CN114331945A - Large flange assembling alignment method based on multiple local images - Google Patents
Large flange assembling alignment method based on multiple local images Download PDFInfo
- Publication number
- CN114331945A CN114331945A CN202011061902.8A CN202011061902A CN114331945A CN 114331945 A CN114331945 A CN 114331945A CN 202011061902 A CN202011061902 A CN 202011061902A CN 114331945 A CN114331945 A CN 114331945A
- Authority
- CN
- China
- Prior art keywords
- flange
- image
- flanges
- local
- gap
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000013519 translation Methods 0.000 claims abstract description 37
- 238000005457 optimization Methods 0.000 claims abstract description 36
- 239000013598 vector Substances 0.000 claims abstract description 13
- 238000000605 extraction Methods 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 7
- 238000009826 distribution Methods 0.000 claims description 5
- 238000012937 correction Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 238000001914 filtration Methods 0.000 claims description 2
- 230000002068 genetic effect Effects 0.000 claims description 2
- 238000011478 gradient descent method Methods 0.000 claims description 2
- 239000011159 matrix material Substances 0.000 claims description 2
- 238000004519 manufacturing process Methods 0.000 abstract description 2
- 238000004364 calculation method Methods 0.000 description 7
- 238000011156 evaluation Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000001154 acute effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Images
Landscapes
- Image Processing (AREA)
Abstract
The invention discloses a large flange assembling and aligning method based on multiple local images, and belongs to the technical field of manufacturing and assembling alignment. The method comprises the following steps: shooting the matching surfaces of the flanges at different positions to obtain a plurality of local images of the two flanges to be assembled; extracting the straight line information and the circle information in each image pair; calculating the gap vector and the symbol of the flange and the corresponding relation between the gap vector and the translation adjustment amount, and constructing an optimization model to calculate the translation adjustment amount of the flange; and correcting the positions of the bolt holes of the flanges through the calculated translation adjustment quantity of the flanges, and modeling to calculate the rotation adjustment quantity of the flanges. The invention can carry out assembly alignment on large flange parts by using a plurality of local view field high-precision images.
Description
Technical Field
The invention belongs to the technical field of manufacturing and assembly alignment, and particularly relates to a large flange assembly alignment method with multiple local images.
Background
The assembly of large flange parts is mainly carried out manually, relatively offset and rotation between flanges often occur in the assembly process to cause large assembly alignment errors, local collision damage can be caused during assembly, and the assembly precision and reliability are affected, so that the translation adjustment amount and the rotation adjustment amount of the flanges to be assembled in alignment need to be calculated to correct the assembly pose of the flanges. The alignment method for directly obtaining the whole image based on the machine vision usually needs to shoot a single high-resolution image or use a low-resolution image geometric laser distance sensor to carry out high-precision measurement on parts, and has high hardware cost and low efficiency.
Disclosure of Invention
In view of the above, the present invention provides a method for aligning a large flange assembly based on multiple local images, which can perform assembly alignment on a large flange part by using multiple local view-field high-precision images.
A large flange assembling alignment method based on multiple local images comprises the following steps:
shooting the matching surfaces of the flanges at different positions to obtain a plurality of local images of the two flanges to be assembled;
step two: extracting the straight line information and the circle information in each image pair;
step three: constructing an optimization model to calculate the translation adjustment quantity of the flange;
step four: and correcting the positions of the bolt holes of the flanges through the translation adjustment quantity of the flanges, and modeling to calculate the rotation adjustment quantity of the flanges.
Further, each image in the first step comprises the edge of the flange and the bolt hole to be paired, wherein two flange images acquired at the same shooting position are combined to generate one image pair, and a plurality of groups of image pairs are formed.
Further, the process of extracting the linear information and the circle information in the second step includes: firstly, carrying out image filtering processing and edge extraction on each image pair, carrying out circle detection on edge points extracted from each image through Hough transform, and fitting a circle by using a least square method to extract the bolt hole position of a flange; and eliminating points near the circle, and performing linear fitting on the rest points by a least square method to obtain the edge position and the direction of the flange.
Further, the process of constructing the optimization model in the third step to calculate the flange translation adjustment amount is as follows:
calculating a linear equation which is obtained by fitting in the ith local view field and is used for representing the position and the direction of the edge of the flange, and calculating the assembly gap Cd of the edge of the flange in the local view field of the flange in the image pair(i)And direction of
Generating gap vectors
By calculating the ith local field of view direction vFOVd (i)And the direction of the gap
The included angle between them, the gap symbol
Where sgn (x) is a sign function, T represents the transpose of the vector (matrix), and the gap is adjusted with flange translation by an amount Δ T ═ Δ x, Δ y)TThe variation is generated, and the adjusted clearance is:
the obtained gaps of the N image pairs are calculated to generate an image gap sequence Cns { Cn(1),Cn(2),...,Cn(N)Establishing an optimization model by taking the optimal gap consistency as an optimization target and the adjustment quantity as an optimization variable:
wherein avg (Cns) is the mean of the set Cns;
obtaining the translational adjustment quantity delta t of the flange by solving the modelopt＝(Δxopt,Δyopt)T。
Further, the process of modeling and calculating the flange rotation adjustment amount in the fourth step is as follows: adjustment of the quantity Δ t by translation of the flangeopt＝(Δxopt,Δyopt)TThe positions of the extracted flange bolt holes needing to be adjusted in each image pair
To perform correction
Position of bolt hole to be paired
And (3) establishing an optimization model by taking the minimum position difference of the corrected bolt hole pairs to be paired as an optimization target and the rotation adjustment quantity of the flange and the distribution radius of the bolt holes as optimization variables:
and solving the model, and calculating to obtain the flange rotation adjustment amount.
Further, in the third step, a gradient descent method is adopted to calculate the translation adjustment amount.
Further, a genetic algorithm is adopted in the fourth step to calculate the rotation adjustment amount.
Has the advantages that:
1. the invention provides a large flange assembly alignment method based on multiple local images, which is used for implementing assembly alignment on large flange parts by using multiple local view field high-precision images, reducing the hardware cost of an assembly system and improving the efficiency.
2. According to the method, the translation error of the large flange part is evaluated by extracting the multiple image gaps and using the consistency of the image gaps as an index, the integral central position of the part does not need to be detected, only a small amount of edge information is needed for calculation, and the complexity of data to be processed is reduced. And the rotation error evaluation is carried out by taking the error of the center distance of the bolt hole pair as an index, so that the actual process requirement of flange assembly is more approximate.
3. The evaluation index of the gap consistency is determined according to the specific assembly task requirement. The variance of the gap sequence can be used for measurement in general.
4. The invention establishes a translation adjustment optimization model and a rotation adjustment optimization model to calculate the translation adjustment and the rotation adjustment of the flange assembly. The optimization model takes the designed error evaluation index as a target function, takes the translational adjustment amount and the rotational adjustment amount as optimization variables, and respectively calculates the translational adjustment amount and the rotational adjustment amount through model solution, so that the large deviation which is possibly generated when the overall center and the direction of the flange are estimated through local image information by using a traditional method is avoided, and the system assembly alignment precision is improved.
5. According to the method, firstly, the translation adjustment quantity optimization model is used for carrying out translation adjustment quantity optimization calculation, and the translation adjustment quantity optimization model is corrected by using the translation adjustment quantity calculation result, so that the influence of translation errors on rotation errors in the assembly alignment process is eliminated, and the method is more suitable for the actual situation of assembly alignment of large flange parts.
6. The calculation method of the translation adjustment quantity optimization model is selected preferentially according to specific application requirements, and a gradient descent-based method can be generally used for rapid calculation.
Drawings
FIG. 1 is a flow chart of an implementation of the alignment method of the present invention;
FIG. 2 is a schematic view of a flange local field of view image acquisition;
FIG. 3 is a schematic diagram of local field edge detection and straight line and circle extraction;
FIG. 4 is a schematic view of a partial field gap;
FIG. 5 is a schematic view of a partial field of view;
FIG. 6 is a schematic view of a bolt hole position correction;
FIG. 7 is a schematic diagram of image and line and circle detection in the embodiment.
Detailed Description
The invention is described in detail below by way of example with reference to the accompanying drawings.
The invention aims to provide an assembly alignment method based on multiple local images for a large flange. According to the method, the integral adjustment quantity of the flange can be estimated through a plurality of local view fields, and therefore assembly alignment is achieved. FIG. 1 shows a flow chart of the method of the present invention. The method comprises the steps of obtaining local images of a plurality of positions of a flange to generate an image pair; processing images of the image pairs to extract straight lines and circles in the image pairs, and representing the edge and the bolt hole of the flange by a sub-table; calculating the gap in the local view field of the flange according to the extracted linear information, and establishing a translation adjustment optimization model for calculating the translation adjustment; and correcting the extracted bolt hole position according to the calculation result of the translation adjustment amount, and establishing a rotation adjustment amount optimization model for calculating the rotation adjustment amount.
As shown in FIG. 2, the method first obtains a plurality of groups of partial images with bolt holes to be paired at the flange mounting surface. The partial field-of-view images acquired by photographing at the same position are combined into one image pair. Sets of pairs of local field-of-view images are generated by taking at a plurality of locations. The flange 1 with the larger diameter and the flange 2 with the smaller diameter are defined.
As shown in fig. 3, each acquired image includes the edge of the flange and the bolt holes to be mated. Wherein the edge of the flange is characterized by being approximate to a straight line, and the bolt holes are characterized by being approximate to a circle. In order to extract the edge information and the bolt hole information in the image, each image is filtered and denoised firstly, and then the image edge is extracted through an edge extraction operator. And fitting the straight line and the circle in the image through a straight line and circle fitting algorithm to extract the information of the straight line and the circle in the image. This operation is repeated, and straight line and circle information extraction is performed on all the obtained image pairs.
As shown in FIG. 4, in one figureThe straight lines corresponding to the flange 1 and the flange 2 extracted from the image pair are respectively L1And L2The equations are respectively A1 (i)x+B1 (i)y+C1 (i)＝0，A2 (i)x+B2 (i)y+C2 (i)0. The flange gap is defined as the vector of the flange 1 edge pointing towards the flange 2 edge.
The reference direction of the gap is:
the intercept vectors of the two lines are:
gap vcd (i)Is the projection of the intercept vector in the reference direction:
v is to becd (i)Normalized to the gap direction vector
The size of the gap is Cd(i). Thereby generating a gap vector
This operation is repeated, and gaps in all image pairs are extracted.
As shown in FIG. 5, the i-th local viewing field is defined as the viewing field direction v relative to the flange centerFOVd (i)With a proper fit, the angle between the gap direction and the field of view direction should be acute, and the gap should meet this requirement at all fields of view. Defining gap symbols in local field of view
Where Sgn (x) is a sign function. The flange 1 is known to be adjusted by the translation amount Δ t ═ (Δ x, Δ y)TAfter the shift is generated, the gap vector after the change is:
generating a new gap size from the gap sign and the gap vector:
with proper assembly, the obtained N sets of image pairs, the gaps of each image pair constitute a sequence Cns { Cn(1),Cn(2),...,Cn(N)Have higher consistency. The sequence identity was evaluated by variance, thus constructing an identity index:
and (3) constructing an optimization model by taking the minimization of the consistency index as an optimization target and the translation adjustment quantity as an optimization variable:
solving the model to obtain the flange translation adjustment quantity delta topt＝(Δxopt,Δyopt)T。
As shown in fig. 6, the error of the centers of the bolt holes to be paired of the two flanges is composed of a translation error and a rotation error. Firstly, correcting the bolt hole position of the flange 1 by solving the obtained translation adjustment quantity of the flange. In the ith image pair, the positions of the bolt holes of the flange 1 are
Position of bolt hole of flange 2
The positions of the bolt holes of the corrected flange 1 are
The error between the corrected bolt holes of the flange 1 and the corrected bolt holes of the flange 2 is mainly caused by the rotation error. Radius R of rotation error distributed through bolt hole(i)Correcting the rotation adjustment quantity theta, and constructing an optimization model by taking the minimum error of the center distance of the bolt holes of each image as an optimization target:
bolt hole distribution radius R(i)Is known, the model is solved to obtain the rotation adjustment quantity thetaopt。
The solution of the optimization model can be solved through different optimization algorithms.
The invention discloses a large flange assembling alignment method based on a coaxial alignment principle, which estimates the whole assembling error of a flange by using a plurality of local view fields and reduces the hardware cost of an assembling system.
Example (b):
the method is used for carrying out assembly alignment on two flanges (a flange 1 and a flange 2) to be assembled and aligned as an embodiment. The diameter of the flange 1 is 400.2mm, and the diameter of the flange 2 is 400 mm. 8 bolt holes with the diameter of 4mm and the distribution radius of 195mm are uniformly distributed on the flange to form bolt hole pairs. The two flanges respectively have a translation position error of 0.11mm, a translation position error of 0.15mm and a direction error of 1 degree (taking the anticlockwise direction as the positive direction).
Step one, acquiring images at 8 local view fields at the flange matching position, acquiring 8 groups of local view field images, and forming an image pair.
And step two, extracting circle information and straight line information in each image pair to obtain the edge position and direction of the flange and the position and direction of the bolt hole, wherein the arrow direction in each figure is a straight line extraction indication, and the cross is a circle extraction indication, as shown in fig. 7.
And step three, extracting the size, direction, view field direction and gap symbol of the gap in each image pair as shown in table 1.
TABLE 1
By extracting parameters, an optimization model is constructed
Solving the model to obtain the translational adjustment quantity delta topt(-30.34, -37.15), the corresponding actual physical adjustment amount is Δ xopt＝-111.25um，Δyopt-136.23um, and standard error (Δ x)opt＝-110um，Δyopt-150um) is 13.83 um.
And step four, correcting the position of the bolt hole through the calculated translation adjustment amount, as shown in table 2.
TABLE 2
Building optimization models
In the model, bolt hole distribution radius R(i)Is in a known range (195 +/-0.5 mm), and the rotation adjustment quantity theta is obtained by solving the modelopt1.0037 ° with a deviation of 0.0037 ° from the standard error of 1 °. The above calculation results demonstrate the effectiveness of the method.
In summary, the above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (7)
1. A large flange assembling alignment method based on multiple local images is characterized by comprising the following steps:
shooting the matching surfaces of the flanges at different positions to obtain a plurality of local images of the two flanges to be assembled;
step two: extracting the straight line information and the circle information in each image pair;
step three: constructing an optimization model to calculate the translation adjustment quantity of the flange;
step four: and correcting the positions of the bolt holes of the flanges through the translation adjustment quantity of the flanges, and modeling to calculate the rotation adjustment quantity of the flanges.
2. The method for assembling and aligning the large flange based on the multiple partial images as claimed in claim 1, wherein each image in the first step comprises the edge of the flange and the bolt hole to be paired, and two flange images obtained from the same shooting position are combined to form one image pair, and a plurality of sets of image pairs are formed.
3. The method for aligning the assembly of the large flange based on the multiple partial images as claimed in claim 2, wherein the step two of extracting the linear information and the circular information comprises: firstly, carrying out image filtering processing and edge extraction on each image pair, carrying out circle detection on edge points extracted from each image through Hough transform, and fitting a circle by using a least square method to extract the bolt hole position of a flange; and eliminating points near the circle, and performing linear fitting on the rest points by a least square method to obtain the edge position and the direction of the flange.
4. The multi-local-image-based large flange assembling alignment method according to claim 3, wherein the process of constructing the optimization model in the third step to calculate the flange translation adjustment amount is as follows:
calculating a linear equation which is obtained by fitting in the ith local view field and is used for representing the position and the direction of the edge of the flange, and calculating the assembly gap Cd of the edge of the flange in the local view field of the flange in the image pair(i)And direction of
Generating gap vectors
By calculating the ith local field of view direction vFOVd (i)And the direction of the gap
The included angle between them, the gap symbol
Where sgn (x) is a sign function, T represents the transpose of the vector (matrix), and the gap is adjusted with flange translation by an amount Δ T ═ Δ x, Δ y)TThe variation is generated, and the adjusted clearance is:
the obtained gaps of the N image pairs are calculated to generate an image gap sequence Cns { Cn(1),Cn(2),...,Cn(N)Establishing an optimization model by taking the optimal gap consistency as an optimization target and the adjustment quantity as an optimization variable:
where avg (Cns) is the mean of the set Cns.
By solving the model, the flange translation adjustment is obtainedIntegral quantity delta topt＝(Δxopt,Δyopt)T。
5. The method for aligning the assembly of the large flange based on the multiple partial images as claimed in claim 4, wherein the modeling in the fourth step is used for calculating the adjustment amount of the flange rotation as follows: adjustment of the quantity Δ t by translation of the flangeopt＝(Δxopt,Δyopt)TThe positions of the extracted flange bolt holes needing to be adjusted in each image pair
To perform correction
Position of bolt hole to be paired
And (3) establishing an optimization model by taking the minimum position difference of the corrected bolt hole pairs to be paired as an optimization target and the rotation adjustment quantity of the flange and the distribution radius of the bolt holes as optimization variables:
and solving the model, and calculating to obtain the flange rotation adjustment amount.
6. The multi-local-image-based large flange assembling alignment method according to claim 1, wherein a gradient descent method is adopted in the third step to calculate the translation adjustment amount.
7. The multi-local-image-based large flange assembly alignment method according to claim 1, wherein a genetic algorithm is adopted in the fourth step to calculate the rotation adjustment amount.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011061902.8A CN114331945A (en) | 2020-09-30 | 2020-09-30 | Large flange assembling alignment method based on multiple local images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011061902.8A CN114331945A (en) | 2020-09-30 | 2020-09-30 | Large flange assembling alignment method based on multiple local images |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114331945A true CN114331945A (en) | 2022-04-12 |
Family
ID=81011935
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011061902.8A Pending CN114331945A (en) | 2020-09-30 | 2020-09-30 | Large flange assembling alignment method based on multiple local images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114331945A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010025486A1 (en) * | 2008-09-03 | 2010-03-11 | Ruag Aerospace Austria Gmbh | Drive device for adjusting components of a spacecraft that are to be oriented |
| CN108876777A (en) * | 2018-06-14 | 2018-11-23 | 重庆科技学院 | A kind of visible detection method and system of wind electricity blade end face of flange characteristic size |
| CN110933270A (en) * | 2019-11-25 | 2020-03-27 | 天津津航技术物理研究所 | Six-degree-of-freedom precision adjustment imaging chip assembly structure |
| CN111047702A (en) * | 2019-12-18 | 2020-04-21 | 成都飞机工业（集团）有限责任公司 | Automatic welding method for flange bent pipe based on binocular vision |
-
2020
- 2020-09-30 CN CN202011061902.8A patent/CN114331945A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010025486A1 (en) * | 2008-09-03 | 2010-03-11 | Ruag Aerospace Austria Gmbh | Drive device for adjusting components of a spacecraft that are to be oriented |
| CN108876777A (en) * | 2018-06-14 | 2018-11-23 | 重庆科技学院 | A kind of visible detection method and system of wind electricity blade end face of flange characteristic size |
| CN110933270A (en) * | 2019-11-25 | 2020-03-27 | 天津津航技术物理研究所 | Six-degree-of-freedom precision adjustment imaging chip assembly structure |
| CN111047702A (en) * | 2019-12-18 | 2020-04-21 | 成都飞机工业（集团）有限责任公司 | Automatic welding method for flange bent pipe based on binocular vision |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109598762B (en) | High-precision binocular camera calibration method | |
| CN111322967B (en) | Centering method for assembly process of stepped shaft and hole | |
| CN109470170B (en) | Stereoscopic vision space circular attitude high-precision measurement method and system based on optimal projection plane | |
| CN102567989A (en) | Space positioning method based on binocular stereo vision | |
| CN108269289A (en) | A kind of two step optimization methods of camera parameter calibration | |
| CN109978991B (en) | Method for rapidly realizing online measurement of complex component clamping pose error based on vision | |
| CN112258587B (en) | Camera calibration method based on gray wolf particle swarm mixing algorithm | |
| CN112150561A (en) | Multi-camera calibration method | |
| CN115055964B (en) | Intelligent assembling method and system based on fuel injection pump | |
| CN110827360A (en) | Photometric stereo measurement system and method for calibrating light source direction thereof | |
| CN114463442A (en) | Calibration method of non-coaxial camera | |
| CN112053405B (en) | Deviation calibration and external parameter correction method for optical axis and rotating shaft of follow-up vision system | |
| CN113781579A (en) | Geometric calibration method for panoramic infrared camera | |
| CN109766903A (en) | A kind of point cloud model SURFACES MATCHING method based on curved surface features | |
| CN116957987A (en) | Multi-eye polar line correction method, device, computer equipment and storage medium | |
| CN114331945A (en) | Large flange assembling alignment method based on multiple local images | |
| CN104424601B (en) | Centering assembly method and device for special-shaped body assembly parts | |
| CN112489141B (en) | Production line calibration method and device for single-board single-image strip relay lens of vehicle-mounted camera | |
| CN113587904A (en) | Target attitude and position measurement method integrating machine vision and laser reference point information | |
| CN113409395B (en) | High-precision detection and positioning method for catheter end | |
| CN111968182B (en) | Calibration method for nonlinear model parameters of binocular camera | |
| CN110458951B (en) | Modeling data acquisition method and related device for power grid pole tower | |
| CN116958264A (en) | Bolt hole positioning and pose estimation method based on three-dimensional vision | |
| CN111652945A (en) | Camera calibration method | |
| CN116563165A (en) | Neutral beam infrared image distortion correction and data processing method |
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 |