CN115130495A - Rolling bearing fault prediction method and system - Google Patents
Rolling bearing fault prediction method and system Download PDFInfo
- Publication number
- CN115130495A CN115130495A CN202210518746.6A CN202210518746A CN115130495A CN 115130495 A CN115130495 A CN 115130495A CN 202210518746 A CN202210518746 A CN 202210518746A CN 115130495 A CN115130495 A CN 115130495A
- Authority
- CN
- China
- Prior art keywords
- particle
- module
- function
- value
- signal
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000005096 rolling process Methods 0.000 title claims abstract description 48
- 239000002245 particle Substances 0.000 claims abstract description 148
- 238000006731 degradation reaction Methods 0.000 claims abstract description 84
- 230000015556 catabolic process Effects 0.000 claims abstract description 76
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 20
- 238000012952 Resampling Methods 0.000 claims abstract description 11
- 238000000605 extraction Methods 0.000 claims abstract description 8
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 7
- 230000006870 function Effects 0.000 claims description 94
- 238000004364 calculation method Methods 0.000 claims description 29
- 238000007781 pre-processing Methods 0.000 claims description 21
- 238000010606 normalization Methods 0.000 claims description 17
- 238000005070 sampling Methods 0.000 claims description 17
- 238000000926 separation method Methods 0.000 claims description 12
- 238000001914 filtration Methods 0.000 claims description 11
- 239000000463 material Substances 0.000 claims description 9
- 238000002156 mixing Methods 0.000 claims description 9
- 238000012804 iterative process Methods 0.000 claims description 8
- 238000012360 testing method Methods 0.000 claims description 7
- 230000008859 change Effects 0.000 claims description 6
- 239000011159 matrix material Substances 0.000 claims description 6
- 230000001351 cycling effect Effects 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 12
- 230000035945 sensitivity Effects 0.000 abstract description 5
- 230000000737 periodic effect Effects 0.000 abstract description 2
- 238000001514 detection method Methods 0.000 abstract 3
- 238000012423 maintenance Methods 0.000 description 11
- 238000011160 research Methods 0.000 description 11
- 230000007246 mechanism Effects 0.000 description 8
- 230000035882 stress Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000013178 mathematical model Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000012885 constant function Methods 0.000 description 1
- 238000012937 correction 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
- 238000013135 deep learning Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000003449 preventive effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000000611 regression analysis Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- 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
- 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
- G06Q50/04—Manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Health & Medical Sciences (AREA)
- General Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Manufacturing & Machinery (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Primary Health Care (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Acoustics & Sound (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention relates to a rolling bearing fault prediction method and a rolling bearing fault prediction system, relates to the field of fault detection, and provides a bearing fault detection algorithm and a bearing fault detection system.A local mean decomposition algorithm is used for decomposing an original vibration signal, an effective PF component is screened out by using a Pearson correlation coefficient, the signal is reconstructed, the interference of an irrelevant noise signal is removed, the characteristic extraction of a bearing degradation process is realized based on the product form of a harmonic-to-noise ratio and a root-mean-square value, the sensitivity of the harmonic-to-noise ratio to periodic impact is fully utilized, and the problem that the root-mean-square value has lower sensitivity to the initial degradation of a bearing is avoided; the state space models of different degradation processes of the bearing are constructed based on the Paris model and the Foreman model, and compared with a single model prediction model, the prediction accuracy of the RUL is improved more effectively; the regularization of the particle filter resampling process is realized based on the Euclidean distance, the diversity of particles is effectively improved, the particle exhaustion phenomenon of basic particle filter is avoided, and the filter estimation precision of the nonlinear state is improved.
Description
Technical Field
The invention relates to the field of bearing fault prediction, in particular to a method and a system for predicting a fault of a rolling bearing.
Background
China highly attaches importance to nuclear safety, definitely aims at 'risk prevention' as a center, and continuously improves the safety utilization level of nuclear facilities. China's marine nuclear power plant generally adopts correction or preventive maintenance, and the annual maintenance volume is huge, and a large amount of spare parts seriously occupy the arrangement space of other resources. Statistically, the maintenance cost accounts for 73% of the total life cycle cost, and the improper maintenance accounts for more than 1/3% of the total maintenance cost, which indicates that the improper maintenance wastes a large amount of human and material resources. According to statistics of the American electric power society, the operating cost can be reduced by more than 20% after predictive maintenance is implemented in the fields of aerospace and the like. Therefore, it is necessary to accelerate the application of predictive maintenance to fall to the ground, thereby reducing the failure rate of the equipment, avoiding unexpected shutdown, and improving the vitality; meanwhile, the maintenance cost can be reduced, and the economical efficiency is improved. Currently, the bottleneck problem of predictive maintenance is how to accurately predict the remaining useful life of a device, and thus allocate the relevant resources based on the life distribution. And the failure prediction is closely related to a plurality of front-end factors such as aging mechanism, sensing and measuring, characteristic parameter analysis, prediction algorithm and the like of the component, and the coupling relation among the factors is complex. Therefore, the failure prediction model must be analyzed and optimized as a whole.
In order to solve the problems, a large number of fault prediction technical researches are carried out by scholars at home and abroad, the research methods can be roughly divided into three categories, the first category is a multivariate statistical analysis method, generally, an RUL prediction result is presented in a conditional failure probability mode according to observation data, and the method is generally characterized in that observation data are fitted by using methods such as probability theory, mathematical statistics and the like under the condition of not depending on any physical mechanism to form an RUL prediction model. However, such methods require an assumption of lifetime distribution, but often have a large gap from the actual situation. The second type adopts a machine learning related algorithm for research, which essentially belongs to pattern regression analysis, and with the rapid development of artificial intelligence and big data technology, learning the degradation pattern of an element from historical data by using machine learning and deep learning becomes possible without establishing a complex physical model; but compared with other methods, the calculation result is difficult to be obeyed due to the property of the black box; meanwhile, the method completely depends on data for modeling, and the degraded data in the practical process is difficult to obtain, so that the application of the method in the RUL prediction is greatly limited. And the third type is to establish a mathematical model in combination with a physical mechanism to describe the fault process of the engine and finally predict the operation state of the engine according to the model. Although there are some disadvantages, such as too high complexity of the model, difficulty in resolving the failure mechanism of some complex devices, etc.; but once the physical mechanism model is established, the analysis result is the highest accuracy in all the methods.
At present, physical process models are widely applied to the aspects of performance degradation, crack propagation and the like of materials. The Paris-Erdoxan (PE) model is one of the most widely applied physical models in the field of mechanical structure materials and is mainly used for describing the crack propagation process. Then, Melgar et al applied the model to the field of failure prediction of aircraft structural materials. In China, Wangjin et al convert a PE model into an empirical model for fault prediction; the Lu-shine takes the bearing inner ring as a research object, and carries out fault prediction on the fatigue wear of the part. The physical model and the formula do not have prediction and recursion capabilities, so the method is combined with a Bayes framework, starts from statistics and a probability model, and researches a fault prediction technology based on regularized particle filtering and a physical mechanism. Due to its demonstrated superiority in addressing non-linear and non-gaussian system behavior, particle filtering algorithms have been widely applied in various fault prediction scenarios. An et al build a mathematical model to track the machine's state of degradation, while unknown parameters are optimized by the PF method. To further reduce the variance of the RUL estimate and computational load, Haque et al propose an auxiliary particle filter for predicting the remaining lifetime of an insulated gate bipolar transistor. Michael Pecht et al implement predictions on circuit system faults based on particle filter algorithms.
Based on the challenges and opportunities, the rolling bearing is taken as a research object in the invention, because most of the operation load of the rotating machinery in the nuclear power plant is transmitted to the base body through the rolling bearing, the residual service life of the nuclear power plant is effectively predicted in order to avoid equipment failure caused by sudden damage of the bearing and further influence the safe and reliable operation of the nuclear power plant; finally, the laboratory is provided with a test bench for researching the rolling bearing, so that the adaptability can be improved on the basis of the prior period, and the requirement of fault prediction can be met. It is worth emphasizing that the research work in the invention has both basic scientific research value and practical application prospect. From the aspect of scientific significance, the invention relates to the multidisciplinary crossing fields of failure mechanism analysis, artificial intelligence technology and the like, and is based on the research of the formation mechanism of the typical failure mode of the rolling bearing and a failure prediction model of regularized particle filtering. In view of technical application, the rolling bearing fault prediction technology based on local mean decomposition and regularization particle filtering has the following characteristics and advantages: (1) aiming at the characteristics of the fault signals of the rolling bearing, a method for combining LMD and Pearson correlation coefficients is provided to extract degradation characteristics. The extracted Mean Square Harmonic Noise Ratio (MSHNR) index can well reflect the degradation trend of the bearing, and has higher sensitivity to each degradation stage of the bearing compared with the traditional indexes such as effective values and the like; (2) the regularization of the particle filter resampling process is realized based on the Euclidean distance, the influence of the time scale of particle updating on the filter estimation is considered, the diversity of particles is effectively improved, and the estimation precision of the nonlinear state is improved. (3) The rolling bearing service life prediction method based on different degradation stages better conforms to the degradation process of the bearing, and compared with a single model prediction method, the prediction accuracy of the RUL can be improved. The invention is expected to realize the breakthrough of the fault prediction technology through the research, and the research result not only can provide a new idea for improving the safety and the economy of the nuclear power system and other complex systems, but also can promote the establishment and the development of a related theoretical system.
Disclosure of Invention
The invention aims to provide a rolling bearing fault prediction method and a rolling bearing fault prediction system, which can improve the prediction precision of bearing faults.
In order to achieve the purpose, the invention provides the following scheme:
a rolling bearing failure prediction method, the prediction method comprising:
step 1: acquiring an original vibration signal;
and 2, step: preprocessing the original vibration signal to obtain a preprocessed vibration signal x (t);
and step 3: based on the preprocessed vibration signal x (t), calculating all local extreme points n of the vibration signal i ；
And 4, step 4: according to the local extreme point n i The average value m of all adjacent local extreme points is calculated i And an envelope estimate a i ；
And 5: all adjacent average value points m i And an envelope estimate a i Are connected to obtain a local mean function m 11 (t) and an envelope estimation function a 11 (t)；
Step 6: from the preprocessed vibration signal x (t) a local mean function m is derived 11 (t) separating to obtain a function h 11 (t) then using the envelope estimation function a 11 Demodulating to obtain a function s 11 (t)；
And 7: combining said function s 11 (t) repeating the calculation of steps 2-6 as a new function until s 1n (t) is a pure frequency modulation function;
and 8: multiplying the envelope functions obtained in the iterative process from the step 2 to the step 7 to obtain an envelope signal, namely: a is a 1 (t)＝a 11 (t)·a 12 (t)···a 1n (t)；
And step 9: will be a pure frequency-modulated signal s 1n (t) envelope function and envelope signal a 1 (t) multiplying to obtain a first product function component: PF (particle Filter) 1 (t)＝s 1n (t)a 1 (t)；
Step 10: mixing PF 1 (t) separating from the original vibration signal x (t) to obtain a new signal u 1 (t) mixing u 1 (t) repeating steps 2-8 as raw data, and repeating for k times until u k (t) is a monotonic function or the number of extreme points is less than 3;
step 11: the original vibration signal x (t) is decomposed into a series of PF components and a residual component through the above iteration of steps 2 to 10, namely:
step 12: calculating the correlation degree of each PF component and the original vibration signal x (t), selecting the PF component which has the correlation degree larger than or equal to a set threshold value and can completely explain the vibration characteristics of the signal, and reconstructing the signal based on the selected PF component;
step 13: on the basis of a reconstructed signal, calculating a root mean square value RMS and a harmonic noise ratio HNR of the reconstructed signal, realizing the feature extraction of the reconstructed signal based on the product MSHNR of the RMS and the HNR, generating an MSHNR degradation curve, and realizing the division of a stable degradation stage and a rapid degradation stage of the rolling bearing by utilizing the instantaneous change rate of each point of the degradation curve;
step 13: in the bearing stable degradation stage, a state space model of the bearing stable degradation stage is built based on a Paris model, and in the rapid degradation stage, a state space model of the rapid degradation stage is built based on a Foreman model;
step 14: designing basic parameters of a particle filter algorithm; the basic parameters comprise: number of particles N, noise variance Q, initial particle variance P, step size t n Resampling threshold N tv And a spatial dimension n;
step 15: initializing a set of particles in a regularized particle filter from a prior distribution p (x) 0 ) Sampling N particles to obtain an initial particle set
The initialization weight of each particle is
Step 16: inputting the test data set into the regularized particle filter as an input, from the importance density function
Obtaining a particle set at a k moment by intermediate sampling
Namely, it is
x k Particle representing time k, y k Representing the observed value at k moment, and calculating the weight of each particle at k moment
After obtaining the weight of each particle, carrying out weight normalization to obtain
And step 17: calculating the effective particle number based on the normalized particle weight
If the number of effective particles is larger than the threshold value N tv Step 16 is executed, and if the value is smaller than the threshold value, step 18 is executed;
step 18: obtaining a system observation value sequence and a sample estimator observation matrix, and taking a sequence with the length of L as an actual observation value of the system
For observed values calculated from each sample value in the sample set, taking the time sequence as [ j ═ k-L +1, k]Each sample in the interval is observed according to the observationCalculating a corresponding observed value by an equation;
calculating Euclidean distance of system observed value sequence and each particle observed value estimated value
The Euclidean distance is measured
Normalized to the (0,1) interval to obtain
Based on
Sample weights are redistributed to
From
Re-sampling to obtain a new particle set after regularization
Wherein,
a set of particles at time k is obtained for the sampling in step 16,
adjusting the normalized weight of each particle in step 16 through step 18;
step 19: by
Calculating a filtering value of the state estimation at the current moment;
step 20: based on the filter value estimated by the state at the current moment, recursion is carried out by adopting the state space model of the bearing stable degradation stage and the state space model of the rapid degradation stage to obtain the rolling bearing fault prediction result at the current moment;
step 21: judging whether the current time reaches the maximum prediction step number t f If yes, the prediction is finished, the algorithm is exited, otherwise, K equals to K +1, and the step 14 is returned to continue the prediction.
Optionally, the average m of adjacent local extreme points i The expression of (a) is as follows:
wherein n is i Representing local extreme points, n i+1 Is represented by the formula i Adjacent local extreme points.
Optionally, the envelope estimation value a i The expression of (a) is as follows:
optionally, function h 11 (t) and function s 11 The expression of (t) is as follows:
h 11 (t)＝x(t)-m 11 (t)
s 11 (t)＝h 11 (t)/a 11 (t)。
optionally, the set threshold is 0.7.
Optionally, the expression of the state space model of the bearing stability degradation stage is as follows:
Optionally, the expression of the state space model of the fast degradation stage is as follows:
Based on the above method of the present invention, the present invention additionally provides a rolling bearing failure prediction system, the system comprising:
the original vibration signal acquisition module is used for acquiring an original vibration signal;
the preprocessing module is used for preprocessing the original vibration signal to obtain a preprocessed vibration signal x (t);
a local extreme point calculation module for calculating all local extreme points n of the vibration signal based on the preprocessed vibration signal x (t) i ；
A local extreme point average value and envelope estimation value calculation module for calculating the local extreme point n i Finding the average m of all adjacent local extreme points i And an envelope estimate a i ；
A local mean function and envelope estimation function determination module for determining all adjacent mean points m i And an envelope estimate a i Are connected to obtain a local mean function m 11 (t) and envelope estimation function a 11 (t)；
A separation and demodulation module for separating the local mean function m from the preprocessed vibration signal x (t) 11 (t) separating to obtain a function h 11 (t) then using the envelope estimation function a 11 Demodulating to obtain a function s 11 (t)；
A rotation module for rotating the function s 11 (t) repeating the calculation of the preprocessing module-separation and demodulation module as a new function until s 1n (t) is a pure frequency modulation function;
an envelope signal calculation module, for multiplying the envelope function obtained by the iteration process of the preprocessing module and the loop module to obtain an envelope signal,namely: a is a 1 (t)＝a 11 (t)·a 12 (t)···a 1n (t)；
A product function component calculation module for calculating the pure frequency-modulated signal s 1n (t) envelope function and envelope signal a 1 (t) multiplying to obtain a first product function component: PF (particle Filter) 1 (t)＝s 1n (t)a 1 (t)；
A separation module for separating the PF 1 (t) separating from the original vibration signal x (t) to obtain a new signal u 1 (t) mixing u 1 (t) repeating the pre-processing module-envelope signal calculation module step as raw data, cycling k times until u k (t) is a monotonic function or the number of extreme points is less than 3;
a decomposition module, for decomposing the original vibration signal x (t) into a series of PF components and a residual component through the above-mentioned preprocessing module-separation module iterative process, namely:
the correlation degree calculation module is used for calculating the correlation degree of each PF component and the original vibration signal x (t), selecting the PF component which has the correlation degree larger than or equal to a set threshold value and can completely interpret the vibration characteristics of the signal, and then reconstructing the signal based on the selected PF component;
the dividing module is used for calculating a root mean square value RMS and a harmonic noise ratio HNR of a reconstructed signal on the basis of the reconstructed signal, realizing the feature extraction of the reconstructed signal based on the product MSHNR of the RMS and the HNR, generating an MSHNR degradation curve, and realizing the division of a stable degradation stage and a rapid degradation stage of the rolling bearing by utilizing the instantaneous change rate of each point of the degradation curve;
the model building module is used for building a state space model of the bearing stable degradation stage based on the Paris model in the bearing stable degradation stage and building a state space model of the rapid degradation stage based on the Foreman model in the rapid degradation stage;
the parameter setting module is used for setting basic parameters of the particle filter algorithm; the basic parameterThe method comprises the following steps: number of particles N, noise variance Q, initial particle variance P, step size t n Resampling threshold N tv And a spatial dimension n;
an initialization module for initializing a set of particles in a regularized particle filter from a prior distribution p (x) 0 ) Sampling N particles to obtain an initial particle set
The initialization weight of each particle is
A normalization module for inputting the test data set as input into the regularized particle filter from the importance density function
Obtaining a particle set at a k moment by intermediate sampling
Namely that
x k Particle representing time k, y k Representing the observed value at the k moment, and simultaneously calculating the weight of each particle at the k moment
After obtaining the weight of each particle, carrying out weight normalization to obtain
A first judgment module for calculating the effective particle number based on the normalized particle weight
If the number of effective particles is larger than the threshold value N tv Executing a normalization module, and if the normalization value is smaller than the threshold value, executing a next module;
a new particle set determining module for obtaining system observation value sequence and sample estimator observation matrix, and taking the sequence with length L as the actual observation value of the system
For observed values calculated from each sample value in the sample set, the time sequence is taken as [ j ═ k-L + 1, k]Calculating corresponding observed values of all samples in the interval according to an observation equation;
calculating a system observation value sequence and Euclidean distance of each particle observation value estimation value
The Euclidean distance is measured
Normalized to the (0,1) interval to obtain
Based on
Reassign the sample weights as
From
Re-sampling to obtain a new particle set after regularization
Wherein,
for the set of particles at time k,
for each granuleThe weight value after sub-normalization;
a filtered value calculation module for calculating a filtered value
Calculating a filtering value of the state estimation at the current moment;
the fault prediction module is used for recursion to obtain a rolling bearing fault prediction result at the current moment by adopting a state space model of the bearing stable degradation stage and a state space model of the rapid degradation stage based on the filter value estimated by the state at the current moment;
a second judging module for judging whether the current time reaches the maximum predicted step number t f If yes, the prediction is finished, the algorithm is exited, otherwise, K is K +1, and the parameter setting module is returned to continue the prediction.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
according to the method, the original vibration signal is decomposed through a local mean decomposition algorithm, the effective PF component is screened out by utilizing the Pearson correlation coefficient, further, the signal is reconstructed, the interference of irrelevant noise signals is removed, and the signal-to-noise ratio of the signal is improved; the characteristic extraction of the bearing degradation process is realized based on the product form of the harmonic-to-noise ratio and the root-mean-square value, the sensitivity of the harmonic-to-noise ratio to periodic impact is fully utilized, the multiplication of the root-mean-square value is realized, the purpose of highlighting the degradation fluctuation trend is achieved, and the problem of low sensitivity of the root-mean-square value to the initial degradation of the bearing is avoided; the state space models of different degradation processes of the bearing are constructed based on the Paris model and the Foreman model, and compared with a single model prediction model, the state space models are more consistent with the degradation process of the bearing, and the prediction precision of the RUL can be effectively improved; the regularization of the particle filter resampling process is realized based on the Euclidean distance, the influence of the time scale of particle updating on the filter estimation is considered, the diversity of particles is effectively improved, the particle exhaustion phenomenon of basic particle filter is avoided, and the filter estimation precision of the nonlinear state is improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
Fig. 1 is a flowchart of a rolling bearing fault prediction method according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide a rolling bearing fault prediction method and a rolling bearing fault prediction system, which can improve the prediction precision of bearing faults.
In order to make the aforementioned objects, features and advantages of the present invention more comprehensible, the present invention is described in detail with reference to the accompanying drawings and the detailed description thereof.
Fig. 1 is a flowchart of a rolling bearing fault prediction method according to an embodiment of the present invention, and as shown in fig. 1, the method includes:
step 1: the method comprises the steps of storing original data obtained by an acceleration sensor and process parameter sensors such as differential pressure, temperature and flow on a rolling bearing of the circulating water pump in calculation through a data acquisition board card, preprocessing and data characteristic engineering are conducted on the acquired original parameters, noise interference and characteristics irrelevant to fault prediction are removed, data normalization and standardization are conducted on the data, and the influence of dimension on subsequent fault prediction is avoided.
And 2, step: based on the preprocessed vibration signal x (t), calculating all local mean value points n of the vibration signal i Simultaneously according to the local extreme point n i Find out allAverage value m of adjacent local extreme points i And an envelope estimate a i I.e. by
And step 3: based on cubic Hermite interpolation, all adjacent mean value points m i And an envelope estimate a i Are connected to obtain a local mean function m 11 (t) and an envelope estimation function a 11 (t)。
And 4, step 4: from the original vibration signal x (t) a local mean function m is derived 11 (t) separating to obtain a function h 11 (t) then using the envelope estimation function a 11 Demodulating to obtain s 11 (t)。
h 11 (t)＝x(t)-m 11 (t)
s 11 (t)＝h 11 (t)/a 11 (t)
And 5: will s 11 (t) repeating the calculation of steps 2-4 as a new function until s 1n (t) is a pure frequency modulation function, i.e. the envelope estimation function is a constant function 1, in practical application, a variation delta can be set, and when 1-delta is more than or equal to a 1n When (t) is less than or equal to 1+ delta, the iteration is terminated.
Step 6: multiplying the envelope functions obtained in the iteration process of the step 2-5 to obtain envelope signals, namely:
a 1 (t)＝a 11 (t)·a 12 (t)···a 1n (t)
then the pure frequency-modulated signal s 1n (t) envelope function and envelope signal a 1 (t) multiplying to obtain a first product function component:
PF 1 (t)＝s 1n (t)a 1 (t)
and 7: mixing PF 1 (t) separating from the original vibration signal x (t) to obtain a new signal u 1 (t) mixing u 1 (t) repeating the steps 2-6 as raw data, repeating the steps k times until u k (t) is a monotonic function or the number of extreme points is less than 3.
And 8: the original vibration signal x (t) is decomposed into a series of PF components and a residual component through the iterative process of steps 2-7, namely:
and step 9: calculating the correlation degree of each PF component with the original vibration signal x (t) based on a Pearson Correlation Coefficient (PCC), selecting a PF with PCC being more than or equal to 0.7 and capable of completely interpreting the vibration characteristics of the signal, and reconstructing the signal based on the selected PF.
Step 10: on the basis of a reconstructed signal, a Root Mean Square (RMS) and a Harmonic Noise Ratio (HNR) of the signal are calculated, feature extraction of the reconstructed signal is realized on the basis of a product MSHNR of the RMS and the HNR, on the basis of an MSHNR degradation curve, an alarm threshold value is set by using u +6 sigma to judge that a bearing enters a degradation stage from normal operation (wherein u and sigma represent the mean value and the standard deviation of an MSHNR index when the bearing is in a normal period), and the instantaneous change rate of each point of the degradation curve is used for realizing the division of a stable degradation stage and a rapid degradation stage of the rolling bearing.
Step 11: in the stable degradation stage of the bearing, a state space model of the degradation process is built based on a Paris model, in the rapid degradation stage, the state space model of the degradation process is built based on a Foreman model, further, on the basis of the selection of the multi-stage state space model, the fault prediction of the rolling bearing is carried out based on the regularized particle filtering, and the models of the stable degradation stage and the rapid degradation stage of the bearing are built as follows:
(1) rolling bearing stable degradation stage-Paris model
The Paris model is:
in which a is a crackLength, N is the number of load cycles, da/dN is the crack propagation rate, C, m is the material parameter, and Δ K is the stress intensity factor amplitude. When the load cycle interval dN of the rolling bearing is sufficiently small, the above equation is approximated in a differential form, and a model of the stable degradation stage of the bearing is obtained as follows:
Δ σ is the stress range at the bearing crack.
(2) Foreman model of rapid degradation stage of rolling bearing
The Foreman model is:
in the formula, R and K c Stress ratio and fracture toughness, respectively. When the load cycle interval dN of the rolling bearing is sufficiently small, the differential term of the crack propagation rate of the above formula of the Forman model can be approximated in a differential form, and at the moment, the fatigue crack rapid propagation model based on the Forman model is as follows:
step 12: the basic parameters of the particle filter algorithm are set, and the non-time-varying parameters required to be set by regularizing the particle filter are shown in the following table.
TABLE 1 particle Filter Algorithm basis parameters
| Number of particles | Variance of noise | Variance of initial particle | Step size | Resampling threshold | Spatial dimension |
| N | Q | P | t n | N tv | n |
Step 13: initializing a set of particles in a regularized particle filter from a prior distribution p (x) 0 ) Sampling N particles to obtain an initial particle set
Initialization weight of each particle
Step 14: inputting the test data set as input into the regularized particle filter as a function of the importance density
(in the formula, x k Particle representing time k, y k Observed value representing k time) to obtain a set of k time particles
Namely that
And simultaneously calculating the weight of each particle at the moment k:
in the formula
The weight of the ith particle at time k,
is a likelihood probability function of the system, which characterizes the state of the system by x k-1 Transfer to x k Degree of similarity with the observed value, determined by the system's observation equation (y for the system's observation equation) k ＝h(x k )+v k Representing the degree to which an observed value reflects the state of the system, where h (-) is an observation function, v k To observe noise).
The state transfer function of the system can be represented by the state equation of the system (the state equation of the system is x) k ＝f(x k-1 )+u k-1 Representing the transition of the system state from a previous state to a subsequent state, where f (-) is the state transition function, u k Process noise). After obtaining the weight of each particle, carrying out weight normalization
Step 15: calculating effective particle number
If greater than the threshold value N tv Step 14 is performed, and if the threshold value is less than the threshold value, step 16 is performed.
Step 16: regularized resampling based on euclidean distance: preparing system observation value sequence and sample estimator observation matrix, and taking sequence with length L as actual observation value of system
For observed values calculated from each sample value in the sample set, taking the time sequence as [ j ═ k-L + 1, k]And calculating corresponding observed values of all samples in the interval according to an observation equation:
euclidean distance based on observation value sequence of Euclidean distance calculation system and observation value estimation value of each particle
Then will be
Normalized to the (0,1) interval to obtain
Based on
The sample weight value of the reassignment is
Then from
In the process of resampling to obtain a regularized new particle set
Wherein,
for the set of particles at time k,
normalizing the weight value of each particle; .
And step 17: by
And (4) calculating a filter value of the state estimation at the current moment, and then recurrently obtaining and outputting a fault prediction result of the rolling bearing at the current moment by the current state space model selected in the step 11.
Step 18: judging whether the current time isWhether the maximum predicted step number t is reached f If yes, the prediction is finished, the algorithm is exited, otherwise, K equals to K +1, and the step 12 is returned to continue the prediction.
Step 19: the actual fault prediction result of the rolling bearing of the circulating water pump is obtained through steps 1-18 based on local mean decomposition and regularized particle filtering, the related result can be referred by maintenance and decision-making personnel, and related measures can be taken in time, so that the safety is ensured, and the economy can be improved.
Based on the above method of the present invention, the present invention additionally provides a rolling bearing failure prediction system, the system comprising:
the original vibration signal acquisition module is used for acquiring an original vibration signal;
the preprocessing module is used for preprocessing the original vibration signal to obtain a preprocessed vibration signal x (t);
a local extreme point calculation module for calculating all local extreme points n of the vibration signal based on the preprocessed vibration signal x (t) i ；
A local extreme point average value and envelope estimation value calculation module for calculating the local extreme point n i Finding the average m of all adjacent local extreme points i And an envelope estimate a i ；
A local mean function and envelope estimation function determination module for determining all adjacent mean points m i And an envelope estimate a i Are connected to obtain a local mean function m 11 (t) and an envelope estimation function a 11 (t)；
A separation and demodulation module for separating the local mean function m from the preprocessed vibration signal x (t) 11 (t) separating to obtain a function h 11 (t) then using the envelope estimation function a 11 Demodulating to obtain a function s 11 (t)；
A rotation module for rotating the function s 11 (t) repeating the calculation of the preprocessing module-separation and demodulation module as a new function until s 1n (t) is a pure frequency modulation function;
an envelope signal calculation module, configured to multiply an envelope function obtained in an iterative process of the preprocessing module and the loop module to obtain an envelope signal, that is: a is 1 (t)＝a 11 (t)·a 12 (t)…a 1n (t)；
A product function component calculation module for calculating the pure frequency-modulated signal s 1n (t) envelope function and envelope signal a 1 (t) multiplying to obtain a first product function component: PF (particle Filter) 1 (t)＝s 1n (t)a 1 (t)；
A separation module for separating the PF 1 (t) separating from the original vibration signal x (t) to obtain a new signal u 1 (t) mixing u 1 (t) repeating the steps of the preprocessing module-envelope signal calculation module as raw data, and cycling k times until u k (t) is a monotonic function or the number of extreme points is less than 3;
and the decomposition module is used for decomposing the original vibration signal x (t) into a series of PF components and a residual component through the iterative process of the preprocessing module-the separation module, namely:
the correlation degree calculation module is used for calculating the correlation degree of each PF component and the original vibration signal x (t), selecting the PF component which has the correlation degree larger than or equal to a set threshold value and can completely interpret the vibration characteristics of the signal, and then reconstructing the signal based on the selected PF component;
the dividing module is used for calculating a root mean square value RMS and a harmonic noise ratio HNR of a reconstructed signal on the basis of the reconstructed signal, realizing the feature extraction of the reconstructed signal based on the product MSHNR of the RMS and the HNR, generating an MSHNR degradation curve, and realizing the division of a stable degradation stage and a rapid degradation stage of the rolling bearing by utilizing the instantaneous change rate of each point of the degradation curve;
the model building module is used for building a state space model of the bearing stable degradation stage based on the Paris model in the bearing stable degradation stage and building a state space model of the rapid degradation stage based on the Foreman model in the rapid degradation stage;
the parameter setting module is used for setting basic parameters of the particle filter algorithm; the basic parameters comprise: number of particles N, noise variance Q, initial particle variance P, step size t n Resampling threshold N tv And a spatial dimension n;
an initialization module for initializing a set of particles in a regularized particle filter from a prior distribution p (x) 0 ) Sampling N particles to obtain an initial particle set
The initialization weight of each particle is
A normalization module for inputting the test data set as input into the regularized particle filter from the importance density function
Obtaining a k-time particle set by intermediate sampling
Namely that
x k Particle representing time k, y k Representing the observed value at the k moment, and simultaneously calculating the weight of each particle at the k moment
After obtaining the weight of each particle, carrying out weight normalization to obtain
A first judgment module for calculating the effective particle number based on the normalized particle weight
If it is validThe number of particles is greater than a threshold value N tv Executing a normalization module, and executing a next module if the normalization is less than the threshold value;
a new particle set determining module for obtaining system observation value sequence and sample estimator observation matrix, and taking the sequence with length L as the actual observation value of the system
For observed values calculated from each sample value in the sample set, the time sequence is taken as [ j ═ k-L + 1, k]Calculating corresponding observed values of all samples in the interval according to an observation equation;
calculating a system observation value sequence and Euclidean distance of each particle observation value estimation value
The Euclidean distance is measured
Normalized to the (0,1) interval to obtain
Based on
Reassign the sample weights as
From
Re-sampling to obtain a new particle set after regularization
Wherein,
for the set of particles at time k,
normalizing the weight value of each particle;
a filtered value calculation module for calculating a filtered value
Calculating a filtering value of the state estimation at the current moment;
the fault prediction module is used for recursion to obtain a rolling bearing fault prediction result at the current moment by adopting a state space model of the bearing stable degradation stage and a state space model of the rapid degradation stage based on the filter value estimated by the state at the current moment;
a second judging module for judging whether the current time reaches the maximum predicted step number t f If yes, the prediction is finished, the algorithm is exited, otherwise, K is K +1, and the parameter setting module is returned to continue the prediction.
In the present specification, the embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system disclosed by the embodiment, the description is relatively simple because the system corresponds to the method disclosed by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (8)
1. A rolling bearing failure prediction method, characterized by comprising:
step 1: acquiring an original vibration signal;
step 2: preprocessing the original vibration signal to obtain a preprocessed vibration signal x (t);
and step 3: based on the preprocessed vibration signal x (t), calculating all local extreme points n of the vibration signal i ；
And 4, step 4: according to the local extreme point n i Finding the average m of all adjacent local extreme points i And an envelope estimate a i ；
And 5: all adjacent mean points m i And an envelope estimate a i Are connected to obtain a local mean function m 11 (t) and an envelope estimation function a 11 (t)；
Step 6: from the preprocessed vibration signal x (t) a local mean function m is derived 11 (t) separating to obtain a function h 11 (t) then using the envelope estimation function a 11 Demodulating to obtain a function s 11 (t)；
And 7: combining said function s 11 (t) repeating the calculation of steps 2-6 as a new function until s 1n (t) is a pure frequency modulation function;
and 8: multiplying the envelope functions obtained in the iterative process from the step 2 to the step 7 to obtain an envelope signal, namely: a is 1 (t)＝a 11 (t)·a 12 (t)···a 1n (t)；
And step 9: will be a pure frequency-modulated signal s 1n (t) envelope function and envelope signal a 1 (t) multiplying to obtain a first product function component: PF (particle Filter) 1 (t)＝s 1n (t)a 1 (t)；
Step 10: mixing PF 1 (t) separating from the original vibration signal x (t) to obtain a new signal u 1 (t) mixing u 1 (t) repeating steps 2-8 as raw data, and repeating for k times until u k (t) is a monotonic function or the number of extreme points is less than 3;
step 11: the original vibration signal x (t) is decomposed into a series of PF components and a residual component through the above iteration of steps 2 to 10, namely:
step 12: calculating the correlation degree of each PF component and the original vibration signal x (t), selecting the PF component which has the correlation degree larger than or equal to a set threshold value and can completely interpret the vibration characteristics of the signal, and then reconstructing the signal based on the selected PF component;
step 13: on the basis of a reconstructed signal, calculating a root mean square value RMS and a harmonic noise ratio HNR of the reconstructed signal, realizing the feature extraction of the reconstructed signal based on the product MSHNR of the RMS and the HNR, generating an MSHNR degradation curve, and realizing the division of a stable degradation stage and a rapid degradation stage of the rolling bearing by utilizing the instantaneous change rate of each point of the degradation curve;
step 13: in the bearing stable degradation stage, a state space model of the bearing stable degradation stage is built based on a Paris model, and in the rapid degradation stage, a state space model of the rapid degradation stage is built based on a Foreman model;
step 14: setting basic parameters of a particle filter algorithm; the basic parameters include: number of particles N, noise variance Q, initial particle variance P, step size t n Resampling threshold N tv And a spatial dimension n;
step 15: initializing a set of particles in a regularized particle filter from a prior distribution p (x) 0 ) N particles are subjected to intermediate sampling to obtain an initial particle set
The initialization weight of each particle is
Step 16: inputting the test data set into the regularized particle filter as an input, from the importance density function
Obtaining a k-time particle set by intermediate sampling
Namely that
x k Particle representing time k, y k Representing the observed value at k moment, and calculating the weight of each particle at k moment
After obtaining the weight of each particle, carrying out weight normalization to obtain
And step 17: calculating the effective particle number based on the normalized particle weight
If the number of effective particles is larger than the threshold value N tv Step 16 is executed, and if the value is smaller than the threshold value, step 18 is executed;
step 18: obtaining a system observation value sequence and a sample estimator observation matrix, and taking a sequence with the length of L as an actual observation value of the system
For observed values calculated from each sample value in the sample set, the time sequence is taken as [ j ═ k-L +1, k]Calculating corresponding observed values of all samples in the interval according to an observation equation;
calculating a system observation value sequence and Euclidean distance of each particle observation value estimation value
The Euclidean distance is measured
Normalized to the (0,1) interval to obtain
Based on
Sample weights are redistributed to
From
Re-sampling to obtain a new particle set after regularization
Wherein,
for the set of particles at time k,
normalizing the weight value of each particle;
step 19: by
Calculating a filtering value of the state estimation at the current moment;
step 20: based on the filter value estimated by the state at the current moment, recursion is carried out by adopting the state space model of the bearing stable degradation stage and the state space model of the rapid degradation stage to obtain the rolling bearing fault prediction result at the current moment;
step 21: judging whether the current time reaches the maximum prediction step number t f If yes, the prediction is finished, the algorithm is exited, otherwise, K equals to K +1, and the step 14 is returned to continue the prediction.
2. Rolling bearing fault prediction method according to claim 1, characterized in that adjacentAverage m of local extreme points i The expression of (a) is as follows:
wherein n is i Representing local extreme points, n i+1 Is represented by a radical and n i Adjacent local extreme points.
3. Rolling bearing fault prediction method according to claim 2, characterized in that the envelope estimation a i The expression of (c) is as follows:
4. rolling bearing fault prediction method according to claim 1, characterized in that function h 11 (t) and function s 11 The expression of (t) is as follows:
h 11 (t)＝x(t)-m 11 (t)
s 11 (t)＝h 11 (t)/a 11 (t)。
5. the rolling bearing failure prediction method according to claim 1, characterized in that the set threshold value is 0.7.
6. The rolling bearing failure prediction method according to claim 1, characterized in that the expression of the state space model of the bearing stability degradation stage is as follows:
7. Rolling bearing fault prediction method according to claim 1, characterized in that the expression of the state space model of the rapid degradation phase is as follows:
8. A rolling bearing failure prediction system, characterized in that the system comprises:
the original vibration signal acquisition module is used for acquiring an original vibration signal;
the preprocessing module is used for preprocessing the original vibration signal to obtain a preprocessed vibration signal x (t);
a local extreme point calculation module for calculating all local extreme points n of the vibration signal based on the preprocessed vibration signal x (t) i ；
A local extreme point average value and envelope estimation value calculation module for calculating the local extreme point n i Finding the average m of all adjacent local extreme points i And an envelope estimate a i ；
A local mean function and envelope estimation function determination module for determining all adjacent mean points m i And an envelope estimate a i Are connected to obtain a local mean function m 11 (t) and envelope estimation function a 11 (t)；
A separation and demodulation module for separating the local mean function m from the preprocessed vibration signal x (t) 11 (t) separating to obtain a function h 11 (t) then using the envelope estimation function a 11 Demodulating to obtain a function s 11 (t)；
A rotation module for rotating the function s 11 (t) repeating the pre-processing module-score as a new functionCalculation of the separation and demodulation module until s 1n (t) is a pure frequency modulation function;
an envelope signal calculation module, configured to multiply an envelope function obtained in an iterative process of the preprocessing module and the loop module to obtain an envelope signal, that is: a is 1 (t)＝a 11 (t)·a 12 (t)···a 1n (t)；
A product function component calculation module for calculating a pure frequency-modulated signal s 1n (t) envelope function and envelope signal a 1 (t) multiplying to obtain a first product function component: PF (particle Filter) 1 (t)＝s 1n (t)a 1 (t)；
A separation module for separating the PF 1 (t) separating from the original vibration signal x (t) to obtain a new signal u 1 (t) mixing u 1 (t) repeating the pre-processing module-envelope signal calculation module step as raw data, cycling k times until u k (t) is a monotonic function or the number of extreme points is less than 3;
and the decomposition module is used for decomposing the original vibration signal x (t) into a series of PF components and a residual component through the iterative process of the preprocessing module-the separation module, namely:
the correlation degree calculation module is used for calculating the correlation degree of each PF component and the original vibration signal x (t), selecting the PF component which has the correlation degree larger than or equal to a set threshold value and can completely interpret the vibration characteristics of the signal, and then reconstructing the signal based on the selected PF component;
the dividing module is used for calculating a root mean square value RMS and a harmonic noise ratio HNR of a reconstructed signal on the basis of the reconstructed signal, realizing the feature extraction of the reconstructed signal based on the product MSHNR of the RMS and the HNR, generating an MSHNR degradation curve, and realizing the division of a stable degradation stage and a rapid degradation stage of the rolling bearing by utilizing the instantaneous change rate of each point of the degradation curve;
the model building module is used for building a state space model of the bearing stable degradation stage based on the Paris model in the bearing stable degradation stage and building a state space model of the rapid degradation stage based on the Foreman model in the rapid degradation stage;
the parameter setting module is used for setting basic parameters of the particle filter algorithm; the basic parameters include: number of particles N, noise variance Q, initial particle variance P, step size t n Resampling threshold N tv And a spatial dimension n;
an initialization module for initializing a set of particles in a regularized particle filter from a prior distribution p (x) 0 ) Sampling N particles to obtain an initial particle set
The initialization weight of each particle is
A normalization module for inputting the test data set as input into the regularized particle filter from the importance density function
Obtaining a k-time particle set by intermediate sampling
Namely, it is
x k Particle representing time k, y k Representing the observed value at k moment, and calculating the weight of each particle at k moment
After obtaining the weight of each particle, carrying out weight normalization to obtain
A first judging module for judging whether the current value is less than the predetermined valueCalculating the effective particle number by the normalized particle weight
If the number of effective particles is larger than the threshold value N tv Executing a normalization module, and executing a next module if the normalization is less than the threshold value;
a new particle set determining module for obtaining system observation value sequence and sample estimator observation matrix, and taking the sequence with length L as the actual observation value of the system
For observed values calculated from each sample value in the sample set, taking the time sequence as [ j ═ k-L +1, k]Calculating corresponding observed values of all samples in the interval according to an observation equation;
calculating a system observation value sequence and Euclidean distance of each particle observation value estimation value
The Euclidean distance is measured
Normalized to the (0,1) interval to obtain
Based on
Sample weights are redistributed to
From
Re-sampling to obtain a new particle set after regularization
Wherein,
for the set of particles at time k,
normalizing the weight value of each particle;
a filtered value calculation module for calculating a filtered value
Calculating a filtering value of the state estimation at the current moment;
the fault prediction module is used for recursion to obtain a rolling bearing fault prediction result at the current moment by adopting a state space model of the bearing stable degradation stage and a state space model of the rapid degradation stage based on the filter value estimated by the state at the current moment;
a second judging module for judging whether the current time reaches the maximum predicted step number t f If yes, the prediction is finished, the algorithm is exited, otherwise, K is K +1, and the parameter setting module is returned to continue the prediction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210518746.6A CN115130495A (en) | 2022-05-12 | 2022-05-12 | Rolling bearing fault prediction method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210518746.6A CN115130495A (en) | 2022-05-12 | 2022-05-12 | Rolling bearing fault prediction method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115130495A true CN115130495A (en) | 2022-09-30 |
Family
ID=83376096
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210518746.6A Pending CN115130495A (en) | 2022-05-12 | 2022-05-12 | Rolling bearing fault prediction method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115130495A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115616082A (en) * | 2022-12-14 | 2023-01-17 | 杭州兆华电子股份有限公司 | Keyboard defect analysis method based on noise detection |
| CN116187982A (en) * | 2023-04-24 | 2023-05-30 | 成都盛锴科技有限公司 | Rail transit vehicle multi-source multi-mode fault risk reasoning and maintenance decision method |
| CN116252820A (en) * | 2023-05-12 | 2023-06-13 | 西南交通大学 | Polygonal quantitative detection method for high-speed train wheels driven by improved frequency domain integration method |
| CN117629636A (en) * | 2023-12-05 | 2024-03-01 | 哈尔滨工程大学 | Health assessment and fault diagnosis method and system for rolling bearing of combustion engine |
| CN118143503A (en) * | 2024-05-08 | 2024-06-07 | 韶关市洁盟超声科技有限公司 | Copper-aluminum substrate welding control method based on pressure detection |
| CN118518998A (en) * | 2024-07-24 | 2024-08-20 | 山东奥斯登房车有限公司 | New energy caravan load insulation fault positioning method and system |
-
2022
- 2022-05-12 CN CN202210518746.6A patent/CN115130495A/en active Pending
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115616082A (en) * | 2022-12-14 | 2023-01-17 | 杭州兆华电子股份有限公司 | Keyboard defect analysis method based on noise detection |
| CN116187982A (en) * | 2023-04-24 | 2023-05-30 | 成都盛锴科技有限公司 | Rail transit vehicle multi-source multi-mode fault risk reasoning and maintenance decision method |
| CN116187982B (en) * | 2023-04-24 | 2023-08-11 | 成都盛锴科技有限公司 | Rail transit vehicle multi-source multi-mode fault risk reasoning and maintenance decision method |
| CN116252820A (en) * | 2023-05-12 | 2023-06-13 | 西南交通大学 | Polygonal quantitative detection method for high-speed train wheels driven by improved frequency domain integration method |
| CN116252820B (en) * | 2023-05-12 | 2023-09-05 | 西南交通大学 | Polygonal quantitative detection method for high-speed train wheels driven by improved frequency domain integration method |
| CN117629636A (en) * | 2023-12-05 | 2024-03-01 | 哈尔滨工程大学 | Health assessment and fault diagnosis method and system for rolling bearing of combustion engine |
| CN117629636B (en) * | 2023-12-05 | 2024-05-24 | 哈尔滨工程大学 | Health assessment and fault diagnosis method and system for rolling bearing of combustion engine |
| CN118143503A (en) * | 2024-05-08 | 2024-06-07 | 韶关市洁盟超声科技有限公司 | Copper-aluminum substrate welding control method based on pressure detection |
| CN118518998A (en) * | 2024-07-24 | 2024-08-20 | 山东奥斯登房车有限公司 | New energy caravan load insulation fault positioning method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115130495A (en) | Rolling bearing fault prediction method and system | |
| CN107657088B (en) | Rolling bearing fault diagnosis method based on MCKD algorithm and support vector machine | |
| CN109271975B (en) | Power quality disturbance identification method based on big data multi-feature extraction collaborative classification | |
| CN109297689B (en) | Large-scale hydraulic machinery intelligent diagnosis method introducing weight factors | |
| Jiang et al. | Fault diagnosis for rolling bearing using a hybrid hierarchical method based on scale-variable dispersion entropy and parametric t-SNE algorithm | |
| CN113420691A (en) | Mixed domain characteristic bearing fault diagnosis method based on Pearson correlation coefficient | |
| CN111413089A (en) | Gear fault diagnosis method based on combination of VMD entropy method and VPMCD | |
| CN106441896A (en) | Characteristic vector extraction method for rolling bearing fault mode identification and state monitoring | |
| CN114363195B (en) | Network flow prediction and early warning method for time and frequency spectrum residual convolution network | |
| CN108074015B (en) | Ultra-short-term prediction method and system for wind power | |
| CN103729444B (en) | The abnormal deviation data examination method of potential relation between a kind of data based on monitoring of equipment | |
| CN113642508A (en) | Bearing fault diagnosis method based on parameter self-adaptive VMD and optimized SVM | |
| CN111459144A (en) | Airplane flight control system fault prediction method based on deep cycle neural network | |
| CN112215405A (en) | Non-invasive type residential electricity load decomposition method based on DANN domain adaptive learning | |
| CN115829157A (en) | Chemical water quality index prediction method based on variational modal decomposition and auto former model | |
| CN102305792A (en) | Nonlinear partial least square optimizing model-based forest carbon sink remote sensing evaluation method | |
| Liu et al. | Research on fault diagnosis of planetary gearbox based on variable multi-scale morphological filtering and improved symbol dynamic entropy | |
| CN117390508B (en) | Hydroelectric generating set signal state identification method based on time-shifting multi-scale cosine similarity entropy | |
| Meng et al. | A gear fault diagnosis method based on improved accommodative random weighting algorithm and BB-1D-TP | |
| CN110222386A (en) | A kind of planetary gear degenerate state recognition methods | |
| Zhao et al. | A feature extraction method based on LMD and MSE and its application for fault diagnosis of reciprocating compressor | |
| CN113297722A (en) | Coastal soil salinity assessment method and system | |
| CN116821828A (en) | Multi-dimensional time sequence prediction method based on industrial data | |
| Han et al. | A novel fault diagnosis method for rotating machinery of imbalanced data | |
| CN114235408B (en) | Bearing fault diagnosis method and system based on improved cascade variational modal decomposition |
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 |