 2018-07-13 | 北京建筑大学 | A kind of rolling bearing method for predicting residual useful life based on sparse coding |
| WO2021000556A1 (en) * | 2019-07-02 | 2021-01-07 | 中国科学院深圳先进技术研究院 | Method and system for predicting remaining useful life of industrial equipment, and electronic device |
| CN114154736A (en) * | 2021-08-20 | 2022-03-08 | 北京工业大学 | Rolling bearing service life prediction method based on similarity matching optimization theory |
| CN114372324A (en) * | 2021-12-23 | 2022-04-19 | 安徽农业大学 | Method and equipment for predicting service degradation track of key parts of rotary mechanical equipment |
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108280424A (en) * | 2018-01-22 | 2018-07-13 | 北京建筑大学 | A kind of rolling bearing method for predicting residual useful life based on sparse coding |
| WO2021000556A1 (en) * | 2019-07-02 | 2021-01-07 | 中国科学院深圳先进技术研究院 | Method and system for predicting remaining useful life of industrial equipment, and electronic device |
| CN114154736A (en) * | 2021-08-20 | 2022-03-08 | 北京工业大学 | Rolling bearing service life prediction method based on similarity matching optimization theory |
| CN114372324A (en) * | 2021-12-23 | 2022-04-19 | 安徽农业大学 | Method and equipment for predicting service degradation track of key parts of rotary mechanical equipment |
Non-Patent Citations (1)
| Title |
|---|
| 张传凯;: "基于EMD降噪和LSTM网络的地铁风机轴承寿命预测", 风机技术, no. 03, 26 June 2020 (2020-06-26) * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116226646A (en) * | 2023-05-05 | 2023-06-06 | 国家石油天然气管网集团有限公司 | Method, system, equipment and medium for predicting health state and residual life of bearing |
| CN117647392A (en) * | 2024-01-30 | 2024-03-05 | 成都三一能源环保技术有限公司 | Downhole drilling instrument scrapping monitoring and early warning system based on data analysis |
| CN117647392B (en) * | 2024-01-30 | 2024-04-09 | 成都三一能源环保技术有限公司 | Downhole drilling instrument scrapping monitoring and early warning system based on data analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114323644B (en) | Gear box fault diagnosis and signal acquisition method and device and electronic equipment | |
| CN111597682B (en) | Method for predicting remaining life of bearing of gearbox of wind turbine | |
| CN101799674B (en) | Method for analyzing service state of numerical control equipment | |
| CN108444696A (en) | A kind of gearbox fault analysis method | |
| CN114239377A (en) | Method and system for evaluating health state of urban rail electromechanical equipment and storage medium | |
| CN111766512A (en) | Generator fault maintenance system and method | |
| CN113947017A (en) | Method for predicting residual service life of rolling bearing | |
| Martin-del-Campo et al. | Dictionary learning approach to monitoring of wind turbine drivetrain bearings | |
| CN103346797A (en) | Real-time compression method for gear fault signals | |
| CN113391207A (en) | Motor fault detection method, medium and system | |
| CN110969185A (en) | Equipment abnormal state detection method based on data reconstruction | |
| CN111814849A (en) | DA-RNN-based wind turbine generator key component fault early warning method | |
| CN113361782B (en) | Photovoltaic power generation power short-term rolling prediction method based on improved MKPLS | |
| CN112597607A (en) | Predictive maintenance method and system for high-speed bearing of wind driven generator | |
| CN115018181B (en) | Online monitoring and life predicting method for light rail unit speed reducer gearbox | |
| CN115018181A (en) | Online monitoring and service life prediction method for speed reducer gear box of light rail unit | |
| Wu et al. | An improved LSTM neural network with uncertainty to predict remaining useful life | |
| CN113435228A (en) | Motor bearing service life prediction and analysis method based on vibration signal modeling | |
| CN116070368A (en) | Method for predicting residual life of high-speed bearing of offshore wind turbine | |
| CN114936494A (en) | Data physical fusion driven high-temperature component reliability evaluation method and system | |
| Zhou et al. | A fatigue life prediction method for the drive system of wind turbine using internet of things | |
| Peng et al. | Remaining useful life prediction for aircraft engines based on grey model | |
| CN113158535A (en) | Wind turbine generator main bearing temperature prediction method based on deep learning | |
| Zhang et al. | Health status characterization and RUL prediction method of escalator bearing based on VMD-GRU | |
| CN214330806U (en) | Wind driven generator system and fault analysis equipment of wind driven generator |
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 | ||
| GR01 | Patent grant |