Feed Control System of Multi-head Milling Machine Based on Workpiece Features

Abstract

The present invention provides a multi-head milling machine feed control system based on workpiece features, which relates to the field of milling processing technology and includes: an information interaction module; a workpiece transmission module; a feature recognition module for performing workpiece feature recognition and obtaining a set of workpiece structural features; a solution matching module for performing feed solution matching and obtaining an initial feed solution set; a fitting relationship establishment module for establishing a functional fitting relationship between processing evaluation parameters and feed control parameters; a feed parameter optimization module for optimizing the feed parameters of the initial feed solution set and obtaining a target feed solution; and a feed control module for performing feed control of a milling component set according to the target feed solution. The present invention solves the technical problem that the traditional multi-head milling machine feed control system is usually controlled according to fixed parameters and lacks precise analysis of processing parameters, resulting in a lack of adaptability and an inability to achieve the best processing effect.

Description

Technical Field

[0001] The invention relates to the technical field of milling processing, and in particular to a multi-head milling machine feed control system based on workpiece features. Background Art

[0002] Milling machine feed control systems are mainly used in milling processing in the manufacturing industry, which usually has strict requirements on the processing accuracy, surface quality and processing efficiency of the workpiece. However, the existing technology still has certain problems with milling machine feed control. On the one hand, traditional systems are mainly based on empirical rules or simple parameter presets, and cannot accurately control parameters according to the specific workpiece characteristics, resulting in instability during the processing process, affecting the surface quality and processing accuracy of the workpiece; on the other hand, traditional systems usually rely on simple image processing technology or manual detection to identify the features of the workpiece. This method is not only inefficient, but also prone to omissions and recognition errors. Summary of the Invention

[0003] This application provides a multi-head milling machine feed control system based on workpiece features, aiming to solve the technical problem that traditional multi-head milling machine feed control systems are usually controlled according to fixed parameters, lack of precise analysis of processing parameters, resulting in lack of adaptability and inability to achieve optimal processing effects.

[0004] The present application discloses a multi-head milling machine feed control system based on workpiece features, the system comprising: an information interaction module, the information interaction module being used to interactively obtain material information and processing requirement information of the target workpiece when the target workpiece is transferred to a first working area, wherein the processing requirement information includes a target milling index; a workpiece transfer module, the workpiece transfer module being used to transfer the target workpiece to a second working area, wherein the second working area comprises a workpiece image acquisition component and a workpiece feature recognition component; a feature recognition module, the feature recognition module being used to perform image acquisition on the target workpiece through the workpiece image acquisition component, perform workpiece feature recognition on the image acquisition result through the workpiece feature recognition component, and obtain a workpiece structural feature set; a solution matching module, the solution matching module being used to, based on the material information and the workpiece structural feature set, Feed schemes are matched in a preset feed scheme library to obtain an initial feed scheme set; a fitting relationship establishment module is used to establish a function fitting relationship between the processing evaluation parameters of the target workpiece and the feed control parameters of the target multi-head milling machine, wherein the processing evaluation parameters correspond to the target milling indicators; a feed parameter optimization module is used to optimize the feed parameters of the initial feed scheme set based on the function fitting relationship and the target milling indicators as the optimization target to obtain a target feed scheme; a feed control module is used to transfer the target workpiece to a third working area, the third working area including a milling component set and a workpiece fixing component, fix the target workpiece according to the workpiece fixing component, and control the feed of the milling component set according to the target feed scheme.

[0005] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0006] Through the information interaction module, the system can accurately and interactively obtain material information and processing requirement information when the workpiece enters the first working area, ensuring the rationality and accuracy of the processing parameters; the feature recognition module performs comprehensive image acquisition and feature recognition on the workpiece through the image acquisition component and the feature recognition component, and accurately obtains the structural feature set of the workpiece; the solution matching module accurately matches the preset feed solution library based on the material information and the workpiece structural feature set, and generates an initial feed solution set to ensure the adaptability of the feed parameters; the fitting relationship establishment module and the feed parameter optimization module optimize the initial feed solution for the target milling index by establishing a function fitting relationship and parameter optimization, obtain the optimal target feed solution, and improve the processing quality and efficiency; the feed control module controls the feed of the milling component according to the target feed solution to ensure the stability and efficiency of the processing process.

[0007] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 A schematic structural diagram of a multi-head milling machine feed control system based on workpiece features is provided for an embodiment of the present application;

[0009] Figure 2 A schematic diagram of the execution flow of an image acquisition result acquisition module in a multi-head milling machine feed control system based on workpiece features is provided for an embodiment of the present application.

[0010] Explanation of the accompanying drawings: information interaction module 10 , workpiece transmission module 20 , feature recognition module 30 , solution matching module 40 , fitting relationship establishment module 50 , feed parameter optimization module 60 , feed control module 70 . DETAILED DESCRIPTION

[0011] The embodiment of the present application solves the technical problem that the traditional multi-head milling machine feed control system is usually controlled according to fixed parameters, lacks precise analysis of processing parameters, resulting in lack of adaptability and inability to achieve optimal processing effects by providing a multi-head milling machine feed control system based on workpiece characteristics.

[0012] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically introduced in conjunction with the drawings in the specification.

[0013] like Figure 1 As shown, an embodiment of the present application provides a multi-head milling machine feed control system based on workpiece features, the system comprising:

[0014] The information interaction module 10 is used to interactively obtain material information and processing requirement information of the target workpiece when the target workpiece is transferred to the first working area, wherein the processing requirement information includes a target milling index.

[0015] When the target workpiece is transferred to the first working area, the sensor in the area detects the arrival of the workpiece, triggering the start of the information interaction module 10, and reading the material identification label or barcode on the target workpiece through a sensor, such as an RFID reader, a barcode scanner, etc. These labels or barcodes pre-store the material information of the workpiece, or query the background database through the unique identification code of the workpiece, such as the serial number, to obtain the material information.

[0016] The processing requirement information is pre-stored in the system's database. The corresponding processing requirement information is extracted from the database through the identification code of the target workpiece. This information includes target milling indicators, such as milling accuracy, surface roughness, etc.

[0017] The workpiece conveying module 20 is used to convey the target workpiece to a second working area, wherein the second working area includes a workpiece image acquisition component and a workpiece feature recognition component.

[0018] After completing information acquisition in the first working area, the target workpiece is moved from the first working area to the second working area. The second working area includes a workpiece image acquisition component and a workpiece feature recognition component. The workpiece image acquisition component includes one or more fixed image acquisition devices, such as a camera, a laser scanner, etc. The workpiece feature recognition component includes a structural feature recognition model for performing image processing on the acquired workpiece image and identifying the structural features of the workpiece.

[0019] The feature recognition module 30 is used to capture an image of the target workpiece through the workpiece image capture component, perform workpiece feature recognition on the image capture result through the workpiece feature recognition component, and obtain a workpiece structural feature set.

[0020] After the feature recognition module 30 receives the signal indicating that the workpiece transfer is complete, it starts the workpiece image acquisition component, automatically focuses and calibrates the workpiece, and then performs image acquisition when ready. Since the workpiece requires multi-angle image acquisition, the system controls the rotary table to rotate the workpiece in a predetermined angle sequence. An image is acquired each time the workpiece rotates one angle, and the surface image of the workpiece is captured at each predetermined angle to ensure that all key surface features of the workpiece are fully covered and the image acquisition results are obtained.

[0021] The collected images are denoised, and image enhancement technology is used to improve the image clarity and contrast. For images collected from multiple angles, image alignment is performed so that the image at each angle can accurately correspond to different parts of the workpiece. A structural feature recognition model is used to identify feature areas in the image to represent the structural features of the workpiece.

[0022] The scheme matching module 40 is used to match feeding schemes in a preset feeding scheme library according to the material information and the workpiece structural feature set, and obtain an initial feeding scheme set.

[0023] Load the preset feed scheme library, which stores feed schemes corresponding to various materials and structural features. These schemes include parameters such as cutting speed, feed rate, tool selection, cutting path, cutting depth, etc.

[0024] According to the received material information, the preset feeding scheme library is traversed to screen out the feeding scheme that matches the material characteristics of the target workpiece to form a preliminary feeding scheme set; according to the workpiece structural feature set, the preset feeding scheme library is screened to find the feeding scheme that matches the specific structural characteristics of the workpiece.

[0025] The results of material information matching and structural feature matching are cross-validated so that the selected scheme meets the requirements of both material and structural features. The cross-validated schemes are combined to generate an initial feed scheme set. Each scheme includes a combination of multiple parameters, such as cutting speed, feed rate, tool selection, cutting path, cutting depth, etc.

[0026] The fitting relationship establishment module 50 is used to establish a functional fitting relationship between the machining evaluation parameters of the target workpiece and the feed control parameters of the target multi-head milling machine, wherein the machining evaluation parameters correspond to the target milling indicators.

[0027] Determine the machining evaluation parameters of the target workpiece, including surface roughness, machining accuracy, machining time, tool wear, etc., corresponding to the target milling indicators, and extract key feed control parameters from the initial feed plan set, including cutting speed, feed rate, cutting depth, cutting path, tool selection, etc., and use these key feed control parameters as constraints to retrieve historical machining data, extract the machining evaluation parameters of the corresponding workpiece and the corresponding feed control parameters as part of the training data set.

[0028] Select a function fitting model, such as linear regression, with the feed control parameters as input variables and the machining evaluation parameters as output variables. Divide the dataset into a training set and a validation set for model training and validation. Use the training set for model training, then adjust the model parameters to minimize the error function. Use the validation set to test the trained model and evaluate its predictive accuracy. Adjust the model parameters based on the evaluation results to further improve model accuracy. Iterate the optimization process to determine the final fitting function form that accurately expresses the relationship between the feed control parameters and the machining evaluation parameters.

[0029] The feed parameter optimization module 60 is used to optimize the feed parameters of the initial feed scheme set according to the function fitting relationship and take the target milling index as the optimization target to obtain the target feed scheme.

[0030] Determine the optimization goal, namely the target milling indicators, including surface roughness, machining accuracy, machining time, tool wear, etc. Based on the fitting relationship model, the target milling indicators are converted into an objective function. The objective function is a function of the feed parameters, aiming to optimize the machining evaluation parameters, for example, minimizing surface roughness or maximizing machining efficiency. Define the constraints of the feed parameters, including upper and lower limits of cutting speed, feed rate, and cutting depth. Select a suitable optimization algorithm, such as genetic algorithm, particle swarm optimization algorithm, simulated annealing algorithm, etc., and initialize the optimization algorithm parameters, including population size, number of iterations, and convergence criteria. The initial set of feed solutions is used as the initial solution set of the optimization algorithm. Based on these initial solutions, the algorithm begins to search for the optimal solution. In each iteration, the optimization algorithm calculates the evaluation value of the current feed solution according to the objective function. Based on the evaluation value, the feed control parameters are adjusted. The optimization algorithm updates the parameter values and gradually approaches the optimal solution. The convergence condition is determined, such as when the number of iterations reaches the upper limit or the change in the evaluation value is less than the set threshold. The judgment result determines whether to continue the iteration.

[0031] After optimization, the candidate optimal solution is finally generated, which is verified to ensure that it meets all constraints and can achieve the expected target milling indicators, ensuring that it can be applied in actual production. After verification, the final target feed solution is output.

[0032] The feed control module 70 is used to transfer the target workpiece to a third working area. The third working area includes a milling component set and a workpiece fixing component. The target workpiece is fixed according to the workpiece fixing component, and the feed of the milling component set is controlled according to the target feeding scheme.

[0033] Transfer the target workpiece to the third working area. Adjust the position and clamping force of the workpiece fixing assembly according to the size and shape of the target workpiece to ensure that the workpiece is stable and immobile. Check the status of the milling assembly, including the wear of the tool, and replace and install the tool if necessary according to the tool selection parameters in the target feed plan. Configure the feed control parameters in the target feed plan, such as cutting speed, feed rate, cutting depth, cutting path, etc. in the CNC system, load the corresponding machining program to match the target feed plan, start the milling assembly, and perform machining operations according to the set feed control parameters. This can accurately control the feed process of the milling assembly and ensure the machining quality and efficiency of the target workpiece.

[0034] Furthermore, if Figure 2 As shown, the system further includes an image acquisition result acquisition module to perform the following operation steps:

[0035] The second working area also includes a rotating table; geometric data of the target workpiece and a first contact surface are obtained, wherein the first contact surface is the initial contact surface between the target workpiece and the rotating table; a rotation angle sequence is generated according to the geometric data, and the rotating table is controlled to rotate the target workpiece according to the rotation angle sequence; after each rotation operation, the target workpiece is image captured by the workpiece image capture component until the first contact surface is reached again, and multiple surface images of the target workpiece are output; the multiple surface images are aligned to obtain the image capture result.

[0036] Specifically, the second working area includes not only a workpiece image acquisition component and a workpiece feature recognition component, but also a rotating table. After the target workpiece is transferred to the second working area, it is placed on the rotating table.

[0037] Sensors installed in the second working area, such as laser scanners and 3D cameras, are used to obtain geometric data of the target workpiece, including the size and shape of the workpiece. The obtained geometric data is processed and analyzed, for example, using 3D software to generate a 3D model of the workpiece and determine the initial contact surface between the target workpiece and the turntable, that is, the initial placement surface of the workpiece on the turntable.

[0038] Based on the acquired geometric data, the structural features of the target workpiece are analyzed, the key surfaces that need to be imaged are determined, the required rotation angle for each key surface is calculated, and a rotation angle sequence is generated. For example, if images of four sides need to be captured, the rotation angle sequence is 90 degrees, 180 degrees, and 270 degrees.

[0039] According to the generated rotation angle sequence, the rotation angle of the rotary table is set, the rotary table is started, and the target workpiece is rotated in sequence according to the set angle sequence.

[0040] After each rotation to the specified angle, the workpiece image acquisition component is used to capture images of the target workpiece, and the surface image data at different angles are recorded. The rotation operation and image acquisition are repeated until the rotary table returns to the initial contact surface position again, and all surface image data during the entire rotation process are recorded.

[0041] The collected images are preprocessed, including denoising and chromatic aberration correction. Image processing algorithms are used to align multiple surface images to ensure that the boundaries and details of each image can be accurately matched. For example, key feature points are detected in each image, and the transformation matrix between images, such as translation, rotation, and scaling, is calculated. The images are aligned to a unified coordinate system, and the aligned images are spliced in the 3D model to generate a complete surface image of the target workpiece.

[0042] Furthermore, the system further includes a surface image acquisition module to perform the following operation steps:

[0043] When the workpiece is first in contact with the first surface, the workpiece image acquisition component is started to acquire an image of the target workpiece to obtain a first surface image; according to the first rotation angle, the rotation table is controlled to rotate the target workpiece to obtain a second contact surface; the second contact surface is compared with the first contact surface, and if the comparison fails, the image acquisition and rotation operations are continued; until the workpiece is in contact with the first surface again, the image acquisition and rotation operations are stopped, and the multiple surface images are output.

[0044] Specifically, when the workpiece is initially at the first contact surface position, the workpiece image acquisition component is activated to capture and record a first surface image at the initial contact surface. Based on the generated rotation angle sequence, a first rotation angle is obtained. The first rotation angle is the first in the rotation angle sequence, i.e., the initial rotation angle corresponding to the initial position. Based on this first rotation angle, the rotary stage is controlled to rotate the target workpiece to a specified angle, so that the workpiece rotates to the second contact surface position.

[0045] The current second contact surface after rotation is compared with the first contact surface. For example, the second contact surface image is compared with the first contact surface image. If a match fails, it indicates that the workpiece has not yet returned to its initial position. The rotary stage is controlled to continue rotating the target workpiece according to the rotation angle sequence, and image acquisition continues at each new rotation position, recording surface image data. After each rotation and image acquisition, the contact surface comparison is performed again until the comparison is successful. When the workpiece is once again in the first contact surface position, the image acquisition and rotation operations are stopped, and the multiple surface images acquired during the entire process are output.

[0046] Furthermore, the system further includes a workpiece structural feature set acquisition module to perform the following operation steps:

[0047] A structural feature recognition model is pre-constructed, wherein the structural feature recognition model includes multiple structural feature recognition channels, the multiple structural feature recognition channels are connected in parallel, and the structural feature recognition model is embedded in the workpiece feature recognition component; feature extraction is performed on the image acquisition results to determine a number of key surface features for workpiece feature recognition; the multiple structural feature recognition channels are matched based on the several key surface features, and synchronous structural feature recognition is performed on the several structural feature recognition channels obtained by matching to obtain a workpiece structural feature set.

[0048] Specifically, multiple structural feature recognition channels are constructed, each channel is used to identify a specific type of structural feature. These channels can be constructed based on machine learning or deep learning algorithms, such as convolutional neural networks (CNN), support vector machines (SVM), etc. Multiple recognition channels are configured in parallel to ensure that the model can simultaneously process outputs from different feature recognition channels.

[0049] The collected images are preprocessed, including denoising and enhancement, to improve image quality. Image processing algorithms, such as edge detection and corner detection, are used to extract preliminary features from the preprocessed images. The extracted preliminary features are analyzed to determine several key surface features for workpiece feature recognition. These key features are representative structural features, such as holes, grooves, edges, etc.

[0050] The extracted key surface features are matched with the identification features of each structural feature recognition channel, and the most suitable recognition channel is selected. For example, similarity is calculated using feature descriptors, and the structural feature recognition channels with the highest similarity are selected. The image data is processed in parallel by the several structural feature recognition channels obtained through matching, and the output results of multiple recognition channels are integrated to generate a set of workpiece structural features.

[0051] Furthermore, the system also includes a structural feature recognition model building module to perform the following steps:

[0052] With the first key surface feature as a constraint, sample collection is performed to obtain a first sample image data set; multiple first associated structural features in the first key surface feature are used to perform feature recognition labeling on the first sample image data set to obtain a first structural feature labeling result set; the first sample image data set and the first structural feature labeling result set are used to perform structural feature recognition channel training to obtain a first structural feature recognition channel; based on the first structural feature recognition channel, the structural feature recognition model of the multiple structural feature recognition channels in parallel is constructed.

[0053] Specifically, first key surface features are randomly selected for sample collection. These features are significant structural features on the workpiece, such as specific holes, grooves, and edges. Specific parameters of the first key surface features, such as size, shape, and location, are clearly defined as constraints for sample collection. A workpiece sample containing the first key surface feature is selected, and an image acquisition device is configured to perform comprehensive image acquisition of the workpiece sample to obtain a first sample image data set.

[0054] Based on the first key surface feature, multiple first associated structural features are determined, and these features include geometric details or related structural features around the first key surface feature. Feature recognition and marking tools, such as image processing software, automatic marking algorithms, etc., are configured, and multiple first associated structural features in the first key surface feature are used to perform feature recognition and marking on the first sample image data set to form a first structural feature marking result set.

[0055] The first sample image data set and the corresponding first structural feature labeling result set are divided into a training data set and a test data set, for example, with a ratio of 80% training and 20% testing. A model type, such as a convolutional neural network (CNN) or a support vector machine (SVM), is selected to train the structural feature recognition channel. The model is trained using the training data set, and the model parameters are optimized using a backpropagation algorithm to minimize the loss function. The model is validated using cross-validation techniques, and the model parameters are adjusted to avoid overfitting and underfitting. The model performance is evaluated using the test data set, and indicators such as accuracy are calculated. Based on the evaluation results, the model is further adjusted and optimized to improve its recognition accuracy. Finally, a trained first structural feature recognition channel model is obtained.

[0056] Similar to the training method of the first channel, different feature labeling result sets and sample image data sets are used to train multiple structural feature recognition channels, design a parallel structural feature recognition model architecture, connect multiple independent structural feature recognition channels in parallel, and fuse the recognition results of multiple channels.

[0057] Furthermore, the system further includes an initial feeding plan set acquisition module to perform the following operation steps:

[0058] Interact with historical milling machine feed control records to obtain a preset feed plan library; traverse the preset feed plan library according to the material information to obtain a first initial feed plan set, wherein the first initial feed plan set includes multiple cutting speed parameters, multiple feed speed parameters, and multiple tool selection parameters mapped to multiple material features in the material information; traverse the preset feed plan library according to the workpiece structure feature set to obtain a second initial feed plan set, wherein the second initial feed plan set includes multiple cutting path parameters, multiple cutting depth parameters, and multiple feed mode parameters mapped to multiple workpiece structure features; combine the first initial feed plan set and the second initial feed plan set to obtain the initial feed plan set.

[0059] Specifically, the system collects historical milling machine feed control records. These records include detailed data on past workpiece machining, such as material type, cutting speed, feed rate, tool selection, and machining results. These records are then organized and cleaned, removing incomplete or erroneous records to ensure data accuracy. The records are then categorized and organized based on workpiece material type and structural characteristics to create a structured feed plan database.

[0060] Receive material information of the target workpiece, including detailed characteristics such as material type, hardness, and heat treatment status. Based on the received material information, search for a feed plan that meets the material information in a preset feed plan library to form a first initial feed plan set. The plans in the first initial feed plan set include key parameters such as cutting speed parameters, feed speed parameters, and tool selection parameters.

[0061] Based on the workpiece structural feature set, a library of preset feed schemes is searched for feed schemes that meet the requirements. These include optimal cutting paths for different structural features, recommended cutting depths for different structural features, and appropriate feed methods such as down milling, up milling, and plunge milling. The selected feed schemes that meet the workpiece structural features are aggregated to form a second set of initial feed schemes. Each plan in this second set includes key parameters such as cutting path parameters, cutting depth parameters, and feed method parameters.

[0062] The first initial feeding scheme set based on material information and the second initial feeding scheme set based on workpiece structural features are integrated to obtain the initial feeding scheme set.

[0063] Furthermore, the system further includes a first initial feeding plan set acquisition module to perform the following operation steps:

[0064] A first material feature is extracted according to the material information, and a first recording material feature is extracted from the preset feeding scheme library; a similarity analysis is performed on the first material feature and the first recording material feature to obtain a first similarity; when the first similarity reaches a preset similarity threshold, a first adding instruction is generated; and according to the first adding instruction, a first initial feeding scheme corresponding to the first recording material feature is added to the first initial feeding scheme set.

[0065] Specifically, a key feature is randomly extracted from the material information as the first material feature, a preset feeding scheme library is accessed, a material feature set corresponding to all records is extracted from the scheme library, and a first recorded material feature is randomly selected from the set.

[0066] A similarity calculation algorithm, such as Euclidean distance, cosine similarity, or Manhattan distance, is selected to calculate the similarity between the first material feature and the first recorded material feature to obtain a similarity score, i.e., a first similarity. A predetermined similarity threshold is set as a criterion for determining whether the material features are similar. The calculated first similarity is compared with the predetermined similarity threshold. If the first similarity reaches or exceeds the predetermined similarity threshold, the first material feature is deemed similar to the first recorded material feature, and a first addition instruction is generated, instructing the feeding scheme corresponding to the first recorded material feature to be added to the first initial feeding scheme set, until all similar material records in the predetermined feeding scheme library are traversed.

[0067] Furthermore, the system further includes a scheme compensation adjustment module to perform the following operation steps:

[0068] During the feed control process, data monitoring is performed to obtain feed control monitoring data, wherein the feed control monitoring data includes cutting speed monitoring data, feed speed monitoring data, and cutting depth monitoring data; the cutting speed parameters, feed speed parameters, and cutting depth parameters of the target feed scheme are extracted; the cutting speed monitoring data, the feed speed monitoring data, and the cutting depth monitoring data are aligned and fitted with the cutting speed parameters, the feed speed parameters, and the cutting depth parameters, respectively, to obtain multiple fitting coefficients; when any fitting coefficient exceeds a preset fitting coefficient interval, a compensation coefficient is generated, and the target feed scheme is compensated and adjusted according to the compensation coefficient.

[0069] Specifically, according to the target feeding plan, the feed control system is started, the target workpiece is processed, and sensors are installed at key positions of the milling machine to monitor the cutting speed, feed speed and cutting depth in real time. Specifically, the cutting speed is monitored in real time, and the speed data of the tool at each moment in the cutting process is recorded; the feed speed is monitored in real time, and the speed data of the workpiece at each moment in the feeding process is recorded; the cutting depth is monitored in real time, and the cutting depth data of the tool in the workpiece at each moment is recorded.

[0070] The control parameters are extracted from the target feed scheme, including cutting speed parameters, feed rate parameters, and cutting depth parameters, which represent the ideal machining parameters.

[0071] The real-time monitoring data and the target parameters are aligned on the time axis so that the monitoring data at each moment corresponds to the target parameters in the same time period. The monitoring data and the target parameters are fitted using a linear regression method. For example, a fitting formula is determined, such as y=ax, where y is the monitoring data, x is the target parameter, and a is the fitting coefficient.

[0072] The cutting speed monitoring data is fitted with the cutting speed parameters to obtain the cutting speed fitting coefficient. Similarly, the feed speed fitting coefficient and the cutting depth fitting coefficient are obtained. Preset interval thresholds are set for the fitting coefficients. These thresholds represent the allowable deviation range. Each fitting coefficient is tested to see if it exceeds the preset interval threshold. If any fitting coefficient exceeds the preset interval threshold, a corresponding compensation coefficient is generated based on the fitting coefficient to adjust the parameters of the target feed plan to compensate for the deviation in actual processing. Specifically, based on the generated compensation coefficient, the cutting speed parameters, feed speed parameters, and cutting depth parameters in the target feed plan are adjusted to form a new feed plan, thereby optimizing feed control.

[0073] In summary, the multi-head milling machine feed control system based on workpiece features provided by the embodiments of the present application has the following technical effects:

[0074] Through the information interaction module, the system can accurately and interactively obtain material information and processing requirement information when the workpiece enters the first working area, ensuring the rationality and accuracy of the processing parameters; the feature recognition module performs comprehensive image acquisition and feature recognition on the workpiece through the image acquisition component and the feature recognition component, and accurately obtains the structural feature set of the workpiece; the solution matching module accurately matches the preset feed solution library based on the material information and the workpiece structural feature set, and generates an initial feed solution set to ensure the adaptability of the feed parameters; the fitting relationship establishment module and the feed parameter optimization module optimize the initial feed solution for the target milling index by establishing a function fitting relationship and parameter optimization, obtain the optimal target feed solution, and improve the processing quality and efficiency; the feed control module controls the feed of the milling component according to the target feed solution to ensure the stability and efficiency of the processing process.

[0075] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multi-head milling machine feed control system based on workpiece features, characterized in that: The multi-head milling machine feed control system based on workpiece features includes: An information interaction module, configured to interactively acquire material information and processing requirement information of the target workpiece when the target workpiece is transferred to the first working area, wherein the processing requirement information includes a target milling index; A workpiece conveying module, wherein the workpiece conveying module is used to convey the target workpiece to a second working area, wherein the second working area includes a workpiece image acquisition component and a workpiece feature recognition component; A feature recognition module, the feature recognition module is used to capture an image of the target workpiece through the workpiece image capture component, perform workpiece feature recognition on the image capture result through the workpiece feature recognition component, and obtain a workpiece structural feature set; A scheme matching module is used to match feeding schemes in a preset feeding scheme library according to the material information and the workpiece structural feature set to obtain an initial feeding scheme set; a fitting relationship establishment module, the fitting relationship establishment module being used to establish a functional fitting relationship between a machining evaluation parameter of a target workpiece and a feed control parameter of a target multi-head milling machine, wherein the machining evaluation parameter corresponds to the target milling index; A feed parameter optimization module, configured to optimize the feed parameters of the initial feed scheme set based on the function fitting relationship and taking the target milling index as the optimization target to obtain a target feed scheme; A feed control module, wherein the feed control module is used to transfer the target workpiece to a third working area, wherein the third working area includes a milling component set and a workpiece fixing component, the target workpiece is fixed according to the workpiece fixing component, and the feed of the milling component set is controlled according to the target feeding scheme.

2. The multi-head milling machine feed control system based on workpiece features as claimed in claim 1, characterized in that: The multi-head milling machine feed control system based on workpiece features further includes an image acquisition result acquisition module to perform the following operation steps: The second working area also includes a rotating table; Acquire geometric data of the target workpiece and a first contact surface, wherein the first contact surface is an initial contact surface between the target workpiece and the rotary stage; generating a rotation angle sequence according to the geometric data, and controlling the rotary table to perform a rotation operation on the target workpiece according to the rotation angle sequence; After each rotation operation, the workpiece image acquisition component acquires an image of the target workpiece until the first contact surface is reached again, and outputs a plurality of surface images of the target workpiece; Perform image alignment on the multiple surface images to obtain the image acquisition result.

3. The multi-head milling machine feed control system based on workpiece features as claimed in claim 2, characterized in that: The multi-head milling machine feed control system based on workpiece features further includes a surface image acquisition module to perform the following operation steps: When the workpiece is at the first contact surface for the first time, the workpiece image acquisition component is activated to acquire an image of the target workpiece to obtain a first surface image; controlling the rotary table to rotate the target workpiece according to the first rotation angle to obtain a second contact surface; comparing the second contact surface with the first contact surface, and if the comparison fails, continuing the image acquisition and rotation operations; When the first contact surface is reached again, the image acquisition and rotation operations are stopped, and the plurality of surface images are output.

4. The multi-head milling machine feed control system based on workpiece features as claimed in claim 1, characterized in that: The multi-head milling machine feed control system based on workpiece features further includes a workpiece structure feature set acquisition module to perform the following operation steps: Pre-building a structural feature recognition model, wherein the structural feature recognition model includes a plurality of structural feature recognition channels, the plurality of structural feature recognition channels are connected in parallel, and the structural feature recognition model is embedded in the workpiece feature recognition component; Performing feature extraction on the image acquisition results to determine several key surface features for workpiece feature recognition; The plurality of structural feature recognition channels are matched based on the plurality of key surface features, and synchronous structural feature recognition is performed on the plurality of structural feature recognition channels obtained by matching to obtain a workpiece structural feature set.

5. The multi-head milling machine feed control system based on workpiece features as claimed in claim 2, characterized in that: The multi-head milling machine feed control system based on workpiece features also includes a structural feature recognition model building module to perform the following operation steps: Taking the first key surface feature as a constraint, performing sample collection to obtain a first sample image data set; Using a plurality of first associated structural features in the first key surface features to perform feature recognition marking on the first sample image data set to obtain a first structural feature marking result set; Using the first sample image data set and the first structural feature labeling result set, performing structural feature recognition channel training to obtain a first structural feature recognition channel; Based on the first structural feature recognition channel, the structural feature recognition model in which the multiple structural feature recognition channels are connected in parallel is constructed.

6. The multi-head milling machine feed control system based on workpiece features as claimed in claim 1, characterized in that: The multi-head milling machine feed control system based on workpiece features further includes an initial feed plan set acquisition module to perform the following operation steps: Interactive historical milling machine feed control records to obtain a library of preset feed plans; Traversing the preset feed scheme library according to the material information to obtain a first initial feed scheme set, wherein the first initial feed scheme set includes a plurality of cutting speed parameters, a plurality of feed speed parameters, and a plurality of tool selection parameters mapped to a plurality of material characteristics in the material information; Traversing the preset feed scheme library according to the workpiece structural feature set to obtain a second initial feed scheme set, wherein the second initial feed scheme set includes a plurality of cutting path parameters, a plurality of cutting depth parameters, and a plurality of feed mode parameters mapped to a plurality of workpiece structural features; The initial feeding scheme set is obtained by combining the first initial feeding scheme set and the second initial feeding scheme set.

7. The multi-head milling machine feed control system based on workpiece features as claimed in claim 6, characterized in that: The multi-head milling machine feed control system based on workpiece features further includes a first initial feed plan set acquisition module to perform the following operation steps: extracting a first material feature according to the material information, and extracting a first recording material feature from the preset feeding scheme library; performing a similarity analysis on the first material feature and the first recording material feature to obtain a first similarity; When the first similarity reaches a preset similarity threshold, generating a first adding instruction; According to the first adding instruction, the first initial feeding scheme corresponding to the first recording material feature is added to the first initial feeding scheme set.

8. The multi-head milling machine feed control system based on workpiece features as claimed in claim 6, characterized in that: The multi-head milling machine feed control system based on workpiece features further includes a scheme compensation adjustment module to perform the following operation steps: Performing data monitoring during the feed control process to obtain feed control monitoring data, wherein the feed control monitoring data includes cutting speed monitoring data, feed speed monitoring data, and cutting depth monitoring data; Extracting cutting speed parameters, feed rate parameters, and cutting depth parameters of the target feed scheme; aligning and fitting the cutting speed monitoring data, the feed speed monitoring data, and the cutting depth monitoring data with the cutting speed parameter, the feed speed parameter, and the cutting depth parameter, respectively, to obtain a plurality of fitting coefficients; When any fitting coefficient exceeds a preset fitting coefficient interval, a compensation coefficient is generated, and a compensation adjustment is performed on the target feeding scheme according to the compensation coefficient.

Images