Bearing data generation method and system based on physical model information migration

Abstract

The invention belongs to the technical field of intelligent bearings, and particularly discloses a bearing data generation method and system based on physical model information migration, and the method comprises the steps: extracting a mean value and a standard deviation of Gaussian distribution of vibration acceleration signal distribution characteristics of a bearing, generating Gaussian white noise, and calculating the contact stiffness between a ball bearing and inner and outer ring raceways; a four-degree-of-freedom dynamic balance equation of the fault ball bearing is constructed and solved according to the contact deformation and the relation between the contact deformation and the load, and a bearing dynamic response signal is obtained; gaussian white noise is migrated to a bearing dynamics dynamic response signal, and bearing fusion data distributed like the actually measured signal is generated; and constructing a target function, carrying out optimization solution on the target function, judging whether a target convergence effect is achieved or not, if so, updating the noise addition control parameters, and outputting bearing fusion data. According to the technical scheme, the bearing fusion data is generated through a small number of actual measurement signals, acquisition of the actual measurement signals is reduced, and manpower and material resources are saved.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent bearings and relates to a bearing data generation method and system based on physical model information migration. Background Art

[0002] As a key component of rotating machinery, bearings have the characteristics of strong load-bearing capacity, high precision, good reliability, friction reduction, noise reduction and wide adaptability. They are widely used in machinery manufacturing, aerospace, automobile industry and other industries.

[0003] Bearings are an indispensable and important part of all kinds of machinery and equipment. When a bearing fails, it will not only affect the normal operation and production efficiency of the equipment, but may also cause serious harm to personnel safety and corporate economy. Therefore, it is very important to detect and deal with bearing failures in a timely manner.

[0004] As an important research direction of high-end bearings, the intelligent research of bearings is particularly important. However, predictive fault diagnosis of bearings requires a large amount of basic data, but in actual engineering applications, the acquisition of bearing data requires a lot of time, financial resources and manpower, resulting in a low intelligence level and poor diagnostic accuracy of early intelligent bearings. Therefore, designing a bearing data generation method has a high practical value for the accumulation of bearing data and subsequent data analysis.

[0005] At present, there are various methods for generating bearing data, including establishing a bearing dynamics model to obtain dynamic response as bearing data or using a generative adversarial network. Inspired by the digital-analog driven method, a joint data and model driven method can be used to generate bearing data, which has application prospects in the field of bearing data generation.

[0006] However, there are still related problems in how to link data with models to generate bearing signal data with smaller information distribution differences. Therefore, researchers in this field are urgently needed to solve such technical problems, promote the development of bearing data generation, and improve the intelligence level of smart bearings. Summary of the invention

[0007] The purpose of the present invention is to provide a bearing data generation method and system based on physical model information migration, which generates bearing fusion data through a small amount of measured signals, reduces the collection of measured signals, and saves manpower and material resources.

[0008] In order to achieve the above object, the basic scheme of the present invention is: a bearing data generation method based on physical model information migration, comprising the following steps:

[0009] S1, collect the vibration acceleration signal of the bearing and extract the mean and standard deviation of its distribution characteristic Gaussian distribution;

[0010] S2, based on the mean and standard deviation of the Gaussian distribution of the distribution characteristics of the vibration acceleration signal, generates Gaussian white noise that is the same as the measured bearing vibration acceleration signal;

[0011] S3, simplify the bearing into a mass-spring-damper system, calculate the contact stiffness between the ball bearing and the inner and outer ring raceways, as well as the relationship between the contact deformation and the load, and construct the four-degree-of-freedom dynamic equilibrium equation of the faulty ball bearing;

[0012] S4, using the Runge-Kutta method, solve the four-degree-of-freedom dynamic equilibrium equation of the faulty ball bearing and obtain the dynamic response signal of the bearing;

[0013] S5, transferring Gaussian white noise to the dynamic response signal of the bearing dynamics to generate bearing fusion data with the same distribution as the measured signal;

[0014] S6, constructing the objective function based on the bearing fusion data and measured signals;

[0015] S7, optimize and solve the objective function to determine whether the target convergence effect is achieved. If so, the noise addition control parameters obtained by the solution are input into the migration process in step S5, and the bearing fusion data is output; otherwise, the noise addition control parameters are automatically selected to step S6.

[0016] The working principle and beneficial effect of this basic solution are as follows: this technical solution generates Gaussian noise with the same distribution as the measured signal, and uses the established dynamic model to simulate the bearing dynamic response data, and finally fuses the Gaussian noise and the bearing dynamic response data to obtain the fused bearing signal, which can be generated by a small amount of measured signals, reducing the collection of measured signals and saving manpower and material resources. At the same time, the generated signal has the same distribution as the measured signal, and the signal is reliable, which solves the problem of difficulty in obtaining bearing data in actual engineering applications.

[0017] Further, step S1 is specifically as follows:

[0018] Extract the distribution characteristics of the bearing vibration acceleration signal and the mean value μ of the Gaussian distribution r and its standard deviation σ r :

[0019]

[0020] Among them, S is the measured sample data; n is the total number of samples, S i Represents the measured sample data of the i-th sample.

[0021] Extract the distribution characteristics of the bearing vibration acceleration signal and the mean value μ of the Gaussian distribution r and its standard deviation σ r , which is convenient for subsequent use.

[0022] Further, step S2 is specifically as follows:

[0023] Through the mean μ r With standard deviation σ r , constraining the noise added to the simulation signal, the added Gaussian noise must satisfy the distribution:

[0024]

[0025] Where f(x) is the probability density function and x is the sample value;

[0026] Generate Gaussian white noise identical to the measured signal:

[0027] N t =μ r +σ r ·Z t

[0028] Among them, N t is the Gaussian noise with the same distribution as the measured signal, Z t is a random variable that follows a standard normal distribution N(0,1).

[0029] Based on the mean μ r With standard deviation σ r , generate Gaussian noise, simple operation.

[0030] Further, step S3 simplifies the bearing into a mass-spring-damper system, calculates the contact stiffness between the ball and the inner and outer ring raceways, and the relationship between the contact deformation and the load, and the specific steps are as follows:

[0031] According to Hertz contact theory, the ball and the bearing raceway are considered as smooth elastic bodies, with only elastic contact deformation and obeying Hooke's theorem. The contact stiffness K′ between the ball and the inner and outer raceways is calculated as follows:

[0032]

[0033] Where k, Γ and ∑ are the ellipse parameters, the first-kind total ellipse integral, and the second-kind total ellipse integral respectively; ∑ρ is the sum of the curvatures of the contact pair, which is obtained by the geometric parameters of the bearing; E * is the equivalent elastic modulus;

[0034] Calculate the total contact stiffness K between the ball and the inner and outer ring raceways:

[0035]

[0036] Where η is the load-deformation index; for ball bearings, η = 1.5; for cylindrical roller bearings, η = 10 / 9; Ki is the inner ring contact stiffness, K o is the inner ring contact stiffness;

[0037] According to Hertz contact deformation theory, the relationship between contact deformation and load is:

[0038] F=Kδ η +Cv

[0039] Among them, F is the Hertz contact force; δ is the radial contact deformation of the contact pair; v is the radial contact velocity of the contact pair; C is the damping coefficient between the inner and outer raceways.

[0040] The bearing is simplified into a mass-spring-damper system, and according to the Hertz contact theory, the ball and the bearing raceway are considered as smooth elastic bodies, with only elastic contact deformation and obeying Hooke's theorem, which is convenient for calculating relevant data.

[0041] Furthermore, the fault of the ball bearing is simplified into a rectangle that runs through the entire raceway width. Considering the fault displacement excitation, the four-degree-of-freedom dynamic equilibrium equation of the faulty ball bearing is constructed as follows:

[0042]

[0043] Among them, m i is the total mass of the inner ring and the supporting shaft; m h K is the total mass of the outer ring and the bearing seat; s and K h are the stiffness of the supporting shaft and the bearing seat respectively; C s , C h are the damping coefficients of the support shaft and the bearing seat respectively; H is the rolling element displacement excitation; x i ,y i and x o ,y o are the vibration displacements of the inner and outer rings in the X and Y directions, respectively; and are the vibration speeds of the inner and outer rings in the X and Y directions, respectively; and are the vibration accelerations of the inner ring and outer ring in the X and Y directions respectively; w y is the external load applied to the inner ring of the bearing in the Y direction; Z is the number of bearing rollers, x is the displacement of the rolling bearing in the X direction, y is the displacement of the rolling bearing in the Y direction, γ is the bearing tolerance, θ j is the position of the jth roller; v j is the contact velocity of the jth rolling element at any angular position;

[0044] λ j The parameters of the jth rolling element in the load zone are:

[0045]

[0046] δ j is the total contact deformation of the jth rolling element at any angular position:

[0047] δ j =(x i -x o )sinθ j +(y i -y o )cosθ j -c

[0048] Among them, c is the radial clearance of the bearing.

[0049] Based on the concentrated spring-mass-damper model of rolling bearing, the fault of ball bearing is simplified as a rectangle running through the entire raceway width. Therefore, the four-degree-of-freedom dynamic equilibrium equation of the faulty ball bearing can be constructed by considering the fault displacement excitation.

[0050] Furthermore, in step S5, the Gaussian white noise is transferred to the dynamic response signal of the bearing dynamics to generate the bearing fusion data S with the same distribution as the measured signal. t ,for:

[0051] S t =S s +λ·N t

[0052] Among them, S s is the dynamic response signal of the bearing dynamics, N t is a noise signal containing the distribution characteristics of the measured signal; λ is a noise addition control parameter, which is used to control the amount of noise added to achieve the addition of noise with a specified signal-to-noise ratio. Based on the signal-to-noise ratio formula, the noise addition control parameter is calculated:

[0053]

[0054] Among them, E s is the energy of the simulated signal; i is the sample number; A is the amount of noise to be added; Refers to 10 to the power of A / 10.

[0055] The generated Gaussian white noise information with the same distribution as the measured signal is migrated to the dynamic response signal of the bearing dynamics to complete the information migration process.

[0056] Further, step S6 constructs the objective function based on the bearing fusion data and the measured signal:

[0057] min(R),R=|RMS(S t )-RMS(S)|

[0058] Among them, R is the bearing fusion data S t The absolute value of the RMS deviation between the measured signal S and the actual signal S; RMS() is a function that calculates the root mean square value RMS of the signal:

[0059]

[0060] Among them, n is the total number of samples, i is the sample number, and x is i is the vibration displacement of the inner ring in the X direction.

[0061] Minimize the bearing fusion data S t The optimal solution of the statistical characteristic value (RMS) between the measured signal S is used to realize automatic update of the noise addition control parameters.

[0062] The present invention also provides a bearing data generation system based on physical model information migration, comprising a data acquisition unit and a processing unit, wherein the data acquisition unit is used to acquire a vibration acceleration signal of the bearing, and an output end of the data acquisition unit is connected to an input end of the processing unit;

[0063] The processing unit executes the method of the present invention to generate bearing data.

[0064] The system has a simple structure and can quickly generate corresponding data of normal and faulty bearings as needed through a small amount of measured signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a comparison diagram of information migration signals of the bearing data generation method based on physical model information migration of the present invention;

[0066] Figure 2 It is a comparison diagram of the probability density distribution difference of the information migration signal of the bearing data generation method based on the physical model information migration of the present invention;

[0067] Figure 3 It is a flow chart of the bearing data generation method based on physical model information migration of the present invention. DETAILED DESCRIPTION

[0068] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0069] In the description of the present invention, it is necessary to understand that the terms "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention.

[0070] In the description of the present invention, unless otherwise specified and limited, it should be noted that the terms "installed", "connected" and "connected" should be understood in a broad sense. For example, it can be a mechanical connection or an electrical connection, or it can be the internal connection between two components. It can be a direct connection or an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to the specific circumstances.

[0071] The present invention discloses a bearing data generation method based on physical model information migration, which can quickly generate normal and fault corresponding data of bearings according to needs in order to solve the problem of difficulty in obtaining bearing data in actual engineering applications. Figure 3 As shown, the bearing data generation method based on physical model information migration includes the following steps:

[0072] S1, collect the vibration acceleration signal of the bearing and extract the mean and standard deviation of its distribution characteristic Gaussian distribution;

[0073] S2, based on the mean and standard deviation of the Gaussian distribution of the distribution characteristics of the vibration acceleration signal, generates Gaussian white noise that is the same as the measured bearing vibration acceleration signal, overcoming the problem of large distribution differences between the bearing dynamic response signal and the measured signal;

[0074] S3, simplify the bearing into a mass-spring-damper system, calculate the contact stiffness between the ball bearing and the inner and outer ring raceways, as well as the relationship between the contact deformation and the load, and construct the four-degree-of-freedom dynamic equilibrium equation of the faulty ball bearing;

[0075] S4, using the Runge-Kutta method, solve the four-degree-of-freedom dynamic equilibrium equation of the faulty ball bearing and obtain the dynamic response signal of the bearing;

[0076] S5, transferring Gaussian white noise to the dynamic response signal of the bearing dynamics to generate bearing fusion data with the same distribution as the measured signal;

[0077] S6, constructing the objective function based on the bearing fusion data and measured signals;

[0078] S7, optimize and solve the objective function to determine whether the target convergence effect is achieved. If it is achieved, the noise addition control parameter λ obtained by the solution is input into the migration process (migration formula) in step S5, and the bearing fusion data is output to realize the automatic regulation of the noise addition amount; otherwise, the noise addition control parameter is automatically selected to step S6. The optimization algorithm (such as the PSO algorithm) will automatically select the noise addition control parameter to be solved for the objective function until the target convergence effect is achieved; this is equivalent to a loop, continuously selecting the noise addition control parameter until the target effect is achieved, jumping out of the loop, and solving the noise addition control parameter.

[0079] In a preferred embodiment of the present invention, step S1 specifically comprises:

[0080] Extract the distribution characteristics of the bearing vibration acceleration signal Gaussian distribution mean (expected value) μ r and its standard deviation σ r :

[0081]

[0082] Among them, S is the measured sample data; n is the total number of samples, S i Represents the measured sample data of the i-th sample.

[0083] In a preferred embodiment of the present invention, step S2 is specifically:

[0084] Through the mean μ r With standard deviation σ r , the noise added to the simulation signal can be constrained. The added Gaussian noise must satisfy the following distribution:

[0085]

[0086] Where f(x) is the probability density function and x is the sample value;

[0087] Generate Gaussian white noise identical to the measured signal:

[0088] N t =μ r +σ r ·Z t

[0089] Among them, N t is the Gaussian noise with the same distribution as the measured signal, Z t is a random variable that follows a standard normal distribution N(0,1).

[0090] In a preferred embodiment of the present invention, a bearing dynamics model is established, and the bearing dynamic response signal is obtained by calculation. This is a simulation process. The dynamics model has the characteristics of universality, and some parameters in the model can be modified to obtain dynamic response signals of different fault types of different bearings, which significantly reduces the cost and time of the signal.

[0091] However, the dynamic response signal is too idealized and cannot simulate the noise in the actual operating environment, resulting in a large distribution difference between the simulated signal and the measured signal, and the dynamic response signal needs to be further processed.

[0092] Step S3 simplifies the bearing into a mass-spring-damper system, calculates the contact stiffness between the ball and the inner and outer ring raceways, and the relationship between the contact deformation and the load. The specific steps are as follows:

[0093] According to Hertz contact theory, the ball and the bearing raceway are considered as smooth elastic bodies, with only elastic contact deformation and obeying Hooke's theorem. The contact stiffness K′ between the ball and the inner and outer raceways is calculated as follows:

[0094]

[0095] Where k, Γ and ∑ are the ellipse parameters, the first kind of total ellipse integral, and the second kind of total ellipse integral respectively; ∑ρ is the sum of the curvatures of the contact pair, which can be obtained through the geometric parameters of the bearing; E * is the equivalent elastic modulus;

[0096] Calculate the total contact stiffness K between the ball and the inner and outer ring raceways:

[0097]

[0098] Where η is the load-deformation index; for ball bearings, η = 1.5; for cylindrical roller bearings, η = 10 / 9; K i is the inner ring contact stiffness, K o is the inner ring contact stiffness;

[0099] According to Hertz contact deformation theory, the relationship between contact deformation and load is:

[0100] F=Kδ η +Cv

[0101] Among them, F is the Hertz contact force; δ is the radial contact deformation of the contact pair; v is the radial contact velocity of the contact pair; C is the damping coefficient between the inner and outer raceways.

[0102] In a preferred embodiment of the present invention, based on the concentrated spring-mass-damping model of the rolling bearing, the fault of the ball bearing is simplified to a rectangle that runs through the entire raceway width. Taking into account the fault displacement excitation, the four-degree-of-freedom dynamic equilibrium equation of the faulty ball bearing is constructed as follows:

[0103]

[0104] Among them, m i is the total mass of the inner ring and the supporting shaft; m h K is the total mass of the outer ring and the bearing seat; s and K h are the stiffness of the supporting shaft and the bearing seat respectively; C s , C h are the damping coefficients of the support shaft and the bearing seat respectively; H is the rolling element displacement excitation; x i ,y i and x o ,y o are the vibration displacements of the inner and outer rings in the X and Y directions, respectively; and are the vibration speeds of the inner and outer rings in the X and Y directions, respectively; and are the vibration accelerations of the inner ring and outer ring in the X and Y directions respectively; w y is the external load applied to the inner ring of the bearing in the Y direction; Z is the number of bearing rollers, x is the displacement of the rolling bearing in the X direction, y is the displacement of the rolling bearing in the Y direction, γ is the bearing tolerance, θ j is the position of the jth roller; v j is the contact velocity of the jth rolling element at any angular position;

[0105] λ j The parameters of the jth rolling element in the load zone are:

[0106]

[0107] δ j is the total contact deformation of the jth rolling element at any angular position:

[0108] δ j =(x i -x o )sinθ j +(y i -y o )cosθ j -c

[0109] Among them, c is the radial clearance of the bearing.

[0110] In a preferred embodiment of the present invention, the established dynamic equation is solved using the Runge-Kutta method to obtain the bearing dynamic response signal S s ,like Figure 1 and Figure 2 As shown, in step S5, the generated Gaussian white noise with the same distribution as the measured signal is migrated to the dynamic response signal of the bearing dynamics, completing the information migration process and generating the bearing fusion data S with the same distribution as the measured signal. t ,for:

[0111] S t =S s +λ·N t

[0112] Among them, S s is the dynamic response signal of the bearing dynamics, N t is a noise signal containing the distribution characteristics of the measured signal; λ is a noise addition control parameter, which is used to control the amount of noise added to achieve the addition of noise with a specified signal-to-noise ratio. Based on the signal-to-noise ratio formula, the noise addition control parameter is calculated:

[0113]

[0114] Among them, E s is the energy of the simulated signal; i is the sample number; A is the amount of noise to be added; Refers to 10 to the power of A / 10.

[0115] In a preferred embodiment of the present invention, the present invention proposes a method for automatically updating the noise addition control parameters, which can be regarded as minimizing the bearing fusion data S t The optimal solution of the statistical characteristic value (RMS) between the measured signal S. Step S6 constructs the objective function based on the bearing fusion data and the measured signal:

[0116] min(R),R=|RMS(S t )-RMS(S)|

[0117] Among them, R is the bearing fusion data S t The absolute value of the RMS deviation between the measured signal S and the actual signal S; RMS() is a function that calculates the root mean square value RMS of the signal:

[0118]

[0119] Among them, n is the total number of samples, i is the sample number, and x is i is the vibration displacement of the inner ring in the X direction.

[0120] This patent uses the particle swarm optimization algorithm (PSO) as an optimization algorithm for automatically updating noise addition control parameters. Compared with other optimization algorithms, the PSO algorithm has the advantages of simple implementation, high accuracy, and fast convergence speed.

[0121] Through the particle swarm optimization algorithm, the noise can be added by adjusting the control parameters so that the bearing simulation signal after information migration not only has a smaller distribution gap with the measured signal, but also has a closer statistical characteristic value, so that more reliable bearing data can be automatically obtained.

[0122] The present invention also provides a bearing data generation system based on physical model information migration, including a data acquisition unit and a processing unit. The data acquisition unit is used to acquire the vibration acceleration signal of the bearing, and the output end of the data acquisition unit is electrically connected to the input end of the processing unit.

[0123] The processing unit executes the method of the present invention to generate bearing data. The system has a simple structure and can quickly generate normal and fault corresponding data of the bearing according to needs through a small amount of measured signals.

[0124] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0125] Although the embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

Claims

1. A bearing data generation method based on physical model information migration, characterized in that: The steps include: S1, collect the vibration acceleration signal of the bearing and extract the mean and standard deviation of its distribution characteristic Gaussian distribution; S2, based on the mean and standard deviation of the Gaussian distribution of the distribution characteristics of the vibration acceleration signal, generates Gaussian white noise that is the same as the measured bearing vibration acceleration signal; S3, simplify the bearing into a mass-spring-damper system, calculate the contact stiffness between the ball bearing and the inner and outer ring raceways, as well as the relationship between the contact deformation and the load, and construct the four-degree-of-freedom dynamic equilibrium equation of the faulty ball bearing; S4, using the Runge-Kutta method, solve the four-degree-of-freedom dynamic equilibrium equation of the faulty ball bearing and obtain the dynamic response signal of the bearing; S5, transferring Gaussian white noise to the dynamic response signal of the bearing dynamics to generate bearing fusion data with the same distribution as the measured signal; S6, constructing the objective function based on the bearing fusion data and measured signals; S7, optimize and solve the objective function to determine whether the target convergence effect is achieved. If so, the noise addition control parameters obtained by the solution are input into the migration process in step S5, and the bearing fusion data is output; otherwise, the noise addition control parameters are automatically selected to step S6.

2. The method for generating bearing data based on physical model information migration according to claim 1, characterized in that: Step S1 is specifically as follows: Extract the distribution characteristics of the bearing vibration acceleration signal and the mean value μ of the Gaussian distribution r and its standard deviation σ r : Among them, S is the measured sample data; n is the total number of samples, S i Represents the measured sample data of the i-th sample.

3. The method for generating bearing data based on physical model information migration according to claim 2, characterized in that: Step S2 is specifically as follows: By the mean μ r With standard deviation σ r , constraining the noise added to the simulation signal, the added Gaussian noise must satisfy the distribution: Where f(x) is the probability density function and x is the sample value; Generate Gaussian white noise identical to the measured signal: N t =μ r +s r ·Z t Among them, N t is the Gaussian noise with the same distribution as the measured signal, Z t is a random variable that follows a standard normal distribution N(0,1).

4. The method for generating bearing data based on physical model information migration according to claim 1, characterized in that: Step S3 simplifies the bearing into a mass-spring-damper system, calculates the contact stiffness between the ball and the inner and outer ring raceways, and the relationship between the contact deformation and the load. The specific steps are as follows: According to Hertz contact theory, the ball and the bearing raceway are considered as smooth elastic bodies, with only elastic contact deformation and obeying Hooke's theorem. The contact stiffness K′ between the ball and the inner and outer raceways is calculated as follows: Where, k, Γ and Σ are the ellipse parameters, the first kind of total ellipse integral, and the second kind of total ellipse integral respectively; 'ρ is the sum of the curvatures of the contact pair, which is obtained by the geometric parameters of the bearing; E * is the equivalent elastic modulus; Calculate the total contact stiffness K between the ball and the inner and outer ring raceways: Where η is the load-deformation index; for ball bearings, η = 1.5; for cylindrical roller bearings, η = 10 / 9; K i is the inner ring contact stiffness, K o is the inner ring contact stiffness; According to Hertz contact deformation theory, the relationship between contact deformation and load is: F=Kδ η +Cv Among them, F is the Hertz contact force; δ is the radial contact deformation of the contact pair; v is the radial contact velocity of the contact pair; C is the damping coefficient between the inner and outer raceways.

5. The method for generating bearing data based on physical model information migration according to claim 4, characterized in that: The fault of the ball bearing is simplified as a rectangle that runs through the entire raceway width. Considering the fault displacement excitation, the four-degree-of-freedom dynamic equilibrium equation of the faulty ball bearing is constructed as follows: Among them, m i is the total mass of the inner ring and the supporting shaft; m h K is the total mass of the outer ring and the bearing seat; s and K h are the stiffness of the supporting shaft and the bearing seat respectively; C s , C h are the damping coefficients of the support shaft and the bearing seat respectively; H is the rolling element displacement excitation; x i ,y i and x o ,y o are the vibration displacements of the inner and outer rings in the X and Y directions, respectively; and are the vibration speeds of the inner and outer rings in the X and Y directions, respectively; and are the vibration accelerations of the inner ring and outer ring in the X and Y directions respectively; w y is the external load applied to the inner ring of the bearing in the Y direction; Z is the number of bearing rollers, x is the displacement of the rolling bearing in the X direction, y is the displacement of the rolling bearing in the Y direction, γ is the bearing tolerance, θ j is the position of the jth roller; v j is the contact velocity of the jth rolling element at any angular position; λ j The parameters of the jth rolling element in the load zone are: δ j is the total contact deformation of the jth rolling element at any angular position: d j =(x i -x o )sinθ j +(y i -y o )cosθ j -c Among them, c is the radial clearance of the bearing.

6. The method for generating bearing data based on physical model information migration according to claim 5, characterized in that: In step S5, Gaussian white noise is transferred to the dynamic response signal of the bearing dynamics to generate bearing fusion data S with the same distribution as the measured signal. t ,for: S t =S s +λ·N t Among them, S s is the dynamic response signal of the bearing dynamics, N t is a noise signal containing the distribution characteristics of the measured signal; λ is a noise addition control parameter, which is used to control the amount of noise added to achieve the addition of noise with a specified signal-to-noise ratio. Based on the signal-to-noise ratio formula, the noise addition control parameter is calculated: Among them, E s is the energy of the simulated signal; i is the sample number; A is the amount of noise to be added; Refers to 10 to the power of A / 10.

7. The method for generating bearing data based on physical model information migration according to claim 1, characterized in that: Step S6 constructs the objective function based on the bearing fusion data and the measured signal: min(R),R=|RMS(S t )-RMS(S)| Among them, R is the bearing fusion data S t The absolute value of the RMS deviation between the measured signal S and the actual signal S; RMS() is a function that calculates the root mean square value RMS of the signal: Among them, n is the total number of samples, i is the sample number, and x is i is the vibration displacement of the inner ring in the X direction.

8. A bearing data generation system based on physical model information migration, characterized in that: It includes a data acquisition unit and a processing unit, wherein the data acquisition unit is used to acquire the vibration acceleration signal of the bearing, and the output end of the data acquisition unit is connected to the input end of the processing unit; The processing unit executes the method according to any one of claims 1 to 7 to generate bearing data.

Images