Automatic radial-axial ring forging control method based on real-time contour extraction

Abstract

The invention belongs to the technical field of machine vision, and particularly relates to a radial-axial ring forging automatic control method based on contour real-time extraction. The automatic control method for controlling the roundness of the ring forge piece is designed according to the technological parameters of the radial-axial ring forging technology and the forge piece deformation mechanism through the machine vision and the contour recognition technology, and real-time automatic control over the ring piece in the intelligent ring forging process can be achieved. The method has good feasibility, and the machining efficiency and the forging quality are effectively improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of machine vision, and particularly relates to a radial-axial ring forging automatic control method based on real-time contour extraction. Background Art

[0002] The radial-axial ring forging technology is the most important processing technology for forming ring-shaped parts. In this process, a ring-shaped metal blank is heated to a high temperature, and a radial-axial ring forging machine is used to process the blank to cause plastic deformation, ultimately achieving the processing goals of increasing the diameter of the ring and reducing the wall thickness. In actual production and processing, the control of the ring forging process relies more on the initially designed process parameters and the experience of engineers. After the size of the ring reaches the target size, a relatively long roundness adjustment stage is required to ensure the shape requirements. However, with the continuous expansion of the application scenarios of ring forgings, higher requirements are put forward for the processing efficiency of ring forging technology and the dimensional shape quality of ring forgings. Therefore, a more scientific and reliable method is needed to control the processing process. Currently, machine vision and image recognition technologies are widely used in the field of intelligent manufacturing and have achieved excellent results with their outstanding feature recognition and process detection capabilities. Therefore, machine vision technology can be applied to the automatic control of intelligent ring forging.

[0003] Xu et al. proposed a mathematical model for the diameter expansion of ring forgings in the radial-axial ring forging process and verified it through finite element simulation to achieve the control of the radial-axial ring forging process. However, this control method has not been popularized and has not been combined with machine vision technology, and still does not belong to automatic control. Husman et al. used machine vision technology to identify the height changes in the radial processing area during the radial-axial ring forging process, but did not control this material flow phenomenon.

[0004] The Allwood team has achieved many results in the field of image recognition and automatic control of the radial-axial ring forging process using a machine vision system. In 2014, the team used a machine vision system to identify the contour of the deformed ring forging during the processing, effectively eliminating the influence of the occluded part and obtaining an accurate ring contour. In 2015, the team used the feature extraction and contour recognition technology of machine vision to automatically control the radial-axial ring forging process with additional marks on the surface of the ring blank. However, this automatic control method relies on surface marks, and surface marks cannot be added to high-temperature forgings, so this method has great limitations. In 2019, the team proposed the law of curvature change in the radial processing area of the radial-axial ring forging process and the process parameters, and effectively verified it using finite elements. However, the team did not perform automatic control for the feature of roundness loss caused by curvature change. Summary of the Invention

[0005] In view of this, to solve the problems existing in the prior art, the present invention provides a radial-axial ring forging automatic control method based on real-time contour extraction.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] A radial-axial ring forging automatic control method based on real-time contour extraction uses an industrial camera as an image data acquisition device to obtain the geometric morphology information of the ring-shaped part in the intelligent ring forging process in real time; takes the geometric morphology information as input data, and obtains a depth image that meets the requirements through a depth prediction model in real time; inputs the depth image and semantic hint points into a semantic segmentation model to obtain a mask of the key parts of the radial-axial ring forging process in real time; performs morphological operations and contour recognition based on the mask result, and quickly obtains a high-accuracy contour map of the key parts of the radial-axial ring forging process through a clustering algorithm; uses the real-time obtained contour map to perform simultaneous automatic control on the axial machining area and the radial machining area respectively, and obtains the corresponding adjusted motion roller motion parameters; uses socket communication to transmit the speed parameters in the computer to the radial-axial ring rolling equipment in real time, and the automatic control process continues until the ring part reaches the target size.

[0008] The existing control methods for the radial-axial ring forging process mainly rely on initial process parameters and manual experience control, and cannot meet the requirements of high-precision and high-efficiency ring part processing. In addition, the existing automatic control methods need to rely on adding additional marks on the surface of the ring forging, and this method is not applicable to the automatic control of high-temperature forgings.

[0009] The present invention uses machine vision and contour recognition technologies, designs an automatic control method for the roundness control of ring forgings according to the process parameters and forging deformation mechanism of the radial-axial ring forging process, and realizes the real-time automatic control of the ring-shaped part in the intelligent ring forging process. This method has good feasibility and effectively improves the processing efficiency and forging quality.

[0010] Further, the automatic control method in the axial machining area is as follows:

[0011] Mark the distance between the outer diameter of the ring part and the edge of the taper roller at the initial machining position, use the real-time obtained contour map to calculate the real-time distance between the outer diameter of the ring part and the edge of the taper roller, and keep the relative position between the outer diameter of the ring part and the edge of the taper roller unchanged during the machining process.

[0012] Further, the automatic control method in the radial machining area is as follows:

[0013] According to the real-time obtained contour map, calculate the outer edge curvatures of the input end and the output end of the ring part in the radial machining area respectively. If the curvature of the output end becomes larger, increase the feed speed of the core roller; if the curvature of the output end decreases, decrease the feed speed of the core roller.

[0014] The present invention uses an industrial camera as an image data acquisition device, and through means such as camera pose optimization and data communication, it can perceive and obtain the geometric morphology information (RGB image) of the ring-shaped part during the intelligent ring forging process in real time. Taking the RGB image as the input data, without additional training and learning, directly through the depth prediction model Depth Anything v2, a depth image that meets the requirements can be obtained in real time. The depth image and semantic cue points are input into the semantic segmentation model Mobile SAM, and the mask (i.e., Mask information) of the key parts of the radial-axial ring forging process can be obtained in real time. Based on the mask result, morphological operations and contour recognition are carried out, and through a clustering algorithm, a contour map of the key parts of the radial-axial ring forging process with high accuracy can be obtained quickly. Using the contour map obtained in real time, automatic control is carried out simultaneously for the radial processing area and the axial processing area respectively.

[0015] In the axial processing area, if the rotational speed of the tapered roller is too high, the ring will shift towards the output side of the axial pass; if the rotational speed of the tapered roller is too low, the ring will shift towards the output side of the radial pass. If the rotational speeds in the two passes are seriously inconsistent, folding of the ring part may occur. If the position of the tapered roller is kept unchanged, the rotational speed of the tapered roller needs to be continuously reduced; if the rotational speed of the tapered roller is kept unchanged, the tapered roller needs to move linearly backward, and the relative position between the tapered roller and the outer diameter of the ring part remains unchanged. Therefore, the present invention marks the distance between the outer diameter of the ring part and the edge of the tapered roller at the initial processing position, calculates the real-time distance between the outer diameter of the ring part and the edge of the tapered roller using the contour map obtained in real time, and the relative position between the outer diameter of the ring part and the edge of the tapered roller should be kept unchanged during the processing. If this distance increases, the radial movement speed of the tapered roller is reduced; if this distance decreases, the radial movement speed of the tapered roller is increased.

[0016] In the radial processing area, the ring part is subjected to the radial rolling force of the main roller and the core roller, so the wall thickness decreases. However, for the wall thickness of the ring part to decrease uniformly in the radial direction, the wall thickness reduction rates of the two parts of the main roller and the core roller need to be kept consistent. If they are inconsistent, it will cause curvature changes at the input end and output end of the ring part in the radial processing area, resulting in loss of roundness during the processing of the ring part and increasing the risk of processing instability. Therefore, the present invention calculates the outer edge curvatures of the input end and output end of the ring part in the radial processing area respectively according to the contour map obtained in real time. If the curvature of the output end becomes larger, the feed speed of the core roller is increased; if the curvature of the output end decreases, the feed speed of the core roller is decreased.

[0017] According to the above control method, real-time automatic control is carried out simultaneously for the radial processing area and the axial processing area during the processing, and the corresponding adjusted motion roller motion parameters are obtained. Using socket communication, the speed parameters in the computer are transmitted to the radial-axial ring rolling equipment in real time, and the automatic control process continues until the ring part reaches the target size. At this time, the feed speed of the core roller is set to 0, and the whole circle is completed to end the processing.

[0018] Therefore, the present invention can perform real-time automatic control of the radial-axial machining process based on the real-time contour recognition map without adding other additional markings. The control method has the advantages of high feasibility, wide applicability, accurate, stable and efficient machining process.

[0019] Furthermore, the specific steps of the radial-axial ring forging automatic control method based on real-time contour extraction include:

[0020] (1) Preparation work: Place the initial blank in the machining area of the ring rolling equipment. Install an industrial camera directly above it to be responsible for obtaining the top view, and install an industrial camera on the side to be responsible for obtaining the front view. Take the initial reference position map of the radial machining area and the axial machining area, perform contour recognition on the initial reference position, and obtain the real-time distance between the outer diameter of the ring and the edge of the conical roller.

[0021] (2) Equipment parameter setting: Adjust each rolling component of the ring rolling equipment to the initial working position, and set the main roller speed, core roller feed speed, conical roller speed, conical roller radial speed, and conical roller axial speed of the ring rolling machine.

[0022] (3) Start-up and operation: Connect the power supply, the ring rolling equipment starts to process the blank, and the two industrial cameras start to obtain RGB images and input the RGB image information to the communication computer.

[0023] (4) Depth map prediction: Taking the RGB image obtained in step (3) as the input, through the Depth Anything v2 model, effectively remove the useless noise information and predict the corresponding depth map data.

[0024] (5) Semantic segmentation of key areas: Taking the depth map obtained in step (4) and the semantic cue points of the key areas as the input, through the Mobile SAM model, predict the masks of key areas such as the ring, conical roller, and core roller.

[0025] (6) Filtering processing: Perform filtering processing on the RGB information of the key areas obtained in step (5) to remove color noise and achieve a smooth effect.

[0026] (7) Edge detection: Perform edge detection on the RGB information processed in step (6). According to the actual working conditions, set the hyperparameters of the edge detection algorithm to obtain clear edge information.

[0027] (8) Useless edge clipping: Perform cutting processing on the edge information obtained in step (7), remove the remaining redundant parts, and only retain the parts of the ring, main roller, core roller, and conical roller to obtain the required contour map. According to the relative position relationship between the ring and the rolling mill, without adding additional surface markings on the forging surface, perform simultaneous automatic control of the axial machining area and the radial machining area respectively.

[0028] (9) Axial machining area control: According to the initial machining position diagram, mark the distance between the outer diameter of the ring workpiece and the edge of the taper roller. During the machining process, the relative position between the outer diameter of the ring workpiece and the edge of the taper roller should be kept unchanged. According to the contour diagram obtained in step (8), during the machining process, if this distance increases, then reduce the radial movement speed of the taper roller; if this distance decreases, then increase the radial movement speed of the taper roller;

[0029] (10) Radial machining area control: According to the contour diagram obtained in step (8), calculate the curvature of the outer edge of the ring workpiece at the input end and the output end of the radial machining area (on the side of the main roller core roller) respectively. If the curvature at the output end becomes larger, then increase the feeding speed of the core roller; if the curvature at the output end decreases, then reduce the feeding speed of the core roller;

[0030] (11) Control data transmission: According to the control methods in step (9) and step (10), adjust the parameters of each moving rolling mill in real time, and transmit the motion parameters in the computer to the ring rolling mill through socket communication to achieve automatic control;

[0031] (12) Process end: When the ring workpiece reaches the required size, set the feeding speed of the core roller to 0, rotate a full circle, and stop the entire process.

[0032] Thus, compared with the current radial-axial ring forging process control method that relies on the surface marks (Mark) of the forging, the present invention utilizes advanced technologies in the field of machine vision to perform contour recognition and feature calculation on the key areas of radial-axial ring rolling. Based on the contour recognition data, through the contour features of the ring workpiece and the positional relationship between the ring workpiece and the rolling mill, the process parameters that cause the deformation characteristics of the forging are inversely analyzed, and the coordinated control optimization of the radial machining area and the axial machining area can be realized.

[0033] Compared with the existing control methods, the present invention does not need to add additional marks on the surface of the forging for positioning and identifying the deformation state of the ring workpiece, but only performs automatic control through the positional relationship between the ring workpiece and the rolling mill obtained by the machine vision contour recognition technology and the real-time contour features of the ring workpiece deformation, which can effectively avoid the influence of the oxide scale of the high-temperature forging on the surface features and has a wide range of application scenarios. At the same time, the automatic control method for the radial machining area and the automatic control method for the axial machining area proposed by the present invention can be independently controlled and coordinated, with excellent application flexibility. The present invention performs real-time automatic control based on the contour recognition technology of machine vision. During the process, the deformation of the ring workpiece is uniform and stable. After the size reaches the requirement, only one full circle of rounding is required to end the machining, which has the advantages of high precision and high efficiency in the process of machining detection and control. In addition, the present invention uses industrial cameras and computers as the implementation equipment, which has the characteristics of high reliability and low cost. Description of the Drawings

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to the provided drawings.

[0035] Figure 1 FIG. is a schematic diagram of the radial-axial ring forging process in the automatic control method of radial-axial ring forging based on real-time contour extraction of the present invention. Among them, 1 is the main roll, 2 is the core roll, 3 is the taper roll, 4 is the deformed ring forging, 5 is the radial processing area, 6 is the axial processing area, and 7 is the non-processing area.

[0036] Figure 2 FIG. is a schematic diagram of the contour extraction route of the deformed ring forging in Embodiment 1 of the present invention.

[0037] Figure 3 FIG. is a schematic diagram of the automatic control method in Embodiment 1 of the present invention. FIG. (a) shows the automatic control parameters of the axial processing area, and FIG. (b) shows the automatic control parameters of the radial processing area. Among them, 1 is the main roll, 2 is the core roll, and 3 is the main roll.

[0038] Figure 4 FIG. is the ring blank in the processing process of Embodiment 1 of the present invention.

[0039] Figure 5 FIG. is the ring with the target size obtained by the automatic control of the radial-axial ring forging process in Embodiment 1 of the present invention. Detailed implementation manners

[0040] The following will clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the protection scope of the present invention.

[0041] Here, the special term "embodiment" does not necessarily mean better or superior to other embodiments as "exemplary". For the performance index tests in the embodiments of this application, unless otherwise specified, the conventional test methods in the art are adopted. It should be understood that the terms described in this application are only used to describe specific implementation manners and are not used to limit the content disclosed in this application.

[0042] Unless otherwise specified, the technical and scientific terms used in this article have the same meanings as those commonly understood by those of ordinary skill in the technical field to which this application belongs; the test methods and technical means not specifically mentioned in this application are all the experimental methods and technical means commonly adopted by those of ordinary skill in the art.

[0043] To better illustrate the content of this application, numerous specific details are given in the following specific embodiments. Those skilled in the art should understand that this application can still be implemented without certain specific details. In the embodiments, some methods, means, instruments, equipment, etc. well-known to those skilled in the art are not described in detail in order to highlight the gist of this application.

[0044] On the premise of no conflict, the technical features disclosed in the embodiments of this application can be combined arbitrarily, and the obtained technical solutions belong to the content disclosed in the embodiments of this application.

[0045] The present invention belongs to the technical field of machine vision, and specifically relates to a radial-axial ring forging automatic control method based on real-time contour extraction. The present invention utilizes machine vision and contour recognition technologies, and designs an automatic control method for the roundness control of ring forgings according to the process parameters and forging deformation mechanism of the radial-axial ring forging process, realizing the real-time automatic control of the ring-shaped parts during the intelligent ring forging process. This method has good feasibility and effectively improves the processing efficiency and forging quality.

[0046] To better understand the present invention, the following embodiments are used to further specifically elaborate on the present invention, but it should not be construed as a limitation of the present invention. For those skilled in the art, some non-essential improvements and adjustments made according to the above-mentioned invention content are also considered to fall within the protection scope of the present invention.

[0047] Embodiment 1

[0048] A radial-axial ring forging automatic control method based on real-time contour extraction, the positions of each component are as Figure 1 shown:

[0049] (1) Preparation work: Place the initial blank ( Figure 4 ) in the processing area of the ring rolling equipment, install an industrial camera directly above it to be responsible for obtaining the top view, and install an industrial camera on the side to be responsible for obtaining the front view, and take the initial reference position maps of the radial processing area and the axial processing area;

[0050] (2) Equipment parameter setting: Adjust each rolling component of the ring rolling equipment to the initial working position, and set the main roller speed, core roller feed speed, cone roller speed, cone roller radial speed, and cone roller axial speed of the ring rolling machine;

[0051] (3) Start and run: Connect the power supply, the ring rolling equipment starts to process the blank, the two industrial cameras start to obtain RGB images, and input the RGB image information to the communication computer;

[0052] (4) Depth map prediction: Using the RGB image obtained in step (3) as the input, the corresponding depth map data is predicted through the Depth Anything v2 model;

[0053] (5) Semantic segmentation of key regions: Using the depth map obtained in step (4) and the semantic cue points of the key regions as the input, the masks of key regions such as the ring part, taper roller, and core roller are predicted through the Mobile SAM model;

[0054] (6) Filtering process: Filter the RGB information of the key regions obtained in step (5) to remove color noise and achieve a smooth effect;

[0055] (7) Edge detection: Perform edge detection on the RGB information processed in step (6). According to the actual working conditions, set the hyperparameters of the edge detection algorithm to obtain clear edge information;

[0056] (8) Useless edge cropping: Cut the edge information obtained in step (7) to remove the remaining redundant parts, and only retain the ring part, main roller, core roller, and taper roller parts to obtain the required contour map ( Figure 2 )

[0057] (9) Axial machining area control: According to the machining initial position map, mark the distance between the outer diameter of the ring part and the edge of the taper roller. During the machining process, the relative position between the outer diameter of the ring part and the edge of the taper roller should be kept unchanged. According to the contour map obtained in step (8), if this distance increases during the machining process, the radial movement speed of the taper roller is reduced; if this distance decreases, the radial movement speed of the taper roller is increased ( Figure 3 )

[0058] (10) Radial machining area control: According to the contour map obtained in step (8), calculate the curvatures of the outer edges of the ring part at the input and output ends of the radial machining area (on the side of the main roller and core roller) respectively. If the curvature at the output end becomes larger, the feed speed of the core roller is increased; if the curvature at the output end decreases, the feed speed of the core roller is reduced ( Figure 3 )

[0059] (11) Control data transmission: According to the control methods in step (9) and step (10), adjust the parameters of each moving roller in real time, and transmit the motion parameters in the computer to the ring rolling mill through socket communication to achieve automatic control;

[0060] (12) Process end: When the ring part reaches the required size, set the feed speed of the core roller to 0, rotate one full circle, and stop the entire process to obtain the ring part with the target size ( Figure 5 )

[0061] The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An automatic control method for radial-axial ring forging based on real-time profile extraction, characterized in that, An industrial camera is used as an image data acquisition device to obtain the geometric shape information of the ring-shaped part in the intelligent ring forging process in real time; Taking the geometric shape information as input data, a depth image that meets the requirements is obtained in real time through a depth prediction model; The depth image and semantic cue points are input into the semantic segmentation model to obtain the mask of the key parts of the radial-axial ring forging process in real time; based on the mask result, morphological operations and contour recognition are performed, and a contour map of the key parts of the radial-axial ring forging process with high accuracy is quickly obtained through a clustering algorithm; using the contour map obtained in real time, the axial processing area and the radial processing area are automatically controlled simultaneously, and the corresponding adjusted motion roller motion parameters are obtained; using socket communication, the speed parameters in the computer are transmitted to the radial-axial ring rolling equipment in real time, and the automatic control process is carried out until the ring part reaches the target size.

2. The method according to claim 1, wherein The automatic control method in the axial processing area is as follows: Mark the distance between the outer diameter of the ring part and the edge of the cone roller at the initial processing position, use the contour map obtained in real time to calculate the real-time distance between the outer diameter of the ring part and the edge of the cone roller, and keep the relative position between the outer diameter of the ring part and the edge of the cone roller unchanged during the processing.

3. The method according to claim 1, wherein The automatic control method in the radial processing area is as follows: According to the contour map obtained in real time, calculate the outer edge curvatures of the input end and the output end of the ring part in the radial processing area respectively. If the curvature of the output end becomes larger, increase the feeding speed of the core roller; if the curvature of the output end decreases, decrease the feeding speed of the core roller.

4. The method according to any one of claims 1 to 3, characterized in that The specific steps include: (1) Preparation work: Place the initial blank in the processing area of the ring rolling equipment, install an industrial camera directly above it to be responsible for obtaining the top view, and install an industrial camera on the side to be responsible for obtaining the front view, and take the initial reference position maps of the radial processing area and the axial processing area; (2) Equipment parameter setting: Adjust each rolling part of the ring rolling equipment to the initial working position, and set the main roller speed, core roller feeding speed, cone roller speed, cone roller radial speed, and cone roller axial speed of the ring rolling machine; (3) Start-up and operation: Turn on the power, the ring rolling equipment starts to process the blank, the two industrial cameras start to obtain RGB images, and input the RGB image information into the communication computer; (4) Depth map prediction: Taking the RGB image obtained in step (3) as input, through the Depth Anything v2 model, the corresponding depth map data is predicted; (5) Semantic segmentation of key areas: Taking the depth map obtained in step (4) and the semantic cue points of the key areas as input, through the Mobile SAM model, the masks Mask of key areas such as the ring part, cone roller, and core roller are predicted; (6) Filtering processing: Perform filtering processing on the RGB information of the key areas obtained in step (5) to remove color noise and achieve a smooth effect; (7) Edge detection: Perform edge detection on the RGB information processed in step (6), and set the hyperparameters of the edge detection algorithm according to the actual working conditions to obtain clear edge information; (8) Useless edge cropping: Perform cutting processing on the edge information obtained in step (7), remove the remaining redundant parts, and only keep the parts of the ring part, main roller, core roller, and cone roller to obtain the required contour map; (9)Axial machining area control: According to the initial machining position diagram, mark the distance between the outer diameter of the ring and the edge of the taper roll. During the machining process, the relative position between the outer diameter of the ring and the edge of the taper roll should be kept unchanged. According to the contour diagram obtained in step (8), if this distance increases during the machining process, reduce the radial movement speed of the taper roll; if this distance decreases, increase the radial movement speed of the taper roll; (10)Radial machining area control: According to the contour diagram obtained in step (8), calculate the curvatures of the outer edges of the rings at the input and output ends of the radial machining area (on the side of the main roll core roll) respectively. If the curvature at the output end becomes larger, increase the feed speed of the core roll; if the curvature at the output end decreases, reduce the feed speed of the core roll; (11)Control data transmission: According to the control methods in step (9) and step (10), adjust the parameters of each moving roll in real time, and transmit the motion parameters in the computer to the ring rolling mill through socket communication to achieve automatic control; (12)Process end: When the ring-shaped part reaches the required size, set the feed speed of the core roll to 0, rotate a full circle, and stop the entire process.

Images