A method and system for flexible intelligent grinding of axial and radial defects of medium and thick plates based on visual measurement
The axial-radial flexible grinding robot system addresses defects in stainless steel plates through precise visual measurement and adaptive grinding, enhancing quality and efficiency while preventing wheel damage.
Patent Information
- Application Number
- CN202411216719.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-02
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-09-02
AI Technical Summary
The production of medium and thick stainless steel plates is hindered by quality defects such as inclusions and scratches, leading to poor repair quality consistency, high material waste, and low production efficiency, with existing robot grinding systems prone to grinding wheel cracking and uneven surface consistency.
An axial-radial flexible grinding robot system using visual measurement and a 3D camera for defect detection, combined with a force compensator and 'return' shaped grinding trajectory, to ensure precise and efficient repair of defects while preventing wheel cracking and improving surface quality.
Enhances product qualification rates, reduces material waste, and improves production efficiency by ensuring consistent and high-quality grinding of stainless steel plates.
Smart Images

Figure CN118720868B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of intelligent grinding of stainless steel medium and thick plates, and in particular relates to a method and system for flexible intelligent grinding of medium and thick plate defects in the axial and radial directions based on visual measurement. Background Art
[0002] The petrochemical and energy industries occupy an important position in the national economy. The key core components of their equipment, such as cold hydrogenation reactors and steam generators, are manufactured in large quantities using stainless steel and nickel-based alloy medium and thick plates. The annual supply of medium and thick plates such as austenitic stainless steel plates, duplex stainless steel plates, nickel-based alloy plates and composite plates in large stainless steel hot rolling mills can reach more than 300,000 tons. During the production process of medium and thick plates, quality defects such as inclusions, pits and scratches are very likely to appear on the surface of the steel plates. The defects caused by the production process need to be repaired and polished as a whole before the finished product, which is the last processing step. At present, manual methods are mainly used to determine, repair and repair defects. The repair quality consistency is poor, the product qualification rate is low, the material waste is serious, and the production efficiency is low. When ordinary robot grinding systems are grinding warped steel plates, it is easy to cause the grinding wheel to crack due to uneven force, and the surface consistency of the steel plate after grinding is poor. Summary of the invention
[0003] In response to the difficulties in surface quality inspection and repair in the production process of medium and thick stainless steel plates, the present invention proposes a method and system for axial-radial flexible intelligent grinding of medium and thick plate defects based on visual measurement. The axial-radial flexible grinding robot is used to realize intelligent inspection and repair of the surface quality of medium and thick stainless steel plates, which can improve the product qualification rate, reduce material waste, and improve production efficiency. It supports the independent, controllable and high-quality development of the medium and thick stainless steel plate industry and enhances the competitiveness and influence of the industry; at the same time, it can also promote the development of the petrochemical and energy industries and improve energy security and economic development levels.
[0004] To achieve the above object, the present invention provides the following solutions:
[0005] A method for flexible intelligent grinding of axial-radial defects of medium and thick plates based on visual measurement, comprising the following steps:
[0006] Step 1: Scan the workpiece to be polished using a 3D camera to obtain defect information of the workpiece to be polished, wherein the workpiece to be polished is a stainless steel medium and thick plate;
[0007] Step 2: Based on the defect information, obtain defect position information of the workpiece to be polished;
[0008] Step 3: Based on the defect location information, plan the grinding sequence and grinding trajectory;
[0009] Step 4: Based on the planned grinding sequence and grinding trajectory, the workpiece is subjected to axial-radial flexible intelligent grinding.
[0010] Preferably, the method for planning the grinding trajectory includes: adopting a "return" - shaped grinding trajectory.
[0011] Preferably, the method for obtaining the grinding depth through "return" - shaped grinding includes:
[0012] ,
[0013] wherein, is the number of grinding times required to reach the depth , is obtained by designing a cutting depth test experiment, and when the number of grinding times is , the obtained grinding depth.
[0014] Preferably, the method for obtaining the grinding width through "return" - shaped grinding includes:
[0015] ,
[0016] wherein, is the width of the grinding wheel, is the number of times of translational movement required by the robot in the direction, is the specified grinding width.
[0017] The present invention also provides a mid - thick plate defect axis - radial flexible intelligent grinding system based on vision measurement, including: a scanning module, a position acquisition module, a planning module, and a grinding module;
[0018] The scanning module is used to scan the workpiece to be ground by using a 3D camera to obtain the defect information of the workpiece to be ground, wherein the workpiece to be ground is a stainless - steel mid - thick plate;
[0019] The position acquisition module is used to obtain the defect position information of the workpiece to be ground based on the defect information;
[0020] The planning module is used to plan the grinding sequence and the grinding trajectory based on the defect position information;
[0021] The grinding module is used to perform axis - radial flexible intelligent grinding on the workpiece to be ground based on the planned grinding sequence and the grinding trajectory.
[0022] Preferably, the process of planning the grinding trajectory includes: adopting a "return" - shaped grinding trajectory.
[0023] Preferably, the process of obtaining the grinding depth through "return" - shaped grinding includes:
[0024] ,
[0025] wherein, The number of grinding times required to achieve the depth is obtained by designing a depth-of-cut test experiment. When the number of grinding times is the obtained grinding depth is
[0026] Preferably, the process of obtaining the grinding width by "return" grinding includes:
[0027] ,
[0028] where is the width of the grinding wheel, is the number of times the robot needs to translate in the direction, and
[0029] is the specified grinding width.
[0030] (1) Aiming at the problem of warping of the plate surface during the production and processing of stainless steel, the designed axis-radial force compensator of the present invention can enable the robot to complete flexible grinding operations, compensate for errors caused by uneven steel plates, improve grinding accuracy, prevent the grinding wheel from cracking due to uneven stress, and improve the service life of the grinding wheel;
[0031] (2) The "return" grinding trajectory designed by the present invention has strong adaptability, can complete the smooth transition of the grinding edge, prevent the generation of new defects due to grinding, and improve the grinding quality of the surface defects of the plate;
[0032] (3) The two sides of the robot grinding in the present invention adopt an arc-down cutting method to contact the steel plate, making the two sides of the grinding smoother. During the down cutting process, the force compensator will change the magnitude of the force in real time, that is, the output force of the compensator when contacting the steel plate is small, and the small force can prevent the grinding pits from being left on the two sides during rigid contact; after the grinding wheel contacts the steel plate, the downward pressure is increased within a short time, and the large pressure can improve the depth of cut of grinding and improve the grinding efficiency;
[0033] (4) The camera used in the present invention is a 3D blue light camera. The blue light camera has the advantages of high precision, fast response, etc., and is not affected by the reflection of the stainless steel surface, preventing the loss of point clouds. The camera is used for visual detection of the type and depth information of defects. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] In order to more clearly illustrate the technical solutions of the present invention, the following briefly introduces the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.
[0035] Figure 1 Flowchart of an axial-radial flexible intelligent grinding method for medium-thick plate defects based on vision measurement according to an embodiment of the present invention;
[0036] Figure 2 Schematic diagram of "return" shaped grinding according to an embodiment of the present invention;
[0037] Figure 3 Structural diagram of an axial-radial flexible grinding robot according to an embodiment of the present invention. Detailed implementation manners
[0038] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings 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. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.
[0039] To make the above objects, features, and advantages of the present invention more obvious and understandable, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.
[0040] Embodiment 1
[0041] The present invention provides an axial-radial flexible intelligent grinding method for medium-thick plate defects based on vision measurement, including the following steps:
[0042] Step 1: Use a 3D camera to scan the workpiece to be ground to obtain the defect information of the workpiece to be ground, where the workpiece to be ground is a stainless steel medium-thick plate;
[0043] Step 2: Based on the defect information, obtain the defect position information of the workpiece to be ground;
[0044] Step 3: Based on the defect position information, plan the grinding sequence and grinding trajectory;
[0045] Step 4: Based on the planned grinding sequence and grinding trajectory, perform axial-radial flexible intelligent grinding on the workpiece to be ground.
[0046] In this embodiment, as Figure 3 shown, the present invention realizes intelligent detection and repair of the surface quality of stainless steel medium-thick plates through an axial-radial flexible grinding robot, specifically including axial-radial force compensation and the planning of the movement trajectory of the axial-radial flexible grinding robot.
[0047] The robot system includes a robot control cabinet, an axial-radial flexible grinding robot, a end flange, an axial-radial force compensator, an electric spindle, and a grinding wheel.
[0048] When the steel plate is warped, the axial-radial flexible grinding robot effectively ensures the grinding quality and increases the service life of the grinding wheel. The floating mechanism of the axial-radial force compensator is installed on the main frame, and floating is achieved through the electrical proportional valve and cylinder. The electrical proportional valve adjusts the air pressure in real time according to the feedback signal of the force sensor to maintain a constant pressure. The cylinder is connected to the air pump to provide an air source. The electric spindle is installed at the end of the axial-radial force compensator, and the steering, speed and other grinding parameters of the electric spindle are controlled by the host computer. The axial-radial flexible force compensator structure can meet the requirements of axial and radial floating, maintain constant force, and improve the accuracy and stability of grinding.
[0049] like Figure 1 As shown, it mainly includes the following steps:
[0050] Step 1: The 3D camera scans the workpiece to be polished and uploads the defect information to the database;
[0051] Step 2: The axial-radial flexible grinding robot reads the defect position information in the database, plans the grinding sequence and grinding trajectory, and starts grinding;
[0052] Step 3: After grinding is completed, the robot returns to the standby position and waits for the next instruction.
[0053] In this embodiment, the axis-radial flexible grinding robot does not communicate directly with the camera. The database is used to store the results obtained by the 3D camera scanning. It is the medium for the axis-radial flexible grinding robot and the camera to exchange data, realize data sharing and exchange, and enable the axis-radial flexible grinding robot to obtain and process the latest scanning data in real time, so that the axis-radial flexible grinding robot can accurately obtain defect information.
[0054] In this embodiment, the axial-radial flexible grinding robot obtains defect location information by reading the data saved by the 3D camera in the database. Main process:
[0055] 1. The 3D camera scans the workpiece and obtains the 3D point cloud data of the workpiece.
[0056] 2. The host computer stores the processed and analyzed point cloud data information in the database, mainly the position coordinate points and depth information.
[0057] 3. The axial-radial flexible grinding robot obtains the defect information on the steel plate by reading the contents in the database, and then plans the grinding sequence and grinding trajectory according to the defect information.
[0058] In this embodiment, after the axial-radial flexible grinding robot obtains the locations of all the defects to be polished from the database, it needs to use an algorithm to sort and plan the order of polishing so that the total distance traveled by the polishing is as small as possible to improve efficiency. Specific steps:
[0059] Step 3.1. After the axial-radial flexible grinding robot obtains all defect position information in the database, it classifies and stores all defect data in several arrays inside the robot. By traversing the array s used to store x position data, the point with the smallest x value in array s is selected, and this point is exchanged with the first point in array s to determine the first defect to be polished. At the same time, all the defect data corresponding to x also need to be exchanged in its corresponding array to maintain the consistency of the defect data;
[0060] Step 3.2. Find the defect closest to the previous defect. This process is achieved by calculating and comparing the distance between the two defect points. The defect point is the center coordinate of the defect, which is obtained by polishing the rectangular frame;
[0061] Step 3.3. After finding the nearest defect point, swap the result with the current defect point. The defect obtained at this time is closest to the previous defect.
[0062] Step 3.4. Repeat steps 3.2 and 3.3 until the last defect point is removed.
[0063] By sorting by distance, the robot can be ensured to grind defects in order and reduce invalid distance during the robot's movement.
[0064] In this embodiment, the grinding trajectory is planned and a "U"-shaped grinding trajectory is adopted, so that the grinding depth of the outer frame is small and the depth of the inner frame is large, ensuring that the depth and size of the superimposed innermost frame meet the requirements. The interval between each frame is determined by the edge width-to-depth ratio to achieve a smooth transition of the grinding edge, prevent excessive grinding, and prevent the edges and corners from being unable to be processed, resulting in new defects, thereby ensuring the stability and consistency of the grinding effect.
[0065] like Figure 2 As shown, the grinding depth is increased by grinding in the shape of a “return” to achieve a smooth transition. The mathematical logic is as follows:
[0066] When starting, the data that needs to be transmitted to the host computer are the three points of the given defect frame grinding range , , in, , is the positioning point in the length direction, and the length is . , is the positioning point in the width direction, and the width is 。The principle of "return" - shaped grinding is to grind different frame shapes, so that the grinding depth of the central frame increases due to the superposition of the grinding times. Therefore, it is necessary to first determine the relationship between the grinding depth and the grinding times, as well as the depth of cut for a single grinding. By designing a depth - of - cut test experiment (determining a single depth of cut by taking the average value of single - time grinding and multiple - time grinding), it can be obtained that when the grinding times is the grinding depth obtained is and the relationship formula for the grinding times required to reach the depth is:
[0067] ,
[0068] If the grinding times at the deepest part of the grinding area is to reach , it should be expanded outward by frame shapes. At this time, the outermost frame is denoted as , where is the number of cycles, and its value range is . The values of its boundary three points , , are:
[0069] ,
[0070] ,
[0071] ,
[0072] ,
[0073] ,
[0074] ,
[0075] Among them, is the set distance between "return" - shaped frames, and taking the value of the grinding wheel width can better increase the universality of the logic.
[0076] The width of the th frame is:
[0077] ,
[0078] is the number of times the robot needs to translate in the direction. By the number of cycles of the internal program of the robot to meet the specified grinding width , and its value is:
[0079] ,
[0080] in, is the grinding wheel width.
[0081] In this embodiment, the method for starting grinding after the grinding sequence and grinding trajectory planning are completed includes: the axial-radial flexible grinding robot moves to the first defect position according to the sorted order, and then determines the specific track of the U-shaped grinding according to the depth information of the first defect, and then the axial-radial flexible grinding robot completes the U-shaped grinding through internal motion instructions. After completion, the axial-radial flexible grinding robot reads the position of the second defect, plans the grinding path, and starts moving. And so on, until the grinding task is completed.
[0082] In this embodiment, the positioning error compensation during grinding is:
[0083] 1. When grinding, The arc value is controlled by the robot's internal motion program. The robot has a turning radius parameter during operation, that is, the three points on the given robot path are , , , the robot from Point When the point moves, due to the influence of the turning radius, before reaching When you click on Point movement, the turning radius is At this time, there will be deviation in the actual position where the grinding wheel contacts the steel plate, and the deviation needs to be compensated to ensure positioning accuracy.
[0084] Turning radius is The trajectory equation of the 1 / 4 arc is
[0085]
[0086] in, .
[0087] When the grinding wheel contacts the steel plate, the downward pressure distance is hour,
[0088]
[0089] So the right deviation
[0090]
[0091] 2. When radial flexibility exists, a certain width deviation will occur after the grinding wheel and the steel plate are pressed tightly together, and deviation compensation is required. The compensation value .in
[0092]
[0093] is the actual grinding width, For the ideal grinding width, is the inclination angle of the steel plate.
[0094] Embodiment 2
[0095] The present invention also provides a medium and thick plate defect axial-radial flexible intelligent grinding system based on visual measurement, comprising: a scanning module, a position acquisition module, a planning module and a grinding module;
[0096] The scanning module is used to scan the workpiece to be polished using a 3D camera to obtain defect information of the workpiece to be polished, wherein the workpiece to be polished is a stainless steel medium and thick plate;
[0097] The position acquisition module is used to acquire defect position information of the workpiece to be polished based on the defect information;
[0098] The planning module is used to plan the grinding sequence and grinding trajectory based on the defect location information;
[0099] The grinding module is used to perform axial-radial flexible intelligent grinding on the workpiece based on the planned grinding sequence and grinding trajectory.
[0100] In this embodiment, the process of planning the grinding trajectory includes: using a "return"-shaped grinding trajectory.
[0101] In this embodiment, the process of obtaining the grinding depth by grinding in the shape of a “Hui” comprises:
[0102] ,
[0103] in, To reach depth The number of grinding times required, In order to design the cutting depth test experiment, the grinding times are The grinding depth obtained is
[0104] In this embodiment, the process of obtaining the grinding width by grinding in the shape of a “return” comprises:
[0105] ,
[0106] in, is the grinding wheel width, Needed by robots The number of direction shifts, For the specified grinding width.
[0107] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.
Claims
1. A flexible intelligent grinding method for axial-radial defects of medium and thick plates based on visual measurement, characterized in that: The following steps are involved: Step 1: Scanning a workpiece to be polished using a 3D camera to obtain defect information of the workpiece to be polished, wherein the workpiece to be polished is a stainless steel medium and thick plate; Step 2: Based on the defect information, obtain defect location information of the workpiece to be polished; Step 3: Based on the defect location information, plan the grinding sequence and grinding trajectory; Step 4: Based on the planned grinding sequence and grinding trajectory, the workpiece is subjected to axial-radial flexible intelligent grinding; The flexible intelligent grinding method is realized by an axial-radial flexible grinding robot. The robot system includes a robot control cabinet, an axial-radial flexible grinding robot, an end flange, an axial-radial force compensator, an electric spindle and a grinding wheel; The floating mechanism of the axial-radial force compensator is installed on the main frame and achieves floating through an electric proportional valve and a pneumatic cylinder. The electric proportional valve adjusts the air pressure in real time based on the feedback signal from the force sensor to maintain constant pressure. The pneumatic cylinder is connected to an air pump to provide an air source. The electric spindle is installed at the end of the axial-radial force compensator, and the direction, speed and grinding parameters of the electric spindle are controlled by the host computer. The axial-radial flexible force compensator structure meets the requirements of axial and radial floating and maintains constant force. In the flexible intelligent polishing method, a 3D camera scans the workpiece to be polished and uploads defect information to a database; The axial-radial flexible grinding robot reads the defect location information in the database, plans the grinding sequence and grinding trajectory, and starts grinding; After grinding is completed, the robot returns to the standby position and waits for the next instruction; The axial-radial flexible grinding robot does not communicate directly with the camera. The database is used to store the results obtained by the 3D camera scanning. It acts as an intermediary for data exchange between the axial-radial flexible grinding robot and the camera, enabling data sharing and exchange, allowing the axial-radial flexible grinding robot to obtain and process the latest scanning data in real time. The axial-radial flexible grinding robot obtains defect location information by reading the data stored by the 3D camera in the database; the process is as follows: The 3D camera scans the workpiece and obtains the three-dimensional point cloud data of the workpiece; The host computer stores the processed and analyzed point cloud data in the database, mainly including position coordinates and depth information; The axial-radial flexible grinding robot obtains defect information on the steel plate by reading the content in the database, and then plans the grinding sequence and grinding trajectory based on the defect information; After the axial-radial flexible grinding robot obtains the locations of all the defects to be polished from the database, it needs to use an algorithm to sort and plan the order of polishing. The specific steps are as follows: Step 3.
1. After the axial-radial flexible grinding robot obtains all defect location information from the database, it categorizes and stores all defect data in several arrays within the robot. It then traverses the array s used to store x-position data, selects the point with the smallest x value in array s, swaps this smallest point with the first point in array s, and identifies this as the first defect to be ground. Simultaneously, all data corresponding to the defect at x in its corresponding array must also be swapped to maintain the consistency of the defect data. Step 3.
2. Find the defect closest to the previous defect. This process is achieved by calculating and comparing the distance between the two defect points. The defect point is the center coordinate of the defect, which is obtained by polishing the rectangular frame. Step 3.
3. After finding the nearest defect point, swap the result with the current defect point. The defect obtained at this time is closest to the previous defect. Step 3.
4. Repeat steps 3.2 and 3.3 until the last defect point is found. Plan the grinding trajectory and adopt a "return" grinding trajectory to achieve a smooth transition by increasing the grinding depth through "return" grinding; When starting up, the data to be transmitted to the host computer are three points within the given defective frame grinding range , , Among them, and are the positioning points in the length direction, and the length is ; and are the positioning points in the width direction, and the width is ; The "return" character grinding is carried out by grinding different frames; due to the superposition of the grinding times, the grinding depth of the central frame increases. First, determine the relationship between the grinding depth and the grinding times, and the feed per cut for single grinding; through designing a feed per cut test experiment, it is obtained that when the grinding times is , the depth obtained by grinding is , and the relationship formula for the grinding times required to reach the depth is: , To make the deepest grinding area reach the number of grinding times , then expand outward A frame, at this time, the outermost frame is recorded as ,in, is the number of cycles, the value range is , the values of the three points on its boundary , , for: , , , , , , in, The distance between the "circular" boxes is set. Setting it to the grinding wheel width can better increase the universality of the logic. No. ( ) box width The values are: , Needed by robots The number of directional translations is determined by the number of program cycles within the robot to achieve the specified grinding width. , whose values are: , in, is the grinding wheel width; The method for starting grinding after the grinding sequence and grinding trajectory planning are completed includes: the axial-radial flexible grinding robot moves to the first defect position according to the sorted order, and then determines the specific U-shaped grinding trajectory based on the depth information of the first defect. After that, the axial-radial flexible grinding robot completes the U-shaped grinding through internal motion instructions; after completion, the axial-radial flexible grinding robot reads the position of the second defect, plans the grinding path, and starts moving; and so on until the grinding task is completed; Positioning error compensation during grinding: When polishing, The arc contacts the steel plate in the direction to form a buffer, preventing the rigid contact from leaving pits on the steel plate surface; the arc value is controlled by the robot's internal motion program. The robot has a turning radius parameter during operation, that is, the three points on the given robot path , , , the robot from Point direction When the point moves, due to the influence of the turning radius, before reaching At this point, it will start to Point motion, the turning radius is At this time, there will be deviation in the actual contact position between the grinding wheel and the steel plate, and the deviation needs to be compensated; Turning radius is The equation of the 1 / 4 arc trajectory is in, ; When the grinding wheel contacts the steel plate, the downward pressure distance is hour, So right deviation ; When radial flexibility exists, a certain width deviation will occur after the grinding wheel and the steel plate are pressed tightly together, and deviation compensation is required. The compensation value is ;in is the actual grinding width, For the ideal grinding width, is the inclination angle of the steel plate.
Citation Information
Patent Citations
Method for eliminating joint tool marks in numerical control grinding and polishing
CN110147075A
Small-curvature polishing and grinding method based on robot
CN111823100A
Intelligent manual-teaching-free grinding and polishing method and system fused with visual depth analysis
CN112757057A
Robot intelligent smooth polishing execution device based on machine vision
CN114888671A
Digital twinning-based weld defect grinding system and grinding method in narrow and small space of ship body
CN117340716A
