A high-precision CNC cutting platform and control method thereof

Abstract

The present invention relates to the technical field of numerical control machine tools, in particular to a high-precision CNC cutting platform and a control method thereof, which obtains a processing feature model diagram of a pre-processed product; compares and analyzes the processing feature model diagram with a standard feature model diagram to determine whether the pre-processed product has an abnormal processing state area; if there is no abnormal processing state area, the CNC cutting platform is controlled to perform batch production based on a preset processing scheme; if there is an abnormal processing state area, a feature analysis is performed on the abnormal processing state area to determine whether there is a cutting size offset or a cutting crack in the abnormal processing state area; if there is a cutting size offset or a cutting crack in the abnormal processing state area, the preset processing scheme is adjusted and optimized. The processing accuracy and production efficiency are significantly improved, and the quality control capability of the product is enhanced, and the intelligentization and automation of the production process are realized.

Description

Technical Field

[0001] The invention relates to the technical field of numerically controlled machine tools, in particular to a high-precision CNC cutting platform and a control method thereof. Background Art

[0002] As modern industry's requirements for product accuracy and efficiency continue to increase, high-precision CNC cutting platform technology has developed rapidly. Traditional CNC cutting platforms have the disadvantages of low control accuracy, low processing efficiency and low intelligence. In order to meet the growing market demand, in recent years, high-precision CNC cutting platform technology has made continuous breakthroughs, and many new technologies and methods have emerged. High-precision CNC cutting platforms usually adopt high-precision machine tool structures, equipped with high-performance servo motors, precision guide rails and high-precision sensors to ensure the accuracy and stability of the cutting path. At the same time, advanced control systems and algorithms can monitor the cutting process in real time and make intelligent adjustments to further improve cutting accuracy and efficiency. In addition, the control method of high-precision CNC cutting platforms has also been continuously developed, such as path planning algorithms based on machine learning, adaptive control algorithms, multi-sensor fusion technology, etc., which can effectively improve cutting accuracy, efficiency and reliability. The application of these technologies enables high-precision CNC cutting platforms to meet the higher requirements of modern industry for product accuracy, efficiency and automation, and promotes the transformation and upgrading of the manufacturing industry. In view of this, the present application proposes a high-precision CNC cutting platform and a control method thereof. Summary of the invention

[0003] The invention overcomes the shortcomings of low control accuracy, low processing efficiency and low intelligence of a traditional CNC cutting platform, and provides a high-precision CNC cutting platform and a control method thereof.

[0004] To achieve the above object, the present invention proposes a control method for a high-precision CNC cutting platform, comprising the following steps:

[0005] Obtaining a preset processing plan for a target product, importing the preset processing plan into a control terminal of a CNC cutting platform, and controlling the CNC cutting platform to perform pre-processing based on the preset processing plan;

[0006] After the pre-processing is completed, the pre-processed product is scanned and detected based on the ultrasonic detector to obtain the processing feature model diagram of the pre-processed product;

[0007] Compare and analyze the processing feature model diagram with the standard feature model diagram to determine whether there is an abnormal processing state area in the pre-processed product; if there is no abnormal processing state area, control the CNC cutting platform to perform batch production based on the preset processing plan;

[0008] If there is an abnormal processing state area, a feature analysis is performed on the abnormal processing state area to determine whether there is a cutting size deviation or a cutting crack in the abnormal processing state area;

[0009] If there is a cutting size deviation or a cutting crack in the abnormal processing state area, the preset processing plan is adjusted and optimized to obtain an adjusted and optimized processing plan, and the CNC cutting platform is controlled to perform batch production according to the adjusted and optimized processing plan.

[0010] Preferably, the processing feature model diagram is compared and analyzed with the standard feature model diagram to determine whether there is an abnormal processing state area in the pre-processed product, specifically:

[0011] Obtaining order requirement information of a target product, determining size requirement information of the target product according to the order requirement information, and constructing a standard feature model diagram of the target product according to the size requirement information;

[0012] Constructing a grid space coordinate system, importing the processing feature model diagram and the standard feature model diagram into the grid space coordinate system, retrieving the processing benchmarks in the two feature model diagrams, and aligning and merging the processing benchmarks of the two feature model diagrams with each other in the grid space coordinate system;

[0013] The grid space coordinate system is divided into a plurality of sub-grid spaces to divide the two feature model images into a plurality of sub-model regions, and the volume values ​​of the corresponding sub-model regions in the two feature model images in each sub-grid space are calculated respectively;

[0014] The overlap of each sub-model area in the two feature model images is calculated according to the volume value of the corresponding sub-model area in the two feature model images in each sub-grid space; and the overlap of each sub-model area is compared with a preset overlap threshold;

[0015] If the overlap of a sub-model area is greater than a preset overlap threshold, the sub-model area is defined as a normal processing state area; if the overlap of a sub-model area is not greater than the preset overlap threshold, the sub-model area is defined as an abnormal processing state area.

[0016] Preferably, if there is an abnormal processing state area, a feature analysis is performed on the abnormal processing state area to determine whether there is a cutting size deviation in the abnormal processing state area, specifically:

[0017] Separating a sub-model region corresponding to the abnormal processing state region from the processing feature model graph to obtain a sub-processing feature model graph of the abnormal processing state region;

[0018] and separating the sub-model region corresponding to the abnormal processing state region from the standard feature model graph to obtain a sub-standard feature model graph of the abnormal processing state region;

[0019] The Canny edge detection algorithm is used to perform edge detection processing on the sub-processing feature model graph and the sub-standard feature model graph respectively, so as to obtain an edge contour graph of the sub-processing feature model graph and an edge contour graph of the sub-standard feature model graph;

[0020] The Hausdorff distance algorithm is used to calculate the overlap between the edge contour map of the sub-processing feature model map and the edge contour map of the sub-standard feature model map;

[0021] If the degree of overlap between the edge contour map of the sub-processing feature model map and the edge contour map of the sub-standard feature model map is not greater than the preset overlap threshold, it indicates that the cutting size deviation occurs in the abnormal processing state area;

[0022] If the degree of overlap between the edge contour map of the sub-processing feature model map and the edge contour map of the sub-standard feature model map is greater than a preset overlap threshold, it indicates that there is no cutting size deviation in the abnormal processing state area.

[0023] Preferably, if there is an abnormal processing state area, a feature analysis is performed on the abnormal processing state area to determine whether there is a cutting crack in the abnormal processing state area, specifically:

[0024] Separating a sub-model region corresponding to the abnormal processing state region from the processing feature model graph to obtain a sub-processing feature model graph of the abnormal processing state region;

[0025] voxelize the sub-processing feature model graph based on a voxelization method to voxelize the sub-processing feature model graph into a plurality of voxel grids;

[0026] Traverse each voxel grid and determine whether each voxel grid has a model graph filler;

[0027] If a certain voxel grid has a model graph filler, the filling state of the voxel grid is recorded as 1; if a certain voxel grid does not have a model graph filler, the filling state of the voxel grid is recorded as -1;

[0028] Repeat the above steps until all voxel grids are traversed; calculate the total number of voxel grids whose filling state is recorded as 1, and calculate the total number of voxel grids whose filling state is recorded as -1;

[0029] Dividing the total number of voxel grids whose filling state is recorded as -1 by the total number of voxel grids whose filling state is recorded as 1, to obtain the crack proportion in the abnormal processing state region;

[0030] If the crack proportion in the abnormal processing state region is greater than a preset proportion threshold, it indicates that cutting cracks have occurred in the abnormal processing state region;

[0031] If the crack proportion in the abnormal processing state region is not greater than a preset proportion threshold, it means that no cutting cracks occur in the abnormal processing state region.

[0032] Preferably, if there is a cutting size deviation in the abnormal processing state area, the preset processing scheme is adjusted and optimized to obtain an adjusted and optimized processing scheme, specifically:

[0033] If there is a cutting size deviation in the abnormal processing state area, the operation log of the CNC cutting platform is obtained, and the actual cutting path when the cutting head performs cutting processing on the abnormal processing state area is extracted according to the operation log;

[0034] and obtaining a preset processing scheme for the target product, and extracting a preset cutting path for the cutting head to perform cutting processing on the abnormal processing state area according to the preset processing scheme;

[0035] Calculating the Euclidean distance between the actual cutting path and the preset cutting path at the same cutting position node;

[0036] Determine the drift amplitude of each cutting position node when the cutting head performs cutting processing on the abnormal processing state area according to the Euclidean distance value between the actual cutting path and the preset cutting path at the same cutting position node;

[0037] According to the drift amplitude of each cutting position node when the cutting head performs cutting processing on the abnormal processing state area, the preset feed amount on each cutting position node in the preset processing plan is adjusted to obtain an adjusted and optimized processing plan.

[0038] Preferably, if there is a cutting crack in the abnormal processing state area, the preset processing scheme is adjusted and optimized to obtain an adjusted and optimized processing scheme, specifically:

[0039] Acquire the combination component characteristic information of the target product, and retrieve the limit cutting force value that the target product can withstand in the big data network according to the combination component characteristic information of the target product;

[0040] Obtaining the operation log of the CNC cutting platform, and extracting the actual cutting force value of each cutting time node when the cutting head performs cutting processing on the abnormal processing state area according to the operation log;

[0041] Compare the actual cutting force value of each cutting time node with the limit cutting force value to determine whether the actual cutting force value of a certain cutting time node is greater than the limit cutting force value;

[0042] If there is a situation where the actual cutting force value of a certain cutting time node is greater than the limit cutting force value, the difference between the actual cutting force value of the cutting time node and the limit cutting force value is calculated to obtain a cutting force deviation value, and the preset feed speed at the corresponding cutting position node in the preset processing plan is adjusted according to the cutting force deviation value to obtain an adjusted and optimized processing plan;

[0043] If the actual cutting force value at each cutting time node is not greater than the limit cutting force value, the operation log extracts the actual cutting temperature value when the cutting head performs cutting processing on the abnormal processing state area;

[0044] Determine whether the actual cutting temperature value is greater than the preset cutting temperature value. If it is, calculate the actual cutting temperature value and the preset cutting temperature value to obtain a cutting temperature deviation value. According to the cutting temperature deviation value, adjust the preset coolant spray amount at the corresponding cutting position node in the preset processing plan to obtain an adjusted and optimized processing plan.

[0045] In addition, to achieve the above-mentioned purpose, the present invention also proposes a high-precision CNC cutting platform, which is applied to a control method of any of the high-precision CNC cutting platforms, comprising:

[0046] The main body of the machine tool provides a platform for cutting and processing, including the machine tool bed, worktable, column and slide;

[0047] The motion system controls the motion of the machine tool, including servo motors, lead screws, guide rails and bearings to ensure the motion accuracy and speed of the machine tool;

[0048] Tool system, including tool magazine, tool changer and tool holder for storing, changing and holding tools;

[0049] The workbench system, including the workbench, fixtures and positioning devices, is used to fix and position the workpiece to ensure the accuracy and stability of the cutting process;

[0050] The CNC controller is responsible for receiving the machining program instructions and controlling the machine tool movement according to the instructions, including the servo motor and tool changer;

[0051] Automatic material handling systems for automatic loading and unloading of processed products as well as automatic supply and recovery;

[0052] Data storage and processing system, used to store and process processing programs, processing data and system parameters.

[0053] In addition, to achieve the above-mentioned purpose, the present invention also proposes a control system for a high-precision CNC cutting platform, characterized in that the control system of the high-precision CNC cutting platform includes a memory and a processor, and the memory stores a control method program for the high-precision CNC cutting platform. When the control method program for the high-precision CNC cutting platform is executed by the processor, any one of the control method steps of the high-precision CNC cutting platform is implemented.

[0054] In addition, to achieve the above-mentioned purpose, the present invention also proposes a computer-readable storage medium, in which a control method program for a high-precision CNC cutting platform is stored. When the control method program for the high-precision CNC cutting platform is executed by a processor, any one of the control method steps for the high-precision CNC cutting platform is implemented.

[0055] The present invention solves the technical defects existing in the background technology, and has the following beneficial effects: obtaining a processing feature model diagram of a pre-processed product; comparing and analyzing the processing feature model diagram with the standard feature model diagram to determine whether there is an abnormal processing state area in the pre-processed product; if there is no abnormal processing state area, the CNC cutting platform is controlled to perform batch production based on a preset processing scheme; if there is an abnormal processing state area, the abnormal processing state area is subjected to feature analysis to determine whether there is a cutting size offset or a cutting crack in the abnormal processing state area; if there is a cutting size offset or a cutting crack in the abnormal processing state area, the preset processing scheme is adjusted and optimized to obtain an adjusted and optimized processing scheme, and the CNC cutting platform is controlled to perform batch production according to the adjusted and optimized processing scheme. Through the present invention, not only the processing accuracy and production efficiency are significantly improved, but also the quality control ability of the product is enhanced, the intelligence and automation of the production process are realized, the intelligent adjustment and optimization of the processing scheme are realized, the processing error is effectively reduced, and the yield rate is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0057] Figure 1 The overall method flow chart of this control method is as follows;

[0058] Figure 2 This is a partial method flow chart of this control method;

[0059] Figure 3 This is a schematic diagram of the first structure of the CNC cutting platform;

[0060] Figure 4 This is a second structural schematic diagram of the CNC cutting platform. DETAILED DESCRIPTION

[0061] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0062] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0063] The present invention proposes a control method for a high-precision CNC cutting platform. Figure 1 As shown, the following steps are included:

[0064] S102: Obtaining a preset processing plan for a target product, importing the preset processing plan into a control terminal of a CNC cutting platform, and controlling the CNC cutting platform to perform pre-processing based on the preset processing plan;

[0065] S104: After the pre-processing is completed, the pre-processed product is scanned and detected based on the ultrasonic detector to obtain a processing feature model diagram of the pre-processed product;

[0066] S106: Compare and analyze the processing feature model diagram with the standard feature model diagram to determine whether there is an abnormal processing state area in the pre-processed product; if there is no abnormal processing state area, control the CNC cutting platform to perform batch production based on the preset processing plan;

[0067] S108: If there is an abnormal processing state area, perform feature analysis on the abnormal processing state area to determine whether there is a cutting size deviation or cutting crack in the abnormal processing state area;

[0068] S110: If there is a cutting size deviation or a cutting crack in the abnormal processing state area, the preset processing plan is adjusted and optimized to obtain an adjusted and optimized processing plan, and the CNC cutting platform is controlled to perform batch production according to the adjusted and optimized processing plan.

[0069] It should be noted that, first of all, the relevant technical personnel formulate a detailed processing plan (i.e., a preset processing plan) according to the design drawings or manufacturing requirements of the target product, including parameters such as cutting path, speed, and force. This preset plan usually contains the expected value of product quality and the guiding principles of the manufacturing process. The above preset processing plan is imported into the control terminal of the CNC (computer numerical control) cutting platform through a specific interface or software. This step is to convert the theoretical processing plan into an action instruction that can be executed by the machine. The CNC cutting platform automatically performs the cutting task according to the imported processing plan to complete the preliminary processing of the product. This process may include multiple steps, such as rough processing, fine processing, etc. After the pre-processing is completed, the product is scanned by an ultrasonic detector to collect detailed information about the cutting surface and form a processing feature model diagram. The ultrasonic detector can provide non-destructive inspection to help obtain the internal structure and surface state of the product. The obtained processing feature model diagram is compared and analyzed with the pre-set standard model to check whether the product meets the expected quality standards. If an abnormality is found, such as cutting size offset or cutting crack, it means that there may be errors or problems in the processing process. Through comparative analysis, the areas where problems may occur in the processing process, that is, the abnormal processing state areas, are identified. For the abnormal processing status areas found, the specific reasons need to be further analyzed, such as whether it is caused by cutting size offset or cutting cracks. According to the analysis results, the original preset processing plan is adjusted and optimized to solve the existing problems. Finally, according to the adjusted and optimized processing plan, the CNC cutting platform is mass-produced again to ensure that all products can meet the expected quality standards. The entire process embodies a closed-loop management from design to production to quality control, aiming to improve production efficiency and product quality through automation and intelligent means, while responding to possible production challenges by continuously optimizing processing plans.

[0070] Preferably, the processing feature model diagram is compared and analyzed with the standard feature model diagram to determine whether there is an abnormal processing state area in the pre-processed product, such as Figure 2 As shown, specifically:

[0071] S202: Acquire order requirement information of a target product, determine size requirement information of the target product according to the order requirement information, and construct a standard feature model diagram of the target product according to the size requirement information;

[0072] S204: constructing a grid space coordinate system, importing the processing feature model graph and the standard feature model graph into the grid space coordinate system, retrieving the processing benchmarks in the two feature model graphs, and aligning and merging the processing benchmarks of the two feature model graphs in the grid space coordinate system;

[0073] S206: dividing the grid space coordinate system into a plurality of sub-grid spaces to divide the two feature model images into a plurality of sub-model regions, and respectively calculating the volume values ​​of the corresponding sub-model regions in the two feature model images in each sub-grid space;

[0074] S208: calculating the overlap of each sub-model region in the two feature model images according to the volume value of the corresponding sub-model region in the two feature model images in each sub-grid space; and comparing the overlap of each sub-model region with a preset overlap threshold;

[0075] S210: If the overlap of a certain sub-model region is greater than a preset overlap threshold, the sub-model region is defined as a normal processing state region; if the overlap of a certain sub-model region is not greater than the preset overlap threshold, the sub-model region is defined as an abnormal processing state region.

[0076] It should be noted that, first of all, according to the requirements of the customer order, the size, shape and other specifications of the target product are clarified. This information is the basis for constructing the standard feature model diagram. Based on the order requirement information, an idealized product model that meets the expected standards is established. This model represents the product characteristics in an ideal state and serves as the benchmark for subsequent comparison. In order to quantify the comparison, a unified spatial coordinate system is introduced to compare the processing feature model diagram and the standard feature model diagram in the same dimension. The processing feature model diagram and the standard feature model diagram are imported into the grid space coordinate system, and by identifying and aligning the processing reference points of the two models, their relative positions in the coordinate system are ensured to be correct for subsequent analysis. The grid space coordinate system is divided into multiple sub-areas, and then the processing feature model diagram and the standard feature model diagram are divided into multiple sub-model areas. This division helps to refine the analysis of the differences between different parts. The volume value is calculated for each sub-model area in the sub-grid space. This is the first step to quantify the model difference. The volume value can be used to intuitively see the entity parts of different areas. According to the volume value in the sub-grid space, the overlap between the processing feature model diagram and the standard feature model diagram in each sub-model area is calculated. The overlap reflects the degree of match between the actual processing and the expected standard. The calculated overlap is compared with the pre-set threshold. If the overlap of a partition exceeds the threshold, it means that the processing result of the area is highly matched with the standard model and is defined as the "normal processing state area". On the contrary, if the overlap is lower than the threshold, it indicates that there is a large difference, and the area is marked as the "abnormal processing state area". Through this step, the difference between the product generated during the CNC cutting process and the expected standard can be detected and evaluated, and the problem areas that may exist in the processing process can be accurately identified, providing a basis for subsequent adjustments and optimizations, thereby improving product quality and production efficiency.

[0077] Preferably, if there is an abnormal processing state area, a feature analysis is performed on the abnormal processing state area to determine whether there is a cutting size deviation in the abnormal processing state area, specifically:

[0078] Separating a sub-model region corresponding to the abnormal processing state region from the processing feature model graph to obtain a sub-processing feature model graph of the abnormal processing state region;

[0079] and separating the sub-model region corresponding to the abnormal processing state region from the standard feature model graph to obtain a sub-standard feature model graph of the abnormal processing state region;

[0080] The Canny edge detection algorithm is used to perform edge detection processing on the sub-processing feature model graph and the sub-standard feature model graph respectively, so as to obtain an edge contour graph of the sub-processing feature model graph and an edge contour graph of the sub-standard feature model graph;

[0081] The Hausdorff distance algorithm is used to calculate the overlap between the edge contour map of the sub-processing feature model map and the edge contour map of the sub-standard feature model map;

[0082] If the degree of overlap between the edge contour map of the sub-processing feature model map and the edge contour map of the sub-standard feature model map is not greater than the preset overlap threshold, it indicates that the cutting size deviation occurs in the abnormal processing state area;

[0083] If the degree of overlap between the edge contour map of the sub-processing feature model map and the edge contour map of the sub-standard feature model map is greater than a preset overlap threshold, it indicates that there is no cutting size deviation in the abnormal processing state area.

[0084] It should be noted that, according to the comparative analysis in the previous stage, the area with abnormal processing status is determined. This part of the area may not conform to the expected standard model due to cutting size offset or other problems. After confirming the abnormal area, the next step is to separate the abnormal area from the overall model, and obtain the abnormal area subset of the processing feature model map (called the sub-processing feature model map) and the abnormal area subset of the standard feature model map (called the sub-standard feature model map). These two sub-model maps only contain information about the abnormal area, which is convenient for subsequent detailed analysis. The sub-processing feature model map and the sub-standard feature model map are processed using the Canny edge detection algorithm. The Canny algorithm is a widely used edge detection algorithm that can effectively identify the edges in the image and form an edge contour map. The purpose of this is to clarify the boundaries of the model for subsequent comparison. Next, the Hausdorff distance algorithm is used to calculate the overlap between the two edge contour maps. The Hausdorff distance is an indicator for measuring the similarity of two sets, especially suitable for the comparison of geometric shapes. It is used to evaluate the similarity between the edge contour of the sub-processing feature model map and the edge contour of the sub-standard feature model map. The calculated overlap is compared with the preset overlap threshold. If the overlap is less than or equal to the threshold, it means that the edge contour of the abnormal area is significantly different from the standard model, that is, there is a cutting size offset; conversely, if the overlap is greater than the threshold, it is considered that there is no cutting size offset in the area. This method can accurately locate and determine whether the size offset occurs during the CNC cutting process, providing a basis for subsequent adjustments and optimizations, thereby improving processing accuracy and product quality.

[0085] Preferably, if there is an abnormal processing state area, a feature analysis is performed on the abnormal processing state area to determine whether there is a cutting crack in the abnormal processing state area, specifically:

[0086] Separating a sub-model region corresponding to the abnormal processing state region from the processing feature model graph to obtain a sub-processing feature model graph of the abnormal processing state region;

[0087] voxelize the sub-processing feature model graph based on a voxelization method to voxelize the sub-processing feature model graph into a plurality of voxel grids;

[0088] Traverse each voxel grid and determine whether each voxel grid has a model graph filler;

[0089] If a certain voxel grid has a model graph filler, the filling state of the voxel grid is recorded as 1; if a certain voxel grid does not have a model graph filler, the filling state of the voxel grid is recorded as -1;

[0090] Repeat the above steps until all voxel grids are traversed; calculate the total number of voxel grids whose filling state is recorded as 1, and calculate the total number of voxel grids whose filling state is recorded as -1;

[0091] Dividing the total number of voxel grids whose filling state is recorded as -1 by the total number of voxel grids whose filling state is recorded as 1, to obtain the crack proportion in the abnormal processing state region;

[0092] If the crack proportion in the abnormal processing state region is greater than a preset proportion threshold, it indicates that cutting cracks have occurred in the abnormal processing state region;

[0093] If the crack proportion in the abnormal processing state region is not greater than a preset proportion threshold, it means that no cutting cracks occur in the abnormal processing state region.

[0094] It should be noted that, according to the analysis results in the previous article, the area with abnormal processing state is determined. This part of the area may not conform to the expected standard model due to cutting cracks or other problems. After confirming the abnormal area, the next step is to separate the abnormal area from the overall model to obtain the abnormal area subset of the processing feature model map (called the sub-processing feature model map). This step ensures that the subsequent analysis is only performed on the abnormal area. The sub-processing feature model map is processed by voxelization. Voxelization is a process of converting a three-dimensional object into a series of small cubes (voxels). By voxelizing the model map into several voxel grids, complex spatial structures can be decomposed into small blocks that are easy to analyze. Traverse each voxel grid to check whether it contains model map fillers (i.e., model entity parts). If there is filler in a voxel grid, its filling state is marked as 1; if the grid is empty, it is marked as -1. This process is essentially counting the number of model entity parts and gap parts. Repeat the above filling state judgment process until all voxel grids are traversed. Then the total number of voxel grids with a filling state of 1 and the total number of voxel grids with a filling state of -1 are calculated, which provides the ratio information of the solid part and the void part of the model. The crack ratio in the abnormal processing state area is obtained by dividing the total number of voxel grids with a filling state of -1 by the total number of voxel grids with a filling state of 1. This ratio reflects the distribution density of cracks in the abnormal area. Finally, the calculated crack ratio is compared with the pre-set ratio threshold. If the crack ratio exceeds the threshold, it means that there are cutting cracks in the abnormal area; otherwise, it means that there are no cutting cracks in the area. In this way, it is possible to quantitatively evaluate whether there are crack problems in the CNC cutting process, providing a scientific basis for subsequent process adjustments and quality control.

[0095] Preferably, if there is a cutting size deviation in the abnormal processing state area, the preset processing scheme is adjusted and optimized to obtain an adjusted and optimized processing scheme, specifically:

[0096] If there is a cutting size deviation in the abnormal processing state area, the operation log of the CNC cutting platform is obtained, and the actual cutting path when the cutting head performs cutting processing on the abnormal processing state area is extracted according to the operation log;

[0097] and obtaining a preset processing scheme for the target product, and extracting a preset cutting path for the cutting head to perform cutting processing on the abnormal processing state area according to the preset processing scheme;

[0098] Calculating the Euclidean distance between the actual cutting path and the preset cutting path at the same cutting position node;

[0099] Determine the drift amplitude of each cutting position node when the cutting head performs cutting processing on the abnormal processing state area according to the Euclidean distance value between the actual cutting path and the preset cutting path at the same cutting position node;

[0100] According to the drift amplitude of each cutting position node when the cutting head performs cutting processing on the abnormal processing state area, the preset feed amount on each cutting position node in the preset processing plan is adjusted to obtain an adjusted and optimized processing plan.

[0101] It should be noted that if it is confirmed that there is a cutting size offset in the abnormal processing state area. This means that in a specific area, there is a deviation between the actual cutting result and the preset cutting path. At this time, the operation log of the CNC cutting platform is obtained, and the actual cutting path of the cutting head when processing the abnormal area is extracted from it. This step is based on the historical operation record to understand the actual behavior of the cutting head. At the same time, referring to the preset processing plan of the target product, the preset cutting path that the cutting head should follow in theory is extracted. This is an ideal path based on the design and expected results. By comparing the positions of the actual path and the preset path at the same cutting position node, the difference between the two is calculated using the Euclidean distance formula. This distance value reflects the deviation between the actual position and the preset position at each node. According to the Euclidean distance value calculated above, the drift amplitude of each cutting position node when the cutting head is processing the abnormal area is analyzed. The node with a larger drift amplitude indicates that the actual cutting effect at this position is more different from the expected. Finally, based on the drift amplitude obtained by analysis, the preset feed amount at each cutting position node in the preset processing plan is adjusted. By reducing or increasing the feed rate, the movement of the cutting head at a specific position can be fine-tuned to compensate for previously identified deviations, thereby optimizing the entire machining process. This process is essentially a feedback loop, identifying problems, analyzing causes and implementing improvements through a data-driven approach, ultimately achieving the goal of improving machining accuracy and efficiency.

[0102] Preferably, if there is a cutting crack in the abnormal processing state area, the preset processing scheme is adjusted and optimized to obtain an adjusted and optimized processing scheme, specifically:

[0103] Acquire the combination component characteristic information of the target product, and retrieve the limit cutting force value that the target product can withstand in the big data network according to the combination component characteristic information of the target product;

[0104] Obtaining the operation log of the CNC cutting platform, and extracting the actual cutting force value of each cutting time node when the cutting head performs cutting processing on the abnormal processing state area according to the operation log;

[0105] Compare the actual cutting force value of each cutting time node with the limit cutting force value to determine whether the actual cutting force value of a certain cutting time node is greater than the limit cutting force value;

[0106] If there is a situation where the actual cutting force value of a certain cutting time node is greater than the limit cutting force value, the difference between the actual cutting force value of the cutting time node and the limit cutting force value is calculated to obtain a cutting force deviation value, and the preset feed speed at the corresponding cutting position node in the preset processing plan is adjusted according to the cutting force deviation value to obtain an adjusted and optimized processing plan;

[0107] If the actual cutting force value at each cutting time node is not greater than the limit cutting force value, the operation log extracts the actual cutting temperature value when the cutting head performs cutting processing on the abnormal processing state area;

[0108] Determine whether the actual cutting temperature value is greater than the preset cutting temperature value. If it is, calculate the actual cutting temperature value and the preset cutting temperature value to obtain a cutting temperature deviation value. According to the cutting temperature deviation value, adjust the preset coolant spray amount at the corresponding cutting position node in the preset processing plan to obtain an adjusted and optimized processing plan.

[0109] It should be noted that the maximum cutting force value that the product can withstand is retrieved from the big data network through the combined component characteristic information of the target product. This information is crucial to ensure the safety and effectiveness of the processing process. The operation log of the CNC cutting platform is obtained, and the actual cutting force value of the cutting head at each cutting time node when processing the abnormal area is extracted from it. The extracted actual cutting force value is compared with the previously determined limit cutting force value. If the actual cutting force value of a certain cutting time node is found to exceed the limit value, it means that there is too much load at this moment, which may cause cracks or other processing quality problems. For those cases where the actual cutting force value exceeds the limit value, the difference between the two (cutting force deviation value) is calculated, and this is used as a basis to adjust the preset feed speed of the corresponding cutting position node in the preset processing plan. By reducing the feed speed, the cutting force can be reduced, excessive load can be avoided, and the risk of cracks can be reduced. If the actual cutting force values ​​of all cutting time nodes are within the safe range, but the actual cutting temperature value exceeds the preset cutting temperature value, it means that cracks may be caused by excessive thermal stress. At this time, by calculating the deviation between the actual cutting temperature value and the preset cutting temperature value, adjust the preset coolant spray volume of the corresponding cutting position node in the preset processing plan. Increasing the coolant spray volume helps to reduce the temperature of the cutting area, reduce thermal stress, and prevent cracks. Through this series of monitoring and adjustment measures, crack problems in the processing process can be effectively prevented and reduced, and processing quality and product performance can be improved. At the same time, it is important to optimize processing parameters in a data-driven way to achieve an efficient, safe, and high-quality production process.

[0110] Furthermore, the control method of the high-precision CNC cutting platform may also include the following steps:

[0111] Precision grade characteristic model diagram of various dimensional accuracy grades corresponding to the prefabricated target product after cutting and processing on the CNC cutting platform;

[0112] Constructing a graph-based network topology diagram, cutting the graph-based network topology diagram into a plurality of branch nodes, and assigning corresponding dimensional accuracy level labels to each branch node based on various dimensional accuracy levels;

[0113] And storing the accuracy level characteristic model diagrams corresponding to various dimensional accuracy levels on corresponding branch nodes in the graph network topology diagram;

[0114] During the batch production process of the CNC cutting platform, the processing feature model diagram of each actual processed target product is obtained, and the overlap between the processing feature model diagram of the actual processed target product and the accuracy level feature model diagram on each branch node in the graph network topology diagram is calculated;

[0115] If the degree of overlap between the processing feature model diagram of the actual processed target product and the accuracy level feature model diagram on each branch node in the graph network topology diagram is not greater than the preset threshold, it means that the actual processed target product does not meet the preset processing requirements, and it will be marked as waste;

[0116] If the degree of overlap between the processing feature model graph of the actual processed target product and the accuracy level feature model graphs on one or more branch nodes in the graph network topology graph is greater than a preset threshold, it means that the actual processed target product meets the preset processing requirements, and the accuracy level feature model graphs with overlaps greater than the preset threshold are sorted to extract the maximum overlap;

[0117] Obtain the dimension accuracy level label of the corresponding branch node on the network topology diagram in the accuracy level feature model diagram with maximum overlap, determine the accuracy level of the actual target product after processing according to the extracted dimension accuracy level label, and move the actual target product after processing to the corresponding accuracy level storage box according to the accuracy level of the actual target product after processing.

[0118] It should be noted that in order to distinguish products with different dimensional accuracy levels, relevant technical personnel have pre-designed a series of corresponding accuracy level feature model diagrams. These model diagrams describe in detail the ideal shape, size and structural characteristics of products at different accuracy levels. Next, a graph network topology diagram is constructed to organize and link the feature model diagrams of various dimensional accuracy levels. The graph network topology diagram consists of multiple branch nodes, each of which represents a specific dimensional accuracy level. Each branch node is assigned a corresponding label according to the dimensional accuracy level it represents, and the feature model diagram of the corresponding accuracy level is stored on the branch node. This step establishes a direct association between the model diagram and the branch node, which facilitates subsequent query and comparison.

[0119] During batch production, each actual processed target product will generate a processing feature model diagram, reflecting its actual shape, size and structural features. The processing feature model diagram of each actual processed target product is compared with the branch nodes in the graph network topology diagram, and the overlap between the two is calculated. The overlap reflects the similarity between the model diagrams and is used to evaluate the matching degree between the actual product and the ideal accuracy grade model. According to the calculated overlap, the accuracy grade of each product is evaluated. If the overlap of all branch nodes is less than the preset threshold, it means that the product does not meet any preset accuracy grade requirements and is regarded as a defective product (waste product). On the contrary, if the overlap of at least one branch node exceeds the preset threshold, it means that the product meets the requirements of a certain accuracy grade. If the overlap of more than one branch node exceeds the preset threshold, the branch node with the largest overlap is selected. According to the dimensional accuracy grade label of the branch node, the final accuracy grade of the product is determined, and it is moved to the corresponding accuracy grade storage box to achieve efficient and automated classified storage. This method can improve production efficiency and ensure the consistency and accuracy of product quality through an automated and intelligent evaluation process.

[0120] Furthermore, the control method of the high-precision CNC cutting platform may also include the following steps:

[0121] Acquire fault characteristic data when various operating faults occur on the CNC cutting platform, and acquire actual operating data of the CNC cutting platform within a preset time period before various operating faults occur;

[0122] A graph neural network is introduced, and the actual operation data of the CNC cutting platform within a preset time period before various operation failures occur is used as the first graph node, and the fault feature data when various operation failures occur on the CNC cutting platform is used as the second graph node to construct a neural network topology graph;

[0123] Obtaining the state transition probability of the CNC cutting platform transferring from a normal state to a fault state under various operating data conditions according to the neural network topology diagram, and constructing a state transition probability matrix according to the state transition probability;

[0124] Introducing a Markov chain, and importing the state transition probability matrix into the Markov chain for encoding learning, and after learning is completed, outputting a Markov chain prediction model;

[0125] During the actual processing of the CNC cutting platform, real-time operation data of the CNC cutting platform is obtained in a preset processing time period, and the real-time operation data is imported into the Markov chain prediction model for prediction to obtain the transition probability of the CNC cutting platform transferring to a fault state;

[0126] If the probability of the CNC cutting platform transferring to a fault state is greater than a preset probability value, the CNC cutting platform is controlled to stop production and a fault warning message is generated.

[0127] It should be noted that the fault feature data of the CNC cutting platform when various operating faults occur, as well as the actual operating data in a period of time before the fault occurs, are collected. These data include but are not limited to the operating parameters, operating modes, maintenance records, environmental conditions, etc. of the equipment, which are the basis for building a prediction model. Using the Graph Neural Network (GNN), the normal operating data before the fault occurs is used as one of the graph nodes, and the fault feature data is used as another type of node to construct a neural network topology graph. GNN can effectively process and learn the complex relationships and dependencies between nodes, which is very important for understanding the operating state and fault evolution process of the CNC cutting platform. By analyzing the neural network topology graph, the probability of the CNC cutting platform transitioning from the normal state to the fault state under different operating conditions is obtained. These probabilities constitute the state transition probability matrix, which is the core component of the prediction model and can quantify the transition risk between different operating states. Markov chain is a probabilistic model used to describe the process of a system randomly transitioning between different states. By importing the state transition probability matrix into the Markov chain, the dynamic evolution process of the CNC cutting platform from the normal state to the fault state can be simulated. Through encoding learning, the model can capture and learn the dynamic characteristics of the system. After learning, the Markov chain prediction model can predict the probability that the CNC cutting platform will transfer to a fault state within a preset time period in the future. Once the model predicts that the transfer probability exceeds the preset threshold (i.e., the fault warning value), the system will respond immediately. During the actual processing of the CNC cutting platform, real-time operation data is collected regularly and input into the Markov chain prediction model. Based on this, the model predicts the probability that the platform will transfer to a fault state in the future. When the predicted transfer probability exceeds the preset value, the system will automatically trigger the shutdown production and generate fault warning information so that preventive measures can be taken in time to avoid potential equipment damage and production interruptions. This method can provide real-time and dynamic fault warnings, so that preventive measures can be taken before the fault actually occurs, thereby reducing production stoppages, reducing maintenance costs, and improving equipment utilization and production efficiency.

[0128] In addition, the present invention also proposes a high-precision CNC cutting platform, which is applied to the control method of any of the high-precision CNC cutting platforms described above, such as Figure 3 , Figure 4 As shown, including:

[0129] The machine tool body 2101 provides a platform for cutting and processing, including a machine tool bed, a workbench, a column and a slide;

[0130] Motion system 2102, which controls the motion of the machine tool, including servo motors, lead screws, guide rails and bearings, to ensure the motion accuracy and speed of the machine tool;

[0131] Tool system 2103, including a tool magazine, a tool changer and a tool holder, for storing, changing and holding tools;

[0132] The workbench system 2104 includes a workbench, a fixture, and a positioning device, which is used to fix and position the workpiece to ensure the accuracy and stability of the cutting process;

[0133] CNC controller 2105, responsible for receiving machining program instructions and controlling the movement of the machine tool according to the instructions, including the servo motor and tool changer;

[0134] Automatic material handling system 2106 for automatic loading and unloading of processed products and automatic supply and recovery;

[0135] Data storage and processing system, used to store and process processing programs, processing data and system parameters.

[0136] In addition, the present invention also proposes a control system for a high-precision CNC cutting platform, characterized in that the control system of the high-precision CNC cutting platform includes a memory and a processor, and the memory stores a control method program for the high-precision CNC cutting platform. When the control method program for the high-precision CNC cutting platform is executed by the processor, any one of the control method steps of the high-precision CNC cutting platform is implemented.

[0137] In addition, the present invention also proposes a computer-readable storage medium, which stores a control method program for a high-precision CNC cutting platform. When the control method program for the high-precision CNC cutting platform is executed by a processor, any one of the control method steps for the high-precision CNC cutting platform is implemented.

[0138] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A control method for a high-precision CNC cutting platform, characterized in that: The following steps are involved: Obtaining a preset processing plan for a target product, importing the preset processing plan into a control terminal of a CNC cutting platform, and controlling the CNC cutting platform to perform pre-processing based on the preset processing plan; After the pre-processing is completed, the pre-processed product is scanned and detected based on the ultrasonic detector to obtain the processing feature model diagram of the pre-processed product; Compare and analyze the processing feature model diagram with the standard feature model diagram to determine whether there is an abnormal processing state area in the pre-processed product; if there is no abnormal processing state area, control the CNC cutting platform to perform batch production based on the preset processing plan; If there is an abnormal processing state area, a feature analysis is performed on the abnormal processing state area to determine whether there is a cutting crack in the abnormal processing state area; If there are cutting cracks in the abnormal processing state area, the preset processing scheme is adjusted and optimized to obtain an adjusted and optimized processing scheme, and the CNC cutting platform is controlled to perform batch production according to the adjusted and optimized processing scheme; If there is a cutting crack in the abnormal processing state area, the preset processing scheme is adjusted and optimized to obtain an adjusted and optimized processing scheme, which is specifically: Acquire the combination component characteristic information of the target product, and retrieve the limit cutting force value that the target product can withstand in the big data network according to the combination component characteristic information of the target product; Obtaining the operation log of the CNC cutting platform, and extracting the actual cutting force value of each cutting time node when the cutting head performs cutting processing on the abnormal processing state area according to the operation log; Compare the actual cutting force value of each cutting time node with the limit cutting force value to determine whether the actual cutting force value of a certain cutting time node is greater than the limit cutting force value; If there is a situation where the actual cutting force value of a certain cutting time node is greater than the limit cutting force value, the difference between the actual cutting force value of the cutting time node and the limit cutting force value is calculated to obtain a cutting force deviation value, and the preset feed speed at the corresponding cutting position node in the preset processing plan is adjusted according to the cutting force deviation value to obtain an adjusted and optimized processing plan; If the actual cutting force value at each cutting time node is not greater than the limit cutting force value, the operation log extracts the actual cutting temperature value when the cutting head performs cutting processing on the abnormal processing state area; Determine whether the actual cutting temperature value is greater than the preset cutting temperature value. If it is, calculate the actual cutting temperature value and the preset cutting temperature value to obtain a cutting temperature deviation value. According to the cutting temperature deviation value, adjust the preset coolant spray amount at the corresponding cutting position node in the preset processing plan to obtain an adjusted and optimized processing plan.

2. The control method of a high-precision CNC cutting platform according to claim 1, characterized in that: Compare and analyze the processing feature model diagram with the standard feature model diagram to determine whether there is an abnormal processing state area in the pre-processed product, specifically: Obtaining order requirement information of a target product, determining size requirement information of the target product according to the order requirement information, and constructing a standard feature model diagram of the target product according to the size requirement information; Constructing a grid space coordinate system, importing the processing feature model diagram and the standard feature model diagram into the grid space coordinate system, retrieving the processing benchmarks in the two feature model diagrams, and aligning and merging the processing benchmarks of the two feature model diagrams with each other in the grid space coordinate system; The grid space coordinate system is divided into a plurality of sub-grid spaces to divide the two feature model images into a plurality of sub-model regions, and the volume values ​​of the corresponding sub-model regions in the two feature model images in each sub-grid space are calculated respectively; The overlap of each sub-model area in the two feature model images is calculated according to the volume value of the corresponding sub-model area in the two feature model images in each sub-grid space; and comparing the overlap of each sub-model region with a preset overlap threshold; If the overlap of a certain sub-model area is greater than a preset overlap threshold, the sub-model area is defined as a normal processing state area; If the overlap of a certain sub-model area is not greater than a preset overlap threshold, the sub-model area is defined as an abnormal processing state area.

3. The control method of a high-precision CNC cutting platform according to claim 1, characterized in that: If there is an abnormal processing state area, the abnormal processing state area is subjected to feature analysis to determine whether there is a cutting crack in the abnormal processing state area, specifically: Separating a sub-model region corresponding to the abnormal processing state region from the processing feature model graph to obtain a sub-processing feature model graph of the abnormal processing state region; voxelize the sub-processing feature model graph based on a voxelization method to voxelize the sub-processing feature model graph into a plurality of voxel grids; Traverse each voxel grid and determine whether each voxel grid has a model graph filler; If a certain voxel grid has a model graph filler, the filling state of the voxel grid is recorded as 1; if a certain voxel grid does not have a model graph filler, the filling state of the voxel grid is recorded as -1; Repeat the above steps until all voxel grids are traversed; calculate the total number of voxel grids whose filling state is recorded as 1, and calculate the total number of voxel grids whose filling state is recorded as -1; Dividing the total number of voxel grids whose filling state is recorded as -1 by the total number of voxel grids whose filling state is recorded as 1, to obtain the crack proportion in the abnormal processing state region; If the crack proportion in the abnormal processing state region is greater than a preset proportion threshold, it indicates that cutting cracks have occurred in the abnormal processing state region; If the crack proportion in the abnormal processing state region is not greater than a preset proportion threshold, it means that no cutting cracks occur in the abnormal processing state region.

4. A high-precision CNC cutting platform, applied to the control method of the high-precision CNC cutting platform according to any one of claims 1 to 3, characterized in that: include: The main body of the machine tool provides a platform for cutting and processing, including the machine tool bed, worktable, column and slide; The motion system controls the motion of the machine tool, including servo motors, lead screws, guide rails and bearings to ensure the motion accuracy and speed of the machine tool; Tool system, including tool magazine, tool changer and tool holder for storing, changing and holding tools; The workbench system, including the workbench, fixtures and positioning devices, is used to fix and position the workpiece to ensure the accuracy and stability of the cutting process; The CNC controller is responsible for receiving the machining program instructions and controlling the machine tool movement according to the instructions, including the servo motor and tool changer; Automatic material handling systems for automatic loading and unloading of processed products as well as automatic supply and recovery; Data storage and processing system, used to store and process processing programs, processing data and system parameters.

5. A control system for a high-precision CNC cutting platform, characterized in that: The control system of the high-precision CNC cutting platform includes a memory and a processor. The memory stores a control method program for the high-precision CNC cutting platform. When the control method program for the high-precision CNC cutting platform is executed by the processor, the control method steps for the high-precision CNC cutting platform as described in any one of claims 1 to 3 are implemented.

6. A computer-readable storage medium, wherein a control method program for a high-precision CNC cutting platform is stored in the computer-readable storage medium. When the control method program for the high-precision CNC cutting platform is executed by a processor, the control method steps for the high-precision CNC cutting platform as described in any one of claims 1 to 3 are implemented.

Images