Tool holder polishing method and device based on sensor data interaction

Abstract

The application provides a tool clamp polishing method and device based on sensor data interaction, and relates to the technical field of tool clamp polishing. The method comprises the following steps: performing basic data interaction of a tool clamp to build a standard data set, performing three-dimensional data scanning on a blank to build a three-dimensional fitting model, matching an identification center point to determine a three-dimensional contour, generating polishing size data, reading device parameters to build a control model, obtaining N control execution strategies, setting speed-quality balance data, performing strategy screening to obtain control parameters, executing polishing control of the tool clamp, performing node image acquisition of polishing, generating auxiliary control information, real-time adjustment, and completing polishing control of the tool clamp according to the adjustment result. The application solves the technical problem that the traditional polishing method mainly relies on manual operation and experience accumulation, and it is difficult to adaptively adjust polishing parameters according to real-time needs, which leads to low efficiency and errors, and cannot accurately and stably achieve the target polishing effect.

Description

Technical Field

[0001] This invention relates to the field of tool polishing technology, and more specifically to a tool polishing method and apparatus based on sensor data interaction. Background Technology

[0002] Tool grinding technology is a crucial aspect of mechanical manufacturing, involving materials science, mechanical engineering, and lean production. Its development is driven by industry demands and technological advancements, and it continues to improve over time. However, current tool grinding methods still have certain drawbacks, and there is room for improvement. Summary of the Invention

[0003] This application provides a tool grinding method and apparatus based on sensor data interaction, aiming to solve the technical problem that traditional grinding methods mainly rely on manual operation and experience accumulation, which makes it difficult to adaptively adjust grinding parameters according to real-time needs, resulting in low efficiency and easy errors, and thus failing to achieve the target grinding effect accurately and stably.

[0004] In view of the above problems, this application provides a tool grinding method and apparatus based on sensor data interaction.

[0005] The first aspect disclosed in this application provides a tool-grinding method based on sensor data interaction. The method includes: performing basic data interaction on the tool and constructing a standard dataset for the tool based on the basic data, wherein the standard dataset includes standard size constraints and surface roughness constraints; performing a three-dimensional data scan on the blank, constructing a three-dimensional fitting model based on the point cloud dataset, matching marker center points, determining a three-dimensional contour based on the three-dimensional fitting model and the marker center points, and generating grinding dimension data based on the three-dimensional contour and the standard dataset; reading equipment parameters of the grinding equipment and constructing a control model based on the equipment parameters; inputting the grinding dimension data and material data into the control model and outputting N control execution strategies; setting speed-mass balance data, performing strategy filtering of the N control execution strategies based on the balance data, and obtaining control parameters; controlling the grinding equipment to perform grinding control of the tool based on the control parameters, and performing nodal image acquisition of grinding through an image acquisition device, and generating auxiliary control information based on the image acquisition results; adjusting the control parameters in real time through the auxiliary control information, and completing the grinding control of the tool based on the real-time adjustment results.

[0006] Another aspect of this application discloses a tool grinding device based on sensor data interaction. The device includes: a data interaction module for performing basic data interaction on the tool and constructing a standard dataset for the tool based on the basic data, wherein the standard dataset includes standard size constraints and surface roughness constraints; a data scanning module for performing three-dimensional data scanning on the blank, constructing a three-dimensional fitting model based on the point cloud dataset, matching marker center points, determining a three-dimensional contour based on the three-dimensional fitting model and the marker center points, and generating grinding dimension data based on the three-dimensional contour and the standard dataset; and a parameter reading module for reading the equipment parameters of the grinding equipment and constructing a standard dataset based on the equipment parameters. The system comprises: a control model; a strategy output module, which inputs the grinding dimension data and material data into the control model and outputs N control execution strategies; a strategy filtering module, which sets speed-mass balance data and filters the N control execution strategies based on the balance data to obtain control parameters; an auxiliary information generation module, which controls the grinding equipment to perform grinding control of the tool clamp based on the control parameters, and performs nodal image acquisition of grinding through an image acquisition device, and generates auxiliary control information based on the image acquisition results; and a grinding control module, which adjusts the control parameters in real time through the auxiliary control information and completes the grinding control of the tool clamp based on the real-time adjustment results.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] This system performs basic data interaction with the tool pliers, constructs a standard dataset including standard dimensional constraints and surface roughness constraints, performs 3D data scanning on the blank, builds a 3D fitting model, matches and identifies center points, determines the 3D contour, generates grinding dimension data, reads equipment parameters, constructs a control model, inputs the control model, outputs N control execution strategies, sets speed-quality balance data, performs strategy filtering to obtain control parameters, controls the grinding equipment to execute the tool pliers' grinding control, performs grinding node image acquisition, generates auxiliary control information, adjusts the control parameters in real time, and completes the tool pliers' grinding control based on the real-time adjustment results. This solves the technical problem of traditional grinding methods relying mainly on manual operation and experience accumulation, which is difficult to adaptively adjust grinding parameters according to real-time needs, resulting in low efficiency, high error rates, and inability to accurately and stably achieve the target grinding effect. It achieves precise control of the grinding process using sensor data interaction and real-time control, and adaptively adjusts grinding parameters, thereby greatly improving the stability and consistency of the grinding effect, achieving the technical effect of improving the flexibility and intelligence of machining.

[0009] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0010] Figure 1 This application provides a schematic flowchart of a tool polishing method based on sensor data interaction for embodiments of the present application;

[0011] Figure 2 This application provides a schematic diagram illustrating a possible process for generating real-time adjustment results in a tool polishing method based on sensor data interaction.

[0012] Figure 3 This application provides a schematic diagram of a possible process for subsequent tool grinding control in a tool grinding method based on sensor data interaction;

[0013] Figure 4 This application provides a possible structural schematic diagram of a tool grinding device based on sensor data interaction for embodiments of the present application.

[0014] Explanation of reference numerals in the attached diagram: Data interaction module 10, data scanning module 20, parameter reading module 30, strategy output module 40, strategy filtering module 50, auxiliary information generation module 60, and polishing control module 70. Detailed Implementation

[0015] This application provides a tool-grinding method based on sensor data interaction, which solves the technical problem that traditional grinding methods mainly rely on manual operation and experience accumulation. This results in low efficiency and error-proneness due to the difficulty in adaptively adjusting grinding parameters according to real-time needs, leading to an inability to accurately and stably achieve the target grinding effect. The method achieves precise control of the grinding process by utilizing sensor data interaction and real-time control, and adaptively adjusts grinding parameters, thereby greatly improving the stability and consistency of the grinding effect, and achieving the technical effect of improving the flexibility and intelligence of machining.

[0016] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0017] Example 1

[0018] like Figure 1 As shown in the figure, this application provides a tool grinding method based on sensor data interaction, the method including:

[0019] Step S100: Perform basic data interaction of the tool clamp and construct a standard dataset of the tool clamp based on the basic data, wherein the standard dataset includes standard size constraints and surface roughness constraints;

[0020] Specifically, basic data related to the tool clamp are collected by sensors, such as three-dimensional shape and position, process materials, and other basic information. Based on the basic data, standard dimensional constraints and surface roughness constraints of the tool clamp are designed and defined. Specifically, by measuring, detecting, and analyzing the dimensions of the above mathematical model, the dimensions and curvature of each part are determined, and surface roughness features are extracted. Corresponding constraints are set, including standard dimensional constraints and surface roughness constraints, to form a complete standard dataset, which provides good data support for the integration of subsequent grinding control.

[0021] Step S200: Perform 3D data scanning on the blank, construct a 3D fitting model based on the point cloud dataset, match the center point of the marker, determine the 3D contour based on the 3D fitting model and the center point of the marker, and generate grinding dimension data based on the 3D contour and the standard dataset.

[0022] Specifically, laser measurement and computer vision technologies are used to perform three-dimensional measurements on the blank, collecting data on its shape, geometric structure, and surface features. This data is then converted into a point cloud dataset. A surface fitting method is employed to process the point cloud dataset, generating a three-dimensional fitting model that accurately represents the shape of the entire blank, and identifying its center points. Using this three-dimensional model and the identified center points, the blank's contour is calculated and generated to further control the edges and angles during grinding. Based on the determined three-dimensional contour and the standard dataset, the contour and physical quantities of the workpiece's surface characteristics are calculated or abstracted, generating grinding dimension data for subsequent processing control.

[0023] Step S300: Read the equipment parameters of the grinding equipment and construct a control model based on the equipment parameters;

[0024] Specifically, the grinding equipment is connected, and sensors, encoders and other devices are used to monitor and learn the grinding equipment in real time. Real-time displacement, rotation angle and load parameters are collected and converted into digital signals. Combined with CAD (computer-aided design or manufacturing) technology, a mathematical model is established based on the equipment parameters to perform motion simulation, predict and adjust the equipment's motion trajectory and processing speed, and thus achieve precise control.

[0025] Step S400: Input the grinding dimension data and material data into the control model, and output N control execution strategies;

[0026] Specifically, the obtained grinding dimension data and material data are input into the control model to form a three-dimensional space and feature description of the workpiece. Based on the input data, the control model constructed above is used for calculation. Algorithms from computer science fields such as AI, machine learning, and deep neural networks are used to process and optimize the input data in the control model, design and generate N strategies, where N is the number of generated strategies, such as five generated strategies, such as edge deburring, surface polishing, and other processing schemes.

[0027] Step S500: Set speed-mass balance data, and based on the balance data, perform strategy filtering for the N control execution strategies to obtain control parameters;

[0028] Specifically, based on the characteristics of the designed machine and workpiece, the processing capacity and objectives of the equipment are analyzed, and the target data for speed-mass balance are determined. Based on the set objectives, a kinematic model of the workpiece and equipment is established, control parameters are calculated, and the corresponding speed and rotational speed are found based on the target balance value. Then, from N control execution strategies, strategies for speed, step distance, etc., are filtered to reduce those that do not conform to the balance data. Based on the control model, the optimal grinding tool, operating method, and motion trajectory are selected to obtain optimal quality and efficiency, thus obtaining the optimal strategy set. For these strategies, on-site monitoring technology and real-time data processing methods are used to monitor and update the control parameters in real time. Combined with factors such as equipment response time and measurement accuracy, the processing parameters are dynamically adjusted, and the optimal control parameters and control commands are output to achieve the optimal effect at that time.

[0029] Step S600: Based on the control parameters, control the grinding equipment to perform grinding control of the tool pliers, and perform nodal image acquisition of grinding through the image acquisition device, and generate auxiliary control information based on the image acquisition results;

[0030] Specifically, based on the obtained control parameters, corresponding control commands are generated and input into the control system of the grinding equipment. The control system then controls the tool clamps in the grinding equipment to perform grinding operations, executes the grinding task according to plan, and records feedback information. Real-time image acquisition devices such as cameras are used to capture images of the workpiece being ground, and the acquired image data is transmitted to an image processing system. Image processing algorithms analyze and process the acquired images, such as filtering, segmentation, and feature extraction, to generate point clouds. Simultaneously, based on motion state analysis, the grinding points are corrected, and auxiliary control information, such as pattern matching and contour measurement, is generated to optimize and monitor the grinding process.

[0031] Step S700: Adjust the control parameters in real time using the auxiliary control information, and complete the grinding control of the tool pliers based on the real-time adjustment results.

[0032] Specifically, the auxiliary control information is analyzed and processed to obtain several effective real-time adjustment parameters, such as tool position, processing speed, and force. Using machine learning and deep neural network algorithms, the real-time adjustment information is combined with historical data. Based on the system's self-learning, it gradually understands the characteristics of the equipment and the appropriate parameters for achieving the desired grinding effect. Furthermore, it generates multiple parameter values ​​according to different grinding task requirements, providing the optimal control parameters for the current grinding task. Using the control system, based on the real-time adjustment information and adaptive learning results, real-time control of the tool clamp is achieved, and the grinding task is completed according to the pre-set grinding path.

[0033] Furthermore, such as Figure 2 As shown, step S700 of this application further includes:

[0034] Step S710: Perform control fitting based on the control parameters, and set the calibration control results of M nodes, where M is a positive integer greater than 2;

[0035] Step S720: After the grinding control of any node is completed, the depth image of the tool clamp is acquired by the image acquisition device to construct the depth image data of the tool clamp node;

[0036] Step S730: Perform grinding control verification based on the calibration control results and depth image data, and generate the real-time adjustment results based on the grinding control verification results.

[0037] Specifically, based on the obtained control parameters, data analysis and digital model establishment are performed. Curve fitting techniques, such as polynomial curve fitting algorithms, are used to fit the original control data into a curve that meets the control requirements. The generated curve is then discretized at equal intervals of M points, where M is a positive integer greater than 2, representing the number of nodes. For example, discretization at equal intervals of 10 points yields 10 nodes. Algorithms such as fuzzy control and PID control are used to calibrate and optimize the discretized nodes. Through training and debugging, M calibration control results are obtained.

[0038] Using technologies such as lenses and lasers, depth images of the tool pliers at each stage of the polishing process are acquired, and the acquired data is transmitted to an image processing system. Based on the acquired data, digital image processing algorithms are used to further analyze and process the depth images, such as brightness calibration, color space conversion, and noise reduction, to obtain high-precision depth image data.

[0039] Based on the calibration control results and depth image data, the grinding control is executed and feedback information is recorded to verify whether the control effect meets the requirements. The data and information recorded during the actual grinding process are analyzed and processed to obtain various indicators of the current grinding task, such as accuracy, speed, and workpiece surface quality. Based on the analysis results, combined with intelligent algorithms and adaptive learning technology, real-time adjustment results are generated, such as control parameter values ​​and grinding path planning, to provide a basis for optimizing subsequent grinding tasks.

[0040] Furthermore, step S710 of this application also includes:

[0041] Step S711: Obtain the acquisition coordinates and acquisition control data of the image acquisition device;

[0042] Step S712: Extract point cloud data from the node tool clamp based on the acquisition coordinates, acquisition control data and the depth image data, and construct a node size verification dataset;

[0043] Step S713: Perform image feature analysis on the depth image data to generate roughness feature recognition results and construct a surface state verification dataset;

[0044] Step S714: Perform control verification of the calibration control results using the node size verification dataset and the surface state verification dataset.

[0045] Specifically, the image acquisition device is accurately positioned in the grinding and processing area using sensors to determine its acquisition angle and range. A coordinate system is established in the acquisition device through methods such as camera intrinsic and extrinsic parameter calibration, and each acquisition point is determined on the three-dimensional coordinate system. According to the scene requirements and actual operation, the acquisition control parameters, such as exposure time, shutter speed, white balance, etc., are adjusted and optimized to obtain high-quality depth image data.

[0046] Based on depth camera technology, point cloud processing algorithms are used to extract and construct point clouds from the depth image data captured by the image acquisition device, based on the acquired coordinates and depth image data. Preprocessing and optimization are performed through point cloud denoising, registration, feature extraction and other processing methods. The point cloud data of all nodes are combined to construct a node size verification dataset, which includes information such as the position, shape and size of each node.

[0047] Based on the acquired depth image data, different surface feature information, such as protrusions, depressions, and scratches, is extracted through digital image processing technology and image feature learning algorithms. Taking into account multiple surface feature information, and using intelligent algorithms and machine learning models, the roughness feature of the surface state is identified and described, generating roughness feature identification results. The roughness feature identification results of all nodes and other relevant information are combined to construct a surface state verification dataset, which includes information such as the location of each node, surface state features, and grinding control parameters.

[0048] Based on the constructed node size verification dataset, the calibration control results are applied to the actual node grinding control process, and feedback information and various index data are recorded. Similarly, based on the constructed surface condition verification dataset, the calibration control results are applied to the actual workpiece grinding control process, and feedback information and various index data are recorded to evaluate the optimization of the workpiece surface condition. By analyzing and processing the feedback information and index data, the effectiveness and feasibility of the calibration control results are evaluated. Combining intelligent algorithms and adaptive learning techniques, grinding control parameters and strategies are adjusted and optimized to further improve grinding control quality and efficiency.

[0049] Furthermore, step S713 of this application also includes:

[0050] Step S7131: Construct the part recognition features of the tool pliers based on the standard dataset;

[0051] Step S7132: Configure the fuzzy attenuation coefficient of the node position according to the M nodes;

[0052] Step S7133: Adjust the feature fuzziness of the part recognition feature according to the fuzziness attenuation coefficient;

[0053] Step S7134: Perform image feature matching of the corresponding node using the part recognition features after blur adjustment, and segment the image into feature matching region and blurry intersection region according to the matching result;

[0054] Step S7135: Obtain the roughness feature recognition result based on the image segmentation result.

[0055] Specifically, computer vision technology is used to extract and analyze features from depth images of workpieces in the surface condition verification dataset. Different features and attributes of each part are extracted, such as contour, texture, and convexity. Based on deep learning and neural network technologies, recognition models for different parts of the tool pliers are established and trained to accurately identify these parts. The extracted surface condition features are then matched with the tool pliers parts, the degree of fit for each part is calculated, the corresponding feature region for each part is determined, and this information is recorded.

[0056] Based on the constructed surface condition verification dataset, computer vision technology and digital image processing algorithms are used to process and analyze depth image data to obtain information such as surface quality, geometric structure, and morphological attributes for each node. Using a fuzzy logic algorithm, combined with the surface condition features and related information of each node, a fuzzy attenuation coefficient for each node position is calculated. This coefficient takes into account various factors, such as surface quality, grinding difficulty, and importance. Specifically, the fuzzy attenuation coefficient for a node position represents the grinding difficulty and priority order of that node; that is, the more difficult or important the node is to grind, the higher its fuzzy attenuation coefficient. This coefficient can be seen as a weighted allocation of the importance of the grinding task for each node.

[0057] By utilizing the fuzzy attenuation coefficient of the constructed node position, the feature value is multiplied by its corresponding weight and normalized. Taking into account various factors such as surface condition, workpiece shape, and accuracy requirements, the weighted feature value of each node is obtained. The weighted feature value is then fuzzy processed and corrected to adapt to the grinding scenario and task requirements.

[0058] Based on the recorded node information and feature values, the position and feature descriptor of each node are determined. The obtained node feature descriptors are matched with the input image data to find the best matching node and its corresponding feature descriptor value. During the matching process, a threshold is set using fuzzy adjustment information to judge the reliability of the matching results. The standard is to select matching results with high reliability. Based on the matching results, the original image is segmented into feature matching regions and fuzzy intersection regions. Feature matching regions refer to those regions with high matching degree and feature descriptors showing a clear dominant role; while fuzzy intersection regions refer to those regions where multiple feature descriptors contribute, and their feature values ​​may have been affected by fuzzy adjustment.

[0059] Based on the characteristics of the parts, a corresponding roughness recognition model is constructed. By learning and classifying the image feature data, the roughness feature values ​​of each region are identified. These feature values ​​are used to evaluate the contact state between the tool and the workpiece surface, helping to control the grinding quality and avoid excessive tool wear.

[0060] Furthermore, step S7135 of this application also includes:

[0061] Step S71351: Based on the segmentation results, perform gap feature identification for each segmented region and record the gap size, number of gaps, and gap distribution coordinates;

[0062] Step S71352: Based on the gap size and the gap distribution coordinates, perform roughness uniformity analysis on the segmented region to generate a first reference parameter for roughness;

[0063] Step S71353: Evaluate the roughness distribution density of the segmented region based on the number of voids and the void distribution coordinates, and generate a second reference parameter for roughness;

[0064] Step S71354: Perform roughness analysis of the tool clamp using the gap size to generate a third reference parameter for roughness;

[0065] Step S71355: Obtain the roughness feature recognition result based on the first reference parameter, the second reference parameter and the third reference parameter.

[0066] Specifically, based on the aforementioned gap feature extraction technology, gaps in various segmented areas around the tool and workpiece are detected and identified, yielding indicators such as the size, quantity, and distribution of the gap areas. These indicators reflect information such as processing quality, tool usage, and workpiece surface condition. For fuzzy and overlapping areas, the results of multiple feature descriptors are merged and adjusted based on statistical methods such as average or maximum values ​​to obtain more accurate and reliable gap features, providing a basis for subsequent grinding control and quality assessment.

[0067] By using the gap size and distribution coordinate information, the corresponding gap density, gap size variance, gap spacing and other indicators are calculated to reflect the roughness distribution on the machined surface. Based on these indicators, combined with the relationship and weight of each gap parameter, the roughness uniformity of each segmented area is preliminarily analyzed and evaluated, and the first reference parameter of roughness is generated.

[0068] By utilizing the number and distribution coordinates of voids, the corresponding void distribution density, void size frequency, and other indicators are calculated to reflect the roughness distribution on the machined surface. Based on these indicators and the calculation results of the void distribution parameters, the roughness distribution in each segmented region is evaluated, and a second reference parameter for roughness is generated.

[0069] Using the extracted gap size information, the surface roughness of the tool pliers is calculated, reflecting the contact quality between the tool pliers and the workpiece surface. Based on these roughness parameters, a third reference parameter for the roughness of the tool pliers is generated.

[0070] The first, second, and third reference parameters obtained earlier are combined according to certain weights to obtain the overall roughness feature recognition result. The weights of different reference parameters can be set according to actual needs to reflect the importance of different parameters for roughness recognition.

[0071] Furthermore, step S700 of this application also includes:

[0072] Step S700-1: Set the warning stop threshold;

[0073] Step S700-2: When the auxiliary control information triggers the warning stop threshold, a stop control command is generated;

[0074] Step S700-3: And re-perform the grinding control plan of the tool pliers according to the stop control command.

[0075] Specifically, based on actual conditions and needs, corresponding warning and stop thresholds are set to reflect the processing quality level during the grinding control process. When the warning and stop threshold is triggered, auxiliary control information is automatically transmitted to the system for analysis and processing. This information includes real-time monitored processing data, machine operating status, and part surface condition, for subsequent judgment and analysis. The system compares and analyzes the collected processing data with the preset warning and stop threshold to determine whether the processing quality under the current state has reached the warning and stop threshold. If the analysis confirms that the warning and stop threshold has been triggered, a stop control command is generated according to the preset program, and the grinding machine is notified to stop the operation according to the safety procedure. At the same time, the operator can be alerted to this matter through the human-machine interface or alarm signals, and appropriate measures can be taken in a timely manner.

[0076] Specifically, the warning and stop threshold can be divided into two parts: the warning threshold and the stop threshold. The warning threshold is set at a low level. When the processing quality reaches this threshold, the system will issue a warning signal or alarm, prompting operators to strengthen monitoring and adjustments to prevent a further decline in processing quality. The warning threshold uses data obtained from offline model training as a standard reference value. Through real-time measurement and monitoring, when the measured value exceeds the preset standard value, the warning mechanism is activated. The stop threshold is set at a high level. When the processing quality reaches this threshold, the system will automatically stop the grinding process to prevent deep grinding from damaging the surface of the parts. The stop threshold is usually determined through data analysis from offline model training, combined with prior experiments and expert experience.

[0077] The generated stop control commands are analyzed and interpreted. These commands typically provide information such as operating parameters, mode settings, and warning limits, which can help identify the cause, eliminate the fault, and determine the next control solution. After determining the cause of the fault, the grinding control plan is updated and adjusted based on the actual situation. This includes optimizing and improving the grinding equipment, cutting fluid, and tool selection to meet new requirements and needs.

[0078] Furthermore, such as Figure 3 As shown, this application also includes:

[0079] Step S810: Record the grinding data of the tool pliers and map the recorded results with the quality inspection results;

[0080] Step S820: When the quality inspection result meets the preset threshold, a polishing control template is generated based on the polishing data, and a template similarity value is set;

[0081] Step S830: Perform subsequent grinding control of the tool pliers based on the grinding control template and the template similarity value.

[0082] Specifically, in the actual production process, sensors and monitoring equipment are used to monitor and record key parameters and changes of the tool during the grinding process, such as contour shape, surface finish, and cutting edge size. Simultaneously, the tool is subject to regular inspection and testing, including related physical properties, mechanical strength, and surface quality, to obtain quality inspection results and ensure that the processing quality and performance meet requirements. The recorded grinding data and quality inspection results are mapped and analyzed. Specifically, data visualization and statistical analysis can be used to compare and verify the grinding data and quality inspection results, thereby identifying the causes of deterioration in processing quality and making targeted adjustments to the operating plan and optimization of equipment performance.

[0083] Manufacturers or users set preset thresholds for quality inspection based on their own needs, including aspects such as surface finish, cutting edge size, and contour shape, to ensure that the product achieves the expected processing effect. Data mining technology is used to process and analyze the recorded grinding data to extract relevant feature information and patterns. Based on the processed grinding data, corresponding grinding control templates are generated, and a similarity threshold is set for each template. These grinding control templates are standardized grinding operation guidelines, including processing parameters, processing procedures, and operating specifications, to improve the accuracy and reproducibility of operations.

[0084] In practice, the aforementioned grinding control template and its similarity value are used, with the control template directly serving as the operation guide. The equipment is controlled and optimized through computer programs and other means. Simultaneously, data obtained from each operation is calculated and compared with the existing control template, and matching is performed based on statistical methods and pattern recognition technology. When the matching degree meets the requirements, the grinding process is controlled through various automated equipment and programs, and subsequent adjustments and optimizations are made according to the actual situation.

[0085] In summary, the tool polishing method and apparatus based on sensor data interaction provided in this application have the following technical effects:

[0086] This system performs basic data interaction with the tool pliers, constructs a standard dataset including standard dimensional constraints and surface roughness constraints, performs 3D data scanning on the blank, builds a 3D fitting model, matches and identifies center points, determines the 3D contour, generates grinding dimension data, reads equipment parameters, constructs a control model, inputs the control model, outputs N control execution strategies, sets speed-quality balance data, performs strategy filtering to obtain control parameters, controls the grinding equipment to execute the tool pliers' grinding control, performs grinding node image acquisition, generates auxiliary control information, adjusts the control parameters in real time, and completes the tool pliers' grinding control based on the real-time adjustment results. This solves the technical problem of traditional grinding methods relying mainly on manual operation and experience accumulation, which is difficult to adaptively adjust grinding parameters according to real-time needs, resulting in low efficiency, high error rates, and inability to accurately and stably achieve the target grinding effect. It achieves precise control of the grinding process using sensor data interaction and real-time control, and adaptively adjusts grinding parameters, thereby greatly improving the stability and consistency of the grinding effect, achieving the technical effect of improving the flexibility and intelligence of machining.

[0087] Example 2

[0088] Based on the same inventive concept as the tool grinding method based on sensor data interaction in the foregoing embodiments, such as Figure 4 As shown, this application provides a tool sharpening device based on sensor data interaction, the device comprising:

[0089] The data interaction module 10 is used to perform basic data interaction of the tool pliers and construct a standard dataset of the tool pliers based on the basic data. The standard dataset includes standard size constraints and surface roughness constraints.

[0090] The data scanning module 20 is used to perform three-dimensional data scanning on the blank, construct a three-dimensional fitting model based on the point cloud dataset, match the center point of the marker, determine the three-dimensional contour based on the three-dimensional fitting model and the center point of the marker, and generate grinding dimension data based on the three-dimensional contour and the standard dataset.

[0091] The parameter reading module 30 is used to read the equipment parameters of the grinding equipment and construct a control model based on the equipment parameters.

[0092] Strategy output module 40 is used to input the grinding size data and material data into the control model and output N control execution strategies;

[0093] The strategy filtering module 50 is used to set speed-mass balance data, and to filter the N control execution strategies based on the balance data to obtain control parameters.

[0094] The auxiliary information generation module 60 is used to control the grinding equipment to perform grinding control of the tool pliers based on the control parameters, and to perform grinding node image acquisition through the image acquisition device, and generate auxiliary control information based on the image acquisition results.

[0095] The grinding control module 70 is used to adjust the control parameters in real time through the auxiliary control information, and to complete the grinding control of the tool pliers based on the real-time adjustment results.

[0096] Furthermore, the device also includes:

[0097] The control fitting module is used to perform control fitting based on the control parameters and set the calibration control results of M nodes, where M is a positive integer greater than 2;

[0098] The depth image acquisition module is used to acquire the depth image of the tool clamp through the image acquisition device after the grinding control of any node is completed, and to construct the depth image data of the tool clamp node.

[0099] The grinding control verification module is used to perform grinding control verification based on the calibration control results and depth image data, and to generate the real-time adjustment results based on the grinding control verification results.

[0100] Furthermore, the device also includes:

[0101] The acquisition and control data acquisition module is used to obtain the acquisition coordinates and acquisition control data of the image acquisition device;

[0102] The point cloud data extraction module is used to extract point cloud data from the node tool clamp based on the acquisition coordinates, acquisition control data and the depth image data, and to construct a node size verification dataset.

[0103] The image feature analysis module is used to perform image feature analysis on the depth image data, generate roughness feature recognition results, and construct a surface state verification dataset;

[0104] The control verification module is used to perform control verification of the calibration control results using the node size verification dataset and the surface state verification dataset.

[0105] Furthermore, the device also includes:

[0106] A part recognition feature construction module is used to construct part recognition features of the tool pliers based on the standard dataset;

[0107] A fuzzy attenuation coefficient configuration module is used to configure the fuzzy attenuation coefficient of the node position according to the M nodes;

[0108] A feature blur adjustment module is used to adjust the feature blur of the part recognition feature according to the blur attenuation coefficient.

[0109] The image feature matching module is used to perform image feature matching of corresponding nodes by recognizing the features of the parts after blur adjustment, and to segment the image into feature matching regions and blurred intersection regions based on the matching results;

[0110] The recognition result acquisition module is used to obtain the roughness feature recognition result based on the image segmentation result.

[0111] Furthermore, the device also includes:

[0112] The gap feature recognition module is used to perform gap feature recognition for each segmented region according to the segmentation result, and record the gap size, gap number, and gap distribution coordinates;

[0113] The uniformity analysis module is used to perform roughness uniformity analysis of the segmented region based on the gap size and the gap distribution coordinates, and generate a first reference parameter for roughness.

[0114] The density evaluation module is used to evaluate the roughness distribution density of the segmented region based on the number of voids and the void distribution coordinates, and generate a second reference parameter for roughness.

[0115] The roughness analysis module is used to perform roughness analysis of the tool clamp based on the gap size, and generate a third reference parameter for roughness;

[0116] The feature recognition result acquisition module is used to obtain the roughness feature recognition result based on the first reference parameter, the second reference parameter and the third reference parameter.

[0117] Furthermore, the device also includes:

[0118] The threshold setting module is used to set the warning stop threshold;

[0119] A stop control command generation module is used to generate a stop control command when the auxiliary control information triggers the warning stop threshold.

[0120] The grinding control planning module is used to re-plan the grinding control of the tool clamp according to the stop control command.

[0121] Furthermore, the device also includes:

[0122] A grinding data recording module is used to record grinding data of the tool pliers and map the recording results to the quality inspection results;

[0123] The control template generation module is used to generate a grinding control template based on the grinding data and set a template similarity value when the quality inspection result meets a preset threshold.

[0124] The subsequent polishing control module is used to control the polishing of the subsequent tool pliers based on the polishing control template and the template similarity value.

[0125] Through the foregoing detailed description of the tool polishing method based on sensor data interaction, those skilled in the art can clearly understand the tool polishing method and apparatus based on sensor data interaction in this embodiment. As for the apparatus disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the method section description.

[0126] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method of pliers grinding based on sensor data interaction, characterized in that, The method comprises: Performing basic data interaction of the tool clamp and constructing a standard data set of the tool clamp according to the basic data, wherein the standard data set comprises standard size constraints and surface roughness constraints; Performing three-dimensional data scanning on the blank, constructing a three-dimensional fitting model according to a point cloud data set, matching an identification center point, determining a three-dimensional profile according to the three-dimensional fitting model and the identification center point, and generating polishing size data according to the three-dimensional profile and the standard data set; Reading device parameters of a polishing device and constructing a control model according to the device parameters; Inputting the polishing size data and material data into the control model and outputting N control execution strategies; Setting speed-mass balance data, performing strategy screening of the N control execution strategies based on the balance data, and obtaining control parameters; Controlling the polishing device to perform polishing control of the tool clamp based on the control parameters, performing node image acquisition of polishing through an image acquisition device, and generating auxiliary control information according to the image acquisition result; Real-time adjusting the control parameters through the auxiliary control information and completing the polishing control of the tool clamp according to the real-time adjustment result; Performing control fitting based on the control parameters and setting M calibration control results of nodes, wherein M is a positive integer greater than 2; When the polishing control of any node is completed, performing depth image acquisition of the tool clamp through the image acquisition device and constructing depth image data of the tool clamp node; Performing polishing control verification according to the calibration control result and the depth image data and generating the real-time adjustment result based on the polishing control verification result; Constructing part recognition features of the tool clamp according to the standard data set; Configuring a fuzzy attenuation coefficient of a node position according to the M nodes; Performing feature fuzzy adjustment of the part recognition features according to the fuzzy attenuation coefficient; Performing image feature matching of the corresponding node through the part recognition features after the fuzzy adjustment and dividing the image into a feature matching area and a fuzzy intersection area according to the matching result; Obtaining the roughness feature recognition result based on the image segmentation result; Performing gap feature recognition of each segmentation area according to the segmentation result and recording gap size, gap quantity, and gap distribution coordinates; Performing roughness uniformity analysis of the segmentation area based on the gap size and the gap distribution coordinates and generating a first reference parameter of roughness; Performing roughness distribution density evaluation of the segmentation area according to the gap quantity and the gap distribution coordinates and generating a second reference parameter of roughness; Performing roughness analysis of the tool clamp through the gap size and generating a third reference parameter of roughness; Obtaining the roughness feature recognition result according to the first reference parameter, the second reference parameter, and the third reference parameter.

2. The method of claim 1, wherein, The method further comprises: Obtaining acquisition coordinates and acquisition control data of the image acquisition device; Performing point cloud data extraction of the node tool clamp according to the acquisition coordinates, the acquisition control data, and the depth image data and constructing node size verification data set; Perform image feature analysis on the depth image data to generate coarse feature recognition results and construct a surface state verification dataset; Perform control verification on the calibration control results through the node size verification dataset and the surface state verification dataset.

3. The method of claim 1, wherein, The method further includes: Setting a pre-warning stop threshold; When the auxiliary control information triggers the pre-warning stop threshold, a stop control instruction is generated; And the polishing control plan of the tool clamp is re-performed according to the stop control instruction.

4. The method of claim 1, wherein, The method further includes: Recording polishing data of the tool clamp and mapping the recording results with quality detection results; When the quality detection results meet a preset threshold, a polishing control template is generated based on the polishing data, and a template similarity value is set; Polishing control of subsequent tool clamps is performed according to the polishing control template and the template similarity value.

5. A tool holder polishing device based on sensor data interaction, characterized by, A tool clamp polishing method based on sensor data interaction for implementing claims 1-4, comprising: A data interaction module for performing basic data interaction of the tool clamp and constructing a standard dataset of the tool clamp according to the basic data, wherein the standard dataset includes standard size constraints and surface roughness constraints; A data scanning module for performing three-dimensional data scanning on the blank, constructing a three-dimensional fitting model according to the point cloud dataset, matching an identified center point, determining a three-dimensional profile according to the three-dimensional fitting model and the identified center point, and generating polishing size data according to the three-dimensional profile and the standard dataset; A parameter reading module for reading device parameters of a polishing device and constructing a control model according to the device parameters; A strategy output module for inputting the polishing size data and material data into the control model and outputting N control execution strategies; A strategy screening module for setting speed-quality balance data, performing strategy screening of the N control execution strategies based on the balance data, and obtaining control parameters; An auxiliary information generation module for controlling the polishing device to perform polishing control of the tool clamp based on the control parameters, performing node image acquisition of polishing through an image acquisition device, and generating auxiliary control information according to image acquisition results; A polishing control module for real-time adjusting the control parameters through the auxiliary control information and completing polishing control of the tool clamp according to real-time adjustment results.

Images