CN117592332B - Digital twinning-based gearbox model high-fidelity method, system and storage medium - Google Patents
Digital twinning-based gearbox model high-fidelity method, system and storage medium Download PDFInfo
- Publication number
- CN117592332B CN117592332B CN202311563535.5A CN202311563535A CN117592332B CN 117592332 B CN117592332 B CN 117592332B CN 202311563535 A CN202311563535 A CN 202311563535A CN 117592332 B CN117592332 B CN 117592332B
- Authority
- CN
- China
- Prior art keywords
- fidelity
- model
- vibration
- gearbox
- algorithm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 67
- 238000012360 testing method Methods 0.000 claims abstract description 37
- 230000008878 coupling Effects 0.000 claims abstract description 35
- 238000010168 coupling process Methods 0.000 claims abstract description 35
- 238000005859 coupling reaction Methods 0.000 claims abstract description 35
- 238000011156 evaluation Methods 0.000 claims abstract description 26
- 238000005457 optimization Methods 0.000 claims abstract description 26
- 238000004364 calculation method Methods 0.000 claims abstract description 21
- 241000282461 Canis lupus Species 0.000 claims abstract description 20
- 230000004927 fusion Effects 0.000 claims abstract description 18
- 239000002245 particle Substances 0.000 claims abstract description 16
- 238000004458 analytical method Methods 0.000 claims abstract description 11
- 230000009467 reduction Effects 0.000 claims abstract description 10
- 238000005094 computer simulation Methods 0.000 claims abstract description 8
- 230000004044 response Effects 0.000 claims description 43
- 238000012544 monitoring process Methods 0.000 claims description 14
- 238000012216 screening Methods 0.000 claims description 7
- 238000013278 delphi method Methods 0.000 claims description 6
- 238000013461 design Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 238000005336 cracking Methods 0.000 claims description 4
- 230000009466 transformation Effects 0.000 claims description 4
- 239000002131 composite material Substances 0.000 claims description 3
- 238000004088 simulation Methods 0.000 abstract description 8
- 230000003044 adaptive effect Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000012937 correction Methods 0.000 description 4
- 239000006247 magnetic powder Substances 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000001133 acceleration Effects 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 238000011835 investigation Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 241000282421 Canidae Species 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 230000001681 protective effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/17—Mechanical parametric or variational design
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
The invention discloses a high-fidelity method, a system and a storage medium for a gear box model based on digital twinning, which firstly establishes a rigid-flexible coupling finite element dynamics twinning model for a gear box in a healthy state and a fault state, and (3) completing modal analysis and dynamic simulation of the gearbox, preliminarily verifying accuracy of the twin model, and arranging sensor measuring points according to the mode. And then, acquiring multi-channel vibration information by using a gearbox vibration test bed, generating a noise-reduced vibration image by combining a Vmd noise reduction method, and performing fidelity calculation on the single-channel simulation and test vibration image by adopting a fusion image recognition algorithm. Finally, parameters of the twin model are optimized based on a multi-objective optimization algorithm of the gray wolf particle swarm, high-fidelity optimization of the model is achieved, and fidelity evaluation is carried out by adopting a multi-sensor channel self-adaptive evaluation method. The multi-channel self-adaptive fidelity evaluation method provided by the scheme of the invention can help to screen multi-fault conditions and optimal fidelity solutions of multi-sensor channels, so that the model fidelity is more comprehensive and accurate.
Description
Technical Field
The invention relates to the technical field of digital twin of equipment, in particular to a high-fidelity method, a high-fidelity system and a storage medium of a gearbox model based on digital twin.
Background
Harbour equipment plays an important role in logistics loading and unloading and transportation operations, and the severe working environment of the harbour equipment often causes problems of degradation, failure and damage of the structure caused by fatigue and abrasion of key components. Aiming at the defects of high labor cost, serious hysteresis and poor diagnosis precision of a gear box 'planned maintenance' inspection mode, the novel physical driving simulation technology can utilize the high synchronization of a digital model and a physical model to realize efficient and accurate 'condition-dependent maintenance' of structural monitoring, wherein the fidelity of the digital model determines the diagnosis accuracy.
The commonly used fidelity method optimizes through single monitoring point data, and does not optimize multiple monitoring points, so that the model fidelity is poor in comprehensiveness; the traditional optimization algorithm based on the particle swarm is easy to fall into a local optimal solution, so that the accuracy of the optimized fidelity is lower.
The multi-channel self-adaptive fidelity evaluation method provided by the invention can help to screen multi-fault conditions and multi-sensor channel fidelity optimal solutions, so that the model fidelity is more comprehensive and accurate. The gray wolf particle swarm multi-objective optimization algorithm adopted by the invention can avoid the particle swarm search from sinking into a local optimal solution, increase the diversity of the particle search, and improve the accuracy of the calculation result, thereby improving the fidelity of the twin model.
Disclosure of Invention
The invention provides a digital twin-based gearbox model high-fidelity method, a digital twin-based gearbox model high-fidelity system and a digital twin-based gearbox model storage medium, and aims to solve the problems of incomplete fidelity and low accuracy disclosed in the background art.
In order to solve the technical problems, the invention adopts the following technical scheme:
A digital twinning-based gearbox model high-fidelity method, comprising:
S01, establishing a rigid-flexible coupling twin model of the gear box by using finite element simulation software, then carrying out modal analysis to obtain dynamic monitoring points, and carrying out dynamic simulation;
S02, based on a vibration test bed of the parallel fixed-axis gearbox, acquiring vibration physical data of a time domain and a frequency domain of a corresponding multi-channel of the gearbox in a normal or multiple fault states;
S03, performing single-channel fusion algorithm fidelity calculation on the vibration physical data based on a Sift algorithm and a PHash algorithm;
s04, constructing a multichannel fidelity self-adaptive comprehensive evaluation index by combining the single-channel fidelity with a preset dynamics comprehensive index;
s05, selecting parameter variables to be corrected in the rigid-flexible coupling twin model of the gear box, and constructing a response surface function of parameters to be corrected and fidelity of the rigid-flexible coupling twin model of the gear box;
s06, solving non-cracking of a response surface function of parameters to be corrected and fidelity of the rigid-flexible coupling twin model of the gear box based on a gray wolf particle swarm fusion multi-objective optimization algorithm, and screening out an optimal solution by combining a multi-channel fidelity self-adaptive comprehensive evaluation index to finish the fidelity optimization of the rigid-flexible coupling twin model of the gear box.
As a possible implementation manner, further, the scheme S01 includes: and establishing a flexible finite element twin model of the gear and the box body by utilizing finite element simulation software, replacing rigid box walls and gear components with flexible components, setting boundary conditions and model parameters, establishing a rigid-flexible coupling twin model of the gear box of the parallel fixed-axis gear box, carrying out modal analysis on the flexible box body, arranging 6 monitoring points at the maximum amplitude to carry out vibration response monitoring, and carrying out dynamic simulation on the rigid-flexible coupling twin model of the gear box to verify the accuracy of the twin model.
As a preferred implementation choice, in the present solution S02, preferably, based on a parallel fixed axis gearbox vibration test stand, vibration physical data of time domain and frequency domain corresponding to 6 channels of the gearbox in normal or multiple fault states is collected, noise reduction processing is performed through a Vmd noise reduction algorithm, and then the signal is converted into a frequency domain signal by using fast fourier transform, and a corresponding test image is generated.
As a preferred implementation choice, in the present scheme S03, preferably, a single-channel fusion algorithm fidelity calculation is performed on the test image corresponding to the vibration physical data based on the Sift algorithm and the PHash algorithm, and the obtained comprehensive fidelity result is used as a final virtual-real signal image fidelity calculation result.
As a preferred implementation option, preferably, the scheme S03 includes:
the formula of the Sift algorithm for calculating the fidelity of the test image corresponding to the vibration physical data is as follows:
ε′d＝sim1(B1,B2)×100％ (1)
Sim 1(B1,B2) represents the similarity value between the vibration frequency domain response image B 1 of the rigid-flexible coupled twin model and the vibration frequency domain response image B 2 of the physical data when the Sift algorithm is used, ε' d represents the fidelity of the two images of the channel under the Sift algorithm, d represents the channel serial number of the sensor, and d= [1,2,3,4,5,6];
The equation for calculating the fidelity of the test image corresponding to the vibration physical data by PHash algorithm is as follows:
ε″d＝sim2(B1,B2)×100％ (2)
Sim 2(B1,B2) represents the similarity value between the vibration frequency domain response image B 1 of the rigid-flexible coupling model and the vibration frequency domain response image B 2 of the physical data when the PHash algorithm is used, and epsilon d represents the fidelity of the two images of the channel under the PHash algorithm;
The calculation formula for the fidelity of the single channel fusion algorithm for vibration response is as follows:
As a preferred implementation choice, preferably, in the scheme S04, combining the single-channel fidelity and the preset dynamics comprehensive index, constructing the multi-channel fidelity adaptive comprehensive evaluation index includes:
Calculating importance coefficients for representing signal indexes under different fault states by using a Delphi method;
establishing a comprehensive index of the sensor by utilizing the time domain characteristic index and combining the index importance coefficient, and evaluating the capability of the sensor for collecting important information;
And constructing a multichannel fidelity self-adaptive comprehensive evaluation index by utilizing the fidelity and the sensor comprehensive index so as to perform self-adaptive evaluation on the multichannel fidelity, and providing a basis for the multichannel model fidelity optimization of the gearbox.
As a preferred implementation choice, preferably, in the present solution S05, selecting a parameter variable to be corrected in the rigid-flexible coupling twin model of the gearbox, and constructing a response surface function of the parameter to be corrected and the fidelity of the rigid-flexible coupling twin model of the gearbox includes:
Adopting a central composite design method to carry out response surface test design; and adopting a second-order polynomial to perform response surface function fitting, wherein the fitting formula is as follows:
wherein a 0 is a constant, a i、bi、cij is a coefficient to be determined of first-order, second-order and second-order cross terms, n is the number of parameters to be corrected, and the number of times of least test is
As a preferred implementation choice, in the present solution S06, the speed and location update formula of the sirius particle swarm fusion multi-objective optimization algorithm is preferably as follows:
Vd(k+1)＝ω·[Vd(k)+C1r1(X1d(k)-Xd(k)) +C2r2(X2d(k)-Xd(k))+C3r3(X3d(k)-Xd(k))] （5）
wherein r1, r2, r3 are random numbers in [0,1], ω is an adjustable inertial weight, its variation range is [0.5,1], and C 1、C2、C3 is calculated as follows:
C＝2·n2 (7)
Due to the variation of the above equation and the introduction of ω, the wolf group position is updated as:
Wherein P 1、P2、P3 represents the position vector of the updated alpha wolf, beta wolf and delta wolf, and P (t+1) represents the position vector determined after optimizing.
Based on the above, the invention also provides a digital twin-based gearbox model high-fidelity system, which is applied to the digital twin-based gearbox model high-fidelity method.
Based on the foregoing, the present invention further provides a computer readable storage medium, where at least one instruction, at least one program, a code set, or an instruction set is stored, where the at least one instruction, at least one program, code set, or instruction set is loaded by a processor and executed to implement a digital twin-based gearbox model high-fidelity method as described above.
By adopting the technical scheme, compared with the prior art, the invention has the beneficial effects that: according to the scheme, the multi-channel self-adaptive fidelity evaluation method is provided, so that multi-fault conditions and optimal fidelity solutions of multiple sensor channels can be screened, and the fidelity of the twin model is more comprehensive and accurate. In addition, the scheme of the invention adopts a multi-target optimization algorithm of the gray wolf particle swarm, so that the particle swarm search is prevented from being trapped into a local optimal solution, the diversity of the particle search is increased, the accuracy of a calculation result is improved, and the fidelity of the twin model is improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic implementation flow chart of a digital twinning-based gearbox model high-fidelity method of the present solution;
FIG. 2 is a flow chart of the scheme gearbox model high-fidelity optimization;
FIG. 3 is a simplified flow chart of the Sift algorithm mentioned in this scheme;
FIG. 4 is a simplified flow chart of the PHash algorithm mentioned in this scheme;
FIG. 5 is a flowchart of the Delphi method determining index importance coefficients;
FIG. 6 is a flow chart of the scheme for adaptive high fidelity assessment via multiple channels;
FIG. 7 is an overall equipment diagram of the gearbox vibration test stand according to the scheme;
FIG. 8 is a diagram of the gearbox vibration test data acquisition device according to the scheme;
FIG. 9 is an overall view of the gear box apparatus of the present solution;
FIG. 10 is a flow chart of the scheme for optimizing the fidelity response surface model by the gray wolf particle swarm fusion multi-objective optimization algorithm.
Detailed Description
The invention is described in further detail below with reference to the drawings and examples. It is specifically noted that the following examples are only for illustrating the present invention, but do not limit the scope of the present invention. Likewise, the following examples are only some, but not all, of the examples of the present invention, and all other examples, which a person of ordinary skill in the art would obtain without making any inventive effort, are within the scope of the present invention.
As shown in fig. 1 and 2, the embodiment provides a digital twin-based gearbox model high-fidelity method, which includes:
s01, establishing a rigid-flexible coupling twin model of the gear box by using finite element simulation software, then carrying out modal analysis to obtain dynamic monitoring points, and carrying out dynamic simulation; selecting a monitoring point position for sensor arrangement through modal analysis, and acquiring a multipoint vibration response by utilizing dynamic simulation;
S02, based on a vibration test bed of the parallel fixed-axis gearbox, acquiring vibration physical data of a time domain and a frequency domain of a corresponding multi-channel of the gearbox in a normal or multiple fault states;
S03, performing single-channel fusion algorithm fidelity calculation on the vibration physical data based on a Sift algorithm and a PHash algorithm;
s04, constructing a multichannel fidelity self-adaptive comprehensive evaluation index by combining the single-channel fidelity with a preset dynamics comprehensive index;
s05, selecting parameter variables to be corrected in the rigid-flexible coupling twin model of the gear box, and constructing a response surface function of parameters to be corrected and fidelity of the rigid-flexible coupling twin model of the gear box;
s06, solving non-cracking of a response surface function of parameters to be corrected and fidelity of the rigid-flexible coupling twin model of the gear box based on a gray wolf particle swarm fusion multi-objective optimization algorithm, and screening out an optimal solution by combining a multi-channel fidelity self-adaptive comprehensive evaluation index to finish the fidelity optimization of the rigid-flexible coupling twin model of the gear box.
The scheme can also optimize by constructing a twin model of the healthy and fault gearbox and comparing the two model optimization results by using a fidelity method, and can mutually verify the accuracy of model establishment and the feasibility of the fidelity optimization method.
In the scheme S01, a flexible finite element twin model of a gear and a box body is established by utilizing finite element simulation software, rigid box walls and gear components are replaced by flexible components, boundary conditions and model parameters are set, a gear box rigid-flexible coupling twin model of a parallel fixed-axis gear box is established, then modal analysis is carried out on the flexible box body, 6 monitoring points are arranged at the maximum amplitude position for vibration response monitoring, and meanwhile dynamic simulation is carried out on the gear box rigid-flexible coupling twin model so as to verify the accuracy of the twin model.
In the scheme S02, based on a vibration test bed of a parallel fixed-axis gearbox, vibration physical data of time domains and frequency domains corresponding to 6 channels of the gearbox in normal or multiple fault states are collected, noise reduction processing is performed through a Vmd noise reduction algorithm, and then fast fourier transformation is utilized to convert the vibration physical data into frequency domain signals, and corresponding test images are generated.
In the scheme S03, a single-channel fusion algorithm fidelity calculation is performed on a test image corresponding to vibration physical data based on a Sift algorithm and a PHash algorithm, and the obtained comprehensive fidelity result is used as a final virtual-real signal image fidelity calculation result.
The scheme S03 includes:
the formula of the Sift algorithm for calculating the fidelity of the test image corresponding to the vibration physical data is as follows:
ε′d＝sim1(B1,B2)×100％ (1)
Sim 1(B1,B2) represents the similarity value between the vibration frequency domain response image B 1 of the rigid-flexible coupled twin model and the vibration frequency domain response image B 2 of the physical data when the Sift algorithm is used, ε' d represents the fidelity of the two images of the channel under the Sift algorithm, d represents the channel serial number of the sensor, and d= [1,2,3,4,5,6];
The equation for calculating the fidelity of the test image corresponding to the vibration physical data by PHash algorithm is as follows:
ε″d＝sim2(B1,B2)×100％ (2)
Sim 2(B1,B2) represents the similarity value between the vibration frequency domain response image B 1 of the rigid-flexible coupling model and the vibration frequency domain response image B 2 of the physical data when the PHash algorithm is used, and epsilon d represents the fidelity of the two images of the channel under the PHash algorithm;
The calculation formula for the fidelity of the single channel fusion algorithm for vibration response is as follows:
in this scheme S04, combining the single-channel fidelity and the preset dynamic comprehensive index, the constructing the multi-channel fidelity adaptive comprehensive evaluation index includes:
Calculating importance coefficients for representing signal indexes under different fault states by using a Delphi method;
establishing a comprehensive index of the sensor by utilizing the time domain characteristic index and combining the index importance coefficient, and evaluating the capability of the sensor for collecting important information;
And constructing a multichannel fidelity self-adaptive comprehensive evaluation index by utilizing the fidelity and the sensor comprehensive index so as to perform self-adaptive evaluation on the multichannel fidelity, and providing a basis for the multichannel model fidelity optimization of the gearbox.
In the scheme S05, selecting a parameter variable to be corrected in the rigid-flexible coupling twin model of the gearbox, and constructing a response surface function of the parameter to be corrected and the fidelity of the rigid-flexible coupling twin model of the gearbox includes:
Adopting a central composite design method to carry out response surface test design; and adopting a second-order polynomial to perform response surface function fitting, wherein the fitting formula is as follows:
wherein a 0 is a constant, a i、bi、cij is a coefficient to be determined of first-order, second-order and second-order cross terms, n is the number of parameters to be corrected, and the number of times of least test is
In the scheme S06, the speed and position update formula of the gray wolf particle swarm fusion multi-objective optimization algorithm is as follows:
Vd(k+1)＝ω·[Vd(k)+C1r1(X1d(k)-Xd(k))+C2r2(X2d(k)-Xd(k))+C3r3(X3d(k)-Xd(k))] (5)
wherein r1, r2, r3 are random numbers in [0,1], ω is an adjustable inertial weight, its variation range is [0.5,1], and C 1、C2、C3 is calculated as follows:
C＝2·n2 (7)
Due to the variation of the above equation and the introduction of ω, the wolf group position is updated as:
Wherein P 1、P2、P3 represents the position vector of the updated alpha wolf, beta wolf and delta wolf, and P (t+1) represents the position vector determined after optimizing.
As an example of implementation, the present solution is further described by the following example:
A digital twin-based gearbox model high-fidelity method schematic diagram shown in figure 1 and a gearbox model high-fidelity optimization flow chart shown in figure 2, the scheme of the embodiment comprises the following steps:
based on finite element dynamics simulation software, a rigid-flexible coupling twin model of the gear box is established in a combined mode, modal analysis and dynamics simulation are carried out, monitoring point positions are selected for sensor arrangement through modal analysis, and multi-point vibration response is obtained through dynamics simulation;
based on a vibration test bed of the parallel fixed-axis gearbox, acquiring physical data of time domains and frequency domains of multiple channels under multiple fault states of the gearbox;
single-channel fidelity is calculated based on integration of a Sift algorithm and a PHash algorithm; combining single-channel fidelity with a preset sensor comprehensive index to construct a multi-channel comprehensive fidelity self-adaptive evaluation index;
Constructing a response surface model of the parameter variable to be corrected and the vibration response fidelity of the twin model;
Based on a gray wolf particle swarm fusion multi-objective optimization algorithm, solving non-cracking of a response surface function of the twin finite element dynamics model, and screening out an optimal solution by combining with a comprehensive fidelity self-adaptive evaluation index to finish fidelity optimization of a rigid-flexible coupling model;
By constructing a twin model of the healthy and fault gearbox, optimizing by using a fidelity method, and comparing the optimized results of the two models, the accuracy of model establishment and the feasibility of the fidelity optimizing method can be mutually verified.
As shown in a Sift algorithm flow chart of FIG. 3, the embodiment scheme comprises the steps of performing differential calculation on local features of a single-channel simulation and test vibration image by using the Sift algorithm, constructing a scale space based on a scale invariant feature transformation algorithm, detecting spatial extreme points, generating key feature point positions, determining key feature point directions, generating descriptors, and matching and comparing the key feature points.
Wherein the scale space construction may be convolved with the image with a gaussian kernel, as:
L(x,y,σ)＝G(x,y,σ)×I(x,y) (10)
Wherein σ represents a spatial scale parameter, and G (x, y, σ) is a gaussian kernel function, as follows:
In order to make the key point have rotation invariance, the main direction of the key point is specified by the gradient direction characteristic of the pixels in the field, wherein the gradient is characterized by the direction with the fastest gray level change of the point, and the gradient mode and the direction of each pixel are calculated as follows:
the euclidean distance is used to define the similarity of feature points. The Euclidean distance calculation formula and the similarity calculation are as follows:
Wherein the number of returned matching feature points is m, the number of non-matching feature points is n, sim 1(A1,A2) represents that the similarity value of the two pictures a 1、A2 is calculated by using the Sift algorithm.
As shown in PHash algorithm flow chart of FIG. 4, the embodiment comprises calculating the difference between the whole feature of the single-channel simulation and the test vibration image by PHash algorithm, compressing and graying the vibration image after inputting the vibration image, calculating DCT coefficient mean value by DCT transformation, connecting the perceived hash value of the generated image, calculating the hamming distance of PHash values of the two images, and obtaining the image similarity.
Wherein the hamming distance calculation is derived from equation (16):
the image similarity is obtained by the formula (17):
Wherein the number of the same values on the same bit of the hash string sequence value of the two pictures is p, the number of the different values on the same bit is q, sim 2(A1,A2) represents that the PHash algorithm is adopted to calculate the similarity value of the two pictures A 1、A2
In the embodiment, the Delphi method is a system data collection method, and collects judgment information of experts to negotiate specific questions in a certain field into unified comments, and investigation and analysis are required for comments of different experts. If the opinion diverges greatly, the questionnaire needs to be placed again until the opinions are unified. In order to ensure the reliability and consistency of the data, the invention needs to investigate and feed back for a plurality of times, and issues a questionnaire to a plurality of experts to perform 3 rounds of investigation.
As shown in the figure 6, the method specifically comprises the following steps: for a fault states, each state consists of b indicators, where b= [1,2,3,4,5,6]. Each expert gives a weight in a fault state, wherein V a,b,t represents the weight given by the t-th expert, where t= [1,2, ], n ], the sum of the weights of the different indicators in the same fault state is 1, as follows:
The average value of each special familia value is used as a final importance coefficient by a Delphi method, and the importance coefficient expression related to X rms、Cq、Ce、Cf、Ip、Cw in the corresponding fault state is as follows:
In order to determine the capability of different sensors for acquiring different important information, firstly, vibration signals are acquired, noise reduction processing is carried out by a digital filter, then, the characteristic values of the same characteristic index and the indexes acquired by different channel sensors are calculated, and the characteristic index coefficients are solved as follows:
Wherein, p represents a characteristic index, p= [1,2,3,4,5,6], R p,q represents a characteristic index coefficient, and G p,q represents a characteristic value of the index. The importance of each channel can be measured by the sensor comprehensive index, and is calculated by the following formula:
Wherein, T q represents the comprehensive index of each sensor, and the larger the index value is, the more important information acquired by the sensor is indicated, and the smaller the index value is, the less important information acquired by the sensor is indicated.
After the comprehensive index of each sensor is calculated, the single-channel comprehensive fidelity is fused with the comprehensive index to obtain the channel self-adaptive comprehensive fidelity evaluation index M q, which is calculated by the following formula:
Mq＝εq·Tq(22)
summing up the adaptive fidelity evaluation indexes of the multiple channels, and solving the maximum value of the adaptive fidelity evaluation indexes, wherein the calculation method comprises the following steps:
And screening the Pareto optimal solution by taking the maximum value of the adaptive comprehensive fidelity evaluation index as a screening basis, and taking the screened correction parameter value and the fidelity as the optimal solution. If a plurality of groups of solutions still exist after screening, the relevant corresponding parameters are replaced to the rigid-flexible coupling simulation model for verification, and the optimal solution with the smallest difference between the simulation fidelity and the calculation fidelity of the model is selected.
In this embodiment, a gearbox vibration test stand is shown in fig. 7-9, comprising,
In the overall equipment diagram of the gearbox vibration test bench shown in fig. 7, 1 represents a bench control equipment, 2 represents a signal transmission equipment, 3 represents a data conversion and acquisition system, 4 represents a high-performance computer processing equipment, and 5 represents a vibration test bench.
In a diagram of the gearbox vibration test data acquisition equipment shown in fig. 8, 1 represents signal vibration data acquisition analysis software, 2 represents a data converter, 3 represents a high-performance computer, 4 represents a load controller, 5 represents a motor controller, 6 represents a vibration data acquisition device, 7 represents a vibration data transmission line, 8 represents a router, 9 represents a switch, and 10 represents a signal network line.
In the overall view of the gear box apparatus shown in fig. 9, 1 denotes a gear box, 2 denotes a laser type rotational speed measuring instrument, 3 denotes a coupling, 4 denotes a motor, 5 denotes a shield, 6 denotes a magnetic powder brake which serves as a load, 7 denotes a vibration acceleration sensor, and 8 denotes a test stand and a base.
The flow for collecting vibration signals is as follows: the sensor is arranged on the surface of the gearbox body and connected with the acquisition card through a signal connecting wire, the acquisition card is connected with the switch and the router through a network cable, one end of the photoelectric speed sensor is connected with the signal converter, the signal converter is connected with the acquisition card through a signal connecting wire, the voltage signal is converted into a square wave signal, and the high-performance computer end is also connected to the switch through the network cable. Therefore, the acquisition card, the vibration acceleration sensor, the photoelectric rotating speed sensor, the router, the switch and the high-performance computer form local area network connection together to share information. The motor control cabinet and the load controller are respectively connected with the motor and the magnetic powder brake, after the protective cover is covered, the motor control cabinet and the load controller switch are sequentially opened, the actual rotation speed of the motor control cabinet is set to be 1500rpm, and the load controller applies torque 2Nm to the magnetic powder brake. Starting a vibration test bed, opening vibration monitoring software to collect data, arranging a digital filter noise reduction module in the acquisition card, carrying out noise reduction treatment on the collected signals, wherein the single collection time length is 5s, the sampling frequency is 2kHz, and collecting the time domain and converting the time domain into a frequency domain signal. And after the collection is finished, closing the motor control cabinet and the load controller, opening the gear box to replace the fault pinion, performing the operation again, and collecting time domain and frequency domain signals under the fault state.
The flow of the gray wolf particle swarm fusion multi-objective optimization algorithm optimization fidelity response surface model is shown in fig. 10, and the specific process comprises the following steps:
1) Setting an initial correction parameter value and a range, taking the maximum fidelity as a target, selecting a proper test method according to the correction parameter and the characteristics of the fidelity target function, fitting and verifying a response surface model, and obtaining the response surface function.
2) Setting the population dimension dim, the iteration number K and the scale popsize, and setting the initial values of the parameters alpha and A, C and the inertia weight omega.
3) And randomly generating the speed and the position of the population in the initial state, and carrying out iterative optimization on the algorithm to realize the population initialization.
4) The first three individuals were ranked according to the gray wolf individual fitness values, α, β and δ, and X α(k)、Xβ(k)、Xδ (k) was defined as their location information.
5) And (3) updating and calculating individual positions of the wolves to obtain new values of alpha, A, C and omega, and returning to the step (3) for recalculation to obtain new alpha, beta and delta.
6) If the maximum number of iterations is reached, the position where the wolf found the optimal solution is considered, and the process goes to step (17). If not, the current position result is used as a potential solution, and the step (14) is returned again to continuously search for a better solution.
7) And outputting an optimal scheme group meeting the end condition, updating the correction parameters of the rigid-flexible coupling model, and verifying the back-substitution response surface model.
The foregoing description is only a partial embodiment of the present invention, and is not intended to limit the scope of the present invention, and all equivalent devices or equivalent processes using the descriptions and the drawings of the present invention or directly or indirectly applied to other related technical fields are included in the scope of the present invention.
Claims (9)
1. A digital twinning-based gearbox model high-fidelity method, which is characterized by comprising the following steps:
S01, establishing a rigid-flexible coupling twin model of the gear box by using finite element simulation software, then carrying out modal analysis to obtain dynamic monitoring points, and carrying out dynamic simulation;
S02, based on a vibration test bed of the parallel fixed-axis gearbox, acquiring vibration physical data of a time domain and a frequency domain of a corresponding multi-channel of the gearbox in a normal or multiple fault states;
S03, performing single-channel fusion algorithm fidelity calculation on the vibration physical data based on a Sift algorithm and a PHash algorithm;
s04, constructing a multichannel fidelity self-adaptive comprehensive evaluation index by combining the single-channel fidelity with a preset dynamics comprehensive index;
s05, selecting parameter variables to be corrected in the rigid-flexible coupling twin model of the gear box, and constructing a response surface function of parameters to be corrected and fidelity of the rigid-flexible coupling twin model of the gear box;
s06, solving non-cracking of a response surface function of parameters to be corrected and fidelity of the rigid-flexible coupling twin model of the gear box based on a gray wolf particle swarm fusion multi-objective optimization algorithm, and screening out an optimal solution by combining a multi-channel fidelity self-adaptive comprehensive evaluation index to finish the fidelity optimization of the rigid-flexible coupling twin model of the gear box.
2. The digital twinning-based gearbox model high-fidelity method of claim 1, wherein S01 comprises: and establishing a flexible finite element twin model of the gear and the box body by utilizing finite element simulation software, replacing rigid box walls and gear components with flexible components, setting boundary conditions and model parameters, establishing a rigid-flexible coupling twin model of the gear box of the parallel fixed-axis gear box, carrying out modal analysis on the flexible box body, arranging 6 monitoring points at the maximum amplitude to carry out vibration response monitoring, and carrying out dynamic simulation on the rigid-flexible coupling twin model of the gear box to verify the accuracy of the twin model.
3. The high-fidelity method of the digital twin-based gearbox model according to claim 2, wherein in S02, based on a parallel fixed-axis gearbox vibration test bed, after vibration physical data of time domains and frequency domains corresponding to 6 channels of the gearbox in normal or multiple fault states are collected, noise reduction processing is carried out through a Vmd noise reduction algorithm, then fast Fourier transformation is utilized to convert the signals into frequency domain signals, and corresponding test images are generated.
4. The high-fidelity method of the gearbox model based on digital twin according to claim 3, wherein in S03, single-channel fusion algorithm fidelity calculation is carried out on the test image corresponding to the vibration physical data based on a Sift algorithm and a PHash algorithm, and the obtained comprehensive fidelity result is used as a final virtual-real signal image fidelity calculation result.
5. The digital twinning-based gearbox model high-fidelity method of claim 4, wherein S03 comprises:
the formula of the Sift algorithm for calculating the fidelity of the test image corresponding to the vibration physical data is as follows:
ε′d＝sim1(B1,B2)×100％ (1)
Sim 1(B1,B2) represents the similarity value between the vibration frequency domain response image B 1 of the rigid-flexible coupled twin model and the vibration frequency domain response image B 2 of the physical data when the Sift algorithm is used, ε' d represents the fidelity of the two images of the channel under the Sift algorithm, d represents the channel serial number of the sensor, and d= [1,2,3,4,5,6];
The equation for calculating the fidelity of the test image corresponding to the vibration physical data by PHash algorithm is as follows:
ε″d＝sim2(B1,B2)×100％ (2)
Sim 2(B1,B2) represents a similarity value between the vibration frequency domain response image B 1 of the rigid-flexible coupled twin model and the vibration frequency domain response image B 2 of the physical data when the PHash algorithm is used, and epsilon d represents fidelity of two images of the channel under the PHash algorithm;
The calculation formula for the fidelity of the single channel fusion algorithm for vibration response is as follows:
6. The digital twin-based gearbox model high-fidelity method according to claim 5, wherein in S04, combining the single-channel fidelity and the preset dynamics comprehensive index to construct the multi-channel fidelity self-adaptive comprehensive evaluation index comprises:
Calculating importance coefficients for representing signal indexes under different fault states by using a Delphi method;
establishing a comprehensive index of the sensor by utilizing the time domain characteristic index and combining the index importance coefficient, and evaluating the capability of the sensor for collecting important information;
And constructing a multichannel fidelity self-adaptive comprehensive evaluation index by utilizing the fidelity and the sensor comprehensive index so as to perform self-adaptive evaluation on the multichannel fidelity, and providing a basis for the multichannel model fidelity optimization of the gearbox.
7. The high-fidelity method of a gear box model based on digital twin according to claim 1, wherein in S05, selecting parameter variables to be corrected in the gear box rigid-flexible coupling twin model, and constructing a response surface function of parameters to be corrected and fidelity of the gear box rigid-flexible coupling twin model comprises:
Adopting a central composite design method to carry out response surface test design; and adopting a second-order polynomial to perform response surface function fitting, wherein the fitting formula is as follows:
wherein a 0 is a constant, a i、bi、cij is a coefficient to be determined of first-order, second-order and second-order cross terms, n is the number of parameters to be corrected, and the number of times of least test is
8. A digital twin-based gearbox model hi-fi system, characterized in that it is applied with a digital twin-based gearbox model hi-fi method as claimed in any of the claims 1 to 7.
9. A computer-readable storage medium, characterized by: the storage medium stores at least one instruction, at least one program, code set, or instruction set, which is loaded by a processor and executed to implement the digital twin-based gearbox model hi-fi method according to one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311563535.5A CN117592332B (en) | 2023-11-21 | 2023-11-21 | Digital twinning-based gearbox model high-fidelity method, system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311563535.5A CN117592332B (en) | 2023-11-21 | 2023-11-21 | Digital twinning-based gearbox model high-fidelity method, system and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117592332A CN117592332A (en) | 2024-02-23 |
| CN117592332B true CN117592332B (en) | 2024-11-01 |
Family
ID=89917705
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311563535.5A Active CN117592332B (en) | 2023-11-21 | 2023-11-21 | Digital twinning-based gearbox model high-fidelity method, system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117592332B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118153227B (en) * | 2024-03-12 | 2024-08-23 | 江苏环欧智能传动设备有限公司 | Simulation analysis method for vibration response of gear box |
| CN118691847B (en) * | 2024-08-23 | 2024-10-29 | 长沙能川信息科技有限公司 | Transformer substation defect detection method, system and storage medium based on positive sample image |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115481564A (en) * | 2022-07-27 | 2022-12-16 | 南京航空航天大学 | Bearing roller dynamic and static contact rigidity measurement method based on digital twin model |
| CN116046384A (en) * | 2023-02-08 | 2023-05-02 | 昆明理工大学 | Digital twinning-based gearbox online fault diagnosis method |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114323644B (en) * | 2022-03-14 | 2022-06-03 | 中国长江三峡集团有限公司 | Gear box fault diagnosis and signal acquisition method and device and electronic equipment |
| CN114925558A (en) * | 2022-04-19 | 2022-08-19 | 南京航空航天大学 | Digital twinning modeling method for bearing test bed |
| CN114970321A (en) * | 2022-04-28 | 2022-08-30 | 长安大学 | Scene flow digital twinning method and system based on dynamic trajectory flow |
| CN115828696A (en) * | 2022-12-12 | 2023-03-21 | 昆明理工大学 | Gear box PHM optimization method and system based on digital twin technology and application |
-
2023
- 2023-11-21 CN CN202311563535.5A patent/CN117592332B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115481564A (en) * | 2022-07-27 | 2022-12-16 | 南京航空航天大学 | Bearing roller dynamic and static contact rigidity measurement method based on digital twin model |
| CN116046384A (en) * | 2023-02-08 | 2023-05-02 | 昆明理工大学 | Digital twinning-based gearbox online fault diagnosis method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117592332A (en) | 2024-02-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117592332B (en) | Digital twinning-based gearbox model high-fidelity method, system and storage medium | |
| CN114323644B (en) | Gear box fault diagnosis and signal acquisition method and device and electronic equipment | |
| CN110428004B (en) | Mechanical part fault diagnosis method based on deep learning under data imbalance | |
| Gao et al. | ASM1D-GAN: An intelligent fault diagnosis method based on assembled 1D convolutional neural network and generative adversarial networks | |
| CN108256556A (en) | Wind-driven generator group wheel box method for diagnosing faults based on depth belief network | |
| CN117828350A (en) | Bearing fault diagnosis method and system for cross-working-condition unbalanced data | |
| Jiang et al. | A fault diagnostic method for induction motors based on feature incremental broad learning and singular value decomposition | |
| CN116127383A (en) | Fault detection method and device, electronic equipment and storage medium | |
| Xue et al. | A novel framework for motor bearing fault diagnosis based on multi-transformation domain and multi-source data | |
| CN114755010A (en) | Rotary machine vibration fault diagnosis method and system | |
| CN112414715B (en) | Bearing fault diagnosis method based on mixed feature and improved gray level symbiosis algorithm | |
| Lee et al. | An explainable deep learning-based prognostic model for rotating machinery | |
| CN116011507A (en) | Rare fault diagnosis method for fusion element learning and graph neural network | |
| CN118194136B (en) | Fault diagnosis method for mechanical equipment | |
| JP6950647B2 (en) | Data determination device, method, and program | |
| CN117473872B (en) | Rolling bearing fault diagnosis method based on FFT (fast Fourier transform) coding and L-CNN (linear-to-CNN) | |
| CN112966560A (en) | Electric spindle fault diagnosis method and device based on deconvolution imaging | |
| CN117932390A (en) | Vibration signal analysis and diagnosis method based on integration of attention mechanism and DCGAN | |
| CN117171544A (en) | Motor vibration fault diagnosis method based on multichannel fusion convolutional neural network | |
| CN114298101A (en) | Dynamic weighing method and system | |
| CN115984606A (en) | Planetary gear fault diagnosis method and system based on CEEMDAN denoising | |
| CN115659132A (en) | Diagnostic method for depth map convolution neural network fusing multi-sensor network | |
| CN115525866A (en) | Deep learning rolling bearing fault diagnosis method and system | |
| CN114048762A (en) | Double-attention-guided rotating machine health assessment method | |
| CN113919388A (en) | Electromechanical equipment fault diagnosis method and device integrating signal frequency spectrum amplitude modulation and deep learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |