Depth hidden variable state space model bearing degradation prediction method and system

Abstract

The invention discloses a deep hidden variable state space model bearing degradation prediction method and system, and the method comprises the steps: correcting the simulation degradation data and real degradation data of a bearing through a data transformation function Box-Cox, fusing a plurality of time domain degradation features to obtain a degradation health index, and obtaining a degradation state representing the degradation rate of the bearing through differential transformation; constructing a prediction model based on a deep hidden variable state space model under a state space model framework; performing feature extraction, fusion and difference on the simulation degradation data and the real data to obtain degradation states of the simulation degradation data and the real data; the method comprises the following steps: acquiring a degradation state pre-training depth hidden variable state space model by using simulation degradation data, initializing the weight of the space model, predicting the degradation state obtained by real data by using a prediction model based on the depth hidden variable state space model, and accumulating the predicted degradation state to obtain a bearing degradation prediction value. The technical problems that an existing process is complex, and a large amount of time and resources are consumed are solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of bearing degradation prediction, and specifically relates to a bearing degradation prediction method and system based on a deep latent variable state space model. Background Art

[0002] Bearings are considered one of the most important components of rotating machinery because of their unique functions. They play a key role in maintaining the normal working position and rotation accuracy of the shaft. They are widely used in many fields such as mechanical processing, transportation, hydropower generation, and aerospace engineering. However, the operating environment of bearings is very complex. They are affected by many factors such as temperature, humidity, and load changes, and are prone to failure. Timely and reasonable predictive maintenance is essential for the normal operation of industrial production. Therefore, accurately predicting the degree of bearing degradation and performing maintenance inspections at the appropriate time are crucial to avoiding business shutdowns and reducing casualties.

[0003] At present, traditional bearing degradation prediction methods mainly rely on methods based on physical models and statistical models. The methods based on physical models are complex and require a lot of time and resources. Most methods based on statistical models rely on limited forms of nonlinear assumptions such as power laws and exponentials, and are often limited to the measurement equations of time series models, which leads to their great limitations in capturing strong nonlinear behaviors and long-term dependencies in bearing degradation data.

[0004] Therefore, researchers began to explore deep learning-based methods to enhance the nonlinear modeling capabilities and adaptability of the model. Deep learning models have achieved some success in time series prediction, but there are still many challenges in bearing degradation prediction. For example: deep learning models cannot provide probabilistic prediction capabilities like statistical models; deep learning models rely heavily on large-scale labeled state data, and have weak learning and generalization modeling capabilities under small samples; it is difficult to model the skewed distribution characteristics and non-constant variance phenomena of degradation data. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a deep latent variable state space model bearing degradation prediction method and system in response to the above-mentioned deficiencies in the prior art, so as to solve the technical problems that the existing methods are complicated and require a lot of time and resources.

[0006] The present invention adopts the following technical solutions: A bearing degradation prediction method using a deep latent variable state space model comprises the following steps: According to the bearing parameters and corresponding working conditions that need to be predicted, a second-order pseudo-cyclostationary bearing fault simulation model based on analytical formula is established to obtain the simulation degradation data of the bearing; The simulated degradation data and the real degradation data were corrected using the data transformation function Box-Cox. The Mahalanobis distance fusion method and the 3-sigma stage division method were used to fuse multiple time-domain degradation features to obtain the degradation health index. The degradation state characterizing the bearing degradation rate was then obtained through differential transformation. In the state space model framework, the recurrent neural network is used as the state transfer equation, and the variational autoencoder is combined as the degradation observation equation to construct a prediction model based on the deep latent variable state space model. The simulated degradation data and real data are subjected to feature extraction, fusion and differentiation to obtain their degradation states. The simulated degradation data is used to obtain the degradation state to pre-train the deep latent variable state space model, initialize the space model weights, and then the degradation state obtained from the real data is predicted using the prediction model based on the deep latent variable state space model. The predicted degradation states are accumulated to obtain the bearing degradation prediction value.

[0007] Preferably, the second-order pseudo-cyclostationary bearing fault simulation model based on the analytical formula is as follows:

[0008] in, T f is the time when the local bearing defect occurs, A i The amplitude modulation gradually increases during the attenuation process, which is composed of the Paris formula, B i is the amplitude distribution of transient pulses that obey the normal distribution, is the pulse attenuation coefficient, r j is the resonant frequency of the rotating system, is the slip time, is the resonant frequency, is the rotation harmonic under variable speed conditions, is the harmonic order of the rotating harmonics, is the occurrence time of the ith pulse, is the amplitude of the rotation harmonics, is Gaussian noise, is the final vibration response of the system in the defect stage.

[0009] Preferably, the vibration model in the health stage is as follows: in, is the resonant frequency, is the rotation harmonic under variable speed conditions, is the harmonic order of the rotating harmonics, is the occurrence time of the ith pulse, is the amplitude of the rotation harmonics, is Gaussian noise.

[0010] Preferably, the obtained simulated degradation data and the real degradation data are corrected using the data transformation function Box-Cox transformation as follows:

[0011] Among them, y is the value after Box-Cox transformation; To adjust the parameters.

[0012] Preferably, the multiple time domain degradation characteristics include: indicators sensitive to overall damage, including the root mean square reflecting signal energy; indicators sensitive to local damage, including the kurtosis reflecting the impact degree; dimensionless indicators that are independent of working conditions and sensitive to damage and failure, including waveform factor and crest factor.

[0013] Preferably, the prediction model based on the deep latent variable state space model is specifically:

[0014] Model Priors:

[0015]

[0016] Model Posterior:

[0017] in, is the joint distribution of the model, which represents the total process of generating predicted values ​​from time 1 to time T. is the degradation prediction value from time 1 to time T, is the hidden variable from time 1 to time T, is the intermediate variable of the transfer distribution from time 1 to time T, is the initial intermediate variable of the recurrent neural network, is the input degradation state from time 1 to time T, is the emission distribution, which represents the process of generating predicted values ​​from latent variables. is the prior distribution, providing the initial hypothesis distribution for the latent variables, To transfer the distribution, we represent the prior distribution of the state at each time step given the previous state and the current input to the model, and learn the temporal correlation in the bearing degradation data.

[0018] Preferably, the degenerate observation equation is:

[0019]

[0020] Among them, z is the latent variable, y is the value after Box-Cox transformation, are the weights and biases of the neural network, is a Gaussian distribution, for t Hidden variables at time steps, Represents the generation process of prior mean and prior variance, which is given by After neural network fitting, are neural network parameters, is the predicted value.

[0021] In a second aspect, an embodiment of the present invention provides a deep latent variable state space model bearing degradation prediction system, comprising: The data module establishes a second-order pseudo-cyclostationary bearing fault simulation model based on an analytical formula according to the bearing parameters to be predicted and the corresponding working conditions, and obtains the simulation degradation data of the bearing; The correction module uses the data transformation function Box-Cox to correct the obtained simulated degradation data and the real degradation data, adopts the Mahalanobis distance fusion method and the 3-sigma stage division method to fuse multiple time-domain degradation features to obtain the degradation health index, and then obtains the degradation state that characterizes the bearing degradation rate through differential transformation; The equation module uses the recurrent neural network as the state transfer equation in the state space model framework and combines the variational autoencoder as the degradation observation equation to build a prediction model based on the deep latent variable state space model. The prediction module extracts, fuses and differentiates the features of the simulated degradation data and the real data to obtain the degradation status of the two; uses the simulated degradation data to obtain the degradation status to pre-train the deep latent variable state space model, initializes the space model weights, and then uses the prediction model based on the deep latent variable state space model to predict the degradation status obtained from the real data; and obtains the bearing degradation prediction value by accumulating the predicted degradation status.

[0022] In a third aspect, a computer device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned deep latent variable state space model bearing degradation prediction method when executing the computer program.

[0023] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, comprising a computer program, which, when executed by a processor, implements the steps of the above-mentioned deep latent variable state space model bearing degradation prediction method.

[0024] In a fifth aspect, a chip comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above-mentioned deep latent variable state space model bearing degradation prediction method when executing the computer program.

[0025] In a sixth aspect, an embodiment of the present invention provides an electronic device, comprising a computer program, which, when executed by the electronic device, implements the steps of the above-mentioned deep latent variable state space model bearing degradation prediction method.

[0026] Compared with the prior art, the present invention has at least the following beneficial effects: A bearing degradation prediction method based on a deep latent variable state space model utilizes the nonlinear mapping capability of a deep neural network and combines the characteristics of uncertainty structure modeling provided by a state space model to improve the nonlinear modeling capability under complex working conditions. It alleviates the problem of data scarcity in the degradation prediction process and introduces the physical mechanism of bearing degradation in the pre-training process to help the model better understand the degradation law and accelerate the model training process. It uses the knowledge provided by the fault model based on the analytical formula as prior knowledge to guide the prediction model to learn the characteristics of the system and guide the model to more accurately capture the state changes of the system when predicting degradation behavior, alleviating the problem of data scarcity, accelerating the model training process and improving the prediction accuracy. By performing Box-cox transformation on the original data, the skewness and non-normality of the data can be eliminated, and the degradation rate can be predicted after differential transformation, effectively improving the long-term prediction capability of the model. It is suitable for the degradation trend prediction of mechanical parts, especially the degradation trend prediction of bearings, providing a promising solution for industrial applications. The degradation trend prediction has high accuracy and can achieve highly accurate prediction effects even for the degradation trend of mechanical parts under actual complex working conditions.

[0027] Furthermore, in the process of predicting the bearing degradation trend, the collected experimental data may contain environmental noise, measurement errors and minor defects. In addition, deep learning models with a large number of neural network layers usually require a large amount of data to train the model. However, in practice, industrial data sets containing bearing degradation states may not provide a sufficient number of samples for comprehensive model training. In order to solve this problem, the vibration signal characteristics of the bearing when healthy and faulty are analyzed, and the second-order pseudo-cyclostationary bearing fault simulation model based on the analytical formula is studied to generate mechanism degradation data containing healthy and faulty states. Using simulation data for pre-training can give full play to the prior knowledge of mechanism data, so that the parameters of the deep latent variable state space model can obtain the bearing degradation characteristics in advance, accelerate model convergence and reduce dependence on real data. This method alleviates the problem of data scarcity and improves prediction accuracy.

[0028] Furthermore, the Box-Cox transformation is a method to adjust the data distribution shape through power function transformation. For degraded data, the distribution difference between simulated data and real data may be due to the inconsistency of noise level, skewness characteristics or dynamic range. Through the Box-Cox transformation, the statistical characteristics of the two types of data can be made closer without changing the core trend of the data, thereby enhancing the representativeness and adaptability of simulated data in model training. This method further optimizes the pre-training effect of the model, enabling the prediction model to capture degradation characteristics more accurately.

[0029] Furthermore, in order to accurately depict the "health-degraded" stage and minimize interference, the original data needs to be processed. At present, various degradation characteristic indicators have been proposed in the field of bearing fault diagnosis, all of which can reflect the bearing degradation in a certain specific mode. However, considering that the characteristic parameters and characteristic frequencies of each component of the tested bearing are unknown, and the damage of each component may appear in a complex manner, the time domain statistical features are selected as the input features of the extraction method. Time domain statistical indicators can effectively reduce the noise present in the original signal, and many indicators themselves have clear physical meanings. We selected 14 indicators to evaluate the health of the bearing, including indicators that are sensitive to overall damage, such as the root mean square reflecting the signal energy, and indicators that are sensitive to local damage, such as the kurtosis reflecting the impact degree. We also selected dimensionless indicators that are independent of the working conditions and sensitive to damage and faults, such as the waveform factor and the crest factor, aiming to comprehensively reflect the bearing degradation caused by various forms of faults.

[0030] It can be understood that the beneficial effects of the second to sixth aspects mentioned above can be found in the relevant description of the first aspect mentioned above, and will not be repeated here.

[0031] In summary, the present invention constructs a bearing degradation prediction model based on a deep latent variable state space model, which not only has a significant improvement in nonlinear modeling capabilities, but also combines the uncertainty structure modeling provided by SSM; in addition, in view of the problem that industrial data sets may not be able to provide a sufficient number of samples, a second-order pseudo-cyclostationary bearing fault simulation model based on an analytical formula is used to provide mechanism degradation information and alleviate the problem of sample scarcity; the introduction of the Box-Cox transformation reduces the difference between the simulation data and the original data, improves the prediction accuracy and robustness of the model, and improves the reliability of life prediction; the use of differential transformation to obtain the degradation state that characterizes the bearing degradation rate, and then the model is trained and predicted to effectively improve the long-term prediction capability of the model.

[0032] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1It is a schematic diagram of the process of the present invention; Figure 2 It is a framework diagram of the deep latent variable state space model of the present invention; Figure 3 It is a graphical model diagram of the deep latent variable state space model of the present invention; Figure 4 It is a framework diagram of the bearing fault model of the present invention; Figure 5 It is a simulation degradation data diagram of the bearing fault model of the present invention; FIG6 is the prediction result of the present invention for the PHM2012 bearing data set, where (a) is Bearing1_1, and (b) is Bearing1_3; Figure 7 A schematic diagram of a computer device provided by an embodiment of the present invention; Figure 8 The block diagram is a chip provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] In the description of the present invention, it should be understood that the terms “include” and “comprises” indicate the presence of described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0036] It should also be understood that the terms used in the present specification are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the present specification and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0037] It should be further understood that the term "and / or" used in the present specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes these combinations. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present invention generally indicates that the associated objects are in an "or" relationship.

[0038] It should be understood that, although the terms first, second, third, etc. may be used to describe preset ranges, etc. in the embodiments of the present invention, these preset ranges should not be limited to these terms. These terms are only used to distinguish preset ranges from each other. For example, without departing from the scope of the embodiments of the present invention, the first preset range may also be referred to as the second preset range, and similarly, the second preset range may also be referred to as the first preset range.

[0039] The word "if" as used herein may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)", depending on the context.

[0040] Various structural schematic diagrams of the embodiments disclosed in the present invention are shown in the accompanying drawings. These figures are not drawn to scale, and some details are magnified and some details may be omitted for the purpose of clear expression. The shapes of various regions and layers shown in the figures and the relative sizes and positional relationships therebetween are only exemplary, and may deviate in practice due to manufacturing tolerances or technical limitations, and those skilled in the art may additionally design regions / layers with different shapes, sizes, and relative positions according to actual needs.

[0041] The present invention provides a bearing degradation prediction method using a deep latent variable state space model. First, under the framework of the state space model, a recurrent neural network is used as a state transfer equation, and a variational autoencoder is combined as a degradation observation equation to construct a deep latent variable state space model. Secondly, according to the bearing parameters to be predicted and the corresponding working conditions, a second-order pseudo-cyclostationary bearing fault simulation model based on an analytical formula is established to obtain the degradation data of the bearing. The simulated degradation data and the real degradation data are corrected using the Box-Cox variation, and a variety of features are fused using the Mahalanobis distance fusion method and the 3-sigma stage division method to obtain a health index. The degradation state characterizing the bearing degradation rate is then obtained through differential transformation. The model is pre-trained using the data obtained from the simulated degradation data, the weights are initialized, and the key features of the degradation data are captured. The model is further trained and predicted using the data obtained from the real bearing degradation data. Compared with existing methods, the present invention can use less historical data, significantly enhance the nonlinear modeling capability, and achieve predictions with higher accuracy and stronger adaptability.

[0042] Example 1 See also Figure 1 The present invention provides a bearing degradation prediction method using a deep latent variable state space model, comprising the following steps: S1, bearing degradation data preprocessing; The collected bearing vibration signals are transformed using the data transformation function Box-Cox to eliminate the skewness and non-normality of the data; multiple time domain degradation features are selected and fused using the Mahalanobis distance fusion method and the 3-sigma stage division method; the fused data are smoothed to generate clear indicators that describe the health and degradation stages; the health indicators are processed using differential transformation to obtain the degradation state that characterizes the bearing degradation rate. The specific steps are as follows: S101, using the data transformation function Box-Cox to transform the collected bearing vibration signal to eliminate the skewness and non-normality of the data; The transformation process is as follows:

[0043] Among them, y is the value after Box-Cox transformation; To adjust the parameters, adjust The value turns the Box-Cox transformation into a logarithmic transformation and a power-law transformation.

[0044] S102, selecting multiple time domain degradation features; Indicators that are sensitive to overall damage, such as the root mean square that reflects the signal energy, and indicators that are sensitive to local damage, such as the kurtosis that reflects the impact degree. Dimensionless indicators that are independent of working conditions and sensitive to damage and failure, such as the form factor and crest factor, are also selected.

[0045] S103, using the Mahalanobis distance fusion method to obtain the degradation health index; The stage division was performed using the 3-sigma criterion, and finally, the fused data obtained using the Mahalanobis distance were smoothed to produce a clear “fitness degradation” indicator.

[0046] S104. Process the health index using differential transformation to obtain a degradation state representing the bearing degradation rate.

[0047] S2. Construct a prediction model based on a deep latent variable state space model; In the framework of traditional statistical learning state space model, the observation equation of the state space model is constructed using variational autoencoder. The latent variables of the variational autoencoder are time-expanded using recurrent neural network, and the prior parameters of the variational autoencoder are updated in sequence according to the output of the recurrent neural network. The time correlation between bearing degradation data is learned, and a degradation model based on the deep latent variable state space model is constructed.

[0048] In the framework of traditional statistical learning state space model, the observation equation of the state space model is constructed using variational autoencoder; The degenerate observation equation is:

[0049]

[0050] in, z is a latent variable, which is relatively independent in VAE; is the predicted value, and its parameters are calculated through the neural network through the latent variable z. are the weights and biases of the neural network.

[0051] The latent variables of the variational autoencoder are time-expanded using a recurrent neural network, and the variational autoencoder prior parameters are updated sequentially according to the output of the recurrent neural network. In this way, the time correlation between bearing degradation is learned. As follows:

[0052] in, Represents the computational process of a recurrent neural network, u is the input degenerate state, h is the middle layer of the recurrent neural network. z is a hidden variable.

[0053] In summary, the joint distribution of the model is expressed as:

[0054] Prior distribution:

[0055]

[0056] Among them, the prior distribution Represents hidden variables z Generated by Gaussian distribution, represents a Gaussian distribution, for t Hidden variables at time steps, prior Represents a priori, is the prior mean, is the prior variance, Represents the generation process of prior mean and prior variance, which is given by After neural network fitting, are neural network parameters.

[0057] Model Posterior:

[0058] Among them, the posterior distribution represents the latent variable from the observed data l With intermediate variables h get, his an intermediate variable in the recurrent neural network that contains past information. The above process is split into independent processes of cyclic iteration to obtain the emission distribution of the tth time step , the prior distribution , transfer distribution , The latent variables in the transfer distribution are generated by Gaussian distribution. represents a Gaussian distribution, is the hidden variable of time step t, prior represents the prior, is the prior mean, is the prior variance, Represents the generation process of prior mean and prior variance, which is given by After neural network fitting, are neural network parameters.

[0059] S3. Model pre-training method integrating domain knowledge.

[0060] According to the bearing parameters and corresponding working conditions, the vibration signals of the bearing in the healthy stage and the defective stage are constructed, and a second-order pseudo-cyclostationary bearing fault simulation model based on the analytical formula under the variable speed condition is formed to obtain the simulation degradation data of the bearing.

[0061] S301, constructing a bearing fault model based on an analytical formula; In the healthy stage, the subtle resonance phenomenon caused by random vibration excitation is ignored, and a vibration model including rotation harmonics and Gaussian noise is constructed; The vibration model is as follows:

[0062] Among them, the harmonic order q(n) of the rotating harmonic is approximately an integer, that is, q(n)≈n. wgn(t) represents Gaussian noise, and R(t) is the rotating harmonic; C n Represents the amplitude of the rotational harmonics.

[0063] The speed R(t) under speed change condition is expressed as follows: The instantaneous shaft phase is derived from this.

[0064]

[0065] In the defect stage, a vibration model including rotation harmonics, repetitive transient pulses of different intensities caused by defects, and constant intensity Gaussian noise is established. Among them, the repetitive pulse is established based on pseudo-cyclostationarity, and the formula is:

[0066] in, is the slip time, which varies between 1% and 3% of the instantaneous duration of the fault pulse, Indicates the natural frequency.

[0067] The Paris formula is used to simulate the degradation process of the bearing. The formula is as follows:

[0068] The unit impulse response A i The time domain product of the amplitude distribution and the pulse sequence d(t) is then convolved with the vibration attenuation function e(t) to generate the final vibration response of the system in the defect stage. The formula is as follows:

[0069] Among them, A i Indicates that the amplitude modulation tends to increase gradually during the attenuation process, B i represents the amplitude distribution of transient pulses that obey the normal distribution, is the attenuation coefficient of the pulse, r j is the resonant frequency of the rotating system.

[0070] By integrating the formulas established for the healthy state and defective state, the analytical formula for the pseudo-cyclostationary failure mechanism model of the bearing under variable speed conditions is derived, as follows:

[0071] in, T f Indicates the time when a local bearing defect occurs, When is the bearing healthy stage, the vibration signal contains only rotation harmonics and Gaussian noise. When is the bearing failure stage, the vibration signal includes repetitive transient pulses of different intensities caused by defects, rotation harmonics, and Gaussian noise of constant intensity; the simulated data with integrated domain knowledge generated by this formula can be used for model pre-training.

[0072] S302: Pre-training of a model integrating domain knowledge.

[0073] The method proposed in the present invention was verified using data obtained from the PHM 2012 forecasting challenge. The specific steps are as follows: S3021. Collect vibration signals of the bearing; S3022. According to the actual bearing parameters and corresponding working conditions given by the data set, a bearing fault simulation model based on an analytical formula is established to simulate and obtain the simulated degradation data of the rolling bearing, such as Figure 5 As shown; S3023. Perform Box-cox transformation on the data obtained from simulation and the PHM2012 bearing degradation data, select 14 time-domain statistical features as input features of the extraction method, use the Mahalanobis distance fusion method and the 3-sigma standard stage division method to fuse the 14 features, and then smooth them to obtain a clear "health degradation" indicator; S3024. Process the health index using differential transformation to obtain a degradation state that characterizes the bearing degradation rate.

[0074] Example 2 The present invention provides a deep latent variable state space model bearing degradation prediction system, which can be used to implement the above-mentioned deep latent variable state space model bearing degradation prediction method. Specifically, the deep latent variable state space model bearing degradation prediction system includes a data module, a correction module, an equation module and a prediction module.

[0075] Among them, the data module establishes a second-order pseudo-cyclostationary bearing fault simulation model based on an analytical formula according to the bearing parameters and corresponding working conditions that need to be predicted, and obtains the simulation degradation data of the bearing; The correction module corrects the obtained simulated degradation data and the real degradation data, integrates multiple time-domain degradation characteristics to obtain the degradation health index, and then obtains the degradation state that characterizes the bearing degradation rate through differential transformation; The equation module uses the recurrent neural network as the state transfer equation in the state space model framework and combines the variational autoencoder as the degradation observation equation to build a prediction model based on the deep latent variable state space model. The prediction module extracts, fuses and differentiates the features of the simulated degradation data and the real data to obtain the degradation status of the two; uses the simulated degradation data to obtain the degradation status to pre-train the deep latent variable state space model, initializes the space model weights, and then uses the prediction model based on the deep latent variable state space model to predict the degradation status obtained from the real data; and obtains the bearing degradation prediction value by accumulating the predicted degradation status.

[0076] Example 3 The present invention provides a terminal device, which includes a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of the deep latent variable state space model bearing degradation prediction method, including: According to the bearing parameters that need to be predicted and the corresponding working conditions, a second-order pseudo-cyclostationary bearing fault simulation model based on analytical formulas is established to obtain the simulated degradation data of the bearing; the obtained simulated degradation data and the real degradation data are corrected, and a variety of time-domain degradation features are integrated to obtain the degradation health index, and then the degradation state characterizing the bearing degradation rate is obtained through differential transformation; in the state space model framework, the recurrent neural network is used as the state transfer equation, and the variational autoencoder is combined as the degradation observation equation to construct a prediction model based on the deep latent variable state space model; the simulated degradation data and the real data are feature extracted, fused, and differentiated to obtain the degradation states of the two; the simulated degradation data is used to obtain the degradation state to pre-train the deep latent variable state space model, the space model weights are initialized, and then the degradation state obtained from the real data is predicted by the prediction model based on the deep latent variable state space model, and the predicted value of bearing degradation is obtained by accumulating the predicted degradation states.

[0077] See also Figure 7 , the terminal device is a computer device. The computer device 60 of this embodiment includes: a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61. When the computer program 63 is executed by the processor 61, the method for calculating the fluid composition in the reservoir transformation wellbore in the embodiment is implemented. To avoid repetition, it is not described one by one here. Alternatively, when the computer program 63 is executed by the processor 61, the functions of each model / unit in the bearing degradation prediction system of the deep latent variable state space model in the embodiment are implemented. To avoid repetition, it is not described one by one here.

[0078] The computer device 60 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art will appreciate that Figure 7 This is only an example of the computer device 60 and does not constitute a limitation of the computer device 60. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.

[0079] The processor 61 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0080] The memory 62 may be an internal storage unit of the computer device 60, such as a hard disk or memory of the computer device 60. The memory 62 may also be an external storage device of the computer device 60, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc., equipped on the computer device 60.

[0081] Furthermore, the memory 62 may include both an internal storage unit of the computer device 60 and an external storage device. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 may also be used to temporarily store data that has been output or is to be output.

[0082] See also Figure 8 The terminal device is an electronic device 600, which is in the form of a general-purpose computing device. The components of the electronic device may include, but are not limited to: at least one processing unit 610, at least one storage unit 620, a bus 630 connecting different platform components (including the storage unit 620 and the processing unit 610), a display unit 640, etc.

[0083] The storage unit stores program codes, which can be executed by the processing unit 610, so that the processing unit 610 performs the steps according to various exemplary embodiments of the present invention described in the above method section of this specification. For example, the processing unit 610 can perform the following steps: Figure 1 Follow the steps shown in .

[0084] The storage unit 620 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 6201 and / or a cache memory unit 6202 , and may further include a read-only memory unit (ROM) 6203 .

[0085] The storage unit 620 may also include a program / utility 6204 having a set (at least one) of program modules 6205, such program modules 6205 including but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment.

[0086] Bus 630 may represent one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

[0087] The electronic device 600 may also communicate with one or more external devices 700 (e.g., keyboards, pointing devices, Bluetooth devices, etc.), may also communicate with one or more devices that enable a user to interact with the electronic device 600, and / or communicate with any device that enables the electronic device 600 to communicate with one or more other computing devices (e.g., routers, modems, etc.). Such communication may be performed via an input / output (I / O) interface 650. Furthermore, the electronic device 600 may also communicate with one or more networks (e.g., a local area network (LAN), a wide area network (WAN), and / or a public network, such as the Internet) via a network adapter 660. The network adapter 660 may communicate with other modules of the electronic device 600 via a bus 630. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 600, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms, etc.

[0088] Example 4 The present invention also provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a terminal device for storing programs and data. It is understandable that the computer-readable storage medium here can include both built-in storage media in the terminal device and extended storage media supported by the terminal device, and can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by a processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that more specific examples (non-exhaustive list) of the computer-readable storage medium here include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0089] Computer readable storage media also include data signals propagated in baseband or as part of a carrier wave, which carry readable program codes. Such propagated data signals can take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable storage medium can also be any readable medium other than a readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, device, or device. The program code contained on the readable storage medium can be transmitted using any appropriate medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination of the above.

[0090] Program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0091] One or more instructions stored in a computer-readable storage medium may be loaded and executed by a processor to implement the corresponding steps of the bearing degradation prediction method of the deep latent variable state space model in the above embodiment; one or more instructions in the computer-readable storage medium are loaded and executed by the processor as follows: According to the bearing parameters that need to be predicted and the corresponding working conditions, a second-order pseudo-cyclostationary bearing fault simulation model based on analytical formulas is established to obtain the simulated degradation data of the bearing; the obtained simulated degradation data and the real degradation data are corrected, and a variety of time-domain degradation features are integrated to obtain the degradation health index, and then the degradation state characterizing the bearing degradation rate is obtained through differential transformation; in the state space model framework, the recurrent neural network is used as the state transfer equation, and the variational autoencoder is combined as the degradation observation equation to construct a prediction model based on the deep latent variable state space model; the simulated degradation data and the real data are feature extracted, fused, and differentiated to obtain the degradation states of the two; the simulated degradation data is used to obtain the degradation state to pre-train the deep latent variable state space model, the space model weights are initialized, and then the degradation state obtained from the real data is predicted by the prediction model based on the deep latent variable state space model, and the predicted value of bearing degradation is obtained by accumulating the predicted degradation states.

[0092] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The components of the embodiments of the present invention described and shown in the drawings here can usually be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0093] The health indicators obtained from the simulated degradation data are used to pre-train the deep latent variable state space model, initialize the model weights and capture the key features of the degradation data. The health indicators of the bearing degradation data in the PHM2012 dataset are divided into the first 60% training data and the last 40% validation data; the pre-trained model is further trained using the training data; when the training converges to a stable level, the validation model is used to predict the bearing degradation data, and the predicted degradation rates are accumulated to obtain the final bearing degradation prediction value; the prediction results of the two datasets Bearing1_1 and Bearing1_3 in the dataset are shown in Figure 6.

[0094] The results show that the prediction effect of the method proposed in this paper is better than that of the statistical model Switching SSM, the deep learning model LSTM and the hybrid model Deepstate.

[0095] In summary, the present invention provides a bearing degradation prediction method and system for a deep latent variable state space model. In terms of the modeling capability of complex nonlinear bearing degradation trends, the dynamic characteristics capture capability of the bearing degradation process is significantly improved, and the nonlinear behavior in the degradation process can be more accurately reflected. At the same time, the method of the present invention has the modeling capability of uncertain factors, and can effectively deal with the uncertainty phenomenon of model parameters caused by random factors, etc. By introducing fault simulation data, the dependence of traditional deep learning models on a large amount of training sample data is overcome, and the model parameters are initialized by simulation data containing mechanism information, which effectively improves the prediction accuracy. Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in one processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other and are not used to limit the protection scope of this application. The specific working process of the units and modules in the above system can refer to the corresponding process in the above method embodiment, which will not be repeated here.

[0096] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0097] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the present invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0098] In the embodiments provided by the present invention, it should be understood that the disclosed devices / terminals and methods can be implemented in other ways. For example, the device / terminal embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0099] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0100] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0101] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0102] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices, and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0103] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0104] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0105] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A bearing degradation prediction method based on a deep latent variable state space model, characterized in that: The following steps are involved: According to the bearing parameters and corresponding working conditions that need to be predicted, a second-order pseudo-cyclostationary bearing fault simulation model based on analytical formula is established to obtain the simulation degradation data of the bearing; The obtained simulated degradation data and real degradation data are corrected, and a variety of time-domain degradation characteristics are integrated to obtain the degradation health index, and then the degradation state representing the bearing degradation rate is obtained through differential transformation; In the state space model framework, the recurrent neural network is used as the state transfer equation, and the variational autoencoder is combined as the degradation observation equation to construct a prediction model based on the deep latent variable state space model. The simulated degradation data and real data are subjected to feature extraction, fusion and differentiation to obtain their degradation states. The simulated degradation data is used to obtain the degradation state to pre-train the deep latent variable state space model, initialize the space model weights, and then the degradation state obtained from the real data is predicted using the prediction model based on the deep latent variable state space model. The predicted degradation states are accumulated to obtain the bearing degradation prediction value.

2. The bearing degradation prediction method based on deep latent variable state space model according to claim 1 is characterized in that: The second-order pseudo-cyclostationary bearing fault simulation model based on the analytical formula is as follows: in, T f is the time when the local bearing defect occurs, A i The amplitude modulation gradually increases during the attenuation process, which is composed of the Paris formula, B i is the amplitude distribution of transient pulses that obey the normal distribution, is the pulse attenuation coefficient, r j is the resonant frequency of the rotating system, is the slip time, is the resonant frequency, is the rotation harmonic under variable speed conditions, is the harmonic order of the rotating harmonics, is the occurrence time of the ith pulse, is the amplitude of the rotation harmonics, is Gaussian noise, is the final vibration response of the system in the defect stage.

3. The bearing degradation prediction method based on deep latent variable state space model according to claim 1 is characterized in that: The vibration model in the healthy stage is as follows: in, is the resonant frequency, is the rotation harmonic under variable speed conditions, is the harmonic order of the rotating harmonics, is the occurrence time of the ith pulse, is the amplitude of the rotation harmonics, is Gaussian noise.

4. The bearing degradation prediction method based on deep latent variable state space model according to claim 1 is characterized in that: The simulated degradation data and the real degradation data are corrected using the data transformation function Box-Cox, as follows: Among them, y is the value after Box-Cox transformation; To adjust the parameters.

5. The bearing degradation prediction method based on deep latent variable state space model according to claim 1 is characterized in that: The various time-domain degradation characteristics include: indicators that are sensitive to global damage, including the root mean square that reflects the signal energy; indicators that are sensitive to local damage, including the kurtosis that reflects the degree of impact; dimensionless indicators that are independent of operating conditions and sensitive to damage and failure, including form factor and crest factor.

6. The bearing degradation prediction method based on deep latent variable state space model according to claim 1 is characterized in that: The prediction model based on the deep latent variable state space model is as follows: Model Priors: Model Posterior: in, is the joint distribution of the model, which represents the total process of generating predicted values ​​from time 1 to time T. is the degradation prediction value from time 1 to time T, is the hidden variable from time 1 to time T, is the intermediate variable of the transfer distribution from time 1 to time T, is the initial intermediate variable of the recurrent neural network, is the input degradation state from time 1 to time T, is the emission distribution, which represents the process of generating predicted values ​​from latent variables. is the prior distribution, providing the initial hypothesis distribution for the latent variables, To transfer the distribution, we represent the prior distribution of the state at each time step given the previous state and the current input to the model, and learn the temporal correlation in the bearing degradation data.

7. The bearing degradation prediction method based on deep latent variable state space model according to claim 6 is characterized in that: The degenerate observation equation is: Among them, z is the latent variable, y is the value after Box-Cox transformation, are the weights and biases of the neural network, is a Gaussian distribution, for t Hidden variables at time steps, Represents the generation process of prior mean and prior variance, which is given by After neural network fitting, are neural network parameters, is the predicted value.

8. A deep latent variable state space model bearing degradation prediction system, characterized in that: include: The data module establishes a second-order pseudo-cyclostationary bearing fault simulation model based on an analytical formula according to the bearing parameters to be predicted and the corresponding working conditions, and obtains the simulation degradation data of the bearing; The correction module corrects the obtained simulated degradation data and the real degradation data, integrates multiple time-domain degradation characteristics to obtain the degradation health index, and then obtains the degradation state that characterizes the bearing degradation rate through differential transformation; The equation module uses the recurrent neural network as the state transfer equation in the state space model framework and combines the variational autoencoder as the degradation observation equation to build a prediction model based on the deep latent variable state space model. The prediction module extracts, fuses, and differentiates the simulated degradation data and the real data to obtain their degradation states. The simulated degradation data is used to obtain the degradation state to pre-train the deep latent variable state space model, initialize the space model weights, and then use the prediction model based on the deep latent variable state space model to predict the degradation state obtained from the real data. The predicted degradation state is accumulated to obtain the bearing degradation prediction value.

9. A computer-readable storage medium storing one or more programs, characterized in that: The one or more programs include instructions, which, when executed by a computing device, cause the computing device to perform the method of any one of claims 1 to 7.

10. A computing device, characterized in that include: One or more processors, a memory and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include steps for executing the method according to any one of claims 1 to 7.

Images