CN114720129A - Rolling bearing residual life prediction method and system based on bidirectional GRU - Google Patents

Rolling bearing residual life prediction method and system based on bidirectional GRU Download PDF

Info

Publication number
CN114720129A
CN114720129A CN202210300479.5A CN202210300479A CN114720129A CN 114720129 A CN114720129 A CN 114720129A CN 202210300479 A CN202210300479 A CN 202210300479A CN 114720129 A CN114720129 A CN 114720129A
Authority
CN
China
Prior art keywords
rolling bearing
degradation index
model
degradation
predicting
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202210300479.5A
Other languages
Chinese (zh)
Other versions
CN114720129B (en
Inventor
张法业
闫星宇
姜明顺
隋青美
张雷
贾磊
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.)
Shandong University
Original Assignee
Shandong 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 Shandong University filed Critical Shandong University
Priority to CN202210300479.5A priority Critical patent/CN114720129B/en
Publication of CN114720129A publication Critical patent/CN114720129A/en
Application granted granted Critical
Publication of CN114720129B publication Critical patent/CN114720129B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
    • G01M13/00—Testing of machine parts
    • G01M13/04—Bearings
    • G01M13/045—Acoustic or vibration analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/048—Activation functions
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/04—Ageing analysis or optimisation against ageing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/082—Learning methods modifying the architecture, e.g. adding, deleting or silencing nodes or connections
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation

Abstract

The invention provides a method and a system for predicting the residual life of a rolling bearing based on a bidirectional GRU (general-purpose unit), which are used for acquiring a vibration signal of the rolling bearing; obtaining a degradation index estimation value of the rolling bearing according to the obtained vibration signal and a preset convolutional neural network model; obtaining a degradation index predicted value according to the degradation index estimated value and a preset first BiGRU model; obtaining a predicted value of the remaining service life according to the degradation index estimated value and a preset second BiGRU model; carrying out state evaluation on the rolling bearing according to the obtained degradation index predicted value and the residual service life predicted value; the method automatically extracts the degradation trend from the original state signals of the bearing, effectively captures the hidden long-term correlation among the time sequence signals, and realizes accurate prediction of the residual life of the bearing.

Description

Rolling bearing residual life prediction method and system based on bidirectional GRU
Technical Field
The invention relates to the technical field of rolling bearing state evaluation, in particular to a rolling bearing residual life prediction method and system based on bidirectional GRU.
Background
The statements in this section merely provide background information related to the present disclosure and may not constitute prior art.
With the rapid development of smart sensing, wireless communication, and computer technologies, Decision Making (DM) is moving towards intelligence, robustness, and adaptation. Residual service life (RUL) prediction is one of the most critical technologies in decision making, so that the failure time can be predicted in advance, a maintenance engineer can conveniently perform qualitative risk analysis, and a corresponding maintenance strategy is made, so that a catastrophic situation is avoided, and the reliability and the safety of mechanical equipment are better ensured.
Rolling bearings are one of the most common but important components in mechanical equipment, and their health will directly affect the safety, reliability and usability of the mechanical equipment. At present, the rolling bearing RUL prediction methods are mainly classified into two types, i.e., model-based methods and data-driven methods. Among them, the model-based method, which infers the future trend of the health state by establishing a physical model, requires a lot of prior knowledge and experience about the study subject, resulting in poor generalization ability. The data-driven method mainly carries out health prediction by modeling historical data without a mathematical model or expert experience of a research object. Therefore, in recent years, a data-driven method is widely used for the remaining life prediction.
The obtaining of the time correlation of the degradation trend and the health prediction are key steps of a data-driven residual life prediction method, and the current mainstream method is to combine a support vector regression machine, an artificial neural network, a deep neural network and other models to predict the residual life on the basis of constructing a degradation index. However, the degradation index construction method based on the artificial extracted features heavily depends on empirical knowledge, most of the studied residual life prediction models at present can only capture the time correlation of data in a single time direction (i.e. forward or backward), and the residual life prediction accuracy needs to be improved.
Disclosure of Invention
In order to solve the defects of the prior art, the invention provides a rolling bearing residual life prediction method and system based on a bidirectional GRU (generalized regression Unit), which can automatically extract a degradation trend from a bearing original state signal, effectively capture the hidden long-term correlation among time sequence signals and realize accurate prediction of the bearing residual life.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention provides a method for predicting the residual life of a rolling bearing based on a bidirectional GRU (ground reference Unit).
A method for predicting the residual life of a rolling bearing based on bidirectional GRUs comprises the following steps:
acquiring a vibration signal of a rolling bearing;
obtaining a degradation index estimation value of the rolling bearing according to the obtained vibration signal and a preset convolutional neural network model;
and giving different weights to the evaluation indexes according to the influence of monotonicity, monotonicity and robustness of the degradation indexes on the residual life prediction result, so as to obtain the degradation index evaluation value integrating the monotonicity, the monotonicity and the robustness.
Obtaining a degradation index predicted value according to the degradation index estimated value and a preset first BiGRU model;
obtaining a predicted value of the remaining service life according to the degradation index estimated value and a preset second BiGRU model;
as an optional implementation, in the training of the preset convolutional neural network model, the adopted degradation index obtaining method is as follows:
the degradation index extraction method based on AVMD-KPCA carries out self-adaptive decomposition on bearing signals to obtain K narrow-band inherent modal component signals and calculates the inherent energy of the K narrow-band inherent modal component signals, and the inherent energy of the obtained narrow-band inherent modal component signals is converted into the degradation index through a KPCA kernel principal component analysis algorithm.
As an optional implementation, the intrinsic energy of the K narrow-band intrinsic mode component signals is compressed in a dimensionality reduction manner by using KPCA, and a kernel function of KPCA is a gaussian kernel function.
As an alternative embodiment, the first principal component extracted by KPCA is used as the degradation indicator estimation value.
As an alternative embodiment, the training of the convolutional neural network includes:
original vibration signal X epsilon R of rolling bearingp×qUsed as input for training convolutional neural network model, the input matrix is formed by dimension a1×b1Is convolved with M convolution kernels, the dimensionality of the convolution layer is (p-a) using the ReLU activation function1+1)×(q-b1+1), the output signature map of the convolutional layer is subsampled in the pooling layer.
As an alternative embodiment, the training data is converted into a plurality of training sample vectors by sliding a time window.
As an optional embodiment, the number of hidden layers and the cells in each hidden layer of the first BiGRU model and the second BiGRU model are optimized using a mesh search method.
The invention provides a rolling bearing residual life prediction system based on bidirectional GRUs.
A bidirectional GRU-based rolling bearing residual life prediction system comprises:
a data acquisition module configured to: acquiring a vibration signal of a rolling bearing;
a degradation indicator estimation module configured to: obtaining a degradation index estimation value of the rolling bearing according to the obtained vibration signal and a preset convolutional neural network model;
a state evaluation module configured to: and obtaining a degradation index evaluation value integrating the trend, the monotonicity and the robustness according to the influence of the monotonicity, the trend and the robustness of the degradation index on the residual life prediction result.
A degradation indicator prediction module configured to: obtaining a degradation index predicted value according to the degradation index estimated value and a preset first BiGRU model;
a remaining useful life prediction module configured to: obtaining a predicted value of the remaining service life according to the degradation index estimated value and a preset second BiGRU model;
a third aspect of the present invention provides a computer-readable storage medium having stored thereon a program which, when executed by a processor, implements the steps in the bidirectional GRU based rolling bearing remaining life prediction method according to the first aspect of the present invention.
A fourth aspect of the present invention provides an electronic device, comprising a memory, a processor and a program stored on the memory and executable on the processor, wherein the processor implements the steps of the method for predicting the remaining life of the bidirectional GRU based rolling bearing according to the first aspect of the present invention when executing the program.
Compared with the prior art, the invention has the beneficial effects that:
1. according to the method and the system for predicting the residual life of the rolling bearing based on the bidirectional GRU, a degradation index predicted value is obtained according to the degradation index estimated value and the preset first BiGRU model; obtaining a predicted value of the remaining service life according to the degradation index estimated value and a preset second BiGRU model; the rolling bearing degradation index estimation method realizes more accurate estimation of the rolling bearing degradation index and more accurate prediction of the residual service life.
2. According to the rolling bearing residual life prediction method and system based on the bidirectional GRU, in the training of the preset convolutional neural network model, firstly, a degradation index extraction method based on AVMD-KPCA is used for carrying out self-adaptive decomposition on bearing signals to obtain K narrow-band inherent modal component signals, the inherent energy of the K narrow-band inherent modal component signals is calculated, and the inherent energy of the obtained narrow-band inherent modal component signals is converted into the degradation index through a KPCA kernel principal component analysis algorithm. And then, using the bearing vibration data as the input of the convolutional neural network model, and using the degradation index as the target output to obtain the convolutional neural network model for extracting the degradation index on line. The subjectivity of obtaining the degradation index is overcome, and the prediction precision of the degradation index is improved.
Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
Fig. 1 is a schematic structural diagram of a method for predicting the remaining life of a bidirectional GRU-based rolling bearing according to embodiment 1 of the present invention.
Fig. 2 is a schematic diagram of a framework of a CNN-based DI estimation method according to embodiment 1 of the present invention.
Fig. 3 is a schematic diagram of the sliding time window processing provided in embodiment 1 of the present invention.
Fig. 4 is a structural diagram of a BiGRU model provided in embodiment 1 of the present invention.
Detailed Description
The invention is further described with reference to the following figures and examples.
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
The embodiments and features of the embodiments of the present invention may be combined with each other without conflict.
Example 1:
the embodiment 1 of the invention provides a method for predicting the residual life of a rolling bearing based on bidirectional GRU (generalized regression unit). firstly, through the complete integration empirical mode decomposition of adaptive noise and dynamic principal component analysis (AVMD-KPCA), a new nonlinear DI (differential index) for training the bearing is extracted, and the degradation process is described according to the real rule of the bearing degradation; then, a CNN model for DI estimation is constructed by learning and capturing the mapping relation between the original vibration signal and the DI of the trained rolling bearing, and the representative features of the CNN model are automatically extracted from the original vibration signal without manually extracting and selecting the features; due to its generalization and robustness, the model can be transferred to other bearings with similar operating conditions without changing the hyper-parameters of the model; once DI is estimated, BiGRU models are built for life prediction, including future DI and RUL prediction.
Specifically, the method comprises the following steps:
s1: and (6) data acquisition. The system is acquired by a multichannel online detection analysis system DH5972N on a rotating mechanical fault simulation platform HFZZ-II.
S2: and (6) extracting DI. Extracting DI from the acquired vibration signal by adopting AVMD-KPCA;
s3: and (4) estimating the DI. Establishing a CNN model, capturing representative characteristics hidden in an original vibration signal, and automatically estimating DI of an online rolling bearing;
s4: and D, DI prediction, namely constructing a BiGRU model, and sending the estimated DI of the online rolling bearing into the trained DI prediction model to perform future DI prediction.
S5: and (3) life prediction, namely constructing a BiGRU model, and sending the estimated DI of the online rolling bearing into a trained life prediction model to predict RUL.
At S1, a data set is obtained from a simulation platform consisting of a series of components that generate vibration signals from run to fault under different operating conditions. The radial force generated by the hydraulic loader was applied to the bearing under test to simulate different operating conditions. The shaft speed is controlled by a motor speed controller. Two accelerometers are used to collect vibration signals with a horizontal and vertical sampling frequency of 20kHz, respectively. The sampling period was set to 1 minute with 1 second of sampling each. Only the horizontal vibration signal is used herein because it contains more information useful for the health degradation of the rolling bearing. The data set included 5 SDK6205 bearings that were tested under three operating conditions. For each case, the first three rolling bearings were used as training bearings and the remainder as test bearings.
In S2, a DI extraction method based on AVMD-KPCA is adopted to carry out self-adaptive decomposition on the bearing signals to obtain K narrow-band inherent modal component signals, and the K narrow-band inherent modal component signals are converted into a degradation index DI through a KPCA kernel principal component analysis algorithm.
The specific steps of the VMD are detailed as follows:
first, for each mode signal uk(t), obtaining its analytic signal by hilbert transform and calculating its single-sided spectrum:
Figure BDA0003565261540000071
next, a center frequency is estimated for each analytic signal
Figure BDA0003565261540000072
And added to the corresponding signal so as to modulate it at a fixed base band, i.e.:
Figure BDA0003565261540000073
finally, the frequency bands of the signal are estimated by solving the gaussian smoothness of the modulated signal, i.e. the L2 norm of its gradient. The variational problem can be expressed as:
Figure BDA0003565261540000074
wherein, { u [ [ u ] ]k}:＝{u1,u2,…uKRepresents the signal ukAll modes of (t) { ω [ omega ] Sk}:＝{ω1,ω2,…ωKRepresents the center frequency of the signal, K is the total number of signal modes,
Figure BDA0003565261540000075
all signal modalities sum.
In order to convert the problem of variation with constraint in the formula (3) into the problem of variation without constraint, a secondary penalty factor alpha and a Lagrange multiplier lambda are introduced; i, formula (3) can convert the augmented Lagrange expression:
Figure BDA0003565261540000081
the variational problem is solved by adopting a multiplicative operator alternating algorithm, and the specific algorithm flow is as follows:
(1) initialization
Figure BDA0003565261540000082
λ1,n＝0
(2) For K1: 1: K, u is updated by solving the optimization problemk。
Figure BDA0003565261540000083
The above problem is converted to the frequency domain using a Parseval/Plancherel Fourier equidistant transform:
Figure BDA0003565261540000084
by omega-omegakInstead of ω in equation (6), the above equation can be converted into:
Figure BDA0003565261540000085
by using the Hermitian symmetry of the real signal in the process of reconstructing the fidelity term, the above formula can be rewritten into an integral form of a non-negative frequency interval, which is specifically as follows:
Figure BDA0003565261540000086
solving this quadratic optimization problem can result in:
Figure BDA0003565261540000091
(3) for K1: 1: K, ω is updated by solving the following optimization problemk：
Figure BDA0003565261540000092
The optimization problem represented by the above equation is converted to the frequency domain for processing in the same processing manner as S2, that is:
Figure BDA0003565261540000093
solving the above equation yields:
Figure BDA0003565261540000094
(4) for a given error decision e, if
Figure BDA0003565261540000095
Terminating the iteration, otherwise, returning to the step (2);
(5) the intrinsic energy e (t) of each narrow band eigenmode component signal is calculated as:
Figure BDA0003565261540000096
and proposes a new index, called spectral cross correlation (SPC), as a selection condition to enable an adaptive selection of the penalty factor α. SPC can assess the degree of modal aliasing for quantification of "overlap". The index SPC is given by the following equation:
Figure BDA0003565261540000097
where F (-) is the Fourier transform of a particular signal or time series.
For each penalty factor α, the Fourier transform of each decomposition mode is computed and its SPC is then retrieved according to equation (14). Then the one closest to the SPC mean is selected and its corresponding penalty factor alpha value is selected as the best value.
Through the algorithm, the vibration signal can be decomposed into K narrow-band inherent modal component signals, and the inherent energy of the signals can be calculated. Then, sending the K sequences consisting of inherent energy to KPCA to perform dimensionality reduction compression, wherein the adopted kernel function is a Gaussian kernel function, and the formula is as follows:
Figure BDA0003565261540000101
the first principal component of the KPCA extraction is chosen herein as DI to describe the degradation process.
At S3, at the convolutional layer, the input is convolved with a set of learnable kernels to obtain a new feature map, as follows:
Figure BDA0003565261540000102
original vibration signal X epsilon R of rolling bearingp×qUsed as input to train the CNN model, and the DI value is used as the target output. The input matrix is formed by dimension a1×b1Is convolved with M convolution kernels.
Using the ReLU activation function, the dimensionality is obtained as (p-a)1+1)×(q-b1+1), the output feature map of the convolutional layer is subsampled in the following pooling layer. Then several convolutional and pooling layers capture representative features from the input raw vibration signal. The fully connected layer then acts as a regression layer, generating a prediction output (DI label). The CNN model is trainedAnd finally, inputting the online vibration signal of the bearing to be tested into the trained CNN model. The CNN model captures representative features directly from the raw vibration signal and can obtain an estimated DI value.
And according to the influence of monotonicity, monotonicity and robustness of the health indexes on the residual life prediction result, giving different weights to the evaluation indexes, and establishing a degradation index evaluation criterion integrating the monotonicity, the monotonicity and the robustness.
Figure BDA0003565261540000111
In the formula, omegaiWhere i is 1,2, and 3 are evaluation index weights, Y (t)k) As an indicator of deterioration, T (T)k) Is a time vector, Vcorr(Y(tk),T(tk) Is a trending value, V, that characterizes the trending of the health indicatormon(Y(tk) Is a monotonic value, which characterizes an increasing or decreasing change in the health indicator, Vrob(Y(tk) Is a robustness value, reflects the tolerance of the health indicator to outliers.
In S4, the GRU model includes 2 gate structures, a reset gate and an update gate, as follows:
Figure BDA0003565261540000112
xtrepresenting input data, ytIs the output of GRU, htRepresenting the output of the GRU unit, r is a reset gate, z is an update gate, r and z together control how the hidden state h is from the frontt-1Calculating to obtain new implicit state htσ denotes sigmoid activation function, WzThe gate weights are updated.
The most basic unit of the BiGRU model consists of a forward-propagating GRU unit and a backward-propagating GRU unit. In a one-way propagation GRU network, the state information is always output from front to back. In the remaining life prediction problem, the output information at the present time may be associated with both the state information at the previous time and the state information at the subsequent time.
The current implicit state information of the BiGRU is input by the current xtImplicit State Forward at time t-1
Figure BDA0003565261540000113
And output of inverted implicit states
Figure 1
The three parts are decided together.
Figure BDA0003565261540000121
The model comprises two steps, a training step and a testing step.
In the training step, a training data set is constructed using the DI values. In order to improve the information quantity of input data and the prediction precision of a model, a sliding time window technology is adopted for constructing a training data set. And continuously sampling by adopting a time window with a fixed length, sliding the time window by one measurement unit each time, and continuously acquiring new samples until the life cycle is finished.
By using a sliding time window processing technique, a training data set can be formed
Figure BDA0003565261540000122
Xt＝[dt,dt+1,…,dt+w]Is the ith training sample vector, where dtWhich represents the normalized DI value of the trained rolling bearing at time t, and w is the length of the time window. y ist＝dt+w+1Is the corresponding tag. By inputting training data, BiGRU can be trained by minimizing the Mean Square Error (MSE) function, expressed as:
Figure BDA0003565261540000123
wherein, ytAnd
Figure BDA0003565261540000124
are actual and predicted labels. T represents the total number of training samples. For fast convergence of the training process, the BiGRU model is equipped with an Adam optimizer that has been proven to be effective in predicting problems.
In the testing step, the DI value of the online test rolling bearing is directly input into the BiGRU model, and the future DI can be predicted.
In S5, another BiGRU model is established to realize RUL prediction for online detection of rolling bearings. The structure of the BiGRU model for RUL prediction is very similar to that of the BiGRU model for future DI prediction.
In the training process, a training data set is constructed using the DI values and the corresponding RUL values. Sliding the time window step by step to obtain input samples ItIn which It＝[dt,dt+1,…dt+l-1]Is the t-th input vector, dt represents the normalized DI value of the rolling bearing at the t moment, and the normalized RUL value o is adoptedtAs an output of the BiGRU model.
Finally, the training set of the BiGRU model for RUL prediction can be expressed as
Figure BDA0003565261540000131
The hyper-parameters of the BiGRU model for RUL prediction, such as the number of hidden layers and the cells in each hidden layer, are optimized based on a grid search technique.
During testing, the estimated DI values are input into the trained BiGRU model, and the corresponding RUL values can be predicted.
FIG. 1 is a schematic diagram of the whole process of the present invention, wherein horizontal vibration signals of a bearing are extracted and converted into degradation indexes DI by an adaptive variational modal decomposition Algorithm (AVMD) and KPCA kernel principal component analysis, a CNN network is trained to extract DI, and original vibration signals X of a rolling bearing belong to Rp×qUsed as input for training the CNN model, and DI value is used as target output, eliminating artificial interference when DI is extracted; and finally, training two BiGRU networks to predict DI and RUL.
Fig. 2 is a block diagram of a CNN neural network. A CNN model was constructed for DI estimation of test rolling bearings. Since the proposed CNN model is versatile and robust, the hyper-parameters of the CNN model are the same in three cases. The model is made up of 7 layers, including two convolutional layers and two max-pooling layers, and three fully-connected layers (F1, F2, and F3). The raw vibration signal of the trained rolling bearing is used as input to the CNN model. In each training sample, 32,400 data points were taken from each original sample, forming a matrix of size 180 × 180. During training, the MSE function is adopted as the loss function of the CNN. After 200 epochs, the optimal model parameters can be obtained by an Adam optimizer.
Fig. 3 is a processing of DI data in training a bidirectional GRU. The GRU model input is in the form of (batch _ size, time _ steps, feature _ nums), where batch _ size refers to the number of samples processed in a batch during model training, time _ steps is a time series step, and feature _ nums is a feature dimension. In order to meet the GRU input requirement, time window sliding is carried out on the original multi-dimensional sensor sequence, and a training sample is constructed. The time window length is time window, which represents the time step of the GRU model, and each time slides forward by one time unit along the time direction, so that a single training sample is a one-dimensional tensor of the time window length, and there is an overlap between two adjacent samples. For the BiGRU that trains the predicted DI, the value corresponding to a time instant after the time window is taken as the label of the sample. For BiGRU to train the predicted RUL, the normalized RUL value o is usedtAs output of the BiGRU model. The performance of bigrus for different window size values is different. When the window size is 5, the BiGRU can achieve the best performance on the bearing. Therefore, in this case, the window size of BiGRU is set to 5.
Fig. 4 is a structural diagram of the BiGRU. After obtaining the estimated DI through the CNN model, a BiGRU model was constructed to predict the future DI of the rolling bearing online test. The number of hidden layers H and the cells K in each hidden layer are two hyper-parameters that control the architecture and topology of the BiGRU model, which has a critical impact on the model performance. These two important hyper-parameters are optimized using grid search techniques. The performance of BiGRU varies when the H and K values are different. It is clear that BiGRU achieves good prediction function when it is composed of 3 hidden layers and 100 neurons per layer. Furthermore, as the H and K values increased, we can see an overall upward trend in training time. This is because larger values of H and K mean that more parameters contained in the BiGRU model need to be optimized.
Example 2:
the embodiment 2 of the present invention provides a bidirectional GRU-based rolling bearing remaining life prediction system, including:
a data acquisition module configured to: acquiring a vibration signal of a rolling bearing;
a degradation indicator estimation module configured to: obtaining a degradation index estimation value of the rolling bearing according to the obtained vibration signal and a preset convolutional neural network model;
a degradation indicator prediction module configured to: obtaining a degradation index predicted value according to the degradation index estimated value and a preset first BiGRU model;
a remaining useful life prediction module configured to: obtaining a predicted value of the remaining service life according to the degradation index estimated value and a preset second BiGRU model;
a state evaluation module configured to: and evaluating the state of the rolling bearing according to the obtained degradation index predicted value and the residual service life predicted value.
The working method of the system is the same as the method for predicting the residual life of the rolling bearing based on the bidirectional GRU provided by the embodiment 1, and the description is omitted here.
Example 3:
embodiment 3 of the present invention provides a computer-readable storage medium on which a program is stored, which when executed by a processor, implements the steps in the bidirectional GRU-based rolling bearing remaining life prediction method according to embodiment 1 of the present invention.
Example 4:
embodiment 4 of the present invention provides an electronic device, which includes a memory, a processor, and a program stored in the memory and executable on the processor, and when the processor executes the program, the processor implements the steps in the method for predicting the remaining life of a bidirectional GRU-based rolling bearing according to embodiment 1 of the present invention.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1. A method for predicting the residual life of a rolling bearing based on bidirectional GRU is characterized by comprising the following steps:
the method comprises the following steps:
acquiring a vibration signal of a rolling bearing;
obtaining a degradation index estimation value of the rolling bearing according to the obtained vibration signal and a preset convolutional neural network model;
obtaining a degradation index predicted value according to the degradation index estimated value and a preset first BiGRU model;
obtaining a predicted value of the remaining service life according to the degradation index estimated value and a preset second BiGRU model;
and evaluating the state of the rolling bearing according to the obtained degradation index predicted value and the residual service life predicted value.
2. The method of predicting the remaining life of a bidirectional GRU based rolling bearing of claim 1, wherein:
in the training of the preset convolutional neural network model, the adopted degradation index acquisition mode is as follows:
the degradation index extraction method based on AVMD-KPCA carries out self-adaptive decomposition on bearing signals to obtain K narrow-band inherent modal component signals and calculates inherent energy, and the inherent energy of the obtained narrow-band inherent modal component signals is converted into the degradation index through a KPCA kernel principal component analysis algorithm.
3. The method of predicting the remaining life of a bidirectional GRU based rolling bearing according to claim 2, characterized in that:
and (3) carrying out dimensionality reduction compression on the inherent energy of the K narrow-band inherent modal component signals by using KPCA, wherein the kernel function of the KPCA is a Gaussian kernel function.
4. The method of predicting the remaining life of a bidirectional GRU based rolling bearing of claim 2, wherein:
and taking the first principal component extracted by KPCA as a degradation index estimated value.
5. The method of predicting the remaining life of a bidirectional GRU based rolling bearing of claim 1, wherein:
the training of the convolutional neural network comprises the following steps:
original vibration signal X epsilon R of rolling bearingp×qAs input for training a convolutional neural network model, the input matrix having a dimension of a1×b1Is convolved by M convolution kernels, the dimensionality of the convolution layer is (p-a) by using the ReLU activation function1+1)×(q-b1+1), the output signature map of the convolutional layer is subsampled in the pooling layer.
6. The method of predicting the remaining life of a bidirectional GRU based rolling bearing of claim 1, wherein:
the training data is converted into a plurality of training sample vectors by sliding a time window.
7. The method of predicting the remaining life of a bidirectional GRU based rolling bearing of claim 1, wherein:
and optimizing the number of hidden layers of the first BiGRU model and the second BiGRU model and the unit in each hidden layer by using a grid searching method.
8. A rolling bearing residual life prediction system based on bidirectional GRU is characterized in that:
the method comprises the following steps:
a data acquisition module configured to: acquiring a vibration signal of a rolling bearing;
a degradation indicator estimation module configured to: obtaining a degradation index estimation value of the rolling bearing according to the obtained vibration signal and a preset convolutional neural network model;
a degradation indicator prediction module configured to: obtaining a degradation index predicted value according to the degradation index estimated value and a preset first BiGRU model;
a remaining useful life prediction module configured to: obtaining a predicted value of the remaining service life according to the degradation index estimated value and a preset second BiGRU model;
a state evaluation module configured to: and evaluating the state of the rolling bearing according to the obtained degradation index predicted value and the residual service life predicted value.
9. A computer-readable storage medium having a program stored thereon, wherein the program, when executed by a processor, implements the steps in the method for predicting the remaining life of a bidirectional GRU based rolling bearing according to any of claims 1-7.
10. An electronic device comprising a memory, a processor and a program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the method for predicting the remaining life of a bidirectional GRU based rolling bearing of any of claims 1-7.
CN202210300479.5A 2022-03-25 2022-03-25 Rolling bearing residual life prediction method and system based on bidirectional GRU Active CN114720129B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210300479.5A CN114720129B (en) 2022-03-25 2022-03-25 Rolling bearing residual life prediction method and system based on bidirectional GRU

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210300479.5A CN114720129B (en) 2022-03-25 2022-03-25 Rolling bearing residual life prediction method and system based on bidirectional GRU

Publications (2)

Publication Number Publication Date
CN114720129A true CN114720129A (en) 2022-07-08
CN114720129B CN114720129B (en) 2023-06-27

Family

ID=82240554

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210300479.5A Active CN114720129B (en) 2022-03-25 2022-03-25 Rolling bearing residual life prediction method and system based on bidirectional GRU

Country Status (1)

Country Link
CN (1) CN114720129B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115438556A (en) * 2022-11-10 2022-12-06 北京理工大学 Method, device and equipment for predicting structural rigidity degradation rate of flexible inflatable aircraft

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120022805A1 (en) * 2009-04-22 2012-01-26 Streamline Automation, Llc Mechanical health monitor apparatus and method of operation therefor
CN103093095A (en) * 2013-01-14 2013-05-08 湖州师范学院 Software failure time forecasting method based on kernel principle component regression algorithm
CN107101813A (en) * 2017-04-26 2017-08-29 河北工业大学 A kind of frame-type circuit breaker mechanical breakdown degree assessment method based on vibration signal
CN110442833A (en) * 2019-06-10 2019-11-12 内蒙古工业大学 One kind assessing Wind turbines health state evaluation method based on various dimensions SCADA data
CN110500217A (en) * 2019-07-23 2019-11-26 南京航空航天大学 Based on can measured data feature common rail for diesel engine system oil pump fault detection method
CN110866314A (en) * 2019-10-22 2020-03-06 东南大学 Method for predicting residual life of rotating machinery of multilayer bidirectional gate control circulation unit network
CN110942103A (en) * 2019-12-05 2020-03-31 桂林电子科技大学 Training method of classifier and computer-readable storage medium
CN112113766A (en) * 2020-09-01 2020-12-22 兰州理工大学 Characteristic extraction method for early damage state of rolling bearing
CN112798280A (en) * 2021-02-05 2021-05-14 山东大学 Rolling bearing fault diagnosis method and system
CN113722985A (en) * 2021-08-12 2021-11-30 武汉科技大学 Method and system for evaluating health state and predicting residual life of aircraft engine

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120022805A1 (en) * 2009-04-22 2012-01-26 Streamline Automation, Llc Mechanical health monitor apparatus and method of operation therefor
CN103093095A (en) * 2013-01-14 2013-05-08 湖州师范学院 Software failure time forecasting method based on kernel principle component regression algorithm
CN107101813A (en) * 2017-04-26 2017-08-29 河北工业大学 A kind of frame-type circuit breaker mechanical breakdown degree assessment method based on vibration signal
CN110442833A (en) * 2019-06-10 2019-11-12 内蒙古工业大学 One kind assessing Wind turbines health state evaluation method based on various dimensions SCADA data
CN110500217A (en) * 2019-07-23 2019-11-26 南京航空航天大学 Based on can measured data feature common rail for diesel engine system oil pump fault detection method
CN110866314A (en) * 2019-10-22 2020-03-06 东南大学 Method for predicting residual life of rotating machinery of multilayer bidirectional gate control circulation unit network
CN110942103A (en) * 2019-12-05 2020-03-31 桂林电子科技大学 Training method of classifier and computer-readable storage medium
CN112113766A (en) * 2020-09-01 2020-12-22 兰州理工大学 Characteristic extraction method for early damage state of rolling bearing
CN112798280A (en) * 2021-02-05 2021-05-14 山东大学 Rolling bearing fault diagnosis method and system
CN113722985A (en) * 2021-08-12 2021-11-30 武汉科技大学 Method and system for evaluating health state and predicting residual life of aircraft engine

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
周裕华: ""滚动轴承的性能退化评估与剩余使用寿命预测方法的研究"", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 *
朱耀楚: ""高压隔膜泵单向阀的故障诊断与性能退化模型研究"", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 *
者娜: ""KPCA和改进SVM在滚动轴承剩余寿命预测中的应用研究"", 《机械设计与制造》 *
赵广谦: ""基于卷积神经网络和双向GRU的滚动轴承剩余寿命预测方法研究"", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115438556A (en) * 2022-11-10 2022-12-06 北京理工大学 Method, device and equipment for predicting structural rigidity degradation rate of flexible inflatable aircraft
CN115438556B (en) * 2022-11-10 2022-12-30 北京理工大学 Method, device and equipment for predicting structural rigidity degradation rate of flexible inflatable aircraft

Also Published As

Publication number Publication date
CN114720129B (en) 2023-06-27

Similar Documents

Publication Publication Date Title
CN111832216A (en) Rolling bearing residual service life prediction method based on EEMD-MCNN-GRU
US11650968B2 (en) Systems and methods for predictive early stopping in neural network training
CN107274011A (en) The equipment state recognition methods of comprehensive Markov model and probability net
Mo et al. Multi-head CNN-LSTM with prediction error analysis for remaining useful life prediction
Ayodeji et al. Causal augmented ConvNet: A temporal memory dilated convolution model for long-sequence time series prediction
CN110757510B (en) Method and system for predicting remaining life of robot
CN108961460B (en) Fault prediction method and device based on sparse ESGP (Enterprise service gateway) and multi-objective optimization
CN112926644A (en) Method and system for predicting residual service life of rolling bearing
CN113723007A (en) Mechanical equipment residual life prediction method based on DRSN and sparrow search optimization BilSTM
CN116562514B (en) Method and system for immediately analyzing production conditions of enterprises based on neural network
CN114266278A (en) Dual-attention-network-based method for predicting residual service life of equipment
CN115587666A (en) Load prediction method and system based on seasonal trend decomposition and hybrid neural network
CN114720129B (en) Rolling bearing residual life prediction method and system based on bidirectional GRU
Regazzoni et al. A physics-informed multi-fidelity approach for the estimation of differential equations parameters in low-data or large-noise regimes
Salah et al. Echo state network and particle swarm optimization for prognostics of a complex system
Esfahani et al. Remaining useful life prognostics based on stochastic degradation modeling: turbofan engine as case study
CN115545334B (en) Land utilization type prediction method and device, electronic equipment and storage medium
CN113779724B (en) Intelligent fault prediction method and system for filling packaging machine
CN109934352B (en) Automatic evolution method of intelligent model
CN116384244A (en) Electromagnetic field prediction method based on physical enhancement neural network
CN113884937B (en) Lithium ion battery residual life prediction method and system based on decomposition integration strategy
US20230297837A1 (en) Method for automated determination of a model compression technique for compression of an artificial intelligence-based model
Kerboua et al. Recurrent neural network optimization for wind turbine condition prognosis
CN113469013A (en) Motor fault prediction method and system based on transfer learning and time sequence
US20230206054A1 (en) Expedited Assessment and Ranking of Model Quality in Machine Learning

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant