Intelligent grinding equipment and method for mechanically automated workpiece processing

Abstract

The present application belongs to the field of automation technology and discloses a mechanical automation workpiece processing intelligent grinding device and method, the method comprising: a workpiece clamping system clamps and fixes the workpiece; a data acquisition and analysis system collects images of the workpiece; a path planning system determines the optimal grinding path through optimization algorithms and simulations based on the weld conditions and the size and characteristics of the grinding tool; an identification and analysis system configures appropriate grinding tools according to different grinding positions; a multi-degree-of-freedom robotic arm drives the grinding assembly to grind the weld bead of the workpiece; the grinding tool is replaced by the grinding assembly; a water-cooled chip removal system effectively removes waste generated during the grinding process to ensure that the equipment does not overheat during operation. The present invention determines the optimal grinding path through optimization algorithms and simulations; configures appropriate grinding tools based on the shape and size of the weld and the planned path; and realizes rapid and automatic replacement of grinding tools through the grinding assembly, thereby improving work efficiency.

Description

Technical Field

[0001] The present application relates to the field of automation technology, and more specifically, to an intelligent grinding device and method for mechanically automated workpiece processing. Background Art

[0002] After the workpiece is welded, the weld needs to be polished. Polishing is a very important step, mainly used to remove burrs, welding slag and other surface defects that may occur during the welding process.

[0003] The document with the prior art publication number CN115709414A provides an automated grinding device for processing products, including a main frame, a workpiece, a double-station loading and unloading device, a six-axis machine shaft robot and a fixture assembly. The double-station loading and unloading device is arranged side by side on the main frame, and the fixture assembly is arranged on the double-station loading and unloading device. The main frame is provided with a dust extraction device for sucking dust and iron filings generated during the processing. The main frame is divided into a grinding area and a worker loading and unloading area. The double-station loading and unloading device is movable between the grinding area and the worker loading and unloading area. When the double-station loading and unloading device moves to the worker loading and unloading area, the workpiece to be processed is loaded and fixed by the fixture assembly and moved to the grinding area for grinding by the six-axis machine shaft robot. After the workpiece is ground, the double-station loading and unloading device is then moved to the worker loading and unloading area for unloading. The invention has the characteristics of simple and reasonable structure, excellent performance, high production efficiency, good grinding effect and low manufacturing cost.

[0004] While the aforementioned prior art solutions can achieve the relevant beneficial effects through existing technical structures, they still have the following drawbacks: 1. Different grinding tools are required for different weld structures, but traditional methods generally rely on manual experience to configure them, which cannot ensure the rationality of the configuration. 2. Traditional grinding tool replacement usually requires manual operation, which is not only time-consuming and labor-intensive, but also may cause equipment damage or reduced grinding quality due to improper operation. The inability to quickly adapt to the grinding requirements of different welds reduces work efficiency and grinding accuracy.

[0005] In view of this, we propose an intelligent grinding device and method for mechanically automated workpiece processing. Summary of the Invention

[0006] 1. Technical problems to be solved

[0007] The purpose of this application is to provide a mechanically automated workpiece processing intelligent grinding equipment and method, which solves the technical problems raised in the above-mentioned background technology, and realizes the determination of the optimal grinding path through the path planning system according to the shape, length, curvature of the weld and the size and characteristics of the grinding tool through optimization algorithm and simulation; identifies various weld structures through the recognition and analysis system, and configures suitable grinding tools according to the shape and size of the weld and the planned path; realizes the rapid and automatic replacement of grinding tools through the grinding assembly, and improves the work efficiency and grinding accuracy.

[0008] 2. Technical solution

[0009] The technical solution of this application provides a mechanical automated workpiece processing intelligent grinding device, including:

[0010] Base: A supporting structure that is strong and stable enough to support the robotic arm and other components above.

[0011] Multi-DOF Robotic Arm: Fixed above the base, the multi-DOF arm features six degrees of freedom, providing ample flexibility and precision. An A-FDC (Axial Force Displace Compliance) axial force compensator is installed at the end of the multi-DOF arm, enabling flexible force control of the tool end according to work requirements and accurately outputting contact force parallel to the arm's axis.

[0012] Control System: A control system is fixed to the base and includes a touchscreen industrial computer, a PLC (Programmable Logic Controller), and a vision system. This system directs the robot's movements, monitors the entire grinding process, and adjusts grinding parameters as needed. The touchscreen industrial computer provides an intuitive human-machine interface, allowing operators to easily monitor and control equipment operations.

[0013] Grinding assembly: fixedly set at the end of the multi-degree-of-freedom robotic arm; the grinding assembly includes grinding wheels, grinding belts and abrasive sheets, etc. Suitable grinding tools are selected according to different weld materials and requirements; different grinding tools can be quickly replaced to adapt to different grinding needs.

[0014] Data acquisition and analysis system: includes sensors and high-definition cameras; obtains information such as weld shape, size and surface quality.

[0015] Path Planning System: Based on the shape, length, curvature of the weld, and the size and characteristics of the grinding tool, the system uses optimization algorithms and simulations to determine the optimal grinding path, thereby improving grinding efficiency and quality and reducing unnecessary material waste and processing time.

[0016] The workpiece clamping system includes a dedicated fixture, a protective cover, and a water-cooled chip removal system. The dedicated fixture clamps and secures the workpiece; the protective cover protects the workpiece during the grinding process, ensuring stability and safety. The water-cooled chip removal system prevents overheating during extended operation and effectively removes waste generated during the grinding process.

[0017] Identification and Analysis System: Identifies various weld structures and configures appropriate grinding tools based on the weld shape, size, and planned path. It also continuously optimizes model parameters based on real-time data monitoring and feedback to improve recognition accuracy and grinding results.

[0018] A temperature sensor is mounted on the grinding assembly to monitor the temperature of the grinding tool in real time, preventing damage or performance degradation caused by overheating. A wired laser sensor is mounted on the multi-degree-of-freedom robotic arm to ensure it can scan the weld seam on the workpiece. Laser scanning captures the three-dimensional shape and dimensions of the weld, providing data support for the path planning system. A pressure sensor is also mounted on the grinding assembly to measure the contact force between the grinding tool and the workpiece, ensuring consistent grinding force and preventing damage to the workpiece caused by excessive force.

[0019] Displacement sensors are installed at the joints of the multi-degree-of-freedom robotic arm to monitor its motion and ensure precise movement along the preset path. Displacement sensors are fixed to the workpiece clamping system to monitor the workpiece's displacement during the grinding process and provide feedback to adjust the grinding path or force.

[0020] As an optional solution of the present invention, the grinding assembly includes: a swivel seat, a rotation drive mechanism, a cover plate, an electric push rod, an electric chuck, a pushing mechanism and a grinding tool;

[0021] A rotation drive mechanism is fixedly arranged on the multi-degree-of-freedom robotic arm;

[0022] A swivel seat is rotatably provided on the multi-degree-of-freedom robotic arm, on which an electric push rod is fixedly provided; an electric chuck is fixedly provided at the end of the movable rod of the electric push rod; the electric chuck is slidably matched with the swivel seat; the position of the electric chuck can be controlled by extending and retracting the electric push rod.

[0023] Several pushing mechanisms are fixedly arranged on the rotating seat;

[0024] A plurality of grinding tools are placed in the rotating seat, and the grinding tools can slide in the rotating seat; the pushing mechanism and the grinding tools correspond one to one.

[0025] The pushing mechanism can drive the grinding tool to move, thereby fixing the grinding tool to the electric chuck.

[0026] A cover plate is detachably fixed on the swivel seat;

[0027] With this technical solution, the push mechanism is controlled to select the desired grinding tool and push it to the electric chuck. The electric chuck grips the pushed grinding tool. The rotary drive mechanism then rotates the swivel and grinding tool to grind the workpiece. The coordination of the electric push rod, electric chuck, push mechanism, and rotary drive mechanism enables automatic selection and replacement of grinding tools, as well as flexible adjustment of the grinding position, improving work efficiency and operational convenience.

[0028] As an optional solution of the present invention, the rotation drive mechanism includes a motor and a driving gear;

[0029] The motor is fixedly arranged at the end of the multi-degree-of-freedom robotic arm, and a driving gear is coaxially fixedly arranged at the output end of the motor;

[0030] A shaft sleeve is fixedly provided on the rotating seat, a driven gear is coaxially fixedly provided on the shaft sleeve, and the driven gear is meshed with the driving gear for transmission;

[0031] Through the above technical solution, the starting motor drives the driving gear to rotate, the driving gear drives the driven gear to rotate, and the driven gear drives the swivel seat and the grinding tool to rotate to perform the grinding operation.

[0032] As an optional solution of the present invention, a plurality of chutes A and chutes B are provided on the swivel seat, and different grinding tools are placed in the chutes A and B;

[0033] The grinding tool's rotating shaft is made of high-strength carbon steel.

[0034] Pushing mechanisms are fixedly arranged in both chute A and chute B.

[0035] The pushing mechanism includes an electric telescopic rod and an electromagnet;

[0036] An electric telescopic rod is fixedly arranged in the chute A and the chute B; an electromagnet is fixedly arranged on the movable rod of the electric telescopic rod;

[0037] With this technical solution, an electric telescopic rod drives the electromagnet, which in turn moves the grinding tool to the electric chuck. The electromagnet is used to hold the grinding tool. When the grinding tool needs to be replaced or positioned, the electric telescopic rod drives the electromagnet to a specific position, which in turn moves the grinding tool to the electric chuck. The electric chuck then secures the grinding tool.

[0038] As an optional solution of the present invention, the data acquisition and analysis system includes:

[0039] Data collection module: collects workpiece dimension data, including welding requirements; collects images of different welding methods and annotates them as reference samples;

[0040] Image acquisition module: includes a high-definition camera and LED lights to capture high-definition images of the workpiece and the welding area;

[0041] Image preprocessing module: preprocesses the collected images, including filtering and denoising, image changes, image enhancement and image restoration;

[0042] Feature extraction module: performs feature extraction on the image after feature extraction to extract features related to the weld bead, including color, shape, and texture;

[0043] Weld bead recognition module: Identifies the shape, size, and position of welds based on the image after feature extraction; identifies the type (such as butt weld, fillet weld, etc.), shape (such as straight line, curved shape, etc.), and size (such as length, width, depth, etc.) of welds;

[0044] Shape recognition: Use shape descriptors such as circularity, rectangularity, Feret diameter, etc. to quantify the shape of the weld bead. Circularity = (4πA) / P 2 Where A is the area of ​​the weld bead, P is the perimeter of the weld bead, and Circularity is the circularity.

[0045] Dimension measurement: Calculate the dimensions of the weld, such as length, width, etc.

[0046]

[0047] Width=Max x (y upper -y lower ); where Length is the length of the weld bead. x, y are the coordinates of the weld bead in the image. (Dx) / (dy) is the derivative of the weld bead curve, representing the weld bead curvature. Width is the width of the weld bead. y upper ,y lower are the coordinates of the upper and lower edges of the weld bead.

[0048] Positioning: Determine the position of the weld in the image, which can be achieved through feature matching or image coordinate transformation.

[0049] X new =T 11 x+T 12 y+T13;

[0050] y new =T 21 x+T 22 y+T23;

[0051] Where x new， y new is the new coordinate after transformation.

[0052] T 11 ,T 12 ,T 13 ,T 21 ,T 22 ,T 23 are the elements of the coordinate transformation matrix.

[0053] Feature fusion: Combine shape, size and position information to identify weld features.

[0054] Feature vector construction: f=[f1,f2,…,f n ]; where f i is the i-th feature, which can be a shape, size or position feature. f is a feature vector containing all extracted features.

[0055] Model optimization module: Use machine learning algorithms to optimize recognition results and improve recognition accuracy. Machine learning model training: Where θ is the model parameter and N is the number of training samples. is the loss function, which measures the difference between the model prediction and the actual value. Ω is the regularization term, which is used to prevent the model from overfitting. i represents the i-th sample. y i is the true label or output of the i-th sample. i The output of the i-th sample predicted by the model.

[0056] The judgment results are evaluated based on the model output and confidence level:

[0057] Confidence = p(f|Weld) / p(f); where Confidence is the confidence level of the recognition result. p(f|Weld) is the posterior probability of a given weld feature. p(f) is the prior probability of the feature.

[0058] Output weld bead recognition results, including shape, size, position and feature description.

[0059] In this technical solution, the weld bead recognition technology is used to describe the shape, size, position, and characteristics of the weld bead, as well as the mathematical calculations and model training during the recognition process. This can achieve accurate quantification and recognition of weld bead characteristics, thereby improving the automation and accuracy of welding quality inspection.

[0060] As an optional solution of the present invention, the feature extraction module performs edge detection by the following method:

[0061] Step 1. Select the smooth function θ(x) as the scaling function. Correspondingly, the first-order and second-order derivatives of the function θ(x) are wavelet functions, which constitute a multi-scale wavelet transform to analyze the characteristics of the image at different scales.

[0062] θ(x)=(8 / 3)[(x+1) 3 u(x+1)-4(x+1 / 2) 3 u(x+1 / 2)+6x 3 u(x)-4(x-1 / 2) 3 u(x-1 / 2)

[0063] +(x-1) 3 u(x-1)];

[0064] Where θ(x) represents the selected smoothing function, which is used as the scaling function in the wavelet transform. u(x) represents the unit step function, which divides the points on the real line into two regions and is often used to define the domain of a function. x represents the independent variable, which represents the coordinate position or scale in the image.

[0065] Step 2: Set the decomposition level and the threshold ε (including zero-crossing detection); ε represents the threshold in wavelet transform, which is used to determine the importance of wavelet coefficients.

[0066] Step 3: Perform wavelet transform on the image and calculate the gradient direction and gradient vector modulus at different scales based on the wavelet coefficients;

[0067] Step 4: Determine the edge: Find the zero crossing points from the wavelet coefficients of each row and column of the image, and find the maximum point of each two adjacent zero crossing points. The point where the maximum value appears in both the row and column is set as the edge point.

[0068] Step 5: According to certain rules, edge points are connected to form boundaries.

[0069] As an optional solution of the present invention, the path planning system includes the following steps for planning the polishing path:

[0070] (1) Detailed analysis of the geometric characteristics of the weld, including its shape, length and curvature. This information is crucial for determining the grinding path because it directly determines the motion trajectory of the grinding tool and the grinding effect. Curvature calculation: Where v is the velocity vector of the weld centerline, is the curl operator. κ is the curvature, which indicates the degree of curvature of the weld centerline.

[0071] (2) Adapt according to the size and characteristics of the grinding tool. The size, shape, material and other characteristics of the grinding tool will directly affect its grinding effect and efficiency. Therefore, the system needs to ensure that the grinding tool matches the geometric characteristics of the weld to achieve the best grinding effect. The matching degree between tool size and weld size: M = 1-[(D tool -D weld ) / D max ]; among them, Dtool is the grinding tool diameter, D weld is the weld diameter, D max is the maximum allowed diameter difference, and M is the degree of matching.

[0072] (3) After obtaining relevant information about the weld and grinding tools, the path planning system uses optimization algorithms and simulations to determine the optimal grinding path. This helps the system predict the grinding effect, time, and material utilization under different grinding paths, thereby selecting the optimal solution.

[0073] Minimize path length: p opt =argmin{∫ tf t0 [|v(t)|dt)]}; where p opt is the optimal path, v(t) is the velocity vector at time t. ∫ tf t0 [|v(t)|dt)] is the integral of the absolute value of the velocity vector from t0 to tf, which represents the path length.

[0074] Processing time prediction: T = Total Path Length / Polishing Speed; where T is the predicted processing time and Polishing Speed ​​is the set polishing advance speed.

[0075] (4) Through the optimization algorithm, a path is found that can meet the grinding requirements while minimizing processing time and material waste. Through simulation, the actual grinding process can be simulated to verify whether the path obtained by the optimization algorithm is effective.

[0076] As an optional solution of the present invention, an optimization algorithm is used to find a path that can meet the grinding requirements while minimizing processing time and material waste, including the following steps:

[0077] (1) Time assessment: clarify the specific indicators such as polishing quality and surface finish that need to be achieved.

[0078] Evaluate the machining time for different paths and seek to minimize it. Consider the amount of material removal that may occur during the grinding process and seek to minimize it.

[0079] (2) Establish a mathematical model: Clarify the specific requirements and constraints for polishing. Select an appropriate mathematical method (such as linear programming) to describe the problem. Establish an objective function to measure the quality of the solution. Use a solution algorithm (such as gradient descent) to find the optimal solution.

[0080] Variables: include the position, speed, direction, etc. of the grinding tool, which change over time to describe the motion trajectory of the grinding tool.

[0081] (3) Select optimization algorithm: Select the optimization algorithm of genetic algorithm and search for the optimal or near-optimal solution.

[0082] (4) Implement simulation: Use simulation software (such as MATLAB Simulink, COMSOL Multiphysics, etc.) to simulate the actual grinding process and verify whether the path derived by the optimization algorithm is effective. The simulation model should be as close as possible to the actual physical process, including the movement of the grinding tool and the material removal rate.

[0083] (5) Iterative optimization: Adjust the parameters of the optimization algorithm or change the mathematical model based on the simulation results to improve the quality of the solution. Repeat the simulation and optimization until a satisfactory solution is found.

[0084] This technical solution enables path planning and optimization during the automatic welding process, which not only improves welding efficiency and quality, but also reduces material waste and processing time.

[0085] As an optional solution of the present invention, the identification and analysis system includes: a grinding tool configuration module and a real-time data monitoring and feedback module;

[0086] Grinding tool configuration module: Automatically select or configure the appropriate grinding tool based on the weld type, shape, size, and planned grinding path. Build a grinding tool library containing a variety of types, sizes, and materials to meet the grinding needs of different welds.

[0087] Real-time Data Monitoring and Feedback Module: This module collects real-time data during the polishing process, such as polishing force, temperature, and vibration. Using data analysis algorithms, it processes and analyzes the collected data to evaluate polishing performance and quality. Based on this feedback, it dynamically adjusts polishing parameters and model parameters to improve recognition accuracy and polishing results.

[0088] The present invention provides a mechanically automated workpiece processing intelligent grinding method, comprising the following steps:

[0089] S1. Clamp and fix the workpiece through the workpiece clamping system and special tooling for the workpiece;

[0090] S2. The data acquisition and analysis system collects images of the workpiece to obtain information such as the shape, size, and surface quality of the weld.

[0091] S3. The path planning system determines the optimal grinding path based on the shape, length, curvature of the weld and the size and characteristics of the grinding tool through optimization algorithms and simulation to improve grinding efficiency and quality.

[0092] S4. The recognition and analysis system configures appropriate grinding tools for different grinding positions based on the shape and size of the weld and the planned path;

[0093] S5. The pushing mechanism pushes the grinding tool to the electric chuck, which then holds the tool. The control system controls the multi-degree-of-freedom robotic arm to drive the grinding assembly and grind the weld of the workpiece according to the planned path.

[0094] S6. When the grinding tool needs to be replaced, the electric push rod first drives the electric chuck back into the swivel seat to release the clamping of the current grinding tool; the electric chuck retreats a certain distance, the front grinding tool is separated from the electric chuck, and the corresponding pushing mechanism takes the front grinding tool back and stores it; then, according to the grinding needs, the pushing mechanism pushes the grinding tool to be used in the next step to the corresponding position of the electric chuck, and the electric push rod drives the electric chuck to advance a certain distance so that the clamping section of the grinding tool is in the electric chuck, and the grinding tool is clamped and fixed by the electric chuck to complete the replacement of the grinding tool; the electric push rod drives the electric chuck and the grinding tool to extend out of the swivel seat; the rotation drive mechanism drives the grinding tool to rotate, and the grinding operation continues;

[0095] S7. While grinding the workpiece, the water-cooled chip removal system effectively removes the waste generated during the grinding process, ensuring that the equipment will not overheat during long-term operation.

[0096] 3. Beneficial effects

[0097] One or more technical solutions provided in the technical solution of this application have at least the following technical effects or advantages:

[0098] 1. The present invention uses a path planning system to determine the optimal grinding path based on the shape, length, curvature of the weld and the size and characteristics of the grinding tool through optimization algorithms and simulations to improve grinding efficiency and quality, reducing unnecessary material waste and processing time.

[0099] 2. Identify various weld structures through the recognition and analysis system, configure appropriate grinding tools according to the shape and size of the weld and the planned path; and continuously optimize model parameters based on real-time data monitoring and feedback to improve recognition accuracy and grinding effect.

[0100] 3. The grinding tool can be quickly and automatically replaced through the grinding assembly. Different grinding tools can be configured for different weld grinding needs, thereby improving work efficiency and grinding accuracy.

[0101] 4. The weld bead recognition technology solution is used to describe the shape, size, position, and characteristics of the weld bead, as well as the mathematical calculations and model training during the recognition process. This can achieve accurate quantification and recognition of weld bead characteristics, thereby improving the automation and accuracy of welding quality inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0102] Figure 1 This is an overall schematic diagram of a mechanically automated workpiece processing intelligent grinding device disclosed in a preferred embodiment of the present application;

[0103] Figure 2 This is a schematic diagram of the coordination of a multi-degree-of-freedom robotic arm and a grinding assembly of a mechanically automated workpiece processing intelligent grinding device disclosed in a preferred embodiment of the present application;

[0104] Figure 3 A schematic diagram of a grinding component of a mechanically automated workpiece processing intelligent grinding device disclosed in a preferred embodiment of the present application;

[0105] Figure 4 This is a schematic diagram of the interior of a grinding component of a mechanically automated workpiece processing intelligent grinding device disclosed in a preferred embodiment of the present application;

[0106] Figure 5 A schematic diagram of a grinding tool of a mechanically automated workpiece processing intelligent grinding device disclosed in a preferred embodiment of the present application;

[0107] Figure 6 A schematic diagram of a push mechanism of a mechanically automated workpiece processing intelligent grinding device disclosed in a preferred embodiment of the present application;

[0108] Figure 7 This is a schematic diagram of the rotary table of the mechanical automated workpiece processing intelligent grinding equipment disclosed in a preferred embodiment of the present application.

[0109] Reference numerals:

[0110] 1. Base; 2. Multi-degree-of-freedom robotic arm; 3. Grinding assembly; 4. Fixed seat; 5. Workpiece clamping system; 31. Turntable; 311. Bushing; 312. Driven gear; 313. Slide A; 314. Slide B; 32. Rotation drive mechanism; 321. Motor; 322. Driving gear; 33. Cover; 34. Electric push rod; 35. Electric chuck; 36. Pushing mechanism; 361. Electric telescopic rod; 362. Electromagnet; 37. Grinding tool. DETAILED DESCRIPTION

[0111] The present application is further described in detail below with reference to the accompanying drawings.

[0112] Reference Figure 1 and Figure 2 , the embodiment of the present application provides a mechanical automated workpiece processing intelligent grinding device, comprising: a base 1, a multi-degree-of-freedom robotic arm 2, a grinding component 3, a fixed seat 4, a workpiece clamping system 5, a control system, a grinding component, a data acquisition and analysis system, a path planning system, a workpiece clamping system and an identification and analysis system;

[0113] Base 1: A supporting structure that is strong and stable enough to support the robotic arm and other components above.

[0114] Multi-DOF Robotic Arm 2: Fixed above Base 1, Multi-DOF Robotic Arm 2 features six degrees of freedom, providing sufficient flexibility and precision. The grinding tool is mounted on the robot, which moves along a pre-set path.

[0115] The end of the multi-degree-of-freedom robotic arm is equipped with an Axial Force Displace Compliance (A-FDC) axial force displacement compensator, which allows flexible force control of the end tool according to work requirements and accurately outputs contact force parallel to the robotic arm's axis. This technology solves the automation challenge of balancing contact surface-sensitive feature processing with rapid contact movement. Multi-degree-of-freedom robotic arm 2 is prior art and is only borrowed in this invention, so it will not be described in detail here.

[0116] Control System: Base 1 is fixed with a control system comprising a touchscreen industrial computer, a PLC (Programmable Logic Controller), and a vision system. This system directs the robot's movements, monitors the entire polishing process, and adjusts polishing parameters as needed. The touchscreen industrial computer provides an intuitive human-machine interface, allowing operators to monitor and control equipment operation. This control system is prior art and is only used in this invention, so it will not be described in detail here.

[0117] Grinding assembly 3: fixedly set at the end of the multi-degree-of-freedom robot arm 2; grinding assembly 3 includes grinding wheels, grinding belts and abrasive sheets, etc. Suitable grinding tools are selected according to different weld materials and requirements; different grinding tools can be quickly replaced to meet different grinding needs.

[0118] Data acquisition and analysis system: including sensors and high-definition cameras;

[0119] Sensors and high-definition cameras: Used to obtain information such as the shape, size, and surface quality of the weld. This data can be used for subsequent path planning and determination of grinding parameters.

[0120] Through the touch screen industrial computer, you can observe fault alarms and diagnostic prompts, current processing wheel type, calculate single-piece cycle time, and count the number of processes during the shift.

[0121] Path Planning System: Based on the shape, length, curvature of the weld, and the size and characteristics of the grinding tool, the system uses optimization algorithms and simulations to determine the optimal grinding path, thereby improving grinding efficiency and quality and reducing unnecessary material waste and processing time.

[0122] Workpiece clamping system 5: including workpiece-specific fixtures, protective covers and water-cooled chip removal system;

[0123] Special tooling for the workpiece is used to clamp and fix the workpiece; the protective cover protects the workpiece during the grinding process, ensuring the stability and safety of the workpiece during the grinding process.

[0124] Water-cooled chip removal system: ensures that the equipment will not overheat during long-term operation and effectively removes waste generated during the grinding process.

[0125] Identification and Analysis System: Identifies various weld structures and configures appropriate grinding tools based on the weld shape, size, and planned path. It also continuously optimizes model parameters based on real-time data monitoring and feedback to improve recognition accuracy and grinding results.

[0126] A temperature sensor is fixedly provided on the grinding component 3 to monitor the temperature of the grinding tool in real time to avoid damage or performance degradation caused by overheating.

[0127] A wired laser sensor is fixed to the multi-degree-of-freedom robotic arm 2 to ensure that it can scan the weld seam on the workpiece. The laser scanning obtains the three-dimensional shape and size information of the weld seam, providing data support for the path planning system.

[0128] A pressure sensor is fixedly provided on the grinding assembly 3 to measure the contact force between the grinding tool and the workpiece, thereby ensuring the stability of the grinding force and avoiding damage to the workpiece caused by excessive force.

[0129] Displacement sensors are installed at the joints of the multi-degree-of-freedom robotic arm 2 to monitor its motion and ensure precise movement along the preset path. A displacement sensor is also fixed to the workpiece clamping system 5 to monitor the workpiece's displacement during the grinding process and provide feedback to adjust the grinding path or grinding force.

[0130] Reference Figure 3 、 Figure 4 and Figure 5 The grinding assembly includes: a rotating seat 31, a rotating driving mechanism 32, a cover plate 33, an electric push rod 34, an electric chuck 35, a pushing mechanism 36 and a grinding tool 37;

[0131] A rotation drive mechanism 32 is fixedly provided on the multi-degree-of-freedom robotic arm 2;

[0132] A swivel seat 31 is rotatably provided on the multi-degree-of-freedom robotic arm 2, and an electric push rod 34 is fixedly provided on the swivel seat 31; an electric chuck 35 is fixedly provided at the end of the movable rod of the electric push rod 34; the electric chuck 35 is slidably fitted with the swivel seat 31; the position of the electric chuck 35 can be controlled by extending and retracting the electric push rod 34.

[0133] Several pushing mechanisms 36 are fixedly arranged on the rotating seat 31;

[0134] A plurality of grinding tools 37 are placed in the rotating seat 31 , and the grinding tools 37 can slide in the rotating seat 31 ; the pushing mechanism 36 and the grinding tools 37 correspond one to one.

[0135] The pushing mechanism 36 can drive the grinding tool 37 to move, thereby fixing the grinding tool 37 to the electric chuck 35 .

[0136] A cover plate 33 is detachably fixed on the rotating seat 31;

[0137] In this technical solution, by controlling the pushing mechanism 36, the grinding tool 37 to be used is selected and pushed to the electric chuck 35. The electric chuck 35 clamps the pushed grinding tool 37. Then the rotation drive mechanism 32 drives the swivel 31 and the grinding tool 37 to rotate to perform the grinding operation on the workpiece. The multi-degree-of-freedom robot arm 2 moves the grinding tool 37 to the position of the workpiece to be ground according to the preset trajectory or the operator's instructions to perform the grinding operation. After the grinding is completed, the electric chuck 35 releases the grinding tool 37, and the pushing mechanism 36 pushes it back to its original position or a new position, ready for the next use. Through the cooperation of the electric push rod, the electric chuck, the pushing mechanism and the rotation drive mechanism, the automatic selection and replacement of the grinding tool and the flexible adjustment of the grinding position are realized, thereby improving the work efficiency and the convenience of operation.

[0138] Reference Figure 6 , the rotation driving mechanism 32 includes a motor 321 and a driving gear 322;

[0139] The motor 321 is fixedly arranged at the end of the multi-degree-of-freedom robotic arm 2, and a driving gear 322 is coaxially fixedly arranged at the output end of the motor 321;

[0140] A shaft sleeve 311 is fixedly provided on the rotating seat 31, and a driven gear 312 is coaxially fixedly provided on the shaft sleeve 311. The driven gear 312 is meshed with the driving gear 322 for transmission.

[0141] In this technical solution, the starting motor 321 drives the driving gear 322 to rotate, the driving gear 322 drives the driven gear 312 to rotate, and the driven gear 312 drives the rotating seat 31 and the grinding tool 37 to rotate to perform the grinding operation.

[0142] Reference Figure 5 and Figure 7 , a plurality of chutes A313 and chutes B314 are provided on the rotating seat 31, and different grinding tools 37 are placed in the chutes A313 and the chutes B314;

[0143] The rotating rod of grinding tool 37 is made of high-strength carbon steel. The high strength of carbon steel enables the rod to withstand the high torque and impact forces generated during the grinding process, ensuring the tool's stability and durability. Since the rod rubs against the workpiece surface during the grinding process, using high-strength carbon steel significantly improves its wear resistance and extends the tool's service life.

[0144] A pushing mechanism 36 is fixedly provided in both the chute A 313 and the chute B 314 .

[0145] The pushing mechanism 36 includes an electric telescopic rod 361 and an electromagnet 362;

[0146] An electric telescopic rod 361 is fixedly provided in the chute A 313 and the chute B 314 ; an electromagnet 362 is fixedly provided on the movable rod of the electric telescopic rod 361 ;

[0147] In this technical solution, an electric telescopic rod 361 drives an electromagnet 362, which in turn moves the grinding tool 37 to the electric chuck 35. The electromagnet 362 is used to attract the grinding tool 37. When the grinding tool needs to be replaced or positioned, the electric telescopic rod 361 drives the electromagnet 362 to a specific position, thereby moving the grinding tool 37 to the electric chuck 35. The electric chuck 35 holds the grinding tool 37 in place, ensuring its stability and safety during the grinding process. This improves the efficiency of grinding tool replacement and reduces manual operation time.

[0148] Furthermore, the data acquisition and analysis system includes:

[0149] Data collection module: collects workpiece dimension data, including welding requirements; collects images of different welding methods and annotates them as reference samples;

[0150] Image acquisition module: includes a high-definition camera and LED lights to capture high-definition images of the workpiece and the welding area;

[0151] Image preprocessing module: preprocesses the collected images, including filtering and denoising, image changes, image enhancement and image restoration;

[0152] Feature extraction module: performs feature extraction on the image after feature extraction to extract features related to the weld bead, including color, shape, and texture;

[0153] Weld bead recognition module: Identifies the shape, size, and position of welds based on the image after feature extraction; identifies the type (such as butt weld, fillet weld, etc.), shape (such as straight line, curved shape, etc.), and size (such as length, width, depth, etc.) of welds;

[0154] Shape recognition: Use shape descriptors such as circularity, rectangularity, Feret diameter, etc. to quantify the shape of the weld bead. Circularity = (4πA) / P 2 Where A is the area of ​​the weld bead, P is the perimeter of the weld bead, and Circularity is the circularity.

[0155] Dimension measurement: Calculate the dimensions of the weld, such as length, width, etc.

[0156]

[0157] Width=Max x (y upper -y lower ); where Length is the length of the weld bead. x, y are the coordinates of the weld bead in the image. (Dx) / (dy) is the derivative of the weld bead curve, representing the weld bead curvature. Width is the width of the weld bead. y upper ,y lower are the coordinates of the upper and lower edges of the weld bead.

[0158] Positioning: Determine the position of the weld in the image, which can be achieved through feature matching or image coordinate transformation.

[0159] X new =T 11 x+T 12 y+T13;

[0160] y new =T 21 x+T 22 y+T23;

[0161] Where x new， y new is the new coordinate after transformation.

[0162] T 11 ,T 12 ,T 13 ,T 21 ,T 22 ,T 23 are the elements of the coordinate transformation matrix.

[0163] Feature fusion: Combine shape, size and position information to identify weld features.

[0164] Feature vector construction: f=[f1,f2,…,f n ]; where f i is the i-th feature, which can be a shape, size or position feature. f is a feature vector containing all extracted features.

[0165] Model optimization module: Use machine learning algorithms to optimize recognition results and improve recognition accuracy. Machine learning model training: Where θ is the model parameter and N is the number of training samples. is the loss function, which measures the difference between the model prediction and the actual value. Ω is the regularization term, which is used to prevent the model from overfitting. i represents the i-th sample. y i is the true label or output of the i-th sample. i The output of the i-th sample predicted by the model.

[0166] The judgment results are evaluated based on the model output and confidence level:

[0167] Confidence = p(f|Weld) / p(f); where Confidence is the confidence level of the recognition result. p(f|Weld) is the posterior probability of a given weld feature. p(f) is the prior probability of the feature.

[0168] Output weld bead recognition results, including shape, size, position and feature description.

[0169] In this technical solution, the weld bead recognition technology is used to describe the shape, size, position, and characteristics of the weld bead, as well as the mathematical calculations and model training during the recognition process. This can achieve accurate quantification and recognition of weld bead characteristics, thereby improving the automation and accuracy of welding quality inspection.

[0170] Furthermore, the feature extraction module performs edge detection by the following method:

[0171] Step 1. Select the smooth function θ(x) as the scaling function. Correspondingly, the first-order and second-order derivatives of the function θ(x) are wavelet functions, which constitute a multi-scale wavelet transform to analyze the characteristics of the image at different scales.

[0172] θ(x)=(8 / 3)[(x+1) 3 u(x+1)-4(x+1 / 2) 3 u(x+1 / 2)+6x 3 u(x)-4(x-1 / 2) 3 u(x-1 / 2)

[0173] +(x-1) 3 u(x-1)];

[0174] Where θ(x) represents the selected smoothing function, which is used as the scaling function in the wavelet transform. u(x) represents the unit step function, which divides the points on the real line into two regions and is often used to define the domain of a function. x represents the independent variable, which represents the coordinate position or scale in the image.

[0175] Step 2: Set the decomposition level and the threshold ε (including zero-crossing detection); ε represents the threshold in wavelet transform, which is used to determine the importance of wavelet coefficients.

[0176] Step 3: Perform wavelet transform on the image and calculate the gradient direction and gradient vector modulus at different scales based on the wavelet coefficients;

[0177] Step 4: Determine the edge: Find the zero crossing points from the wavelet coefficients of each row and column of the image, and find the maximum point of each two adjacent zero crossing points. The point where the maximum value appears in both the row and column is set as the edge point.

[0178] Step 5: According to certain rules, edge points are connected to form boundaries.

[0179] Furthermore, the path planning system includes the following steps for polishing path planning:

[0180] (1) Detailed analysis of the geometric characteristics of the weld, including its shape, length and curvature. This information is crucial for determining the grinding path because it directly determines the motion trajectory of the grinding tool and the grinding effect. Curvature calculation: Where v is the velocity vector of the weld centerline, is the curl operator. κ is the curvature, which indicates the degree of curvature of the weld centerline.

[0181] (2) Adapt according to the size and characteristics of the grinding tool. The size, shape, material and other characteristics of the grinding tool will directly affect its grinding effect and efficiency. Therefore, the system needs to ensure that the grinding tool matches the geometric characteristics of the weld to achieve the best grinding effect. The matching degree between tool size and weld size: M = 1-[(D tool -D weld ) / D max ]; among them, D tool is the grinding tool diameter, D weld is the weld diameter, D max is the maximum allowed diameter difference, and M is the degree of matching.

[0182] (3) After obtaining relevant information about the weld and grinding tools, the path planning system uses optimization algorithms and simulations to determine the optimal grinding path. This helps the system predict the grinding effect, time, and material utilization under different grinding paths, thereby selecting the optimal solution.

[0183] Minimize path length: p opt =argmin{∫ tf t0 [|v(t)|dt)]}; where p optis the optimal path, v(t) is the velocity vector at time t. ∫ tf t0 [|v(t)|dt)] is the integral of the absolute value of the velocity vector from t0 to tf, which represents the path length.

[0184] Processing time prediction: T = Total Path Length / Polishing Speed; where T is the predicted processing time and Polishing Speed ​​is the set polishing speed. Total Path Length is the length of the weld bead that needs to be polished.

[0185] (4) Through the optimization algorithm, a path is found that can meet the grinding requirements while minimizing processing time and material waste. Through simulation, the actual grinding process can be simulated to verify whether the path obtained by the optimization algorithm is effective.

[0186] Furthermore, an optimization algorithm is used to find a path that can meet the grinding requirements while minimizing processing time and material waste, including the following steps:

[0187] (1) Time assessment: clarify the specific indicators such as polishing quality and surface finish that need to be achieved.

[0188] Evaluate the machining time for different paths and seek to minimize it. Consider the amount of material removal that may occur during the grinding process and seek to minimize it.

[0189] (2) Establish a mathematical model: Clarify the specific requirements and constraints for polishing. Select an appropriate mathematical method (such as linear programming) to describe the problem. Establish an objective function to measure the quality of the solution. Use a solution algorithm (such as gradient descent) to find the optimal solution.

[0190] Variables: include the position, speed, direction, etc. of the grinding tool, which change over time to describe the motion trajectory of the grinding tool.

[0191] Constraints: including grinding requirements, tool characteristics, and material properties;

[0192] Polishing requirements: such as surface finish, polishing depth, etc.

[0193] Tool characteristics: such as size, shape, power, etc.

[0194] Material properties: such as hardness, wear resistance, etc.

[0195] Objective function: Minimize processing time and minimize material waste.

[0196] Objective function: J(x) = w1*T process +w2*V removal ;

[0197] Constraints: gi(x)≤0,i=1,…,m;

[0198] Where J(x) is the objective function, which is used to measure the quality of the solution and usually needs to be minimized or maximized. x is the decision variable vector, which contains all the variables that need to be optimized, such as the position, speed, and direction of the grinding tool. w1 and w2 are weight coefficients used to balance the importance of different parts of the objective function. process is the processing time, which represents the total time required to complete the grinding. removal is the material removal amount, representing the volume or mass of material removed during the grinding process. gi(x) is the functional form of the i-th constraint, used to ensure that the solution meets all requirements. m is the number of constraints, indicating how many constraints are considered in the model.

[0199] (3) Select optimization algorithm: Select the optimization algorithm of genetic algorithm and search for the optimal or near-optimal solution.

[0200] (4) Implement simulation: Use simulation software (such as MATLAB Simulink, COMSOL Multiphysics, etc.) to simulate the actual grinding process and verify whether the path derived by the optimization algorithm is effective. The simulation model should be as close as possible to the actual physical process, including the movement of the grinding tool and the material removal rate.

[0201] (5) Iterative optimization: Adjust the parameters of the optimization algorithm or change the mathematical model based on the simulation results to improve the quality of the solution. Repeat the simulation and optimization until a satisfactory solution is found.

[0202] This technical solution enables path planning and optimization during the automatic welding process, which not only improves welding efficiency and quality, but also reduces material waste and processing time.

[0203] Furthermore, the identification and analysis system includes: a grinding tool configuration module and a real-time data monitoring and feedback module;

[0204] The Grinding Tool Configuration Module automatically selects or configures the appropriate grinding tool based on the weld type, shape, and size, as well as the planned grinding path. A grinding tool library is established, containing a variety of grinding tools of various types, sizes, and materials to meet the grinding needs of different welds. Based on the recognition results and grinding requirements, the optimal grinding tool is automatically selected and its parameters (such as speed and feed rate) are adjusted. The Path Planning Module, using a path planning algorithm, calculates the optimal grinding trajectory and speed based on the weld shape and size to ensure grinding quality and efficiency. The Path Planning Module works in conjunction with the Grinding Tool Configuration Module to ensure that the grinding tool efficiently operates along the planned path. After identifying the weld structure, the recognition and analysis system automatically configures the appropriate grinding tool based on the weld shape and size and the planned grinding path. The system then matches the data in the tool library to select the most suitable grinding tool for the weld, such as a grinding wheel, grinding head, or polishing pad, and adjusts its parameters (such as speed and feed rate) to suit the specific grinding requirements.

[0205] Real-time Data Monitoring and Feedback Module: This module collects data from the grinding process, such as grinding force, temperature, and vibration, in real time. Data analysis algorithms are used to process and analyze the collected data to evaluate the grinding effect and quality. Based on the feedback data, grinding parameters and model parameters are dynamically adjusted to improve recognition accuracy and grinding results. During the grinding process, the recognition and analysis system collects various data, such as grinding force, temperature, and vibration, in real time and monitors and analyzes this data. If the data is abnormal or does not meet expectations, the system will immediately issue an alarm and take appropriate measures, such as reducing the rotation speed or pausing grinding, to avoid damage to the workpiece.

[0206] The present invention provides a mechanically automated workpiece processing intelligent grinding method, comprising the following steps:

[0207] S1, clamping and fixing the workpiece by the workpiece clamping system 5 and the workpiece special tooling;

[0208] S2. The data acquisition and analysis system collects images of the workpiece to obtain information such as the shape, size, and surface quality of the weld.

[0209] S3. The path planning system determines the optimal grinding path based on the shape, length, curvature of the weld and the size and characteristics of the grinding tool through optimization algorithms and simulation to improve grinding efficiency and quality.

[0210] S4. The recognition and analysis system configures appropriate grinding tools for different grinding positions based on the shape and size of the weld and the planned path;

[0211] S5: The pushing mechanism 36 pushes the grinding tool 37 to be used to the electric chuck 35, and the electric chuck 35 clamps the pushed grinding tool 37. The control system controls the multi-degree-of-freedom robot arm 2 to drive the grinding assembly 3 to grind the weld of the workpiece according to the planned path;

[0212] S6. When the grinding tool needs to be replaced, the electric push rod 34 first drives the electric chuck 35 to retreat into the swivel seat 31, and releases the clamping of the current grinding tool 37; the electric chuck 35 retreats a certain distance, and the front grinding tool 37 is separated from the electric chuck 35, and the corresponding pushing mechanism 36 takes the front grinding tool 37 back and stores it; then, according to the grinding needs, the pushing mechanism 36 pushes the grinding tool 37 to be used in the next step to the corresponding position of the electric chuck 35, and the electric push rod 34 drives the electric chuck 35 to advance a certain distance so that the clamping section of the grinding tool 37 is in the electric chuck 35, and the grinding tool 37 is clamped and fixed by the electric chuck 35, completing the replacement of the grinding tool 37; the electric push rod 34 drives the electric chuck 35 and the grinding tool 37 to extend out of the swivel seat 31; the rotation drive mechanism 32 drives the grinding tool 37 to rotate, and the grinding operation continues;

[0213] S7. While grinding the workpiece, the water-cooled chip removal system effectively removes the waste generated during the grinding process, ensuring that the equipment will not overheat during long-term operation.

[0214] The operating principle of the present invention's automated workpiece processing intelligent grinding equipment is as follows: the workpiece is clamped and fixed by a workpiece clamping system 5 and dedicated workpiece fixtures; a data acquisition and analysis system captures images of the workpiece to obtain information such as the weld's shape, size, and surface quality. A path planning system uses optimization algorithms and simulations to determine the optimal grinding path based on the weld's shape, length, and curvature, as well as the size and characteristics of the grinding tool, to improve grinding efficiency and quality. The identification and analysis system configures appropriate grinding tools for different grinding locations based on the weld's shape and size and the planned path; and a pushing mechanism 36 pushes the desired grinding tool 37 to the electric chuck 35, which then grips the pushed grinding tool 37. The control system controls the multi-degree-of-freedom robotic arm 2 to drive the grinding assembly 3 to grind the weld of the workpiece according to the planned path; when the grinding tool needs to be replaced, the electric push rod 34 first drives the electric chuck 35 to retreat into the swivel seat 31, releasing the clamping of the current grinding tool 37; the electric chuck 35 then retreats a certain distance, the front grinding tool 37 is separated from the electric chuck 35, and the corresponding pushing mechanism 36 takes the front grinding tool 37 back and stores it; then, according to the grinding needs, the pushing mechanism 36 pushes the grinding tool 37 to be used in the next step to the corresponding position of the electric chuck 35, The electric push rod 34 drives the electric chuck 35 to advance a certain distance, so that the clamping section of the grinding tool 37 is in the electric chuck 35, and the grinding tool 37 is clamped and fixed by the electric chuck 35 to complete the replacement of the grinding tool 37; the electric push rod 34 drives the electric chuck 35 and the grinding tool 37 to extend out of the turntable 31; the rotation drive mechanism 32 drives the grinding tool 37 to rotate and continue the grinding operation; while the workpiece is being ground, the water-cooled chip removal system effectively removes the waste generated during the grinding process to ensure that the equipment will not overheat during long-term operation.

[0215] The present invention uses a path planning system to determine the optimal grinding path based on the shape, length, curvature of the weld and the size and characteristics of the grinding tool through optimization algorithms and simulations to improve grinding efficiency and quality. It reduces unnecessary material waste and processing time. The recognition and analysis system identifies various weld structures and configures appropriate grinding tools based on the shape and size of the weld and the planned path; and continuously optimizes model parameters based on real-time data monitoring and feedback to improve recognition accuracy and grinding effects. The grinding tool 37 can be quickly and automatically replaced through the grinding assembly, and different grinding tools can be configured for different weld grinding needs to improve work efficiency and grinding accuracy. The technical solution for weld identification is used to describe the shape, size, position, characteristics of the weld, as well as the mathematical calculations and model training during the identification process. It can achieve accurate quantification and identification of weld characteristics, thereby improving the automation and accuracy of welding quality detection.

[0216] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for intelligent grinding of mechanically automated workpieces, characterized in that: The following steps are involved: S1. Clamp and fix the workpiece through the workpiece clamping system and special tooling for the workpiece; S2, the data acquisition and analysis system collects images of the workpiece to obtain the shape and size information of the weld; S3. The path planning system determines the optimal grinding path based on the shape and size of the weld and the size and shape of the grinding tool through optimization algorithms and simulations; S4. The recognition and analysis system configures grinding tools for different grinding positions according to the shape and size of the weld and the planned path; S5. The pushing mechanism pushes the grinding tool to be used to the electric chuck, which clamps the grinding tool; the multi-degree-of-freedom robotic arm drives the grinding assembly to grind the weld of the workpiece according to the planned path; S6. When the grinding tool needs to be replaced, the grinding tool is replaced through the grinding assembly, and then the grinding operation is continued; S7. While grinding the workpiece, the water-cooled chip removal system effectively removes the waste generated during the grinding process, ensuring that the equipment will not overheat during long-term operation; Step S3 includes the following steps: S31. Detailed analysis of the weld geometry, including its shape, length, and curvature; curvature calculation: Where v is the velocity vector of the weld centerline, × is the curl operator; κ is the curvature, which indicates the degree of bending of the weld centerline; S32, adapt according to the size and characteristics of the grinding tool; the matching degree between the grinding tool size and the weld size: M = 1-[(D tool -D weld ) / D max ]; among them, D tool is the grinding tool diameter, D weld is the weld diameter, D max is the maximum diameter difference allowed, M is the matching degree; S33. Determine the optimal grinding path using optimization algorithms and simulations; predict the grinding effects and times under different grinding paths, and select the optimal solution; S34. An optimization algorithm is used to find a path that meets the grinding requirements while minimizing processing time and material waste. Simulation can be used to simulate the actual grinding process to verify whether the path obtained by the optimization algorithm is effective. Step S33 includes the following contents: Minimize path length: p opt =argmin{∫ tf t0 [|v(t)|dt)]}; where p opt is the optimal path, v(t) is the velocity vector at time t; ∫ tf t0 [|v(t)|dt)] is the integral of the absolute value of the velocity vector from t0 to tf, representing the path length; Processing time prediction: T = Total Path Length / Polishing Speed; where T is the predicted processing time, Polishing Speed ​​is the set polishing advance speed, and Total Path Length is the length of the weld bead that needs to be polished.

2. The intelligent grinding method for mechanically automated workpiece processing according to claim 1, characterized in that: Step S34 includes the following steps: S341, Time Assessment: Identify the polishing quality and surface finish indicators that need to be achieved; evaluate the processing time under different paths and seek to minimize it; consider the amount of material removal that may occur during the polishing process and seek to minimize it; S342. Establish a mathematical model: clarify the specific requirements and constraints of polishing, select appropriate mathematical methods, establish the objective function, and find the optimal solution through the solution algorithm; Objective function: J(x) = w1*T process +w2*V removal ; Constraints: gi(x)≤0,i=1,…,m; Where J(x) is the objective function, which is used to measure the quality of the solution; x is the decision variable vector, which contains all the variables to be optimized; w1 and w2 are weight coefficients, which are used to balance the importance of different parts of the objective function; T process is the processing time, which means the total time required to complete the grinding; V removal is the material removal amount, which represents the volume or mass of material removed during the grinding process; gi(x) is the functional form of the i-th constraint; m is the number of constraints, which indicates how many constraints need to be considered in the model; S343, selecting an optimization algorithm to search for an optimal or near-optimal solution; S344. Implement simulation: Use simulation software to simulate the actual polishing process to verify whether the path obtained by the optimization algorithm is effective; S345, Iterative Optimization: Adjust the parameters of the optimization algorithm or change the mathematical model according to the simulation results to improve the quality of the solution.

3. The method for mechanically automated workpiece processing and intelligent polishing according to claim 1, characterized in that: The data acquisition and analysis system includes: Data collection module: collects workpiece dimension data, including welding requirements; collects images of different welding methods and annotates them as reference samples; Image acquisition module: includes a high-definition camera and LED lights to capture high-definition images of the workpiece and the welding area; Image preprocessing module: preprocesses the collected images, including filtering and denoising, image changes, image enhancement and image restoration; Feature extraction module: performs feature extraction on the image after feature extraction to extract features related to the weld bead, including color, shape and texture; Weld bead recognition module: identifies the shape, size and position of the weld based on the image after feature extraction; identifies the type, shape and size of the weld; Model optimization module: Use machine learning algorithms to optimize recognition results and improve recognition accuracy; Machine learning model training: Where θ is the model parameter; N is the number of training samples; l is the loss function, which measures the difference between the model prediction and the actual value; Ω is the regularization term used to prevent the model from overfitting; i represents the i-th sample; y i is the true label or output of the i-th sample; y^ i The output of the i-th sample predicted by the model; The judgment results are evaluated based on the model output and confidence level: Confidence = p(f|Weld) / p(f); where Confidence is the confidence level of the recognition result; p(f|Weld) is the posterior probability of a given weld feature; and p(f) is the prior probability of the feature. Output weld bead recognition results, including shape, size, position and feature description.

4. The method for mechanically automated workpiece processing and intelligent polishing according to claim 1, characterized in that: Weld bead recognition module: identifies the shape, size and position of the weld bead based on the image after feature extraction; Includes the following: Shape recognition: Quantify the shape of the weld bead using circularity; Circularity = (4πA) / P 2 ; Where A is the area of ​​the weld bead, P is the perimeter of the weld bead; Circularity is the circularity; Size Measurements: Width=Max x (y upper -y lower ); where Length is the length of the weld bead; x, y are the coordinates of the weld bead in the image; (Dx) / (dy) is the derivative of the weld bead curve, which represents the weld bead curvature; Width is the width of the weld bead; y upper ,y lower are the coordinates of the upper and lower edges of the weld bead; Positioning: Determine the position of the weld in the image, which can be achieved through feature matching or image coordinate transformation; X new =T 11 x+T 12 y+T13;y new =T 21 x+T 22 y+T23; where x new， y new is the new coordinate after transformation; T 11 ,T 12 ,T 13 ,T 21 ,T 22 ,T 23 is the element of the coordinate transformation matrix; Feature fusion: Combining shape, size and position information to identify the characteristics of the weld; feature vector construction: f = [f1, f2, ..., f n ]; where f i is the i-th feature; f is the feature vector containing all extracted features.

5. The method for mechanically automated workpiece processing and intelligent polishing according to claim 1, characterized in that: The grinding assembly includes: a swivel seat, a rotation drive mechanism, a cover plate, an electric push rod, an electric chuck, a pushing mechanism and a grinding tool; A rotation drive mechanism is fixedly arranged on the multi-degree-of-freedom robotic arm; A swivel seat is rotatably provided on the multi-degree-of-freedom robotic arm, and an electric push rod is fixedly provided on the swivel seat; an electric chuck is fixedly provided at the end of the movable rod of the electric push rod; the electric chuck is slidably matched with the swivel seat; and a plurality of pushing mechanisms are fixedly provided on the swivel seat; A plurality of grinding tools are placed in the rotating seat, and the grinding tools can slide in the rotating seat; the pushing mechanism and the grinding tools correspond one to one; the pushing mechanism can drive the grinding tools to move; a cover plate is detachably fixed on the rotating seat.

6. The method for mechanically automated workpiece processing and intelligent polishing according to claim 5, characterized in that: The rotation drive mechanism includes a motor and a driving gear; The motor is fixedly arranged at the end of the multi-degree-of-freedom robotic arm, and a driving gear is coaxially fixedly arranged at the output end of the motor; A shaft sleeve is fixedly arranged on the rotating seat, a driven gear is coaxially fixedly arranged on the shaft sleeve, and the driven gear is meshed with the driving gear for transmission.

7. The method for mechanically automated workpiece processing and intelligent polishing according to claim 6, characterized in that: The swivel seat is provided with multiple chutes A and B, and different grinding tools are placed in chutes A and B; A pushing mechanism is fixedly provided in both chute A and chute B; The pushing mechanism includes an electric telescopic rod and an electromagnet; An electric telescopic rod is fixedly arranged in the chute A and the chute B; an electromagnet is fixedly arranged on the movable rod of the electric telescopic rod.

Images