CN220760124U - Turbine rotor defect automatic detection system based on vision - Google Patents

Abstract

The utility model discloses a vision-based automatic detection system for defects of a turbine rotor, wherein a feeding conveying device is arranged on a rack; the first vision component is arranged above the feeding conveying device; the first manipulator is used for transferring the turbine rotor and is positioned at one side of the feeding conveying device; a first movement and detection device for detecting the blade and shaft end of the turbine rotor transferred by the first manipulator; the second moving and detecting device is used for detecting the radial dimension of the turbine rotor transferred by the first manipulator and the welding position, and an included angle is formed between the second moving and detecting device and the arrangement direction of the first moving and detecting device; the industrial personal computer is electrically connected with the feeding conveying device, the first visual assembly, the first manipulator, the first moving and detecting device and the second moving and detecting device respectively. The utility model can realize automatic detection of the turbine rotor and automatic sorting of defective products.

Description

Turbine rotor defect automatic detection system based on vision

Technical Field

The utility model relates to the technical field of turbine rotors, in particular to an automatic detection system for defects of a turbine rotor based on vision.

Background

With the continuous increase of the quantity of domestic automobiles, the turbocharger is used as an important part of the small-displacement engine of the automobile, and the market scale of the turbocharger is also continuously expanding. The turbine rotor is the most important of the turbocharger, and is the key for compressing air and improving the power of the automobile.

As shown in fig. 7, the turbine rotor is composed of a turbine and a rotating shaft welded and fixed with the turbine. The turbine in the turbine rotor has the characteristics of complex curved surface, thin wall, torsion, variable cross section and easy deformation, and meanwhile, the processing requirement precision is high, the working environment temperature of the turbine rotor can reach 800-900 ℃, and the defects in the production and processing process can greatly influence the quality of the whole turbocharger.

Defects on turbine rotors are currently detected mainly by manual or semiautomatic means, which have the following drawbacks:

(1) The manual inspection or semi-automatic detection has the advantages of large labor capacity and low efficiency of quality inspection; the working time influences the concentration of workers and the defect judgment quality; the detection quality lacking consistency is unstable in detection efficiency and high in operation cost.

(2) The staff with insufficient experience cannot effectively judge the defects; the manual examination depends on experience, the professional is lack, the culture of the professional skill is slow, and the flow of the personnel is large.

(3) Too fine defects are not detectable by human eyes; the detection method of the fixed mode cannot adapt to new defects, the accuracy and the omission factor have large difference; the defect index and quality inspection index of the product cannot be tracked and evaluated in real time, and the index changes are used for helping to improve the quality; the source cannot be traced to the problem product.

Currently, existing automatic detection technology for defects of a turbine rotor is basically binocular vision measurement, namely, a non-contact camera is adopted to perform global scanning on the defects. Although the measurement accuracy is high, there are still many problems, such as the pursuit of global scanning, the large focal length of the lens, the high cost of the whole set of vision system and the large installation volume; part of the measuring system can only realize high-precision measurement under the condition of small depth of field, otherwise, the precision is greatly reduced; the existing measuring system can only solve one problem independently, and if all defects need to be treated, a plurality of systems are needed to form a production line, so that the investment is large.

Disclosure of Invention

The utility model provides a vision-based automatic detection system for defects of a turbine rotor.

Vision-based turbine rotor defect automatic detection system, including the frame, still include:

the feeding conveying device is arranged on the frame;

the first visual assembly is used for detecting the position of the turbine rotor and is arranged above the feeding conveying device;

the first manipulator is used for transferring the turbine rotor and is positioned at one side of the feeding conveying device;

a first movement and detection device for detecting the blade and shaft end of the turbine rotor transferred by the first manipulator;

the second moving and detecting device is used for detecting the radial dimension of the turbine rotor transferred by the first manipulator and the welding position, and forms an included angle with the arrangement direction of the first moving and detecting device, so that the second moving and detecting device and the first moving and detecting device are distributed on different side parts of the first manipulator;

the industrial personal computer is electrically connected with the feeding conveying device, the first visual assembly, the first manipulator, the first moving and detecting device and the second moving and detecting device respectively.

The utility model has the following characteristics:

(1) The utility model overcomes the defect that the traditional operation mode adopts manual inspection or semi-automatic detection, realizes the rapid measurement of the defects of the turbine rotor by utilizing a multi-station detection visual system and a high-precision module, eliminates the defect of personnel experience judgment, and can realize high-speed, rapid and accurate measurement.

(2) The vision system is formed by overlapping a plurality of sets of cameras, can realize detection of various defects in one set of equipment, and meanwhile, has a reject channel for unqualified products, so that the product detection level is effectively improved.

(3) The clamping mechanism, such as the first clamping rotating mechanism, can be suitable for holding different products in different sizes, realizes flexible production of the production line, and has strong adaptability.

(4) The utility model has the advantages of good capacity expansion due to the increase of the later product detection types, and strong secondary development capability.

(5) The utility model can detect not only the rotor of the turbocharger, but also the miniaturized parts with the shaft blades, including the rotor of the small aviation turbojet engine, the small marine propeller and the like.

Drawings

FIG. 1 is a schematic diagram of a vision-based automatic detection system for a defect in a turbine rotor.

Fig. 2 is a schematic view of a part of the component shown in fig. 1 hidden.

Fig. 3 is a schematic diagram showing the first and second clamping and rotating mechanisms at different detection positions from fig. 2.

Fig. 4 is a perspective view of the first clamping and rotating mechanism.

Fig. 5 is a sectional view of the first clamping and rotating mechanism.

Fig. 6 is a structural view of the second clamping and rotating mechanism.

Fig. 7 is a structural view of a turbine rotor.

The reference symbols in the drawings: the feeding and conveying device comprises a frame 1, a feeding and conveying device 2, a first motor 2a, a first bracket 2b, a belt transmission mechanism 2c, a first vision component 3, a first manipulator 4, an industrial personal computer 5, a first linear driver 6, a first clamping and rotating mechanism 7, a second motor 7a, a first bearing 7b, a first mounting plate 7c, a first fixing seat 7d, an air chuck 7e, a central hole 7f, a second vision component 8, a third vision component 9, a second linear driver 10, a second clamping and rotating mechanism 11, a first moving frame 11a, a third linear driver 11b, a rotating driver 11c, a second moving frame 11d, a center 11e, a second bearing 11f, a fourth vision component 12, a fifth vision component 13, a second manipulator 14, a qualified product blanking mechanism 15 and a defective product blanking mechanism 16.

Detailed Description

The present utility model will be described in detail with reference to the accompanying drawings.

As shown in fig. 1 to 6, the vision-based automatic detection system for the defects of the turbine rotor of the present utility model comprises a frame 1, a feeding and conveying device 2, a first vision component 3, a first manipulator 4, a first moving and detecting device, a second moving and detecting device, and an industrial personal computer 5, wherein the industrial personal computer 5 is respectively and electrically connected with the feeding and conveying device 2, the first vision component 3, the first manipulator 4, the first moving and detecting device, and the second moving and detecting device, and all detection results are recorded in the industrial personal computer 5, and the detection results and the qualification rate are displayed in real time on a display screen on site. The relationship between each part is described in detail below:

the frame 1 is of a frame structure, a top cover is arranged on the frame, dark opaque glass is arranged on the periphery of the frame, a working plane is arranged in the middle of the frame, and a working space is formed by the top cover and the working plane. The industrial personal computer 5 can be hung on the frame 1 and used for displaying the working condition of the system and serving as a product information storage tool.

The feeding conveying device 2 is arranged on the frame 1, the feeding conveying device 2 comprises a first motor 2a, a first bracket 2b and a belt transmission mechanism 2c, the belt transmission mechanism 2c is arranged on the first bracket 2b, and the first motor 2a is connected with the belt transmission mechanism 2 c. The turbine rotor is placed on the belt drive 2c, the first motor 2a is operated, the first motor 2a provides power for the belt drive 2c, the belt in the belt drive 2c is moved, and the turbine rotor on the belt is moved. When an operator places the turbine rotor on the belt transmission mechanism 2c, the turbine rotor is placed along the axial direction of the belt pulley in the belt transmission mechanism 2c as much as possible, so that the turbine rotor is placed at orderly intervals, and the placed turbine rotor is not messy. Because the weights at the two ends of the turbine rotor are inconsistent, under the influence of gravity, when the turbine rotor is positioned on the feeding conveying device 2, the included angle between the axial direction of the turbine rotor and the horizontal direction is smaller than 20 degrees.

The first vision subassembly 3 is used for detecting turbine rotor position, and the first vision subassembly 3 sets up the top at material loading conveyor 2, and first manipulator 4 is used for conveying turbine rotor, and first manipulator 4 is located one side at material loading conveyor 2. The first manipulator 4 is installed on the frame 1, and the first manipulator 4 is linked with the feeding conveying device 2, the first moving and detecting device and the second moving and detecting device, and the first manipulator 4 is a five-axis manipulator, so that grabbing and overturning of the turbine rotor can be realized, and the turbine rotor can be placed at a corresponding position after grabbing.

When the turbine rotor is conveyed to the lower part of the first vision component 3 by the feeding conveying device 2, the first vision component 3 photographs the turbine rotor, and the first vision component 3 adopts an industrial CCD camera and a CMOS camera.

After the first vision component 3 gathers the image of the turbine in the turbine rotor, the first vision component 3 transmits positional information to the industrial computer 5, and the material loading conveyor 2 stops working, at this moment, by the work of the first manipulator 4 of industrial computer 5 control, after the first manipulator 4 obtains the material loading signal, snatch the action, snatch the position and be the biggest outer fringe of turbine rotor blade root, switch the axial of turbine rotor simultaneously in the rotation to high accuracy module 4 in-process, make the axial of turbine rotor switch to vertical direction (the upper and lower direction of frame 1).

The first moving and detecting device detects the blade and shaft end of the turbine rotor transferred by the first manipulator 4, the first moving and detecting device comprises a first linear driver 6, a first clamping and rotating mechanism 7, a second visual component 8 and a third visual component 9, the first linear driver 6 is connected with the first clamping and rotating mechanism 7 to drive the first clamping and rotating mechanism 7 to move, in this embodiment, the first linear driver 6 adopts a high-precision moving module, the high-precision moving module is a high-precision linear moving module, the high-precision moving module is installed on the frame 1, a servo motor controlled by a closed loop is adopted as a driving mode, a high-precision screw rod is installed inside, and in this embodiment, the high-precision module is preferably a T3-level screw rod module or a linear motor.

The first clamping and rotating mechanism 7 comprises a second motor 7a, a first bearing 7b, a first mounting plate 7c, a first fixing seat 7d and a pneumatic chuck 7e, wherein a mounting hole is formed in the first mounting plate 7c, a part of the first bearing 7b is matched in the mounting hole, an output shaft of the second motor 7a penetrates through the first mounting plate 7c and the first bearing 7b and then is fixed with the first fixing seat 7d, the first fixing seat 7d is matched with the other part of the first bearing 7b, the pneumatic chuck 7e is fixed with the first fixing seat 7d, and the second motor 7a or the first mounting plate 7c is connected with the first linear driver 6. Since the air chuck 7e has the function of automatic loosening and clamping, it is applicable to the rotating shafts of the turbine rotors with different diameters.

The first robot 4 transfers the turbine rotor located on the feeding conveyor 2 to the first clamping and rotating mechanism 7, and during the transfer, since the first robot 4 adjusts the axial direction of the turbine rotor to the vertical direction, the first robot 4 inserts the rotation shaft of the turbine rotor into the center hole 7f formed by the air chuck 7e of the first clamping and rotating mechanism 7, and the air chuck 7e clamps the rotation shaft of the turbine rotor. The second motor 7a operates, and the second motor 7a drives the first holder 7d to rotate, and the air chuck 7e rotates along with the first holder 7d, so that the turbine rotor clamped by the air chuck 7e also rotates along with it.

Because the first clamping and rotating mechanism 7 is connected with the high-precision moving module, and the high-precision moving module is controlled by the industrial personal computer 5 to linearly move on the frame 1 to reach a specified detection position, the first clamping and rotating mechanism 7 drives the turbine rotor clamped by the first clamping and rotating mechanism to move to the detection position of the second vision component 8 or the third vision component 9 under the driving of the high-precision moving module for detection by the cooperation of a vision system.

When the first clamping and rotating mechanism 7 moves to the detection position of the second visual component 8, the axial direction of the second visual component 8 is parallel to the axial direction of the first clamping and rotating mechanism 7, and when the first clamping and rotating mechanism 7 moves to the detection position of the third visual component 9, an included angle is formed between the axial direction of the third visual component 9 and the axial direction of the first clamping and rotating mechanism 7.

The first clamping and rotating mechanism 7 has four working positions, the first working position is used for receiving the turbine rotor transferred by the first manipulator 4, the first clamping and rotating mechanism 7 moves from the first working position to the second working position under the driving of the first linear driver 6, the second working position is used for being matched and detected by the first clamping and rotating mechanism 7 and the second vision component 8, the second working position is positioned at the downstream of the first working position, the first clamping and rotating mechanism 7 moves from the second working position to the third working position under the driving of the first linear driver 6, the third working position is positioned at the downstream of the second working position, the first clamping and rotating mechanism 7 and the third vision component 9 are matched and detected, the first clamping and rotating mechanism 7 moves from the third working position to the fourth working position under the driving of the first linear driver 6, the fourth working position is positioned at the downstream of the third working position, the fourth working position is used for waiting for transferring the turbine rotor on the first clamping and rotating mechanism 7 to the second moving and detecting device, the third working position is positioned at the unloading position on the unloading device, and the first clamping and rotating mechanism 7 is reversely moved from the first working position to the fourth working position after the first linear driver 6 is driven to the fourth working position.

When the first linear driver 6 drives the first clamping and rotating mechanism 7 to move and stay at the position where the first clamping and rotating mechanism 7 and the second visual component 8 are matched for detection, the second motor 7a works, the second motor 7a drives the first fixing seat 7d to rotate, the air chuck 7e rotates along with the first fixing seat 7d, so that the turbine rotor clamped by the air chuck 7e also rotates along with the rotation, in the process, the second visual component 8 shoots downwards from the top of the turbine rotor, the second visual component 8 acquires pictures of turbines in the plurality of turbine rotors, the second visual component 8 provides the acquired images for the industrial personal computer 5, and the industrial personal computer judges whether the outer diameter of a turbine rotor blade and the size of the shaft end at the top are qualified or not, and the shooting speed of the second visual component 8 can reach 3 seconds/single chip. After the detection of the detection position of the second vision component 8 is completed, the first linear driver 6 drives the first clamping and rotating mechanism 7 to linearly move to a position matched with the detection of the third vision component 9.

The first linear actuator 6 drives the first clamping and rotating mechanism 7 to move and stay at the position detected by the third vision component 9 of the first clamping and rotating mechanism 7, and the third vision component 9 is in a matched state with the surface of the turbine blade of the turbine rotor. Then the second motor 7a works, the second motor 7a drives the first fixing seat 7d to rotate, the air chuck 7e rotates along with the first fixing seat 7d, so that the turbine rotor clamped by the air chuck 7e also rotates along with the first fixing seat, in the process, the third vision component 9 shoots downwards from the side part of the turbine rotor, the third vision component 9 collects pictures of the surfaces of turbine blades in a plurality of turbine rotors, the third vision component 9 provides the collected pictures to the industrial personal computer 5, the industrial personal computer judges whether the surfaces of the turbine blades have defects such as scratches, fleshy and concave, and the like, the highest shooting speed of a camera of the third vision component 9 is 4 frames/second, and the highest imaging speed of the camera of the third vision component 9 is 40 seconds/single chip.

The second vision component 8 and the third vision component 9 are composed of a special measuring light source and a camera, can accurately identify defect measurement, and can realize image acquisition of the axial top surface, the outer contour, the blade surface and the like of the turbine rotor through matching with the first clamping and rotating mechanism 7. The special measuring light source is a multicolor annular light source and has various adjustable light colors.

When the third vision assembly 9 completes the detection, the first linear driver 6 drives the first clamping and rotating mechanism 7 to move and reach a fourth working position, and in this working position, the air chuck 7e on the first clamping and rotating mechanism 7 is released, and at the same time, the first manipulator 4 grabs the turbine rotor, so that after the turbine rotor is separated from the first clamping and rotating mechanism 7, the first manipulator 4 transfers the turbine rotor to the second movement and detection device, and in the transfer process, the first manipulator 4 adjusts the axial direction of the turbine rotor from the vertical direction to the horizontal direction (the left-right direction of the frame 1).

The second moving and detecting device clamps the turbine rotor transferred by the first manipulator 4, detects the radial dimension of the turbine rotor transferred by the first manipulator 4 and the welding position, and forms an included angle with the arrangement direction of the first moving and detecting device, so that the second moving and detecting device and the first moving and detecting device are distributed on different side parts of the first manipulator 4.

The second moving and detecting device comprises a second linear driver 10, a second clamping and rotating mechanism 11, a fourth visual component 12 and a fifth visual component 13, wherein the second linear driver 10 adopts a high-precision moving module, the high-precision moving module is a high-precision linear moving module, the high-precision moving module is arranged on the frame 1, a servo motor controlled by a closed loop is adopted as a drive, a high-precision screw rod is arranged inside the high-precision moving module, and in the embodiment, the high-precision module is preferably a T3-level screw rod module or a linear motor. The positioning precision of the high-precision module is +/-0.01 mm, and the detection of the horizontal direction of the turbine rotor can be met.

The second linear driver 10 is connected with the second clamping and rotating mechanism 11 to drive the second clamping and rotating mechanism 11 to move, the second linear driver 10 is a high-precision linear module, the second clamping and rotating mechanism 11 comprises a first moving frame 11a, a third linear driver 11b, a rotary driver 11c, a second moving frame 11d, a center 11e and a second bearing 11f, the first moving frame 11a is connected with the second linear driver 10, the second bearing 11f is arranged at one end of the first moving frame 11a, the third linear driver 11b is arranged at the other end of the first moving frame 11a, the third linear driver 11b is a high-precision linear module, the second moving frame 11d is connected with the third linear driver 11b, the rotary driver 11c is fixed with the second moving frame 11d, the rotary driver 11c is preferentially connected with the rotary driver 11c by a motor, and an accommodating space 11g for accommodating a workpiece is formed between the center 11e and the second bearing 11 f.

After the first robot 4 transfers the turbine rotor located on the first holding and rotating mechanism 7 to the accommodating space 11g (during the transfer, the first robot 4 adjusts the axial direction of the turbine rotor from the vertical direction to the horizontal direction), the first robot 4 makes one end of the turbine rotor (the end of the rotating shaft of the turbine rotor) cooperate with the second bearing 11f, the third linear driver 11b operates to drive the second moving frame 11d to move linearly, the rotary driver 11c and the tip 11e follow the second moving frame 11d to move linearly, the tip 11e abuts against the other end of the turbine rotor, the turbine rotor is held between the second bearing 11f and the tip 11e, the rotary driver 11c operates, and the rotary driver 11c drives the turbine rotor to rotate through the tip 11 e.

The fourth vision subassembly 12 is located one side of second centre gripping rotary mechanism 11, and when second centre gripping rotary mechanism 11 moved to the testing position of fourth vision subassembly 12 and stopped, at this moment, second centre gripping rotary mechanism 11 drive turbine rotor is rotatory, gathers turbine rotor's image from the lateral part through fourth vision subassembly 12, and fourth vision subassembly 12 provides industrial computer 5 with the image of gathering, and supply industrial computer 5 to judge whether the radial full-size of turbine rotor is qualified. The photographing speed of the fourth vision module 12 can be up to 5 seconds/piece. The light source of the fourth vision component 12 can be automatically switched to different color system light sources such as white light, red light and the like so as to complete detection of all the sizes.

The fifth visual component 13 is located at the other side of the second clamping and rotating mechanism 11, after the fourth visual component 12 completes detection, the second linear driver 10 drives the second clamping and rotating mechanism 11 to stay when moving to the detection position of the fifth visual component 13, the second clamping and rotating mechanism 11 drives the turbine rotor to rotate, the rotation angle speed is 3-10 degrees/second, images of welding positions of the turbine and the rotating shaft in the turbine rotor are acquired from the other side through the fifth visual component 13, the fifth visual component 13 provides the acquired images to the industrial personal computer 5, and the industrial personal computer 5 judges whether the welding of the turbine and the rotating shaft in the turbine rotor has defects of cracks, too high surfacing and the like. The fifth vision component 13 is a laser measuring instrument, and can detect whether the welding has defects such as cracks, too high build-up welding and the like. The laser measuring instrument uses the COMS lens with ultra-high sensitivity, the detection width is not more than 40mm, and the detection range of the turbine rotor can be met.

The utility model further comprises a second mechanical arm 14, a qualified product blanking mechanism 15 and a defective product blanking mechanism 16, wherein the second mechanical arm 14 is arranged on the first side part of the second moving and detecting device, and the qualified product blanking mechanism 15 and the defective product blanking mechanism 16 are arranged on the second side part of the second moving and detecting device. After the detection of the fifth vision component 13 is completed, if the industrial personal computer 5 judges that the turbine rotor is qualified, the second manipulator 14 is controlled to transfer the turbine rotor clamped on the second clamping and rotating mechanism 11 to the qualified product blanking mechanism 15, and if the turbine rotor is unqualified, the second manipulator 14 is controlled to transfer the turbine rotor clamped on the second clamping and rotating mechanism 11 to the defective product blanking mechanism 16.

Claims (7)

1. Vision-based turbine rotor defect automatic detection system, including frame (1), its characterized in that still includes:

the feeding conveying device (2), the feeding conveying device (2) is arranged on the frame;

the first visual assembly (3) is used for detecting the position of the turbine rotor, and the first visual assembly (3) is arranged above the feeding conveying device (2);

the first manipulator (4) is used for transferring the turbine rotor, and the first manipulator (4) is positioned at one side of the feeding conveying device (2);

a first movement and detection device for detecting the blade and shaft end of the turbine rotor transferred by the first manipulator (4);

the second moving and detecting device is used for detecting the radial dimension of the turbine rotor transferred by the first manipulator (4) and the welding position, and forms an included angle with the arrangement direction of the first moving and detecting device, so that the second moving and detecting device and the first moving and detecting device are distributed on different side parts of the first manipulator (4);

the industrial personal computer (5), the industrial personal computer (5) is respectively connected with the feeding conveying device (2), the first visual component (3), the first manipulator (4), the first moving and detecting device and the second moving and detecting device.

2. The vision-based turbine rotor defect automatic detection system according to claim 1, wherein the feeding and conveying device (2) comprises a first motor (2 a), a first bracket (2 b) and a belt transmission mechanism (2 c), the belt transmission mechanism (2 c) is arranged on the first bracket (2 b), and the first motor (2 a) is connected with the belt transmission mechanism (2 c).

3. The vision-based turbine rotor defect automatic detection system according to claim 1, wherein the first moving and detecting device comprises a first linear driver (6), a first clamping and rotating mechanism (7), a second vision component (8) and a third vision component (9), the first linear driver (6) is connected with the first clamping and rotating mechanism (7) for driving the first clamping and rotating mechanism (7) to move, when the first clamping and rotating mechanism (7) moves to a detection position of the second vision component (8), the axial direction of the second vision component (8) is parallel to the axial direction of the first clamping and rotating mechanism (7), and when the first clamping and rotating mechanism (7) moves to a detection position of the third vision component (9), the axial direction of the third vision component (9) forms an included angle with the axial direction of the first clamping and rotating mechanism (7).

4. A vision-based automatic detection system for defects of a turbine rotor according to claim 3, characterized in that the first clamping and rotating mechanism (7) comprises a second motor (7 a), a first bearing (7 b), a first mounting plate (7 c), a first fixing seat (7 d) and a pneumatic chuck (7 e), wherein a mounting hole is formed in the first mounting plate (7 c), a part of the first bearing (7 b) is matched in the mounting hole, an output shaft of the second motor (7 a) passes through the first mounting plate (7 c) and the first bearing (7 b) and then is fixed with the first fixing seat (7 d), the first fixing seat (7 d) is matched with the other part of the first bearing (7 b), and the pneumatic chuck (7 e) is fixed with the first fixing seat (7 d), and the second motor (7 a) or the first mounting plate (7 c) is connected with the first linear driver (6).

5. The vision-based turbine rotor defect automatic detection system according to claim 1, wherein the second movement and detection device comprises a second linear driver (10), a second clamping and rotation mechanism (11), a fourth vision component (12) and a fifth vision component (13), the second linear driver (10) is connected with the second clamping and rotation mechanism (11) to drive the second clamping and rotation mechanism (11) to move, when the second clamping and rotation mechanism (11) moves to a detection position of the fourth vision component (12), the fourth vision component (12) is positioned on one side of the second clamping and rotation mechanism (11), and when the second clamping and rotation mechanism (11) moves to a detection position of the fifth vision component (13), the fifth vision component (13) is positioned on the other side of the second clamping and rotation mechanism (11).

6. The vision-based turbine rotor defect automatic detection system according to claim 5, wherein the second clamping and rotating mechanism (11) comprises a first moving frame (11 a), a third linear driver (11 b), a rotary driver (11 c), a second moving frame (11 d), a center (11 e) and a second bearing (11 f), the first moving frame (11 a) is connected with the second linear driver (10), the second bearing (11 f) is disposed at one end of the first moving frame (11 a), the third linear driver (11 b) is disposed at the other end of the first moving frame (11 a), the second moving frame (11 d) is connected with the third linear driver (11 b), the rotary driver (11 c) is fixed with the second moving frame (11 d), the center (11 e) is connected with the rotary driver (11 c), and an accommodating space for accommodating a workpiece is formed between the center (11 e) and the second bearing (11 f).

7. The vision-based turbine rotor defect automatic detection system according to any one of claims 1 to 6, further comprising a second manipulator (14), a qualified product blanking mechanism (15), and a defective product blanking mechanism (16), wherein the second manipulator (14) is disposed on a first side of the second moving and detecting device, and the qualified product blanking mechanism (15) and the defective product blanking mechanism (16) are disposed on a second side of the second moving and detecting device.