CN117073586A - Device and method for detecting parallelism of mechanical shaft of coaxial double-shaft turntable - Google Patents

Device and method for detecting parallelism of mechanical shaft of coaxial double-shaft turntable Download PDF

Info

Publication number
CN117073586A
CN117073586A CN202311338094.9A CN202311338094A CN117073586A CN 117073586 A CN117073586 A CN 117073586A CN 202311338094 A CN202311338094 A CN 202311338094A CN 117073586 A CN117073586 A CN 117073586A
Authority
CN
China
Prior art keywords
axis
projection image
image sequence
angle
rotation
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.)
Granted
Application number
CN202311338094.9A
Other languages
Chinese (zh)
Other versions
CN117073586B (en
Inventor
栾彬
栾恭勇
栾延淇
李钰
王海英
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Qingdao Mylange Intelligent Manufacturing Co ltd
Original Assignee
Qingdao Mylange Intelligent Manufacturing Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Qingdao Mylange Intelligent Manufacturing Co ltd filed Critical Qingdao Mylange Intelligent Manufacturing Co ltd
Priority to CN202311338094.9A priority Critical patent/CN117073586B/en
Publication of CN117073586A publication Critical patent/CN117073586A/en
Application granted granted Critical
Publication of CN117073586B publication Critical patent/CN117073586B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00—Measuring arrangements characterised by the use of optical techniques
    • G01B11/26—Measuring arrangements characterised by the use of optical techniques for measuring angles or tapers; for testing the alignment of axes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Biophysics (AREA)
  • Evolutionary Computation (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Software Systems (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Physics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Geometry (AREA)
  • Image Analysis (AREA)

Abstract

The invention relates to the technical field of parallelism measurement, and discloses a device and a method for detecting the parallelism of a mechanical shaft of a coaxial double-shaft turntable, wherein the method for detecting the parallelism of the mechanical shaft of the coaxial double-shaft turntable comprises the following steps: the emitting light source is arranged at the center of the A-axis surface; rotating the B axis clockwise n times to generate a first projection image sequence; synchronously rotating the A axis and the B axis clockwise for n times to generate a second projection image sequence; rotating the A axis counterclockwise n times to generate a third projection image sequence; inputting the first projection image sequence, the second projection image sequence and the third projection image sequence into a neural network model, and outputting a value representing the included angle between the B axis and the reference axis and a value representing the included angle between the A axis and the reference axis; the invention can accurately judge the parallelism of the A axis and the B axis of the coaxial double-shaft turntable, acquire the deviation angle value and assist the A axis and the B axis to be adjusted to be horizontal.

Description

Device and method for detecting parallelism of mechanical shaft of coaxial double-shaft turntable
Technical Field
The invention relates to the technical field of parallelism measurement, in particular to a device and a method for detecting the mechanical axis parallelism of a coaxial double-shaft turntable.
Background
The rotating parallelism of the mechanical shaft of the double-shaft turntable affects the mechanical service life and the performance of applied equipment, the grating of the existing laser measurement method is narrow in width and short in length, the whole shaft surface cannot be covered under the condition of the shaft surface of a large plane, the whole plane is observed based on a local plane, and the error is large.
The method comprises the steps of measuring the rotation parallelism of a mechanical axis of a double-shaft turntable by using a projection method, defining two mechanical axes of the mechanical axis of the double-shaft turntable as an A axis and a B axis, setting a rectangular diaphragm on the A axis surface, carrying out central projection on a projection light source to the B axis surface through the rectangular diaphragm to generate a projection rectangle, calculating the difference value between the side length of the projection rectangle under the assumption that the A axis surface is parallel to the B axis surface and the side length of the actual projection rectangle according to the parameters such as the incidence angle of the projection light source, the space angle of the rectangular diaphragm, the central distance of the A axis and the central distance of the B axis, carrying out multiple measurements by rotating the B axis, calculating the average value according to the multiple calculated difference values, and judging that the rotation parallelism does not meet the condition if the average value is larger than a set threshold value. However, if the axis a is parallel to the axis B, the difference of projection calculation will be generated similarly if the axis a is not parallel to the axis B, and misjudgment of the parallelism of the axis a and the axis B will be caused according to the method.
Disclosure of Invention
The invention provides a method for detecting the parallelism of a mechanical shaft of a coaxial double-shaft turntable, which solves the technical problem that a common diaphragm projection method in the related art can cause misjudgment of the parallelism of an A shaft and a B shaft.
The invention provides a method for detecting the parallelism of a mechanical shaft of a coaxial double-shaft turntable, which comprises the following steps:
s101, defining two mechanical axes of a coaxial double-shaft turntable as an A axis and a B axis, arranging an emission light source at the center of a plane of the A axis, arranging a diaphragm between the A axis and the B axis, wherein the plane of the diaphragm is parallel to the plane of the A axis, and the vertical distance between the diaphragm and the plane of the A axis is H;
step S102, rotating the B axis clockwise for n times, wherein the rotation angle of the B axis rotating once is 360 degrees/n, and acquiring an image of the B axis surface once after each rotation is finished, so as to generate a first projection image sequence;
step S103, rotating the A axis and the B axis for n times in a clockwise synchronous manner, wherein the rotation angle of one rotation is 360 degrees/n, and acquiring an image of the B axis surface once after each rotation is finished, so as to generate a second projection image sequence;
step S104, arranging an emission light source at the center of the B axis surface, arranging a diaphragm between the A axis and the B axis, wherein the plane of the diaphragm is parallel to the plane of the B axis, and the vertical distance between the diaphragm and the B axis surface is H;
step S105, rotating the A axis anticlockwise for n times, wherein the rotation angle of the A axis for one time is 360 degrees/n, and acquiring an image of the A axis surface once after each rotation is finished, so as to generate a third projection image sequence;
step S106, inputting the first projection image sequence, the second projection image sequence and the third projection image sequence into a neural network model, and outputting a value representing the included angle between the B axis and the reference axis and a value representing the included angle between the A axis and the reference axis.
Further, the value of H needs to satisfy that the projection of the diaphragm on the B-axis surface does not exceed the range of the B-axis surface.
Further, the first sequence of projection images is expressed as:wherein->Projection image representing the B-axis plane at the first angle of rotation of the B-axis, +.>A projection image of the B-axis surface at the nth angle of rotation of the B-axis is shown.
Further, the methodThe second projection image sequence is expressed as:wherein->Projection image representing the B-axis plane when the first angle of the a-axis and the B-axis is rotated synchronously,/for the first angle>A projection image of the B-axis surface when the n-th angles of the a-axis and the B-axis are rotated synchronously is shown.
Further, the third sequence of projection images is expressed as:wherein->Projection image representing the A-axis plane at a first angle of rotation of the A-axis, +.>A projection image of the a-axis surface at the nth angle of rotation of the a-axis is shown.
Further, the diaphragm is rectangular or isosceles trapezoid in shape.
Further, the images in the projected image sequence are subjected to binarization processing, the gray value of the background is 0, and the gray value of the projection is 255.
Further, the neural network model includes: the first hiding layer, the second hiding layer, the first full-connection layer and the second full-connection layer;
the first hidden layer includes: a first combiner, a second combiner, a first convolution layer, a second convolution layer;
the first combiner of the first hidden layer is used for combining the projection images corresponding to the sequence units of the first projection image sequence and the second projection image sequence to obtain a fourth projection image sequence;
the fourth projection image sequence is expressed as:wherein->Representing a projection image of the B-axis plane at the first angle of rotation of the B-axis and a projection image of the B-axis plane at the first angle of simultaneous rotation of the A-axis and the B-axis, < >>A projection image of the B-axis surface at the nth angle of rotation of the B-axis and a projection image of the B-axis surface at the nth angle of synchronous rotation of the a-axis and the B-axis are shown;
the second combiner of the first hidden layer is used for combining the projection images corresponding to the sequence units of the second projection image sequence and the third projection image sequence to obtain a fifth projection image sequence;
the fifth projection image sequence is expressed as:wherein->Representing a projection image of the B-axis plane when the first angle of the A-axis and the B-axis is rotated simultaneously with a projection image of the A-axis plane when the first angle of the A-axis is rotated, < >>A projection image of the B-axis surface when the n-th angles of the a-axis and the B-axis are rotated simultaneously with a projection image of the a-axis surface when the n-th angles of the a-axis are rotated;
2 channels of a first convolution layer of a t first hidden layer respectively input a projection image of a B axis surface when rotating a t angle of a B axis corresponding to a t sequence unit of a fourth projection image sequence and a projection image of a B axis surface when synchronously rotating the t angles of an A axis and a B axis, and output a t first feature vector;
2 channels of a second convolution layer of a t first hidden layer respectively input a projection image of a B axis surface when the t angle of an A axis and a B axis is synchronously rotated corresponding to a t sequence unit of a fifth projection image sequence and a projection image of an A axis surface when the t angle of the A axis is rotated, and output a t second feature vector;
the t-th LSTM unit of the second hidden layer inputs the t-th first feature vector and the t-th second feature vector respectively, the n-th LSTM unit of the second hidden layer inputs the first full-connection layer and the second full-connection layer, the first full-connection layer outputs a value representing an included angle between the B axis and the reference axis, and the second full-connection layer outputs a value representing an included angle between the A axis and the reference axis.
Further, the axes of the shaft holes of the shaft seat of the A shaft and the shaft seat of the B shaft serve as reference shafts; the method for adjusting the A axis and the B axis comprises the following steps: the flange of the A shaft is arranged to adjust in the direction vertical to the plane of the platform when the A shaft is adjusted, and the flange of the B shaft is arranged to adjust in the direction vertical to the plane of the platform when the B shaft is adjusted;
when the included angle between the A axis and the reference axis is defined to be a negative value, the end, close to the platform, of the A axis is adjusted to swing downwards, and when the included angle is a positive value, the end, close to the platform, of the A axis is adjusted to swing upwards;
and when the included angle between the B axis and the reference axis is defined to be negative, the end, close to the platform, of the B axis is adjusted to swing downwards, and when the included angle is positive, the end, close to the platform, of the A axis is adjusted to swing upwards.
Further, the invention provides a device for detecting the parallelism of a mechanical shaft of a coaxial double-shaft turntable, which comprises: the first projection image sequence generation module is used for setting an emission light source at the center of an A-axis surface, rotating a B-axis clockwise for n times, and acquiring an image of the B-axis surface once after each rotation is completed, so as to generate a first projection image sequence;
the second projection image sequence generating module is used for setting the emitting light source at the center of the A-axis surface, rotating the A-axis and the B-axis for n times in a clockwise synchronous manner, acquiring an image of the B-axis surface once after each rotation is completed, and generating a second projection image sequence;
the third projection image sequence generating module is used for setting the emitting light source at the center of the B axis surface, rotating the A axis anticlockwise for n times, acquiring an image of the A axis surface once after each rotation is finished, and generating a third projection image sequence;
and the prediction module is used for inputting the first projection image sequence, the second projection image sequence and the third projection image sequence into the neural network model and outputting a value representing the included angle between the B axis and the reference axis and a value representing the included angle between the A axis and the reference axis.
The invention has the beneficial effects that: the parallelism of the A axis and the B axis of the coaxial double-shaft turntable can be accurately judged, the deviation angle value is obtained, and the A axis and the B axis are assisted to be adjusted to be horizontal.
Drawings
FIG. 1 is a flow chart of a method for detecting mechanical axis parallelism of a coaxial biaxial turntable of the present invention;
FIG. 2 is a schematic view of a biaxial turntable of the present invention;
FIG. 3 is a schematic diagram of a biaxial turntable of the present invention;
fig. 4 is a schematic block diagram of a mechanical axis parallelism detecting device of a coaxial biaxial turntable of the present invention.
Detailed Description
The subject matter described herein will now be discussed with reference to example embodiments. It is to be understood that these embodiments are merely discussed so that those skilled in the art may better understand and implement the subject matter described herein and that changes may be made in the function and arrangement of the elements discussed without departing from the scope of the disclosure herein. Various examples may omit, replace, or add various procedures or components as desired. In addition, features described with respect to some examples may be combined in other examples as well.
As shown in fig. 1-3, a method for detecting the parallelism of a mechanical shaft of a coaxial double-shaft turntable comprises the following steps:
s101, defining two mechanical axes of a coaxial double-shaft turntable as an A axis and a B axis, arranging an emission light source at the center of a circle of an A axis surface, arranging a diaphragm in the middle of the A axis and the B axis, wherein the plane of the diaphragm is parallel to the plane of the A axis, and the vertical distance between the diaphragm and the A axis surface is H (the arrangement of H needs to meet the condition that the projection of the diaphragm on the B axis surface does not exceed the range of the B axis surface);
step S102, rotating the B axis clockwise for n times, wherein the rotation angle of the B axis rotating once is 360 degrees/n, and acquiring an image of the B axis surface once after each rotation is finished, so as to generate a first projection image sequence;
the first sequence of projection images is represented as:wherein->Projection image representing the B-axis plane at the first angle of rotation of the B-axis, +.>A projection image of the B-axis surface at the nth angle of rotation of the B-axis;
step S103, rotating the A axis and the B axis for n times in a clockwise synchronous manner, wherein the rotation angle of one rotation is 360 degrees/n, and acquiring an image of the B axis surface once after each rotation is finished, so as to generate a second projection image sequence;
the second projection image sequence is expressed as:wherein->Projection image representing the B-axis plane when the first angle of the a-axis and the B-axis is rotated synchronously,/for the first angle>A projection image showing the B-axis surface when the n-th angles of the a-axis and the B-axis are rotated synchronously;
step S104, arranging an emission light source at the center of the B axis surface, arranging a diaphragm between the A axis and the B axis, wherein the plane of the diaphragm is parallel to the plane of the B axis, and the vertical distance between the diaphragm and the B axis surface is H;
step S105, rotating the A axis anticlockwise for n times, wherein the rotation angle of the A axis for one time is 360 degrees/n, and acquiring an image of the A axis surface once after each rotation is finished, so as to generate a third projection image sequence;
the third projection image sequence is expressed as:wherein->Projection image representing the A-axis plane at a first angle of rotation of the A-axis, +.>A projection image of the a-axis surface at the nth angle of rotation of the a-axis;
in the above-described embodiments, the diaphragm has a rectangular or isosceles trapezoid shape.
In the above embodiment, the unit for capturing an image is disposed at the center of the diaphragm, and the shooting direction is consistent with the line connecting the light source and the center of the diaphragm.
Further, the acquired image is a gray image, and if a non-gray image is acquired, gray processing is needed to regenerate the projection image sequence.
Further, under the condition that the axial planes of the A axis and the B axis are not parallel, the parallelism of the B axis can be reflected by simply observing the change between the adjacent projection images of the first projection sequence, and the parallelism of the A axis can be reflected by simply observing the change between the adjacent projection images of the second projection sequence; however, if the axes of the a axis and the B axis are not parallel, such observation results are greatly affected, and the resulting error increases. Therefore, a mode is added to acquire a third projection image sequence, so that the parallelism information of an A axis and a B axis and the axis plane parallelism information can be acquired from the three projection image sequences through joint observation, interference of non-parallel axis planes is eliminated through information subtraction, and the work is difficult to be completed through general calculation; the parallelism information reflected by the projection of the projection image sequence is extracted through training and learning of the neural network model constructed according to the observation requirements, interference of nonparallel axial planes is eliminated, and a result accurately representing the included angle is output.
The projection changes when the axes of the A axis and the B axis are not parallel, but the projection changes have different performances from the projection changes caused by the parallelism of the A axis and the B axis.
In one embodiment of the invention, the images in the sequence of projection images are all background-free, i.e. only the axial plane of the a-axis or the B-axis and the projections on the axial plane remain in the image.
In one embodiment of the invention, the images in the projected image sequence are binarized, the gray value of the background is 0, and the gray value of the projection is 255.
Step S106, inputting the first projection image sequence, the second projection image sequence and the third projection image sequence into a neural network model, and outputting a value representing the included angle between the B axis and the reference axis and a value representing the included angle between the A axis and the reference axis;
the neural network model includes: the first hiding layer, the second hiding layer, the first full-connection layer and the second full-connection layer;
the first hidden layer includes: a first combiner, a second combiner, a first convolution layer, a second convolution layer;
the first combiner of the first hidden layer is used for combining the projection images corresponding to the sequence units of the first projection image sequence and the second projection image sequence to obtain a fourth projection image sequence;
the fourth projection image sequence is expressed as:wherein->Representing a projection image of the B-axis plane at the first angle of rotation of the B-axis and a projection image of the B-axis plane at the first angle of simultaneous rotation of the A-axis and the B-axis, < >>A projection image of the B-axis surface at the nth angle of rotation of the B-axis and a projection image of the B-axis surface at the nth angle of synchronous rotation of the a-axis and the B-axis are shown;
the second combiner of the first hidden layer is used for combining the projection images corresponding to the sequence units of the second projection image sequence and the third projection image sequence to obtain a fifth projection image sequence;
the fifth projection image sequence is expressed as：Wherein->Representing a projection image of the B-axis plane when the first angle of the A-axis and the B-axis is rotated simultaneously with a projection image of the A-axis plane when the first angle of the A-axis is rotated, < >>A projection image of the B-axis surface when the n-th angles of the a-axis and the B-axis are rotated simultaneously with a projection image of the a-axis surface when the n-th angles of the a-axis are rotated;
2 channels of a first convolution layer of a t first hidden layer respectively input a projection image of a B axis surface when rotating a t angle of a B axis corresponding to a t sequence unit of a fourth projection image sequence and a projection image of a B axis surface when synchronously rotating the t angles of an A axis and a B axis, and output a t first feature vector;
2 channels of a second convolution layer of a t first hidden layer respectively input a projection image of a B axis surface when the t angle of an A axis and a B axis is synchronously rotated corresponding to a t sequence unit of a fifth projection image sequence and a projection image of an A axis surface when the t angle of the A axis is rotated, and output a t second feature vector;
the t LSTM unit of the second hiding layer inputs a t first feature vector and a t second feature vector respectively, the n LSTM unit of the second hiding layer inputs a first full-connection layer and a second full-connection layer, the first full-connection layer outputs a value representing an included angle between the B axis and the reference axis, and the second full-connection layer outputs a value representing an included angle between the A axis and the reference axis;
in one embodiment of the present invention, the first full connection layer outputs a probability value corresponding to a first class label expressed as:wherein->Respectively represent B1~m discrete values of the included angle between the axis and the reference axis, the second full-connection layer outputs a probability value corresponding to a second class label, and the second class label is expressed as:wherein->1~m discrete values respectively representing the included angles of the A axis and the reference axis;
in one embodiment of the present invention, the axes of the shaft holes of the shaft seats of the a-shaft and the B-shaft are used as reference shafts;
in the specific adjusting method, because the offset of the A axis and the B axis is caused by the movement of the platform and the gravity center of the platform is deviated to the bottom surface or the top surface of the platform, the offset of the A axis and the B axis is along the direction vertical to the surface of the platform, the A axis and the B axis are firstly rotated to zero position before the adjustment, and when the A axis and the B axis are rotated to zero position, if the offset does not exist, the top surface of the platform is parallel to the horizontal plane;
the flange of the A shaft is arranged to adjust in the direction vertical to the plane of the platform when the A shaft is adjusted, and the flange of the B shaft is arranged to adjust in the direction vertical to the plane of the platform when the B shaft is adjusted;
when the included angle between the A axis and the reference axis is defined to be a negative value, the end, close to the platform, of the A axis is adjusted to swing downwards, and when the included angle is a positive value, the end, close to the platform, of the A axis is adjusted to swing upwards;
and when the included angle between the B axis and the reference axis is defined to be negative, the end, close to the platform, of the B axis is adjusted to swing downwards, and when the included angle is positive, the end, close to the platform, of the A axis is adjusted to swing upwards.
The second hidden layer includes n LSTM cells connected in series.
The a-axis surface refers to one end surface of the a-axis opposite to the B-axis, and the B-axis surface refers to one end surface of the B-axis opposite to the a-axis.
As shown in fig. 4, in one embodiment of the present invention, there is provided a mechanical axis parallelism detecting apparatus of a coaxial biaxial turntable, comprising: the first projection image sequence generating module 201 is configured to set the emission light source at the center of the a-axis surface, rotate the B-axis n times clockwise, rotate the B-axis once by 360 °/n, and acquire an image of the B-axis surface once after each rotation is completed, so as to generate a first projection image sequence;
the second projection image sequence generating module 202 is configured to set the emission light source at the center of the a-axis surface, rotate the a-axis and the B-axis n times in a clockwise synchronous manner, and acquire an image of the B-axis surface once after each rotation is completed, so as to generate a second projection image sequence;
the third projection image sequence generating module 203 is configured to set the emission light source at the center of the B-axis surface, rotate the a-axis n times counterclockwise, rotate the a-axis once by 360 °/n, and acquire an image of the a-axis surface once after each rotation is completed, so as to generate a third projection image sequence;
the prediction module 204 is configured to input the first projection image sequence, the second projection image sequence, and the third projection image sequence into the neural network model, and output a value representing an angle between the B axis and the reference axis and a value representing an angle between the a axis and the reference axis.
The embodiment has been described above with reference to the embodiment, but the embodiment is not limited to the above-described specific implementation, which is only illustrative and not restrictive, and many forms can be made by those of ordinary skill in the art, given the benefit of this disclosure, are within the scope of this embodiment.

Claims (10)

1. The method for detecting the mechanical axis parallelism of the coaxial double-shaft turntable is characterized by comprising the following steps of:
s101, defining two mechanical axes of a coaxial double-shaft turntable as an A axis and a B axis, arranging an emission light source at the center of a plane of the A axis, arranging a diaphragm between the A axis and the B axis, wherein the plane of the diaphragm is parallel to the plane of the A axis, and the vertical distance between the diaphragm and the plane of the A axis is H;
step S102, rotating the B axis clockwise for n times, wherein the rotation angle of the B axis rotating once is 360 degrees/n, and acquiring an image of the B axis surface once after each rotation is finished, so as to generate a first projection image sequence;
step S103, rotating the A axis and the B axis for n times in a clockwise synchronous manner, wherein the rotation angle of one rotation is 360 degrees/n, and acquiring an image of the B axis surface once after each rotation is finished, so as to generate a second projection image sequence;
step S104, arranging an emission light source at the center of the B axis surface, arranging a diaphragm between the A axis and the B axis, wherein the plane of the diaphragm is parallel to the plane of the B axis, and the vertical distance between the diaphragm and the B axis surface is H;
step S105, rotating the A axis anticlockwise for n times, wherein the rotation angle of the A axis for one time is 360 degrees/n, and acquiring an image of the A axis surface once after each rotation is finished, so as to generate a third projection image sequence;
step S106, inputting the first projection image sequence, the second projection image sequence and the third projection image sequence into a neural network model, and outputting a value representing the included angle between the B axis and the reference axis and a value representing the included angle between the A axis and the reference axis.
2. The method for detecting the mechanical axis parallelism of the coaxial double-shaft turntable according to claim 1, wherein the value of H is required to meet the requirement that the projection of the diaphragm on the B-axis surface does not exceed the range of the B-axis surface.
3. The method for detecting mechanical axis parallelism of a coaxial biaxial turntable according to claim 1, wherein the first projection image sequence is expressed as:wherein->Projection image representing the B-axis plane at the first angle of rotation of the B-axis, +.>A projection image of the B-axis surface at the nth angle of rotation of the B-axis is shown.
4. The method for detecting mechanical axis parallelism of a coaxial biaxial turntable according to claim 1, wherein the second projection image sequence is expressed as:wherein->Projection image representing the B-axis plane when the first angle of the a-axis and the B-axis is rotated synchronously,/for the first angle>A projection image of the B-axis surface when the n-th angles of the a-axis and the B-axis are rotated synchronously is shown.
5. The method for detecting mechanical axis parallelism of a coaxial biaxial turntable according to claim 1, wherein the third projection image sequence is expressed as:wherein->Projection image representing the A-axis plane at a first angle of rotation of the A-axis, +.>A projection image of the a-axis surface at the nth angle of rotation of the a-axis is shown.
6. The method for detecting the mechanical axis parallelism of the coaxial double-shaft turntable according to claim 1, wherein the diaphragm is rectangular or isosceles trapezoid in shape.
7. The method for detecting the mechanical axis parallelism of the coaxial biaxial turntable according to claim 1, wherein the images in the projected image sequence are binarized, the gray value of the background is 0, and the gray value of the projection is 255.
8. The method for detecting the mechanical axis parallelism of the coaxial double-axis turntable according to claim 1, wherein the neural network model comprises: the first hiding layer, the second hiding layer, the first full-connection layer and the second full-connection layer;
the first hidden layer includes: a first combiner, a second combiner, a first convolution layer, a second convolution layer;
the first combiner of the first hidden layer is used for combining the projection images corresponding to the sequence units of the first projection image sequence and the second projection image sequence to obtain a fourth projection image sequence;
the fourth projection image sequence is expressed as:wherein->Representing a projection image of the B-axis plane at the first angle of rotation of the B-axis and a projection image of the B-axis plane at the first angle of simultaneous rotation of the A-axis and the B-axis, < >>A projection image of the B-axis surface at the nth angle of rotation of the B-axis and a projection image of the B-axis surface at the nth angle of synchronous rotation of the a-axis and the B-axis are shown;
the second combiner of the first hidden layer is used for combining the projection images corresponding to the sequence units of the second projection image sequence and the third projection image sequence to obtain a fifth projection image sequence;
the fifth projection image sequence is expressed as:wherein->Representing a projection image of the B-axis plane when the first angle of the A-axis and the B-axis is rotated simultaneously with a projection image of the A-axis plane when the first angle of the A-axis is rotated, < >>B axis representing the nth angle of synchronous rotation of A and B axesA projection image of the plane and a projection image of the a-axis plane at the nth angle of rotation of the a-axis;
2 channels of a first convolution layer of a t first hidden layer respectively input a projection image of a B axis surface when rotating a t angle of a B axis corresponding to a t sequence unit of a fourth projection image sequence and a projection image of a B axis surface when synchronously rotating the t angles of an A axis and a B axis, and output a t first feature vector;
2 channels of a second convolution layer of a t first hidden layer respectively input a projection image of a B axis surface when the t angle of an A axis and a B axis is synchronously rotated corresponding to a t sequence unit of a fifth projection image sequence and a projection image of an A axis surface when the t angle of the A axis is rotated, and output a t second feature vector;
the t-th LSTM unit of the second hidden layer inputs the t-th first feature vector and the t-th second feature vector respectively, the n-th LSTM unit of the second hidden layer inputs the first full-connection layer and the second full-connection layer, the first full-connection layer outputs a value representing an included angle between the B axis and the reference axis, and the second full-connection layer outputs a value representing an included angle between the A axis and the reference axis.
9. The method for detecting the parallelism of mechanical shafts of a coaxial biaxial turntable according to claim 1, wherein the axes of the shaft holes of the shaft seat of the a-shaft and the shaft seat of the B-shaft are used as reference shafts;
the method for adjusting the A axis and the B axis comprises the following steps: the flange of the A shaft is arranged to adjust in the direction vertical to the plane of the platform when the A shaft is adjusted, and the flange of the B shaft is arranged to adjust in the direction vertical to the plane of the platform when the B shaft is adjusted;
when the included angle between the A axis and the reference axis is defined to be a negative value, the end, close to the platform, of the A axis is adjusted to swing downwards, and when the included angle is a positive value, the end, close to the platform, of the A axis is adjusted to swing upwards;
and when the included angle between the B axis and the reference axis is defined to be negative, the end, close to the platform, of the B axis is adjusted to swing downwards, and when the included angle is positive, the end, close to the platform, of the A axis is adjusted to swing upwards.
10. The utility model provides a coaxial biax revolving stage mechanical axis depth of parallelism detection device which characterized in that includes: the first projection image sequence generation module is used for setting an emission light source at the center of an A-axis surface, rotating a B-axis clockwise for n times, and acquiring an image of the B-axis surface once after each rotation is completed, so as to generate a first projection image sequence;
the second projection image sequence generating module is used for setting the emitting light source at the center of the A-axis surface, rotating the A-axis and the B-axis for n times in a clockwise synchronous manner, acquiring an image of the B-axis surface once after each rotation is completed, and generating a second projection image sequence;
the third projection image sequence generating module is used for setting the emitting light source at the center of the B axis surface, rotating the A axis anticlockwise for n times, acquiring an image of the A axis surface once after each rotation is finished, and generating a third projection image sequence;
and the prediction module is used for inputting the first projection image sequence, the second projection image sequence and the third projection image sequence into the neural network model and outputting a value representing the included angle between the B axis and the reference axis and a value representing the included angle between the A axis and the reference axis.
CN202311338094.9A 2023-10-17 2023-10-17 Device and method for detecting parallelism of mechanical shaft of coaxial double-shaft turntable Active CN117073586B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311338094.9A CN117073586B (en) 2023-10-17 2023-10-17 Device and method for detecting parallelism of mechanical shaft of coaxial double-shaft turntable

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311338094.9A CN117073586B (en) 2023-10-17 2023-10-17 Device and method for detecting parallelism of mechanical shaft of coaxial double-shaft turntable

Publications (2)

Publication Number Publication Date
CN117073586A true CN117073586A (en) 2023-11-17
CN117073586B CN117073586B (en) 2024-01-09

Family

ID=88706494

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311338094.9A Active CN117073586B (en) 2023-10-17 2023-10-17 Device and method for detecting parallelism of mechanical shaft of coaxial double-shaft turntable

Country Status (1)

Country Link
CN (1) CN117073586B (en)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS5983116A (en) * 1982-11-02 1984-05-14 Olympus Optical Co Ltd Adjustment device for optical system
EP0270768A2 (en) * 1986-12-10 1988-06-15 Dr. Johannes Heidenhain GmbH Incrementor absolute revolution sensor with clamping means
EP0419916A2 (en) * 1989-09-26 1991-04-03 Krauss-Maffei Aktiengesellschaft Device to examine the parallelism of two axes
JP2001144690A (en) * 1999-11-15 2001-05-25 Sony Corp Optical axis correction device and adjustment method for adjustment mechanism section
CN205718879U (en) * 2016-06-27 2016-11-23 上海嘉强自动化技术有限公司 A kind of detection correction beam direction alignment high accuracy quasistatic system
CN108151672A (en) * 2017-12-28 2018-06-12 华中科技大学 A kind of axis geometric tolerance measuring instrument based on projection imaging
CN211234319U (en) * 2019-12-18 2020-08-11 西安理工大学 Three-dimensional axis error detector based on image
CN115553794A (en) * 2022-10-31 2023-01-03 中山大学 Three-dimensional Compton scattering imaging method and system based on parallel hole collimator
WO2023024658A1 (en) * 2021-08-23 2023-03-02 苏州大学 Deep video linkage feature-based behavior recognition method
CN116124070A (en) * 2023-04-04 2023-05-16 青岛迈朗格智能制造有限公司 Quality test bench for vehicle spare and accessory parts

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS5983116A (en) * 1982-11-02 1984-05-14 Olympus Optical Co Ltd Adjustment device for optical system
EP0270768A2 (en) * 1986-12-10 1988-06-15 Dr. Johannes Heidenhain GmbH Incrementor absolute revolution sensor with clamping means
EP0419916A2 (en) * 1989-09-26 1991-04-03 Krauss-Maffei Aktiengesellschaft Device to examine the parallelism of two axes
JP2001144690A (en) * 1999-11-15 2001-05-25 Sony Corp Optical axis correction device and adjustment method for adjustment mechanism section
CN205718879U (en) * 2016-06-27 2016-11-23 上海嘉强自动化技术有限公司 A kind of detection correction beam direction alignment high accuracy quasistatic system
CN108151672A (en) * 2017-12-28 2018-06-12 华中科技大学 A kind of axis geometric tolerance measuring instrument based on projection imaging
CN211234319U (en) * 2019-12-18 2020-08-11 西安理工大学 Three-dimensional axis error detector based on image
WO2023024658A1 (en) * 2021-08-23 2023-03-02 苏州大学 Deep video linkage feature-based behavior recognition method
CN115553794A (en) * 2022-10-31 2023-01-03 中山大学 Three-dimensional Compton scattering imaging method and system based on parallel hole collimator
CN116124070A (en) * 2023-04-04 2023-05-16 青岛迈朗格智能制造有限公司 Quality test bench for vehicle spare and accessory parts

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
陈平;刘文异;宋骐羽;张丽娜;姜云翔;: "基于光学自准直原理的喷管角度偏差测量方法研究", 宇航计测技术, no. 04 *

Also Published As

Publication number Publication date
CN117073586B (en) 2024-01-09

Similar Documents

Publication Publication Date Title
CN105092607B (en) Spherical optics element surface flaw evaluation method
US11244436B2 (en) Method for inspecting mounting state of component, printed circuit board inspection apparatus, and computer readable recording medium
CN109031654A (en) A kind of adaptive optics bearing calibration and system based on convolutional neural networks
JP7028333B2 (en) Lighting condition setting method, equipment, system and program, and storage medium
US11715190B2 (en) Inspection system, image discrimination system, discrimination system, discriminator generation system, and learning data generation device
CN101889196B (en) Feature analyzing apparatus
CN105303615A (en) Combination method of two-dimensional stitching and three-dimensional surface reconstruction of image
CN104537707A (en) Image space type stereo vision on-line movement real-time measurement system
CN108961184A (en) A kind of bearing calibration of depth image, device and equipment
JP2021515885A (en) Methods, devices, systems and programs for setting lighting conditions and storage media
US10755477B2 (en) Real-time face 3D reconstruction system and method on mobile device
JP2020060879A (en) Learning device, image generator, method for learning, and learning program
CN110827360B (en) Photometric stereo measurement system and method for calibrating light source direction thereof
CN116625258A (en) Chain spacing measuring system and chain spacing measuring method
CN117073586B (en) Device and method for detecting parallelism of mechanical shaft of coaxial double-shaft turntable
CN111738971B (en) Circuit board stereoscopic scanning detection method based on line laser binocular stereoscopic vision
CN115187612A (en) Plane area measuring method, device and system based on machine vision
CN107833210B (en) Appearance detection method and system based on neural network
Legarda et al. Comparison and error analysis of the standard pin-hole and Scheimpflug camera calibration models
KR102148817B1 (en) System and method for detecting bad contact lens
CN114241059B (en) Synchronous calibration method for camera and light source in photometric stereo vision system
CN112782201B (en) Lobster eye optical device calibration method based on X-ray focusing image
CN116888620A (en) Deep learning based image enhancement for incremental manufacturing
JP4806767B2 (en) Strain measuring method and apparatus
CN112884744A (en) Detection method and device, detection equipment and storage medium

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
GR01 Patent grant