CN114004306A - Equipment fault evaluation system and method based on multi-dimensional data of Internet of things - Google Patents

Equipment fault evaluation system and method based on multi-dimensional data of Internet of things Download PDF

Info

Publication number
CN114004306A
CN114004306A CN202111325074.9A CN202111325074A CN114004306A CN 114004306 A CN114004306 A CN 114004306A CN 202111325074 A CN202111325074 A CN 202111325074A CN 114004306 A CN114004306 A CN 114004306A
Authority
CN
China
Prior art keywords
fault
motor
bearing
temperature
vibration
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
CN202111325074.9A
Other languages
Chinese (zh)
Inventor
邓嘉明
侯跃恩
叶忠文
邓文锋
曾军
陈文飞
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.)
Jiaying University
Original Assignee
Jiaying University
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 Jiaying University filed Critical Jiaying University
Priority to CN202111325074.9A priority Critical patent/CN114004306A/en
Publication of CN114004306A publication Critical patent/CN114004306A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
    • G06F18/256—Fusion techniques of classification results, e.g. of results related to same input data of results relating to different input data, e.g. multimodal recognition
    • 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
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (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)
  • Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)

Abstract

The invention discloses an equipment fault evaluation system based on multi-dimensional data of the Internet of things, which is used for collecting and analyzing multi-dimensional data such as mixed sound, axial vibration, radial vibration, bearing temperature and the like in the working process of a motor so as to comprehensively evaluate the fault of the motor. The motor maintenance danger assessment coefficient of the motor which is displayed by the multi-dimensional data and used for continuing to work is obtained by acquiring the multi-dimensional data of the motor such as noise, vibration, temperature and the like and analyzing the multi-dimensional data, so that the danger degree of the motor which is used continuously after the motor is in failure is reflected, the quantitative display of the danger degree is realized, the motor failure is conveniently and comprehensively assessed and predicted, and the functions of early warning and maintenance are achieved.

Description

Equipment fault evaluation system and method based on multi-dimensional data of Internet of things
Technical Field
The invention belongs to the technical field of multidimensional data processing, and relates to an equipment fault evaluation system and method based on multidimensional data of the Internet of things.
Background
An electric machine, commonly known as a "motor", refers to an electromagnetic device that converts or transmits electric energy according to the law of electromagnetic induction, and the electric machine mainly functions in a circuit to generate driving torque as a power source for electrical appliances or various machines, and a generator mainly functions in a circuit to convert mechanical energy into electric energy.
At present, only the state of a motor is monitored and managed in the use process of the motor, which is generally represented as recording and storing the operation parameters of the motor, the failure in the operation process of the motor is not predicted, so that the accuracy and the reliability of the failure prediction in the use process of the motor are poor, meanwhile, the motor is interfered by various failures in the use process of the motor, the service life of the motor is influenced, but the judgment of the current motor failure is judged only by manual experience, experimental data are not available, the motor failure cannot be predicted in advance, the motor failure cannot be accurately evaluated, the service life of the motor which continuously works cannot be predicted according to the current failure of the motor, the defects of motor monitoring and failure evaluation are caused, the service life of the motor is shortened, and the motor cannot be maintained in time.
Disclosure of Invention
The invention aims to provide an equipment fault evaluation system based on multi-dimensional data of the Internet of things, which solves the problems in the prior art.
The purpose of the invention can be realized by the following technical scheme:
an equipment fault evaluation system based on multi-dimensional data of the Internet of things comprises a noise source division module, a vibration detection module, a bearing information detection module, a frequency spectrum characteristic classification processing module, a fault chain training interference module, a fault prediction judgment module and a multi-dimensional data evaluation module;
the noise source division module collects mixed sound in the working process of the motor in real time, and performs frequency spectrum analysis on the mixed sound in the working process of the motor by adopting Fourier transform to obtain the frequency spectrum characteristics of the separated bearing noise and the frequency spectrum characteristics of mechanical noise generated when the motor works;
the vibration detection module adopts an eddy current type displacement sensor and respectively collects the axial vibration amplitude and the radial vibration amplitude of the motor in real time; the bearing information detection module adopts a temperature sensor, is arranged on the bearing and is used for collecting the temperature of the surface of the bearing;
the frequency spectrum characteristic classification processing module receives the frequency spectrum characteristics of the bearing noise separated by the noise source distinguishing module and the frequency spectrum characteristics of the mechanical noise generated when the motor works, and compares the frequency spectrum characteristics of the bearing noise and the frequency spectrum characteristics of the mechanical noise with the frequency spectrum characteristics of the bearing noise under different bearing noise fault levels and the frequency spectrum characteristics of the mechanical noise under different mechanical noise fault levels in sequence to screen out the bearing noise fault level corresponding to the separated bearing noise and the mechanical noise fault level corresponding to the mechanical noise;
the method comprises the steps that a fault prediction judging module extracts an axial vibration amplitude and a radial vibration amplitude of a motor, a time domain vibration signal diagram is established according to the obtained axial vibration amplitude and the radial vibration amplitude, time domain vibration signals in the time domain vibration signal diagram are analyzed through Fourier transform, the frequency and the phase of axial vibration and the frequency and the phase of radial vibration are obtained, basic parameters of axial vibration and basic parameters of radial vibration are sequentially analyzed one by one, whether the motor vibration is abnormal or not is preliminarily predicted, the abnormal degree of the motor vibration is judged, the temperature of the surface of a bearing is received, the temperature of the surface of the bearing is drawn into a temperature change curve, the maximum speed of temperature rise of the surface of the bearing is counted, the accumulated time when the temperature of the bearing is larger than a preset temperature W, and the temperature of the surface of the bearing after the temperature of the bearing is larger than the preset temperature W are counted, and the estimated coefficient of a motor analysis shaft is determined through the abnormal degree of the motor vibration and the related parameter information of the temperature of the bearing (ii) a
The multidimensional data evaluation module extracts the bearing noise fault level and the mechanical noise fault level screened by the frequency spectrum characteristic classification processing module, sequentially screens a bearing fault evaluation coefficient corresponding to the bearing noise fault level and a mechanical fault evaluation coefficient corresponding to the mechanical noise fault level which are mapped with the bearing noise fault level and the mechanical noise fault level according to the bearing noise fault level and the mechanical noise fault level, acquires the motor vibration abnormal degree, the motor shaft-holding prediction coefficient and the bearing surface temperature after the bearing temperature is higher than a preset temperature W which are analyzed by the fault prediction judgment module, and predicts a motor maintenance danger evaluation coefficient of the current motor which continuously works by adopting a multidimensional data evaluation model.
Preferably, the method for determining the abnormal degree of vibration of the motor includes the following steps:
step 1, extracting the amplitude, frequency and phase of axial vibration and radial vibration;
step 2, judging whether the amplitude of the axial vibration is larger than k times of the amplitude of the radial vibration, if so, marking the risk factor of the amplitude of the motor vibration as lambda 1, and obtaining a preliminary numerical value of 1.32 through experiments, otherwise, marking as lambda 2, and obtaining a preliminary numerical value of 0.586 through experiments;
step 3, analyzing the ratio v between the axial vibration frequency f1 and the radial vibration frequency f2, and analyzing the phase difference psi (2 pi f) between the axial vibration phase and the radial vibration phase1T+w1)-(2πf2T + w2), T being time, w1 and w2 being the initial phase of axial vibration and the initial phase of radial vibration, respectively;
step 4, combining the data statistics in the step 2 and the step 3 to count the abnormal vibration coefficient
Figure BDA0003346654850000021
r is λ 1 or λ 2, v ═ f1/f 2.
Preferably, the motor shaft holding prediction coefficient
Figure BDA0003346654850000031
The calculation formula of (a) is as follows:
Figure BDA0003346654850000032
eta 1 represents a proportionality coefficient of motor shaft-holding caused by vibration, eta 2 represents a proportionality coefficient of motor shaft-holding caused by bearing temperature, eta 1+ eta 2 is 1, and beta1Expressed as the associated disturbance coefficient, beta, of the motor vibration anomaly caused by the bearing temperature2Expressed as the associated interference coefficient of the bearing temperature rise caused by the abnormal vibration of the motor, T is expressed as the accumulated time length of the bearing temperature being greater than the preset temperature W, TPreparation ofWhen the preset bearing temperature is higher than the upper limit of the preset temperature WLength, t1 and t2 respectively indicate a time point corresponding to the bearing temperature being equal to the preset temperature W and a time point corresponding to the temperature being greater than the preset temperature W, t2 being greater than t1, smaxExpressed as the maximum speed of temperature rise of the bearing surface, and W' expressed as the bearing surface temperature after the bearing temperature is greater than a preset temperature W.
Preferably, the multidimensional data evaluation module is based on comprehensive evaluation of data acquired by various sensors after processing, and the multidimensional data evaluation model is
Figure BDA0003346654850000033
a1, a2, a3 and a4 are respectively weight coefficients corresponding to bearing noise, mechanical noise, motor shaft seizure and bearing temperature fault types, a1+ a2+ a3+ a4 is 1, X and Y are respectively a bearing fault evaluation coefficient and a mechanical fault evaluation coefficient, E is a motor abnormal vibration coefficient, n is 4, T is T, andpreparation ofFor an upper limit duration for which the predetermined bearing temperature is greater than the predetermined temperature WmaxIs the maximum temperature that can be tolerated by the bearing surface,
Figure BDA0003346654850000034
is the accumulated amount of the bearing surface temperature over time after the bearing temperature is greater than the preset temperature W,
Figure BDA0003346654850000035
for the interference coefficient associated with the aj-th fault category to the ai-th fault category, i is 1,2,3,4, i.e. a1, a2, a3, a4 are bearing noise, mechanical noise, motor shaft seizure and bearing temperature anomaly, respectively, and when i is j,
Figure BDA0003346654850000036
equal to 0.
Preferably, the equipment fault evaluation system further includes a fault chain training interference module, the fault chain training interference module obtains the occurrence frequency of each fault type in the motor fault type set a { a1, a 2.,. ai.,. am } during training time, performs normalization analysis on the occurrence frequency of each fault type, and analyzes the weight of each fault type
Figure BDA0003346654850000037
And counting the interference influence times C among the fault types according to the occurrence sequence of the fault typesai→ajTo count the correlation interference coefficient between each fault category
Figure BDA0003346654850000041
XaiThe number of occurrences of the ai fault category over the training duration.
Preferably, the fault chain training interference module performs clustering processing on the interference influence times among fault types by adopting a clustering analysis method, and analyzes the correlation interference coefficient among fault types, and specifically comprises the following steps;
s1, after each fault type is simulated and trained for K times, the weight of each fault type in the training process is counted, and the weight of each fault type is equal to the ratio of the number of times of the fault type appearing in the K times of training to the sample training number of times K;
s2, primarily screening Z fault categories as clustering centers;
s3, establishing an objective function
Figure BDA0003346654850000042
Z is the number of clustering centers, Z is the number of fault categories,
Figure BDA0003346654850000043
the correlation interference influence degree between the fault class of the sample at the d-th time and the g-th clustering center is delta, which is the sum of the weights corresponding to all fault classes simulated and trained in the step S1, pdgDistance between the d sample fault test and the g cluster center, qdWeighting the fault type corresponding to the d sample fault test;
s4, respectively deducing an associated interference influence matrix and a clustering center iteration formula of the target function by adopting a Lagrange multiplier method:
Figure BDA0003346654850000044
Dgis the g fault speciesClass-corresponding cluster center, RgWeights corresponding to the training sample fault types to be classified;
s5, screening out the correlation interference coefficient between each fault type in the correlation interference influence matrix and the clustering center, and establishing a fault chain for each fault type with the correlation interference coefficient larger than 0.
Preferably, the equipment fault evaluation system further comprises a predictive tracking damage module, wherein the predictive tracking damage module is used for extracting a motor maintenance risk evaluation coefficient in the current motor working state, which is obtained by analysis of the multidimensional data evaluation module, and calculating the motor fault surge acceleration according to the motor maintenance risk evaluation coefficient in the current motor working state and the motor maintenance risk evaluation coefficient in the interval duration t3
Figure BDA0003346654850000045
And tracking and predicting the service life of the motor maintained by the continuous work of the motor according to the current motor fault surge coefficient
Figure BDA0003346654850000046
Gt3Maintaining a risk assessment factor for the motor at time t3, GmaxA risk assessment factor is maintained for the maximum motor allowed for the motor.
Preferably, the equipment fault assessment method based on the multidimensional data of the internet of things comprises the following specific steps:
s1, collecting the bearing temperature and the mixed noise in the motor operation process, and separating the mixed noise to obtain the bearing noise and the mechanical noise;
s2, screening the bearing noise and the mechanical noise to obtain the bearing noise fault level and the mechanical noise fault level;
s3, collecting the axial vibration amplitude and the radial vibration amplitude of the motor to establish a time domain vibration signal diagram, and analyzing the basic parameters of the axial vibration and the basic parameters of the radial vibration through the time domain vibration signal diagram;
s4, extracting basic parameters of axial vibration and radial vibration in the step S3 to judge the abnormal degree of the motor vibration;
s5, processing the temperature of the bearing surface to obtain the maximum speed of the temperature rise of the bearing surface, the accumulated time length of the bearing temperature being greater than the preset temperature W and the temperature of the bearing surface after the bearing temperature is greater than the preset temperature W;
s6, analyzing the seizure prediction coefficient of the motor by combining the data in the step S4 and the step S5, and judging the motor maintenance danger evaluation coefficient under the condition that the motor continues to work by adopting a multi-dimensional data evaluation model.
The invention has the beneficial effects that:
the motor maintenance risk assessment coefficient is maintained to the motor that this application was through collecting multidimensional data such as the noise of motor, vibration and temperature and carry out the analysis of multidimensional data to the motor that obtains multidimensional data to show continues to work, has embodied the dangerous degree of continuing to use after the motor trouble, and realizes the quantization show to dangerous degree, is convenient for carry out comprehensive assessment and prediction to the motor trouble, reaches early warning in advance and maintains the effect.
This application is through carrying out integrated processing to motor vibration and bearing temperature to predict out motor vibration and bearing temperature and produce the possibility that the motor embraced the axle, can embrace the axle to the motor in advance and predict and handle, reduce the probability that the motor embraced the axle, and then improve the durable degree of motor, avoid the motor to embrace the damage that the axle caused to the motor.
According to the method and the device, the various fault types causing the motor faults are subjected to cluster analysis to obtain the correlation interference coefficients among the various fault types, the quantitative degree of the correlation degree is provided for the mutual influence judgment among the faults, and then the reliable correlation basis is provided for the motor fault evaluation process.
The motor maintenance risk evaluation coefficients of two different time points are obtained, the motor fault surge acceleration is analyzed according to the motor maintenance risk evaluation coefficients of the two different time points, the service life of the motor maintained according to the current motor fault surge coefficient continuous work is predicted, the service life of the motor in the non-maintenance state is predicted, the accuracy of motor service life prediction is improved, and personnel are promoted to maintain motor faults in time.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to 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 one
An equipment fault evaluation system based on multi-dimensional data of the Internet of things comprises a noise source division module, a vibration detection module, a bearing information detection module, a frequency spectrum characteristic classification processing module, a fault chain training interference module, a fault prediction judgment module and a multi-dimensional data evaluation module.
The motor is referred to equipment in this application, and the trouble kind of motor is many, and the trouble that the motor took place mainly derives from trouble kinds such as bearing noise, mechanical noise, motor armful of axle and bearing temperature anomaly, consequently this application focuses on the analysis of above four kinds of trouble, can protect the fault situation that the maintainer judged the motor and carry out failure assessment and prediction through the screening to above four kinds of trouble.
The noise source area division module collects mixed sound in the working process of the motor in real time, and performs frequency spectrum analysis on the mixed sound in the working process of the motor by adopting Fourier transform to obtain frequency spectrum characteristics of the separated bearing noise and frequency spectrum characteristics of mechanical noise generated when the motor works, wherein the frequency spectrum characteristics comprise amplitude, power and phase.
The vibration detection module adopts an eddy current type displacement sensor, the axial vibration amplitude and the radial vibration amplitude of the motor are respectively collected in real time, and the collected axial vibration amplitude and the collected radial vibration amplitude are sent to the fault prediction judgment module.
The bearing information detection module adopts a temperature sensor, is arranged on a bearing, collects the temperature of the surface of the bearing and sends the collected temperature of the surface of the bearing to the failure prediction judgment module.
The frequency spectrum characteristic classifying and processing module receives the frequency spectrum characteristic of the bearing noise separated by the noise source distinguishing module and the frequency spectrum characteristic of the mechanical noise generated when the motor works, and compares the frequency spectrum characteristic of the bearing noise and the frequency spectrum characteristic of the mechanical noise with the frequency spectrum characteristic of the bearing noise under different bearing noise fault levels and the frequency spectrum characteristic of the mechanical noise under different mechanical noise fault levels in sequence respectively to screen out the bearing noise fault level corresponding to the separated bearing noise and the mechanical noise fault level corresponding to the mechanical noise, so that the fault level classifying and processing corresponding to the sound in the frequency spectrum characteristic are realized, and the qualitative processing of the fault level is realized.
The method comprises the steps that a fault prediction judging module extracts an axial vibration amplitude and a radial vibration amplitude of a motor, a time domain vibration signal diagram is established according to the obtained axial vibration amplitude and the radial vibration amplitude, time domain vibration signals in the time domain vibration signal diagram are analyzed through Fourier transform, the frequency and the phase of axial vibration and the frequency and the phase of radial vibration are obtained, basic parameters of axial vibration and basic parameters of radial vibration are sequentially analyzed one by one, whether the motor vibration is abnormal or not is preliminarily predicted, the abnormal degree of the motor vibration is judged, the temperature of the surface of a bearing is received, the temperature of the surface of the bearing is drawn into a temperature change curve, the maximum speed of temperature rise of the surface of the bearing is counted, the accumulated time when the temperature of the bearing is larger than a preset temperature W, and the temperature of the surface of the bearing after the temperature of the bearing is larger than the preset temperature W are counted, and the estimated coefficient of a motor analysis shaft is determined through the abnormal degree of the motor vibration and the related parameter information of the temperature of the bearing The motor shaft holding prediction coefficient is used for reflecting the possibility that the motor is held under the common influence of motor vibration and bearing surface temperature received by the motor, the prior art only knows that the motor shaft holding is related to the motor vibration and the bearing temperature, but the motor vibration and the bearing surface temperature cannot be combined to carry out the prediction and evaluation of the motor shaft holding, and the comprehensive prediction judgment that the motor vibration and the bearing temperature are combined is realized, the possibility that the motor shaft holding is held can be relatively accurately analyzed, the prior art is prevented from only depending on manual experience, the motor cannot be predicted in advance, the possibility that the motor shaft holding is carried out is reduced, and the durability of the motor is improved.
When the motor vibrates abnormally, the motor bearing is indirectly caused to break down, so that the motor fault can be intuitively and accurately mastered by quantitatively judging the abnormal degree of the motor vibration.
The method for judging the abnormal degree of the motor vibration comprises the following specific steps:
step 1, extracting the amplitude, frequency and phase of axial vibration and radial vibration;
step 2, judging whether the amplitude of the axial vibration is larger than k times of the amplitude of the radial vibration, if so, marking the risk factor of the amplitude of the motor vibration as lambda 1, and obtaining a preliminary numerical value of 1.32 through experiments, otherwise, marking as lambda 2, and obtaining a preliminary numerical value of 0.586 through experiments;
step 3, analyzing the ratio v between the axial vibration frequency f1 and the radial vibration frequency f2, and analyzing the phase difference psi (2 pi f) between the axial vibration phase and the radial vibration phase1T+w1)-(2πf2T + w2), T being time, w1 and w2 being the initial phase of axial vibration and the initial phase of radial vibration, respectively;
step 4, combining the data statistics in the step 2 and the step 3 to count the abnormal vibration coefficient
Figure BDA0003346654850000071
r is lambda 1 or lambda 2, v is f1/f2, the abnormal vibration coefficient reflects the abnormal degree of the motor vibration, and the larger the abnormal vibration coefficient is, the larger the abnormal degree of the motor vibration is, the higher the possibility of axle seizure caused by the abnormal vibration of the motor is.
Wherein, the motor shaft holding pre-estimated coefficient
Figure BDA0003346654850000072
The calculation formula of (a) is as follows:
Figure BDA0003346654850000073
eta 1 represents a proportionality coefficient of motor shaft-holding caused by vibration, eta 2 represents a proportionality coefficient of motor shaft-holding caused by bearing temperature, eta 1+ eta 2 is 1, and beta1Expressed as the associated disturbance coefficient, beta, of the motor vibration anomaly caused by the bearing temperature2Expressed as the associated interference coefficient of the bearing temperature rise caused by the abnormal vibration of the motor, T is expressed as the accumulated time length of the bearing temperature being greater than the preset temperature W, TPreparation ofThe upper limit duration that the preset bearing temperature is greater than the preset temperature W is represented, t1 and t2 represent a time point corresponding to the bearing temperature being equal to the preset temperature W and a time point corresponding to the temperature being greater than the preset temperature W, t2 is greater than t1, and s is greater thanmaxExpressed as the maximum speed of temperature rise of the bearing surface, and W' expressed as the bearing surface temperature after the bearing temperature is greater than a preset temperature W.
The multidimensional data evaluation module extracts the bearing noise fault level and the mechanical noise fault level screened by the frequency spectrum characteristic classification processing module, sequentially screens a bearing fault evaluation coefficient corresponding to the bearing noise fault level and a mechanical fault evaluation coefficient corresponding to the mechanical noise fault level which are mapped with the bearing noise fault level and the mechanical noise fault level according to the bearing noise fault level and the mechanical noise fault level, acquires the motor vibration abnormal degree, the motor shaft holding prediction coefficient and the bearing surface temperature after the bearing temperature is higher than a preset temperature W which are analyzed by the fault prediction judgment module, predicts the current fault hazard degree of the motor by adopting a multidimensional data evaluation model, obtains a motor maintenance hazard evaluation coefficient G for the continuous work of the current motor, reflects the hazard degree of the continuous use of the motor after the motor breaks down so as to quantitatively display the hazard degree, the multidimensional data evaluation module adopts multidimensional detection data to comprehensively detect the motor, so that comprehensive data analysis is carried out by combining faults of bearing temperature, mechanical noise, bearing noise, motor shaft seizure and the like, comprehensive evaluation of motor faults is realized, and the harm degree of continuous work of the motor to the motor under the current motor fault can be optimally displayed.
The bearing noise fault level and the bearing fault evaluation coefficient are mapped with each other, the mechanical noise fault level and the mechanical fault evaluation coefficient are also mapped with each other, and the bearing fault evaluation coefficient and the mechanical fault evaluation coefficient respectively represent the probability of the motor fault under the bearing noise fault level and the probability of the motor fault under the mechanical noise level.
The multidimensional data evaluation module is used for realizing the acquisition and analysis of multidimensional data based on the comprehensive evaluation of the data acquired by various sensors after the data are processed so as to improve the evaluation accuracy of motor faults, and the multidimensional data evaluation model is
Figure BDA0003346654850000081
a1, a2, a3 and a4 are respectively weight coefficients corresponding to bearing noise, mechanical noise, motor shaft seizure and bearing temperature fault types, a1+ a2+ a3+ a4 is 1, X and Y are respectively a bearing fault evaluation coefficient and a mechanical fault evaluation coefficient, E is a motor abnormal vibration coefficient, n is 4, T is T, andpreparation ofFor an upper limit duration for which the predetermined bearing temperature is greater than the predetermined temperature WmaxIs the maximum temperature that can be tolerated by the bearing surface,
Figure BDA0003346654850000082
is the accumulated amount of the bearing surface temperature over time after the bearing temperature is greater than the preset temperature W,
Figure BDA0003346654850000083
for the interference coefficient associated with the aj-th fault category to the ai-th fault category, i is 1,2,3,4, i.e. a1, a2, a3, a4 are bearing noise, mechanical noise, motor shaft seizure and bearing temperature anomaly, respectively, and when i is j,
Figure BDA0003346654850000084
equal to 0.
Example two
And preliminarily analyzing the association degree among the fault types according to the sequence of the occurrence of the fault types, namely designing a fault chain training interference module.
The fault chain training interference module acquires the occurrence frequency of each fault type in a motor fault type set A { a1, a 2.,. ai.,. as, am } under the training duration, performs normalization analysis on the occurrence frequency of each fault type, and analyzes the weight of each fault type
Figure BDA0003346654850000091
And counting the interference influence times C among the fault types according to the occurrence sequence of the fault typesai→ajTo count the correlation interference coefficient between each fault category
Figure BDA0003346654850000092
XaiThe number of occurrences of the ai fault category over the training duration.
EXAMPLE III
Because the related faults are causally related before and after, when a certain fault occurs, another fault which is related mutually also occurs, in order to improve the accuracy of motor fault evaluation, a mean value clustering algorithm is adopted to cluster the interference influence times among fault types to obtain the related interference coefficients among the fault types, and compared with the second embodiment, the statistics of the related interference coefficients among the fault types is higher in accuracy, and artificial subjective factors are eliminated.
Simulating K sample fault tests for each fault type, and conveniently counting the times of the ai fault type causing the aj fault type of the motor when the ai fault type occurs to the motor so as to obtain the interference influence times C among the fault typesai→aj,Cai →aj≤K。
The fault chain training interference module adopts a clustering analysis method to cluster the interference influence times among fault types and analyzes the correlation interference coefficient among the fault types, and the method specifically comprises the following steps;
s1, after each fault type is simulated and trained for K times, the weight of each fault type in the training process is counted, and the weight of each fault type is equal to the ratio of the number of times of the fault type appearing in the K times of training to the sample training number of times K;
s2, primarily screening Z fault categories as clustering centers;
s3, establishing an objective function
Figure BDA0003346654850000093
Z is the number of cluster centers, Z isThe number of the types of the faults,
Figure BDA0003346654850000094
the correlation interference influence degree between the fault class of the sample at the d-th time and the g-th clustering center is delta, which is the sum of the weights corresponding to all fault classes simulated and trained in the step S1, pdgDistance between the d sample fault test and the g cluster center, qdWeighting the fault type corresponding to the d sample fault test;
s4, respectively deducing an associated interference influence matrix and a clustering center iteration formula of the target function by adopting a Lagrange multiplier method:
Figure BDA0003346654850000101
Dgfor the cluster center corresponding to the g-th fault category, RgWeights corresponding to the training sample fault types to be classified;
s5, screening out the correlation interference coefficient between each fault type in the correlation interference influence matrix and the clustering center, and establishing a fault chain for each fault type with the correlation interference coefficient larger than 0.
The mean value clustering algorithm is adopted to cluster the interference influence times among all fault types, so that the correlation interference degree among all fault types can be accurately analyzed, the quantitative embodiment of the correlation degree is provided for the mutual influence judgment among the faults, reliable correlation interference factors are provided for the motor fault evaluation, and the accuracy of the motor fault evaluation is further improved.
Example four
The prediction tracking damage module is used for extracting a motor maintenance risk evaluation coefficient in the current motor working state obtained by analysis of the multidimensional data evaluation module, and calculating the motor fault surge acceleration according to the motor maintenance risk evaluation coefficient in the current motor state and the motor maintenance risk evaluation coefficient under the interval duration t3
Figure BDA0003346654850000102
And tracking and predicting the service life of the motor maintained by the continuous work of the motor according to the current motor fault surge coefficient
Figure BDA0003346654850000103
Gt3Maintaining a risk assessment factor for the motor at time t3, GmaxAnd maintaining the danger evaluation coefficient for the maximum motor allowed by the motor, wherein the larger the motor maintenance danger evaluation coefficient of the motor in the state of the motor is, the higher the danger possibility of the motor for continuous use is indicated.
EXAMPLE five
An equipment fault assessment method based on multi-dimensional data of the Internet of things comprises the following specific steps:
s1, collecting the bearing temperature and the mixed noise in the motor operation process, and separating the mixed noise to obtain the bearing noise and the mechanical noise;
s2, screening the bearing noise and the mechanical noise to obtain the bearing noise fault level and the mechanical noise fault level;
s3, collecting the axial vibration amplitude and the radial vibration amplitude of the motor to establish a time domain vibration signal diagram, and analyzing the basic parameters of the axial vibration and the basic parameters of the radial vibration through the time domain vibration signal diagram;
s4, extracting basic parameters of axial vibration and radial vibration in the step S3 to judge the abnormal degree of the motor vibration;
s5, processing the temperature of the bearing surface to obtain the maximum speed of the temperature rise of the bearing surface, the accumulated time length of the bearing temperature being greater than the preset temperature W and the temperature of the bearing surface after the bearing temperature is greater than the preset temperature W;
and S6, analyzing the seizing prediction coefficient of the motor by combining the data in the step S4 and the step S5, and judging the motor maintenance danger evaluation coefficient under the continuous work of the current motor by adopting a multi-dimensional data evaluation model so as to reflect the damage to the motor caused by the continuous work of the motor.
The foregoing is merely exemplary and illustrative of the principles of the present invention and various modifications, additions and substitutions of the specific embodiments described herein may be made by those skilled in the art without departing from the principles of the present invention or exceeding the scope of the claims set forth herein.

Claims (8)

1. The utility model provides an equipment trouble evaluation system based on thing networking multidimension degree data which characterized in that: the system comprises a noise source division module, a vibration detection module, a bearing information detection module, a frequency spectrum characteristic classification processing module, a fault chain training interference module, a fault prediction judgment module and a multi-dimensional data evaluation module;
the noise source division module collects mixed sound in the working process of the motor in real time, and performs frequency spectrum analysis on the mixed sound in the working process of the motor by adopting Fourier transform to obtain the frequency spectrum characteristics of the separated bearing noise and the frequency spectrum characteristics of mechanical noise generated when the motor works;
the vibration detection module adopts an eddy current type displacement sensor and respectively collects the axial vibration amplitude and the radial vibration amplitude of the motor in real time; the bearing information detection module adopts a temperature sensor, is arranged on the bearing and is used for collecting the temperature of the surface of the bearing;
the frequency spectrum characteristic classification processing module receives the frequency spectrum characteristics of the bearing noise separated by the noise source distinguishing module and the frequency spectrum characteristics of the mechanical noise generated when the motor works, and compares the frequency spectrum characteristics of the bearing noise and the frequency spectrum characteristics of the mechanical noise with the frequency spectrum characteristics of the bearing noise under different bearing noise fault levels and the frequency spectrum characteristics of the mechanical noise under different mechanical noise fault levels in sequence to screen out the bearing noise fault level corresponding to the separated bearing noise and the mechanical noise fault level corresponding to the mechanical noise;
the method comprises the steps that a fault prediction judging module extracts an axial vibration amplitude and a radial vibration amplitude of a motor, a time domain vibration signal diagram is established according to the obtained axial vibration amplitude and the radial vibration amplitude, time domain vibration signals in the time domain vibration signal diagram are analyzed through Fourier transform, the frequency and the phase of axial vibration and the frequency and the phase of radial vibration are obtained, basic parameters of axial vibration and basic parameters of radial vibration are sequentially analyzed one by one, whether the motor vibration is abnormal or not is preliminarily predicted, the abnormal degree of the motor vibration is judged, the temperature of the surface of a bearing is received, the temperature of the surface of the bearing is drawn into a temperature change curve, the maximum speed of temperature rise of the surface of the bearing is counted, the accumulated time when the temperature of the bearing is larger than a preset temperature W, and the temperature of the surface of the bearing after the temperature of the bearing is larger than the preset temperature W are counted, and the estimated coefficient of a motor analysis shaft is determined through the abnormal degree of the motor vibration and the related parameter information of the temperature of the bearing (ii) a
The multidimensional data evaluation module extracts the bearing noise fault level and the mechanical noise fault level screened by the frequency spectrum characteristic classification processing module, sequentially screens a bearing fault evaluation coefficient corresponding to the bearing noise fault level and a mechanical fault evaluation coefficient corresponding to the mechanical noise fault level which are mapped with the bearing noise fault level and the mechanical noise fault level according to the bearing noise fault level and the mechanical noise fault level, acquires the motor vibration abnormal degree, the motor shaft-holding prediction coefficient and the bearing surface temperature after the bearing temperature is higher than a preset temperature W which are analyzed by the fault prediction judgment module, and predicts a motor maintenance danger evaluation coefficient of the current motor which continuously works by adopting a multidimensional data evaluation model.
2. The system for evaluating the equipment fault based on the multidimensional data of the internet of things according to claim 1, wherein: the method for judging the abnormal degree of the motor vibration comprises the following specific steps:
step 1, extracting the amplitude, frequency and phase of axial vibration and radial vibration;
step 2, judging whether the amplitude of the axial vibration is larger than k times of the amplitude of the radial vibration, if so, marking the risk factor of the amplitude of the motor vibration as lambda 1, and obtaining a preliminary numerical value of 1.32 through experiments, otherwise, marking as lambda 2, and obtaining a preliminary numerical value of 0.586 through experiments;
step 3, analyzing the ratio v between the axial vibration frequency f1 and the radial vibration frequency f2, and analyzing the phase difference psi (2 pi f) between the axial vibration phase and the radial vibration phase1T+w1)-(2πf2T + w2), T being time, w1 and w2 being the initial phase of axial vibration and the initial phase of radial vibration, respectively;
step 4, combining the data statistics in the step 2 and the step 3 to count the abnormal vibration coefficient
Figure FDA0003346654840000021
r is λ 1 or λ 2, v ═ f1/f 2.
3. The system for evaluating the equipment fault based on the multidimensional data of the internet of things according to claim 2, wherein: the motor shaft holding pre-estimated coefficient
Figure FDA0003346654840000022
The calculation formula of (a) is as follows:
Figure FDA0003346654840000023
eta 1 represents a proportionality coefficient of motor shaft-holding caused by vibration, eta 2 represents a proportionality coefficient of motor shaft-holding caused by bearing temperature, eta 1+ eta 2 is 1, and beta1Expressed as the associated disturbance coefficient, beta, of the motor vibration anomaly caused by the bearing temperature2Expressed as the associated interference coefficient of the bearing temperature rise caused by the abnormal vibration of the motor, T is expressed as the accumulated time length of the bearing temperature being greater than the preset temperature W, TPreparation ofThe upper limit duration that the preset bearing temperature is greater than the preset temperature W is represented, t1 and t2 represent a time point corresponding to the bearing temperature being equal to the preset temperature W and a time point corresponding to the temperature being greater than the preset temperature W, t2 is greater than t1, and s is greater thanmaxExpressed as the maximum speed of temperature rise of the bearing surface, and W' expressed as the bearing surface temperature after the bearing temperature is greater than a preset temperature W.
4. The system of claim 3, wherein the system comprises: the multi-dimensional data evaluation module is based on the comprehensive evaluation of data acquired by various sensors after the data are processed, and the multi-dimensional data evaluation model is
Figure FDA0003346654840000024
a1, a2, a3 and a4 are respectively weight coefficients corresponding to bearing noise, mechanical noise, motor shaft seizure and bearing temperature fault types, a1+ a2+ a3+ a4 is 1, X and Y are respectively a bearing fault evaluation coefficient and a mechanical fault evaluation coefficient, E is a motor abnormal vibration coefficient, n is 4, T is T, andpreparation ofFor an upper limit duration for which the predetermined bearing temperature is greater than the predetermined temperature WmaxIs the maximum temperature that can be tolerated by the bearing surface,
Figure FDA0003346654840000031
is the accumulated amount of the bearing surface temperature over time after the bearing temperature is greater than the preset temperature W,
Figure FDA0003346654840000032
for the interference coefficient associated with the aj-th fault category to the ai-th fault category, i is 1,2,3,4, i.e. a1, a2, a3, a4 are bearing noise, mechanical noise, motor shaft seizure and bearing temperature anomaly, respectively, and when i is j,
Figure FDA0003346654840000033
equal to 0.
5. The system of claim 4, wherein the system comprises: the equipment fault evaluation system further comprises a fault chain training interference module, the fault chain training interference module acquires the occurrence frequency of each fault type in a motor fault type set A { a1, a2, · ai, · as,. am } under training duration, performs normalization analysis on the occurrence frequency of each fault type, and analyzes the weight of each fault type
Figure FDA0003346654840000034
And counting the interference influence times C among the fault types according to the occurrence sequence of the fault typesai→ajTo count the correlation interference coefficient between each fault category
Figure FDA0003346654840000035
XaiThe number of occurrences of the ai fault category over the training duration.
6. The system of claim 4, wherein the system comprises: the fault chain training interference module adopts a clustering analysis method to cluster the interference influence times among fault types and analyzes the correlation interference coefficient among the fault types, and the method specifically comprises the following steps;
s1, after each fault type is simulated and trained for K times, the weight of each fault type in the training process is counted, and the weight of each fault type is equal to the ratio of the number of times of the fault type appearing in the K times of training to the sample training number of times K;
s2, primarily screening Z fault categories as clustering centers;
s3, establishing an objective function
Figure FDA0003346654840000036
Z is the number of clustering centers, Z is the number of fault categories,
Figure FDA0003346654840000037
the correlation interference influence degree between the fault class of the sample at the d-th time and the g-th clustering center is delta, which is the sum of the weights corresponding to all fault classes simulated and trained in the step S1, pdgDistance between the d sample fault test and the g cluster center, qdWeighting the fault type corresponding to the d sample fault test;
s4, respectively deducing an associated interference influence matrix and a clustering center iteration formula of the target function by adopting a Lagrange multiplier method:
Figure FDA0003346654840000041
Dgfor the cluster center corresponding to the g-th fault category, RgWeights corresponding to the training sample fault types to be classified;
s5, screening out the correlation interference coefficient between each fault type in the correlation interference influence matrix and the clustering center, and establishing a fault chain for each fault type with the correlation interference coefficient larger than 0.
7. The system for evaluating the equipment fault based on the multidimensional data of the Internet of things according to claim 5 or 6, wherein: the equipment fault evaluation system further comprises a prediction tracking damage module, wherein the prediction tracking damage module is used for extracting a motor maintenance risk evaluation coefficient in the current motor working state, which is obtained by analysis of the multidimensional data evaluation module, and calculating the motor fault surge acceleration according to the motor maintenance risk evaluation coefficient in the current motor state and the motor maintenance risk evaluation coefficient under the interval duration t3
Figure FDA0003346654840000042
And tracking and predicting the service life of the motor maintained by the continuous work of the motor according to the current motor fault surge coefficient
Figure FDA0003346654840000043
Gt3Maintaining a risk assessment factor for the motor at time t3, GmaxA risk assessment factor is maintained for the maximum motor allowed for the motor.
8. An equipment fault assessment method based on multi-dimensional data of the Internet of things is characterized by comprising the following steps: the method comprises the following specific steps:
s1, collecting the bearing temperature and the mixed noise in the motor operation process, and separating the mixed noise to obtain the bearing noise and the mechanical noise;
s2, screening the bearing noise and the mechanical noise to obtain the bearing noise fault level and the mechanical noise fault level;
s3, collecting the axial vibration amplitude and the radial vibration amplitude of the motor to establish a time domain vibration signal diagram, and analyzing the basic parameters of the axial vibration and the basic parameters of the radial vibration through the time domain vibration signal diagram;
s4, extracting basic parameters of axial vibration and radial vibration in the step S3 to judge the abnormal degree of the motor vibration;
s5, processing the temperature of the bearing surface to obtain the maximum speed of the temperature rise of the bearing surface, the accumulated time length of the bearing temperature being greater than the preset temperature W and the temperature of the bearing surface after the bearing temperature is greater than the preset temperature W;
s6, analyzing the seizure prediction coefficient of the motor by combining the data in the step S4 and the step S5, and judging the motor maintenance danger evaluation coefficient under the condition that the motor continues to work by adopting a multi-dimensional data evaluation model.
CN202111325074.9A 2021-11-10 2021-11-10 Equipment fault evaluation system and method based on multi-dimensional data of Internet of things Pending CN114004306A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111325074.9A CN114004306A (en) 2021-11-10 2021-11-10 Equipment fault evaluation system and method based on multi-dimensional data of Internet of things

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111325074.9A CN114004306A (en) 2021-11-10 2021-11-10 Equipment fault evaluation system and method based on multi-dimensional data of Internet of things

Publications (1)

Publication Number Publication Date
CN114004306A true CN114004306A (en) 2022-02-01

Family

ID=79928475

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111325074.9A Pending CN114004306A (en) 2021-11-10 2021-11-10 Equipment fault evaluation system and method based on multi-dimensional data of Internet of things

Country Status (1)

Country Link
CN (1) CN114004306A (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113777488A (en) * 2021-09-14 2021-12-10 中国南方电网有限责任公司超高压输电公司昆明局 State evaluation method and device for valve cooling main pump motor and computer equipment
CN114500235A (en) * 2022-04-06 2022-05-13 深圳粤讯通信科技有限公司 Communication equipment safety management system based on Internet of things
CN114810513A (en) * 2022-06-24 2022-07-29 江苏奥派电气科技有限公司 Wind power generator bearing vibration fault intelligent monitoring system based on 5G communication
CN115453267A (en) * 2022-09-15 2022-12-09 北京京能清洁能源电力股份有限公司北京分公司 Fault diagnosis system for electric power information system
CN115982552A (en) * 2022-12-19 2023-04-18 萍乡学院 Electronic signal processing method and system
CN116781864A (en) * 2023-06-21 2023-09-19 浙江宏远智能科技有限公司 Factory operation state remote monitoring system and method based on Internet of things data acquisition

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113777488A (en) * 2021-09-14 2021-12-10 中国南方电网有限责任公司超高压输电公司昆明局 State evaluation method and device for valve cooling main pump motor and computer equipment
CN113777488B (en) * 2021-09-14 2023-12-12 中国南方电网有限责任公司超高压输电公司昆明局 State evaluation method and device for valve cooling main pump motor and computer equipment
CN114500235A (en) * 2022-04-06 2022-05-13 深圳粤讯通信科技有限公司 Communication equipment safety management system based on Internet of things
CN114810513A (en) * 2022-06-24 2022-07-29 江苏奥派电气科技有限公司 Wind power generator bearing vibration fault intelligent monitoring system based on 5G communication
CN115453267A (en) * 2022-09-15 2022-12-09 北京京能清洁能源电力股份有限公司北京分公司 Fault diagnosis system for electric power information system
CN115982552A (en) * 2022-12-19 2023-04-18 萍乡学院 Electronic signal processing method and system
CN115982552B (en) * 2022-12-19 2023-10-20 萍乡学院 Electronic signal processing method and system
CN116781864A (en) * 2023-06-21 2023-09-19 浙江宏远智能科技有限公司 Factory operation state remote monitoring system and method based on Internet of things data acquisition
CN116781864B (en) * 2023-06-21 2024-02-23 浙江宏远智能科技有限公司 Factory operation state remote monitoring system and method based on Internet of things data acquisition

Similar Documents

Publication Publication Date Title
CN114004306A (en) Equipment fault evaluation system and method based on multi-dimensional data of Internet of things
US6088658A (en) Statistical pattern analysis methods of partial discharge measurements in high voltage insulation
CA2379625C (en) System and method for conditioned-based monitoring of a bearing assembly
Subrahmanyam et al. Using neural networks for the diagnosis of localized defects in ball bearings
CN106323452B (en) A kind of detection method and detection device of equipment abnormal sound
JP3321487B2 (en) Device / equipment diagnosis method and system
KR102141391B1 (en) Failure data management method based on cluster estimation
US20150346066A1 (en) Asset Condition Monitoring
WO2010094915A1 (en) Method and apparatus for monitoring and analyzing vibrations in rotary machines
CN111964909A (en) Rolling bearing operation state detection method, fault diagnosis method and system
CN117134503B (en) State monitoring method and system for large-scale power supply device
CN112711850A (en) Unit online monitoring method based on big data
KR20210006832A (en) Method and apparatus for machine fault diagnosis
CN116383633A (en) Method and system for detecting faults of machine-made sand vibrating screen through multi-factor comprehensive analysis
CN116561514A (en) Method, system, device and medium for diagnosing faults of vehicle hub bearing unit
Ginart et al. Automated feature selection for embeddable prognostic and health monitoring (PHM) architectures
CN107702919A (en) Sliding bearing lubricating state monitoring method based on sound emission
CN117435908A (en) Multi-fault feature extraction method for rotary machine
CN117391443A (en) Dust removal equipment state monitoring and early warning method and system
CN112380782A (en) Rotating equipment fault prediction method based on mixed indexes and neural network
CN117370879A (en) Real-time online fault diagnosis method and system for wind turbine gearbox
CN115655731A (en) Diesel engine state monitoring method and device and storage medium
Song et al. Rolling Bearing Fault Diagnosis Under Different Severity Based on Statistics Detection Index and Canonical Discriminant Analysis
KR20220096793A (en) Method and system for comprehensive diagnosis defects in rotating machinery
Bachir et al. Application of principal component analysis approach in gas turbine defect diagnosis

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