An AI vision recognition-based intelligent fixing tool for special-shaped part machining

Abstract

This invention discloses an intelligent fixed fixture for processing irregularly shaped parts based on AI visual recognition, belonging to the technical field of auxiliary irregularly shaped part processing devices. It includes a material carrier, a recognition mechanism, and a positioning mechanism. The material carrier is connected to a machine tool. The recognition mechanism includes a placement platform and a smart camera, both fixedly mounted on the side of the material carrier, with the smart camera positioned above it. Two positioning mechanisms are symmetrically arranged on the top of the material carrier. Each positioning mechanism includes a first slide rail, a second slide rail, a movable slide block, a support base, and a positioning pin. Both the first and second slide rails are horizontally slidably connected to the material carrier. The top and bottom of the movable slide block are slidably connected to the first and second slide rails, respectively. This fixture automatically matches the type of irregularly shaped part through AI visual recognition and drives the positioning mechanism to adjust the position of the positioning pin in multiple dimensions, adapting to various irregularly shaped parts with different positioning hole positions without manual fixture replacement.

Description

Technical Field

[0001] This invention relates to the technical field of auxiliary irregular part processing devices, specifically to an intelligent fixed fixture for irregular part processing based on AI visual recognition. Background Technology

[0002] In the fields of aerospace, automobile manufacturing, and medical devices, the demand for processing irregularly shaped parts (such as engine blades, irregularly shaped gearbox housings, and customized orthopedic implants) is increasing. Moreover, these irregularly shaped parts are mostly made from die forgings. Before processing, two positioning holes need to be pre-drilled on the die forgings. The positioning pins of the tooling cooperate with the positioning holes to achieve fixation, providing a stable reference for subsequent machine tool processing.

[0003] Traditional tooling uses locating pin fixtures designed to fix the positions of locating holes for irregularly shaped parts of a specific specification. However, in current manufacturing, different models and specifications of irregularly shaped parts have different pre-drilled locating hole positions, and the thickness and overall dimensions of these parts also vary. When machining irregularly shaped parts with different locating hole positions, it is necessary to manually disassemble the original fixture and replace it with a new, customized fixture. This process is not only time-consuming but also requires the manufacture of a separate fixture for each type of irregularly shaped part, resulting in high tooling manufacturing costs, large inventory pressure, and an inability to meet the needs of multi-variety, small-batch production, severely restricting production efficiency. Summary of the Invention

[0004] This invention provides an intelligent fixing fixture for processing irregularly shaped parts based on AI visual recognition. It can solve the problem in the current manufacturing industry where the positioning hole positions, thicknesses, and overall dimensions of irregularly shaped parts of different models and specifications vary. Traditional fixtures require the design of fixed positioning pins for a single specification of irregularly shaped part, and manual disassembly and replacement of customized fixtures are required when processing different irregularly shaped parts. This process is time-consuming and requires the separate manufacture of multiple fixtures, resulting in high tooling manufacturing costs, inability to meet the needs of multi-variety, small-batch production, and constraints on production efficiency.

[0005] A smart fixture for machining irregularly shaped parts based on AI visual recognition is disclosed. This fixture is mounted on a machine tool and includes a loading platform, a recognition mechanism, and a positioning mechanism. The loading platform is connected to the machine tool. The recognition mechanism includes a placement platform and a smart camera, both fixedly mounted on the side of the loading platform, with the smart camera positioned above the platform. Two sets of positioning mechanisms are symmetrically arranged on the top of the loading platform. Each positioning mechanism includes a first slide rail, a second slide rail, a movable slide block, a support base, and a positioning pin. The first and second slide rails are horizontally slidably connected to the loading platform, and the sliding directions of the first and second slide rails are perpendicular to their projections on the same horizontal plane. The top and bottom of the movable slide block are slidably connected to the first and second slide rails, respectively. The support base is vertically slidably connected to the movable slide block, and the support base has clearance holes. The positioning pin is fixedly connected to the movable slide block and located within the clearance holes.

[0006] According to one embodiment of the present invention, the positioning mechanism further includes a first guide post and a first lead screw. The first guide post is fixedly connected to the loading platform, and the first lead screw is rotatably connected to the loading platform. One end of the first slide rail has a first sliding hole that slides with the first guide post, and the other end of the first slide rail has a first threaded hole that threadedly connects with the first lead screw. The positioning mechanism further includes a first motor, which is fixedly connected to the loading platform, and the output end of the first motor is coaxially fixedly connected to the first lead screw.

[0007] According to one embodiment of the present invention, the positioning mechanism further includes a second guide post and a second lead screw. The second guide post is fixedly connected to the material carrier, and the second lead screw is rotatably connected to the material carrier. One end of the second slide rail has a second sliding hole that slides with the second guide post, and the other end of the second slide rail has a second threaded hole that threadedly connects with the second lead screw. The positioning mechanism also includes a second motor, which is fixedly connected to the material carrier, and the output end of the second motor is coaxially fixedly connected to the second lead screw.

[0008] According to one embodiment of the present invention, the positioning mechanism further includes an electric push rod, the bottom of the movable slide is provided with a slot, the top of the movable slide is provided with a through hole communicating with the slot, the bottom of the support base is slidably connected to the through hole and extends downward into the slot, the electric push rod is fixedly connected to the movable slide, and the telescopic end of the electric push rod is fixedly connected to the bottom of the support base.

[0009] According to one embodiment of the present invention, the machine tool is provided with a guide rail groove, and a mating key embedded in the guide rail groove is fixedly provided at the bottom of the loading table. The identification mechanism also includes a protective cover, which is provided on the outside of the intelligent camera and fixedly connected to the loading table.

[0010] According to one embodiment of the present invention, a slot is provided on the top of the movable slide, and a plug that mates with the slot is fixedly provided on the bottom of the positioning pin. The plug has an external thread on its outer side, and an internal thread that mates with the plug is provided on the inner side of the slot.

[0011] The advantages of this invention compared to the prior art are: Traditional tooling requires customized, fixed locating pin fixtures for the different positioning hole positions of various irregularly shaped parts. Changing workpieces necessitates disassembling and replacing these fixtures, which is time-consuming and labor-intensive. This new tooling, however, uses AI visual recognition to automatically match the type of irregularly shaped part and drives the positioning mechanism to adjust the locating pin position in multiple dimensions. It can adapt to various irregularly shaped parts with different positioning hole positions without manual fixture changes, making it particularly suitable for processing diverse, small-batch irregularly shaped parts. This significantly reduces tooling changeover costs and time. It reduces reliance on human experience, avoids the uncertainties of manual operation, shortens pre-processing preparation time, and improves overall production efficiency.

[0012] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0013] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a three-dimensional structural diagram of an intelligent fixture for processing irregularly shaped parts based on AI visual recognition.

[0014] Figure 2 This is a three-dimensional structural diagram of the identification mechanism in this invention.

[0015] Figure 3 This is a three-dimensional structural diagram of the positioning mechanism in this invention.

[0016] Figure 4 This is a three-dimensional structural cross-sectional view of the positioning mechanism in this invention.

[0017] Figure 5 This is a three-dimensional structural diagram of the movable slide in this invention.

[0018] Figure 6 This is a three-dimensional structural cross-sectional view of the movable slide in this invention.

[0019] The reference numerals in the figures include: 1. Machine tool; 2. Irregularly shaped parts; 3. Carrying platform; 4. Identification mechanism; 5. Positioning mechanism; 6. Material placement platform; 7. Smart camera; 8. First slide rail; 9. Second slide rail; 10. Movable slide; 11. Support base; 12. Positioning pin; 13. Clearance hole; 14. First guide post; 15. First lead screw; 16. First motor; 17. Second guide post; 18. Second lead screw; 19. Second motor; 20. Electric push rod; 21. Guide rail groove; 22. Matching key; 23. Slot; 24. Insert block; 25. Protective cover. Detailed Implementation

[0020] The specific embodiments of the present invention will be described in detail below, but it should be understood that the scope of protection of the present invention is not limited to the specific embodiments.

[0021] like Figures 1 to 6 As shown, an intelligent fixture for processing irregularly shaped parts based on AI visual recognition is mounted on a machine tool 1. The fixture includes a loading platform 3, a recognition mechanism 4, and a positioning mechanism 5. The loading platform 3 is connected to the machine tool 1. The recognition mechanism 4 includes a placement platform 6 and an intelligent camera 7, both fixedly mounted on the side of the loading platform 3, with the intelligent camera 7 positioned above the placement platform 6. Two sets of positioning mechanisms 5 are symmetrically arranged on the top of the loading platform 3. Each positioning mechanism 5 includes a first slide rail 8, a second slide rail 9, and a movable... The movable slide 10, support base 11, and positioning pin 12 are provided. The first slide rail 8 and the second slide rail 9 are both horizontal and slidably connected to the material platform 3. The sliding direction of the first slide rail 8 and the sliding direction of the second slide rail 9 are perpendicular to each other in the same horizontal plane. The top and bottom of the movable slide 10 are slidably connected to the first slide rail 8 and the second slide rail 9, respectively. The support base 11 is vertical and slidably connected to the movable slide 10. The support base 11 has a clearance hole 13. The positioning pin 12 is fixedly connected to the movable slide 10 and is located in the clearance hole 13.

[0022] First, the irregular part 2 (a forging with pre-drilled positioning holes) to be processed is placed on the placement platform 6 on the side of the loading platform 3. At this time, the smart camera 7 fixed above the placement platform 6 starts working, collecting image information such as the appearance outline and positioning hole position of the irregular part 2, and transmitting the data to the background AI algorithm module. The AI ​​algorithm quickly identifies the specific category of the current irregular part 2 through feature matching (comparing the positioning hole position parameters of different irregular parts 2 in the database), and calculates the standard coordinates corresponding to the positioning holes of this type of irregular part 2 (i.e., the target position that the positioning pin 12 needs to reach).

[0023] In one specific embodiment, a smart camera captures the raw image data stream of the irregularly shaped part above the material loading platform. After the ambient stray light interference is attenuated by a protective cover, the analog signal is quantized into a digital pixel matrix by an image acquisition card on the side of the loading platform. This pixel matrix is ​​then filtered by median filtering to eliminate salt-and-pepper noise, and the outline of the irregularly shaped part and the edges of the positioning holes are extracted using the Canny edge detection operator to form a binarized contour map. Based on this, a connected component labeling algorithm is used to separate the positioning hole region from the main body region of the irregularly shaped part, and the centroid coordinates of the positioning hole region are calculated to obtain the initial position of the positioning hole in the pixel coordinate system.

[0024] The centroid coordinates of the positioning hole obtained from the above processing link are normalized and cross-correlated with the standard template in the database. When the matching response value exceeds a preset threshold, the irregular part category is confirmed, and the offset vector of the positioning hole in the image plane is output. This offset vector needs to be corrected by the camera intrinsic parameter matrix and distortion coefficients, and converted into two-dimensional position coordinates in the world coordinate system with the center of the material placement platform as the origin.

[0025] Formula ① is used to correlate the grayscale value of each pixel in the positioning hole area with its spatial position, and to calculate the position of the centroid of the positioning hole in the pixel coordinate system: in, The row coordinates of the centroid of the positioning hole in the pixel coordinate system; Let be the column coordinate of the centroid of the positioning hole in the pixel coordinate system; i is the row index of the pixel, i∈[1,M]; j is the column index of the pixel, j∈[1,N]; M is the total number of rows of the image (height resolution); N is the total number of columns of the image (width resolution). This represents the grayscale value of the pixel located in the i-th row and j-th column after binarization. ∈{0,255}, obtained by median filtering and Canny edge detection of the original image captured by the smart camera.

[0026] Gray values ​​of pixels in the positioning hole area captured by the smart camera Extracting the centroid coordinates of the positioning hole and Together, they form the centroid coordinate output of formula ①.

[0027] Formula ② is used to convert the centroid coordinates in the pixel coordinate system to three-dimensional position coordinates in the world coordinate system: in, , , These are the coordinates of the positioning hole in the horizontal, depth, and vertical directions in the world coordinate system, respectively. and These are the focal lengths of the smart camera in the row and column directions, respectively. and These are the physical pixel dimensions of the smart camera in the row and column directions, respectively. and The centroid coordinates are output by formula ①; and These are the pixel coordinates of the main point of the smart camera in the row and column directions, respectively. This is the radial distortion correction function. The radial distance from the pixel to the principal point. ,in and Radial distortion coefficient; This is the estimated initial height of the irregularly shaped part on the loading table.

[0028] Obtained from formula ① and As the input for formula ②, substitute into and The item is output after focal length scaling and distortion correction. , , .

[0029] Formula ③ is used to rigidly register the world coordinates of the positioning holes output by Formula ② with the database template coordinates, calculate the optimal rotation matrix and translation vector, and minimize the mean square error between the two sets of coordinates: Where R is the rotation matrix to be solved. ,satisfy (Orthogonality constraint) and (Determinant constraint, ensuring it is a pure rotation matrix); t is the translation vector to be solved, t∈R^3; k is the index number of the positioning hole, k∈{1,2}; K is the total number of positioning holes, K=2; The world coordinates of the k-th positioning hole are output from formula ②. ; The coordinates of the standard template for the k-th positioning hole are stored in the database. ; represents the squared Euclidean norm of the vector; st indicates that it is bound to.

[0030] The Q^((k)) obtained from formula ② is used as the input term for formula ③, and substituted into... The terms, after least squares solution under orthogonal constraints, output R and t. This problem can be solved using singular value decomposition: for and Calculate the centroids separately and decenter them. After constructing the covariance matrix, perform SVD decomposition to obtain the optimal rotation matrix R, and then calculate the translation vector t.

[0031] Formula ④ is used to determine the thickness H of the irregular part and the depth of the positioning hole. Calculate the vertical target height of the support base. Vertical target height of the positioning pin : in, The target vertical height of the top surface of the support base; H represents the reference vertical height of the top surface of the movable slide; H represents the thickness of the irregular part. The target vertical height of the top of the positioning pin; Δ is the depth of the positioning hole; Δ is the safety margin for the top of the positioning pin to extend beyond the top surface of the irregular part, Δ∈[1,3] mm.

[0032] R and t obtained from formula ③ are used to align the coordinates of the positioning hole with the template coordinates (this alignment result is used to determine the horizontal target position of the positioning pin, which is an input of formula ⑤). The aligned coordinates are then combined with H and Together they form the input and output of formula ④. and .

[0033] Formula ⑤ is used to convert the horizontal target coordinates into the target rotation angle of the first motor and the second motor: in, The target rotation angle of the first motor; The target rotation angle for the second motor; The lead of the first lead screw is the distance the first slide rail moves when the first lead screw rotates one revolution, which is determined by the thread parameters of the first lead screw. The lead of the second lead screw is the distance the second slide rail moves when the second lead screw rotates one revolution. and These are the target coordinates of the positioning pin in the first and second directions in the horizontal plane after registration and correction according to formula ③. and These are the actual coordinates of the positioning pin in the first and second directions at the current moment, respectively, which are fed back in real time by displacement sensors installed on the loading platform.

[0034] The R and t obtained from formula ③ are used to... Correction to and The result obtained from formula ④ and Determine the vertical adjustment target; these two factors together form the input and output of formula ⑤. and .

[0035] Formula 6 is used to convert the target height of the support base into the target extension length of the electric actuator: in, The target extension length of the electric actuator; This represents the actual extension length of the electric actuator at the current moment. The target height of the support base is output by formula ④; This represents the actual height of the support at the current moment.

[0036] Obtained from formula ④ As the input to formula ⑥, substitute into Item, combined Output .

[0037] Secondly, the positioning mechanism 5 performs multi-dimensional adjustments based on the recognition results. Since the two sets of positioning mechanisms 5 are symmetrically arranged on the top of the loading platform 3, and each set of positioning mechanisms 5 has adjustment capabilities in three different directions, the movable slide 10 can slide along the mutually perpendicular first slide rail 8 and second slide rail 9 to achieve adjustment at any position in the horizontal plane; in the vertical direction, the support seat 11 can slide up and down along the movable slide 10 to adapt to irregularly shaped parts 2 of different thicknesses. After the target coordinate command generated by the AI ​​algorithm is transmitted to the positioning mechanism 5, the movable slide 10 first moves along the first slide rail 8 and the second slide rail 9, adjusting the positioning pin 12 to the horizontal corresponding position of the positioning hole; the support seat 11 then slides up and down according to the thickness of the irregularly shaped part 2, matching the height of the positioning pin 12 with the height of the positioning hole, ultimately aligning the positioning pin 12 with the two positioning holes on the forging part.

[0038] Next, the positioning hole can be inserted into the positioning pin 12, while the support seat 11 avoids the positioning pin 12 through the clearance hole 13, ensuring stable support for the irregular part 2. The support seat 11 has a large contact area with the bottom surface of the irregular part 2, and the bottom surface of the irregular part 2 is subjected to uniform force during processing, which has a good effect on preventing stress deformation of the forging.

[0039] After the positioning pin 12 is inserted into the positioning hole, the AI ​​algorithm drives the positioning pins 12 of the two sets of positioning mechanisms 5 to move towards the center of the irregular part 2 according to the size parameters of the irregular part 2. This causes the positioning pin 12 to apply a symmetrical horizontal clamping force to the irregular part 2 through the positioning hole, so that the irregular part 2 is firmly fixed on the loading table 3. The tooling completes the fixing action and can be linked with the processing machine tool 1 to start the subsequent processing process.

[0040] Traditional tooling requires customized, fixed positioning pins 12 fixtures for different irregularly shaped parts 2, necessitating disassembly and fixture replacement when changing workpieces, which is time-consuming and labor-intensive. This new tooling, however, automatically matches the type of irregularly shaped part 2 using AI visual recognition and drives the positioning mechanism 5 to adjust the position of the positioning pins 12 in multiple dimensions. It can adapt to various irregularly shaped parts 2 with different positioning hole positions without manual fixture replacement, making it particularly suitable for processing multiple varieties and small batches of irregularly shaped parts 2, significantly reducing tooling changeover costs and time. It reduces reliance on human experience, avoids the uncertainty of manual operation, shortens pre-processing preparation time, and improves overall production efficiency.

[0041] According to one embodiment of the present invention, the positioning mechanism 5 further includes a first guide post 14 and a first lead screw 15. The first guide post 14 is fixedly connected to the material platform 3, and the first lead screw 15 is rotatably connected to the material platform 3. One end of the first slide rail 8 has a first sliding hole that slides with the first guide post 14, and the other end of the first slide rail 8 has a first threaded hole that threadedly connects with the first lead screw 15. The positioning mechanism 5 further includes a first motor 16, which is fixedly connected to the material platform 3, and the output end of the first motor 16 is coaxially fixedly connected to the first lead screw 15.

[0042] After the AI ​​algorithm generates the horizontal target coordinates of the positioning pin 12, the first motor 16, fixed to the loading platform 3, starts, and its output drives the first lead screw 15, which is coaxially fixed, to rotate. Since one end of the first slide rail 8 is threadedly engaged with the first lead screw 15 through the first screw hole, and the other end is slidably engaged with the first guide post 14 fixed to the loading platform 3 through the first sliding hole, the rotational motion of the first lead screw 15 is converted into the linear motion of the first slide rail 8 along the first guide post 14, which in turn drives the movable slide block 10 to move, realizing the position adjustment of the positioning pin 12 in this direction. Finally, it cooperates with the adjustment in the direction of the second slide rail 9 to complete the calibration of the horizontal position of the positioning pin 12.

[0043] According to one embodiment of the present invention, the positioning mechanism 5 further includes a second guide post 17 and a second lead screw 18. The second guide post 17 is fixedly connected to the material platform 3, and the second lead screw 18 is rotatably connected to the material platform 3. One end of the second slide rail 9 has a second sliding hole that slides with the second guide post 17, and the other end of the second slide rail 9 has a second threaded hole that threadedly connects with the second lead screw 18. The positioning mechanism 5 also includes a second motor 19, which is fixedly connected to the material platform 3, and the output end of the second motor 19 is coaxially fixedly connected to the second lead screw 18.

[0044] When the AI ​​algorithm outputs the target position command for the positioning pin 12 in the direction of the second slide rail 9, the second motor 19 fixed to the loading platform 3 starts, driving the coaxial second lead screw 18 to rotate. One end of the second slide rail 9 is threadedly engaged with the second lead screw 18 through the second screw hole, and the other end is slidably engaged with the second guide post 17 fixed to the loading platform 3 through the second sliding hole. The rotation of the second lead screw 18 is converted into linear movement of the second slide rail 9 along the second guide post 17, driving the movable slide block 10 to move. In coordination with the adjustment in the direction of the first slide rail 8, the positioning pin 12 is adjusted to the horizontal position corresponding to the positioning hole.

[0045] According to one embodiment of the present invention, the positioning mechanism 5 further includes an electric push rod 20, the bottom of the movable slide 10 is provided with a slot, the top of the movable slide 10 is provided with a through hole communicating with the slot, the bottom of the support base 11 is slidably connected to the through hole and extends downward into the slot, the electric push rod 20 is fixedly connected to the movable slide 10, and the telescopic end of the electric push rod 20 is fixedly connected to the bottom of the support base 11.

[0046] Once the AI ​​algorithm calculates the vertical target height of the positioning pin 12 based on the thickness of the irregular part 2, the electric push rod 20, fixed to the movable slide 10, is activated. The movable slide 10 has a slot at its bottom and a through hole at its top that communicates with the slot. The bottom of the support base 11 extends through the through hole into the slot and is fixedly connected to the telescopic end of the electric push rod 20, with the support base 11 slidingly engaging with the through hole. When the support base 11 rises or falls, its clearance hole 13 always avoids the positioning pin 12 fixed to the movable slide 10, while the top of the support base 11 contacts the bottom of the irregular part 2, adapting to irregular parts 2 of different thicknesses, ensuring stable support for the workpiece, and matching the height of the positioning pin 12 with the height of the positioning hole.

[0047] According to one embodiment of the present invention, the machine tool 1 is provided with a guide rail groove 21, and the bottom of the loading table 3 is fixedly provided with a mating key 22 embedded in the guide rail groove 21. After the tooling completes the fixing of the irregular part 2, the machine tool 1 can slide along the guide rail groove 21 through the mating key 22 to accurately transport the fixed irregular part 2 to the machine tool processing area, realizing the linkage positioning between the tooling and the machine tool 1. The identification mechanism 4 also includes a protective cover 25, which is installed on the outside of the intelligent camera 7 and fixedly connected to the loading table 3. During the operation of the tooling (such as chip splashing or coolant spraying during processing), it prevents impurities from contacting the camera lens, and at the same time avoids damage to the camera from external impact, ensuring that the camera can always clearly capture image information of the irregular part 2.

[0048] According to one embodiment of the present invention, the top of the movable slide 10 is provided with a slot 23, and the bottom of the positioning pin 12 is fixedly provided with a plug 24 that mates with the slot 23. The outer side of the plug 24 is provided with an external thread, and the inner side of the slot 23 is provided with an internal thread that mates with the plug 24. When it is necessary to replace the positioning pin 12 (e.g., due to wear of the positioning pin 12 or to adapt to positioning holes of different diameters), the plug 24 can be removed from the slot 23 by rotating the positioning pin 12 in the reverse direction, thus completing the disassembly and replacement of the positioning pin 12.

[0049] The above-disclosed embodiments are merely a few specific examples of the present invention. However, the embodiments of the present invention are not limited thereto, and any variations that can be conceived by those skilled in the art should fall within the protection scope of the present invention.

Claims

1. An intelligent fixed fixture for machining irregularly shaped parts based on AI visual recognition, the fixture being mounted on a machine tool (1), characterized in that, The system includes a loading platform (3), an identification mechanism (4), and a positioning mechanism (5). The loading platform (3) is connected to the machine tool (1). The identification mechanism (4) includes a placement platform (6) and a smart camera (7). The placement platform (6) and the smart camera (7) are both fixedly installed on the side of the loading platform (3), and the smart camera (7) is located above the placement platform (6). The positioning mechanism (5) has two sets, which are symmetrically arranged on the top of the loading platform (3). The positioning mechanism (5) includes a first slide rail (8), a second slide rail (9), a movable slide (10), a support base (11), and a positioning pin. (12) The first slide rail (8) and the second slide rail (9) are both in a horizontal state and are slidably connected to the loading platform (3). The sliding direction of the first slide rail (8) and the sliding direction of the second slide rail (9) are perpendicular to the projection of the same horizontal plane. The top and bottom of the movable slide (10) are slidably connected to the first slide rail (8) and the second slide rail (9) respectively. The support base (11) is in a vertical state and is slidably connected to the movable slide (10). The support base (11) has a clearance hole (13). The positioning pin (12) is fixedly connected to the movable slide (10) and is located in the clearance hole (13).

2. The intelligent fixing fixture for processing irregularly shaped parts based on AI visual recognition as described in claim 1, characterized in that, The positioning mechanism (5) further includes a first guide post (14) and a first lead screw (15). The first guide post (14) is fixedly connected to the material platform (3), and the first lead screw (15) is rotatably connected to the material platform (3). One end of the first slide rail (8) is provided with a first sliding hole that slides with the first guide post (14), and the other end of the first slide rail (8) is provided with a first threaded hole that threadedly connects with the first lead screw (15).

3. The intelligent fixing fixture for processing irregularly shaped parts based on AI visual recognition as described in claim 2, characterized in that, The positioning mechanism (5) also includes a first motor (16), which is fixedly connected to the material platform (3), and the output end of the first motor (16) is fixedly connected to the first lead screw (15) on the same axis.

4. The intelligent fixing fixture for processing irregularly shaped parts based on AI visual recognition as described in claim 1, characterized in that, The positioning mechanism (5) further includes a second guide post (17) and a second lead screw (18). The second guide post (17) is fixedly connected to the material platform (3), and the second lead screw (18) is rotatably connected to the material platform (3). One end of the second slide rail (9) is provided with a second sliding hole that slides with the second guide post (17), and the other end of the second slide rail (9) is provided with a second threaded hole that threadedly connects with the second lead screw (18).

5. The intelligent fixing fixture for processing irregularly shaped parts based on AI visual recognition as described in claim 4, characterized in that, The positioning mechanism (5) also includes a second motor (19), which is fixedly connected to the loading platform (3), and the output end of the second motor (19) is fixedly connected to the second lead screw (18) on the same axis.

6. The intelligent fixing fixture for processing irregularly shaped parts based on AI visual recognition as described in claim 1, characterized in that, The positioning mechanism (5) also includes an electric push rod (20). The bottom of the movable slide (10) is provided with a slot, and the top of the movable slide (10) is provided with a through hole that communicates with the slot. The bottom of the support base (11) is slidably connected to the through hole and extends downward into the slot. The electric push rod (20) is fixedly connected to the movable slide (10), and the telescopic end of the electric push rod (20) is fixedly connected to the bottom of the support base (11).

7. The intelligent fixing fixture for processing irregularly shaped parts based on AI visual recognition as described in claim 1, characterized in that, The machine tool (1) is provided with a guide rail groove (21), and the bottom of the loading platform (3) is fixedly provided with a mating key (22) embedded in the guide rail groove (21).

8. The intelligent fixing fixture for processing irregularly shaped parts based on AI visual recognition as described in claim 1, characterized in that, The top of the movable slide (10) is provided with a slot (23), and the bottom of the positioning pin (12) is fixedly provided with a plug (24) that cooperates with the slot (23).

9. The intelligent fixing fixture for processing irregularly shaped parts based on AI visual recognition as described in claim 8, characterized in that, The outer side of the insert (24) is provided with an external thread, and the inner side of the slot (23) is provided with an internal thread that mates with the insert (24).

10. The intelligent fixing fixture for processing irregularly shaped parts based on AI visual recognition as described in claim 1, characterized in that, The identification mechanism (4) also includes a protective cover (25), which is placed on the outside of the smart camera (7) and is fixedly connected to the material carrier (3).

Images