Search

CN-121980954-A - Digital twin tooth surface wear degradation morphology prediction method

CN121980954ACN 121980954 ACN121980954 ACN 121980954ACN-121980954-A

Abstract

The invention provides a digital twin tooth surface wear degradation morphology prediction method which is applied to the field of gear transmission systems and comprises the steps of constructing a gear transient wear model constrained physical information neural network wear depth prediction model to obtain a model for predicting a new tooth profile after gear wear and a rough morphology model for predicting tooth surface evolution after wear; the method comprises the steps of constructing a gear wear time-varying meshing stiffness calculation model to obtain corrected wear gear time-varying meshing stiffness, constructing a gear system dynamics model to obtain a gear wear degradation transmission system simulation dynamics response by combining a lumped parameter method, correcting tooth surface dynamic meshing force by calibrating gear bearing support stiffness and damping, and finally obtaining corrected tooth surface wear depth and corrected tooth profile and tooth surface morphology.

Inventors

  • XIAO HUIFANG
  • LI ZEDONG
  • DAI GUANGHAO
  • WANG MENGQI
  • LI HUI

Assignees

  • 北京科技大学
  • 中国船舶集团有限公司第七〇三研究所

Dates

Publication Date
20260505
Application Date
20260210

Claims (9)

  1. 1. A method for predicting digital twin tooth surface wear degradation morphology, the method comprising: S1, constructing a gear transient wear model constraint physical information neural network wear depth prediction model based on basic parameters of a gear transmission system, and obtaining a model for predicting a new tooth profile after gear wear and a rough morphology model for predicting tooth surface evolution after wear; S2, constructing a gear wear time-varying engagement stiffness calculation model based on a model for predicting a new tooth profile of the worn gear and a rough morphology model for predicting tooth surface evolution after wear, and obtaining corrected wear gear time-varying engagement stiffness; s3, constructing a gear system dynamics model based on the corrected time-varying meshing stiffness of the worn gear and a lumped parameter method, and obtaining a gear wear degradation transmission system simulation dynamics response; s4, obtaining actual measurement signals, simulating dynamic response and calibrating gear bearing supporting rigidity k and damping c by a multi-objective optimization algorithm based on the sensor; S5, based on the calibrated supporting rigidity k and damping c of the gear bearing, obtaining a digital twin model of the high-fidelity gear system, and dynamically engaging force on the tooth surface Correcting based on the tooth surface dynamic meshing force after correction And obtaining the corrected tooth surface abrasion depth and the corrected tooth profile and tooth surface morphology.
  2. 2. The method for predicting the wear degradation profile of a digital twin tooth surface according to claim 1, wherein the basic parameters in S1 include a structural parameter, a working condition parameter and a profile parameter of the tooth surface.
  3. 3. The method for predicting the degradation morphology of the digital twin tooth surface wear according to claim 2, wherein the constructing a physical information neural network wear depth prediction model constrained by the gear transient wear model based on the basic parameters of the gear transmission system in S1, to obtain a model for predicting a new tooth profile after gear wear and a rough morphology model for predicting tooth surface evolution after wear comprises: s101, constructing a tooth surface three-dimensional transient Archard abrasion model as a formula (1) based on Archard abrasion model, tooth surface morphology three-dimensional characteristics and tooth surface inter-engagement dynamic meshing force; ;(1) predicted wear depth for t N cycles of wear; t N is the current engagement period; Is the wear coefficient under t N cycles of wear; Tooth surface contact pressure at t N cycles of wear; Tooth surface meshing point relative speed at t N cycles of wear; ζ is a grid divided along the tooth profile direction; η is a grid divided along the tooth width direction; s102, constructing a physical information neural network under the constraint of a wear mechanism based on a tooth surface three-dimensional transient Archard wear model, and constructing a physical information neural network loss function under the constraint of the wear mechanism; and S103, obtaining a model for predicting a new tooth profile of the gear after abrasion and a rough morphology model for predicting tooth surface evolution after abrasion based on the physical information neural network and the corresponding loss function under the constraint of the abrasion mechanism.
  4. 4. The method for predicting the wear-degradation profile of a digital twin tooth surface according to claim 3, The model for predicting the new tooth profile after gear wear in S103 is formula (2): ;(2) A new tooth profile containing a morphology under t N period wear, The tooth profile containing the morphology in the period of t N-1 , For the predicted wearing depth under the wearing of the period t N , ζ is the grid divided along the tooth profile direction, η is the grid divided along the tooth width direction, t N is the current meshing period, t N-1 is the last meshing period of the current meshing period; the rough morphology model for predicting the tooth surface evolution after abrasion in S103 is formula (3): ;(3) is a rough appearance of tooth surface evolution after t N period abrasion, Polyfit () is a polynomial fitting function.
  5. 5. The method for predicting the wear-degradation profile of a digital twin tooth surface according to claim 4, wherein the constructing a gear wear-time-varying engagement stiffness calculation model based on the model for predicting the new tooth profile after gear wear and the rough profile model for predicting the evolution of the tooth surface after wear in S2, and obtaining the corrected wear gear time-varying engagement stiffness comprises: S201, defining a gear wear time-varying meshing stiffness calculation formula: when the gear wear-time-varying engagement stiffness is the single-tooth zone engagement stiffness, the gear wear-time-varying engagement stiffness calculation formula of the single-tooth zone is formula (4): ;(4) Gear wear time-varying meshing stiffness for a single tooth zone; consider the contact stiffness of the rough appearance to the gear engagement interface for the single tooth area; bending stiffness of the pinion gear being a single tooth zone; The shear stiffness of the pinion gear being a single tooth zone; Axial compression stiffness of the pinion gear for the single tooth region; Bending stiffness of the large gear which is a single tooth area; The shear stiffness of the large gear which is a single tooth area; axial compression stiffness of the large gear in the single tooth area; the basic deformation stiffness of the pinion gear being a single tooth zone; the basic deformation rigidity of the large gear in the single tooth area; when the gear wear-time-varying engagement stiffness is the double-tooth zone, the gear wear-time-varying engagement stiffness calculation formula of the double-tooth zone is formula (5): ;(5) Gear wear time-varying meshing stiffness for the double tooth zone; The subscript i=1, 2, when i=1, is the 1 st pair of gears in the double-toothed region, and when i=2, is the 2 nd pair of gears in the double-toothed region; Consider the contact stiffness of the coarse morphology on the gear engagement interface for the ith pair of gears in the double-tooth area; bending stiffness of the ith pair of pinions being double-tooth zones; The shear stiffness of the ith pair of pinions, which is the double-tooth zone; Axial compression stiffness of the ith pair of pinions in the double-tooth zone; Bending rigidity of the ith pair of large gears in the double-tooth zone; the shear stiffness of the ith pair of large gears in the double-tooth zone; Axial compression rigidity of the ith pair of large gears in the double-tooth zone; A basic deformation stiffness of the ith pair of pinion gears being a double tooth zone; the basic deformation rigidity of the ith pair of large gears in the double-tooth zone; S202, constructing a gear wear lower tooth profile error model based on gear wear time-varying meshing stiffness: When the tooth profile of the double-tooth zone is changed into the meshing stiffness, the tooth profile of the double-tooth zone is error Formula (6): ;(6) a depth of wear for the pinion gear in the first pair of gears; The depth of wear for the large gear in the first pair of gears; The tooth surface roughness of the pinion gear in the first pair of gears is obtained; the tooth surface roughness of the large gear in the first pair of gears is obtained; the depth of wear of the pinion gear in the second pair of gears; the depth of wear for the large gear in the second pair of gears; The gear surface roughness of the pinion gear in the second pair of gears is obtained; the tooth surface roughness of the large gear in the second pair of gears is obtained; When the single-tooth zone is changed in meshing rigidity, only the first pair of gears exist in the single-tooth zone, and the tooth profile error of the single-tooth zone Is formula (7): ;(7) s203, correcting the time-varying meshing stiffness of the gear based on the gear wear lower tooth profile error model: modified gear time-varying meshing stiffness of double-tooth zone Formula (8): ;(8) the time-varying meshing stiffness of the first pair of gears in the double-tooth zone; The time-varying meshing stiffness of the second pair of gears in the double-tooth area is obtained; F is tooth surface contact static meshing force; modified gear time-varying meshing stiffness of single tooth zone Formula (9): ;(9) Equation (10) is derived based on equation (8) and equation (9): ;(10) The gear becomes engaged stiffness for the corrected worn gear.
  6. 6. The method for predicting the wear-degradation morphology of the digital twin tooth surface according to claim 5, wherein the step of constructing a gear system dynamics model based on the corrected wear gear time-varying meshing stiffness combined with a lumped parameter method in the step S3, and obtaining a gear wear-degradation transmission system simulation dynamics response comprises: S301, time-varying meshing stiffness based on corrected worn gear Calculating dynamic mesh forces in a gear drive system , The expression of (2) is formula (11): ;(11) Wherein, the In order to dynamically transfer the errors in the error, For the first derivative of the dynamic transfer error, In order to engage the damping, As a nonlinear tooth flank clearance function; S302, based on dynamic meshing force Adopting a lumped parameter method to establish a dynamic model of the gear transmission system to obtain a simulated dynamic response of the gear wear degradation transmission system 。
  7. 7. The digital twin tooth wear degradation morphology prediction method according to claim 6, wherein the expression of the dynamic transmission error in S301 is formula (12): ;(12) Wherein, the Is the base radius of the pinion gear, Is the base circle radius of the large gear, For the angular displacement of the pinion gear, For the angular displacement of the gearwheel, For the oscillating displacement of the pinion in the x-direction, For the oscillating displacement of the gearwheel in the x-direction, For the oscillating displacement of the pinion in the y direction, For the oscillating displacement of the gearwheel in the y direction, To account for static transmission errors of wear.
  8. 8. The method of predicting digital twin tooth surface wear degradation morphology of claim 7, wherein the sensor-based acquisition of measured signals, simulation of dynamic response, and calibration of gear bearing support stiffness k and damping c in S4 by a multi-objective optimization algorithm comprises: S401 obtaining actual measurement signals based on sensors And measure the signal And S302 Evaluation is performed by pearson correlation coefficient PCC, wherein the evaluation formula of pearson correlation coefficient PCC is formula (13): ;(13) Length of the kinetic response sample; the j-th position simulation signal The value of the one of the values, = 1, 2, 3, ..., N sample ; Is the j-th position measured signal The value of the one of the values, = 1, 2, 3, ..., N sample ; The average value of the simulation signal at the j-th position; The average value of the j-th position actual measurement signal; S402, calibrating bearing support rigidity k and damping c by using a multi-objective gray wolf optimization algorithm based on an evaluation formula of a pearson correlation coefficient PCC, and defining dynamics model optimization parameters as follows Setting the optimization objective function as formula (14): ;(14) the function is optimized for the time domain similarity objective, The function is optimized for the frequency domain similarity objective, and the FFT is a fourier transform.
  9. 9. The method for predicting the wear degradation morphology of the digital twin tooth surface according to claim 8, wherein the step S5 is based on the calibrated gear bearing support rigidity k and the damping c to obtain a digital twin model of the high-fidelity gear system, and the dynamic meshing force of the tooth surface is obtained Correcting based on the tooth surface dynamic meshing force after correction Obtaining the corrected tooth surface wear depth and the corrected tooth profile and tooth surface morphology includes: Based on the bearing support rigidity k and the damping c, obtaining corrected ; Correcting the tooth surface contact pressure in the tooth surface transient wear model based on the formula (15): ;(15) After substituting the corrected tooth surface contact pressure into step S101 for the corrected tooth surface contact pressure, the corrected tooth surface wear depth and the corrected tooth profile, a H , are obtained based on steps S101 to S103, and a H is the hertz contact half width.

Description

Digital twin tooth surface wear degradation morphology prediction method Technical Field The invention relates to the field of gear transmission systems, in particular to a digital twin tooth surface wear degradation morphology prediction method. Background Gear systems, as an integral mechanical transmission component in high-end equipment, often operate in harsh and complex environments. With long running, the tooth surface inevitably wears, which in turn causes changes in the tooth surface topography and the dynamic response of the system. Monitoring methods based on vibration signals are widely used to monitor dynamic response changes of gear systems caused by gear wear, however, wear damage characteristics of gear contact interfaces are not available. For high-end equipment, it is difficult and impractical to extract the tooth surface morphology features of the gear contact interface periodically, so that it is of great significance to construct the mapping relation between the tooth surface morphology evolution characteristic under the gear wear degradation and the dynamic response of the gear system. Aiming at a gear wear modeling method, the prior art takes Chinese patent No. 120541989A as an example, discloses a quasi-static straight gear wear numerical analysis method, a dynamic wear coefficient is calculated by utilizing a mixed lubrication model, a gear wear model is established by means of an Arcard formula, and a quasi-static wear value is calculated. Taking Chinese patent No. 120892912A as an example, a vibration signal-based working condition following gear wear monitoring and predicting method is disclosed, an Archard wear correction model based on a thermal elastic flow lubrication theory is established, and gear engagement interface oil film pressure is used for replacing tooth surface contact pressure. However, the techniques disclosed in these patent documents rely on the calculation of static contact pressure and wear coefficient in the Archard wear model for the gear wear depth, ignoring the influence of dynamic meshing force along with the meshing period and dynamic change of the wear coefficient on the profile error and the profile evolution of the tooth surface in the dynamic evolution of the gear wear. In order to accurately obtain parameters required by tooth surface wear depth calculation, a high-fidelity gear wear degradation model is constructed by utilizing a digital twin technology. The digital twin technology is to learn the physical rule by using an intelligent algorithm through mutual interaction between the physical body and the virtual body, so as to generate analog data close to real data. Taking Chinese patent No. 120609560A as an example, a digital twin unbalanced fault diagnosis method and system of a gear transmission system are disclosed, a gear high-fidelity dynamics model is constructed as a gear virtual body, and different fault twin signals of gears are generated by combining internal excitation of different fault time-varying meshing rigidities of the gears. Taking the patent CN119026266A as an example, the invention discloses a method for constructing a digital twin system of a fatigue test gear transmission system, which uses a rigid-flexible coupling gear system dynamics model to obtain a simulated dynamics model, combines CycleGAN with a simulation signal and an actual measurement signal to generate anti-learning generation gear system twin data, and provides a sample for gear fault diagnosis. However, the technical methods disclosed in these patent documents focus on twin data generation of different fault types of gears, and modeling techniques of virtual body tooth surface damage degradation and real-time interaction of physical body vibration signals in gear wear degradation are not yet available. In summary, the prior art ignores the influence of the dynamic meshing force variation of the tooth surface on the tooth profile error and the tooth surface morphology evolution in the gear wear degradation, and lacks the bridge for establishing the system vibration response and the tooth surface morphology evolution under the gear wear degradation, so that a gear transmission system digital twin tooth surface wear degradation morphology prediction method is urgently needed. Disclosure of Invention In order to solve the above problems, the present invention provides a method for predicting a digital twin tooth surface wear degradation profile, which improves accuracy of tooth surface wear depth prediction, and realizes virtual detection of tooth surface profile in gear wear degradation in a gear transmission system, thereby evaluating a damaged state of a tooth surface, and specifically includes: A digital twin tooth surface wear degradation profile prediction method, the method comprising: S1, constructing a gear transient wear model constraint physical information neural network wear depth prediction model based on basic parameters of a gear transmission s