A method and system for identifying tool wear of CNC machine tools

Abstract

The present application relates to the field of tool wear identification, and in particular to a method and system for identifying tool wear of numerical control machine tools, the method comprising: using an image of tool projection before processing as a first image, and using an image of tool projection after processing as a second image; calculating the thickness wear of the tool according to the first image and the second image; calculating the edge wear of the tool; obtaining the corresponding standard cutting speed through a matching model according to the thickness wear and the edge wear; wherein, extracting the first edge in the first image and the second edge in the second image, calculating the average value of the Euclidean distance between the pixels closest to each pixel of the second edge in the first edge; calculating the image entropy ratio of the image entropy of the pixel on the first edge to the image entropy of the pixel on the second edge; obtaining the thickness wear according to the average value of the Euclidean distance and the image entropy ratio. The present application has the effect of optimizing the cutting speed according to the degree of tool wear.

Description

Technical Field

[0001] The present application relates to the field of tool wear identification, and in particular to a method and system for identifying tool wear of CNC machine tools. Background Art

[0002] Intelligent CNC machine tool cutting refers to the process of machining workpieces using CNC machine tools, which involves many aspects, including the selection of cutting parameters, planning of cutting paths, etc. Intelligent CNC machine tools can achieve efficient, precise and automated cutting through integrated sensors, control systems and software algorithms.

[0003] During the cutting process of the cutting tool in the CNC machine tool, the life cycle of the cutting tool and its wear state have a direct impact on the processing quality and efficiency. The wear of the cutting tool can be analyzed by collecting cutting tool images. For example, the Chinese patent application with publication number CN118096760A discloses a cutting tool wear detection method, device and readable storage medium, the method comprising: when the cutting tool and the spindle are in a preset measurement posture, collecting a full-angle side view of the cutting tool; determining the blade feature map of the cutting tool based on the full-angle side view; processing the blade feature maps of adjacent frames in sequence to determine the similarity data of the cutting tool; and determining the wear state of the cutting tool based on the similarity data.

[0004] The use and wear process of the tool can be divided into three stages: initial wear stage, stable wear stage and rapid wear stage. Within a certain range, slight tool wear will not seriously affect the use of the tool. This is because a certain wear allowance is usually taken into account when designing the tool to ensure good processing performance in the initial and stable wear stages. However, once entering the rapid wear stage, the performance of the tool will decline rapidly, which will not only affect the processing accuracy and surface quality, but may also cause tool damage, increase processing costs and downtime.

[0005] Tools with different degrees of wear have very different sensitivities to cutting speeds. Appropriate cutting speed adjustments can delay or optimize the wear process to a certain extent. Tool wear is a dynamic process. As the degree of wear increases, the cutting performance gradually decreases. In the initial wear stage, the tool surface is relatively smooth and the cutting force is relatively low. Appropriate high-speed cutting helps reduce thermal effects and contact time, delaying wear. However, as the wear enters the regular or even rapid wear stage, the increase in cutting speed will aggravate the wear, resulting in a significant increase in cutting force, and may even cause tool failure. Therefore, timely adjusting the cutting speed according to the wear conditions is extremely critical to maintaining tool life and processing quality. Therefore, how to optimize the cutting speed according to the degree of tool wear has become an urgent problem to be solved. Summary of the invention

[0006] In order to solve the technical problem of how to identify the degree of tool wear and optimize the cutting speed according to the degree of tool wear, the present application provides a method and system for identifying tool wear of a CNC machine tool.

[0007] In a first aspect, the present application provides a method for identifying tool wear of a CNC machine tool, which adopts the following technical solution:

[0008] A method for identifying tool wear of a numerical control machine tool comprises the steps of: taking an image of a tool projection before processing as a first image, taking an image of a tool projection after processing as a second image, wherein the light source in the first image and the second image is the same, the position and placement angle of the tool are the same, and the light source is tilted; calculating the thickness wear of the tool according to the first image and the second image; the thickness wear is calculated by the following method: extracting the projection of the edge of the tool blade in the first image as the first edge, extracting the projection of the edge of the tool blade in the second image as the second edge, calculating the average value of the Euclidean distance between the pixels closest to each pixel of the second edge in the first edge; calculating the image entropy ratio of the image entropy of the pixel points on the first edge to the image entropy of the pixel points on the second edge; obtaining the thickness wear according to the average value of the Euclidean distance and the image entropy ratio; calculating the edge wear of the tool according to the torque of the tool axis; constructing a data set according to the thickness wear and edge wear in historical data, labeling the data set, the label is a standard cutting speed, and training a neural network model according to the labeled data set to obtain a matching model; inputting the thickness wear and edge wear of real-time data into the matching model, and outputting the corresponding standard cutting speed.

[0009] This application reflects the thickness change of the tool by identifying the change of the projected edge in the tool projection image; the blade of the tool will wear evenly and become thinner as the use time increases. If the thickness change of the blade is directly identified through the image, the recognition accuracy is low due to the small degree of change. The change in the edge position of the tool projection can magnify the thickness change of the above-mentioned blade. The projection of the blade should have a clear edge in the initial state. As the wear of the tool appears and increases, the edge position and clarity of the projection will change. The deformation of the blade of the tool is reflected by the change of the torque of the tool; the cutting speed is adjusted dynamically according to the thickness change and deformation of the tool to reduce the frequent replacement due to excessive wear of the tool, reasonably distribute the wear of the tool, reduce the occurrence of local excessive wear, and extend the service life of the tool. Maintain stable cutting conditions to ensure processing accuracy and surface quality.

[0010] Optionally, the calculation formula for the image entropy of the pixel on the first edge is:

[0011] ,in, represents the image entropy of the pixel on the first edge, represents the gray level, Indicates grayscale The probability of appearing on the first edge; the image entropy of the pixel points on the second edge is calculated in the same way as the image entropy calculation method of the pixel points on the first edge.

[0012] Image entropy can reflect the grayscale distribution of pixels on the edge. When the blade wear is small, the edge of the blade projection should be clearer, and the number of grayscale levels of the edge of the tool projection is small. When the blade wear increases, the edge of the blade projection should be more blurred, and the number of grayscale levels of the edge of the tool projection is large.

[0013] Optionally, the calculation formula for the image entropy of the pixel on the first edge is:

[0014] ,in, represents the image entropy of the pixel on the first edge, represents the gray level, Indicates grayscale The probability of appearing on the first edge; the image entropy of the pixel points on the second edge is calculated in the same way as the image entropy calculation method of the pixel points on the first edge, Represents the accumulated value of the gray level present on the first edge.

[0015] If the grayscale of the pixels on the edge decreases uniformly (possibly due to uniform wear or external light interference), but the image entropy is large, there is a defect. A factor of grayscale value can be added to the calculation of edge image entropy to overcome this defect.

[0016] Optionally, calculate the edge wear of the tool using the formula:

[0017] ,in, express The edge wear of the moment, Indicates from time 1, time 2 to The torque sequence of the force on the tool axis at each moment, Indicates the minimum value of the force torque in the torque sequence, Indicates the maximum value of the force torque in the sequence.

[0018] Calculate from the first moment to the The minimum and maximum values ​​of all torque values ​​at the moment are obtained by dividing the minimum value by the maximum value. The closer the ratio is to 1, the more stable the force on the tool during cutting and the lower the wear. If the ratio is far from 1, it means that the force on the tool fluctuates greatly during cutting and the wear is higher.

[0019] Optionally, calculate the edge wear of the tool using the formula:

[0020] ,in, express The edge wear of the moment, Indicates from time 1, time 2 to The torque sequence of the force on the tool axis at each moment, Indicates the minimum value of the force torque in the torque sequence, Indicates the maximum value of the force torque in the torque sequence, represents the median in the torque series, Indicates the number of torque values ​​between the median and the minimum value, Indicates the number of torque values ​​between the maximum value and the median value.

[0021] The closer the ratio is to 1, The closer it is to 1, the smaller the change in torque value and the more regular the torque value distribution, the more stable the tool is. The ratio of the maximum and minimum torque of the tool and the torque distribution are combined to comprehensively judge the stability of the force applied to the tool during operation, so as to quantify the degree of tool wear. Taking multiple factors into consideration, the quantification result of the degree of wear is more accurate.

[0022] Optionally, the calculation of the edge wear of the tool also includes the following steps: the step of denoising the collected initial torque sequence to obtain a torque sequence, the denoising step includes: obtaining the initial torque sequence and initial amplitude sequence of the tool; calculating the synchronization of the initial torque sequence and the amplitude sequence, when the synchronization is less than or equal to the synchronization threshold, deleting the corresponding force torque and vibration amplitude to obtain the torque sequence and amplitude sequence.

[0023] The influence of abnormal vibration data on the calculation results caused by non-wear reasons (such as the fixture is not clamped when holding the workpiece, the tool is installed loosely, etc.) is reduced. When the synchronization is greater than the synchronization threshold, it means that the torque sequence and the amplitude sequence are highly synchronized, that is, they rise synchronously, fall synchronously, or fluctuate synchronously. On the contrary, the synchronization is poor. Deleting this part of the data can improve the accuracy of the calculation.

[0024] Optionally, the synchronization is calculated by: supplementing the shorter sequence of the torque sequence and the amplitude sequence to make the torque sequence and the amplitude sequence equal in length; and taking the Pearson correlation coefficient of the torque sequence and the amplitude sequence as the synchronization.

[0025] The Pearson correlation coefficient is a statistical indicator used to measure the strength of the linear correlation between two variables. Between 0 and 1, where 1 indicates a perfect positive correlation, that is, when one sequence increases, the other sequence also increases accordingly. Indicates a perfect negative correlation, that is, when one series increases, the other decreases.

[0026] In the second aspect, the present application provides a CNC machine tool tool wear identification system, which adopts the following technical solution:

[0027] A tool wear identification system for a numerically controlled machine tool comprises: a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a tool wear identification method for a numerically controlled machine tool is implemented.

[0028] The above-mentioned CNC machine tool tool wear identification method is generated into a computer program and stored in a memory so as to be loaded and executed by a processor. Thus, a system is made based on the memory and the processor for easy use.

[0029] This application has the following technical effects:

[0030] 1. The thickness change of the tool is reflected by identifying the change of the projection edge in the tool projection image; the deformation of the tool blade is reflected by the change of the tool torque; the cutting speed is adjusted dynamically according to the thickness change and deformation of the tool to reduce the frequent replacement caused by excessive tool wear, reasonably distribute the tool wear, reduce the occurrence of local excessive wear, and extend the tool life. Maintain stable cutting conditions to ensure processing accuracy and surface quality.

[0031] 2. As the tool is used for a longer time, the blade will wear out evenly and become thinner. If the thickness change of the blade is directly recognized through images, the recognition accuracy is low due to the small degree of change. The change in the edge position of the tool projection can magnify the thickness change of the blade. The projection of the blade should have a clear edge in the initial state. As the wear of the tool appears and increases, the edge position and clarity of the projection will change. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] By reading the detailed description below with reference to the accompanying drawings, the above and other purposes, features and advantages of the exemplary embodiments of the present application will become easily understood. In the accompanying drawings, several embodiments of the present application are shown in an exemplary and non-restrictive manner, and the same or corresponding numbers represent the same or corresponding parts.

[0033] Figure 1 It is a method flow chart of a method for identifying tool wear of a CNC machine tool according to an embodiment of the present application.

[0034] Figure 2 This is a simplified diagram showing the position relationship of a tool, light and a camera in a method for identifying tool wear for a CNC machine tool according to an embodiment of the present application.

[0035] Figure 3 This is a method flow chart of step S2 in a method for identifying tool wear of a CNC machine tool according to an embodiment of the present application.

[0036] Explanation of the accompanying drawings: 1. blade body; 2. light; 3. camera; 4. projection area; 5. window; 6. edge segment. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present application.

[0038] It should be understood that when the terms "first", "second", etc. are used in the claims, specification and drawings of the present application, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the specification and claims of the present application indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their collections.

[0039] The present application embodiment discloses a method for identifying tool wear of a CNC machine tool, referring to Figure 1 , including steps S1 to S4, which are specifically as follows:

[0040] S1: The image projected by the tool before processing is used as the first image, and the image projected by the tool after processing is used as the second image, wherein the light source in the first image and the second image is the same, and the position and placement angle of the tool are the same, and the light source is tilted.

[0041] In one embodiment, referring to Figure 2 , set an inclined light source, under the action of the light 2 of the light source, a projection area 4 is generated under the blade 1 (tool), set a camera 3, so that the camera 3 takes an image including the projection area 4, set a window 5, so that the edge segment 6 of the image area including the projection area 4 in the window 5 is used as the projection of the edge of the tool blade. Ensure that the light source intensity and angle corresponding to the first image and the second image, and the tool setting position and angle are the same to avoid errors caused by different external factors such as light sources or tools.

[0042] In one embodiment, the window 5 can be set as a rectangular window, and the length and width of the window 5 can be set according to user needs. For example, the length of the window 5 can be 10 cm and the width can be 20 cm. The position of the window 5 is fixed.

[0043] S2: Calculate the thickness wear of the tool according to the first image and the second image.

[0044] Reference Figure 3 The calculation method of thickness wear includes steps S20 to S22, which are as follows:

[0045] S20: extracting the projection of the edge of the tool blade in the first image as the first edge, extracting the projection of the edge of the tool blade in the second image as the second edge, and calculating the average value of the Euclidean distances between the pixel points in the first edge that are closest to the pixel points of the second edge.

[0046] Specifically, for each pixel point in the second edge , find the distance between the first edge The nearest pixel,the distance between pixels can be determined by calculating the Euclidean distance,between all pairs of points, that is: ,in, Represents pixel To Pixel Distance, pixels is a pixel on the second edge, pixel is a pixel on the first edge, Pixel The horizontal axis of Pixel The vertical coordinate of Pixel The horizontal axis of Pixel The vertical coordinate of .

[0047] For each Find the smallest Corresponding The average distance is obtained by summing up all the nearest distances and dividing by the number of pixels on the second edge. The average distance is the average of the Euclidean distances between the pixels on the first edge that are closest to each pixel on the second edge.

[0048] S21: Calculate the image entropy ratio of the image entropy of the pixel points on the first edge to the image entropy of the pixel points on the second edge.

[0049] Image entropy can reflect the grayscale distribution of pixels on the edge. When the blade wear is small, the edge of the blade projection should be clearer, and the number of grayscale levels of the edge of the tool projection is small. When the blade wear increases, the edge of the blade projection should be more blurred, and the number of grayscale levels of the edge of the tool projection is large.

[0050] In one embodiment, the calculation formula of the image entropy of the pixel point on the first edge is:

[0051] ,in, represents the image entropy of the pixel on the first edge, represents the gray level, Indicates grayscale The probability of appearing on the first edge; the image entropy of the pixel points on the second edge is calculated in the same way as the image entropy calculation method of the pixel points on the first edge.

[0052] When the blade wear is small, the edge of the blade projection is clearer, the number of gray levels is small, and the distribution of gray levels is more concentrated, so the image entropy is small. When the blade wear increases, the edge of the blade projection is more blurred, the number of gray levels is large, and the distribution is more dispersed, so the image entropy is larger.

[0053] However, in the above embodiment, if the grayscale of the pixels on the edge decreases uniformly (possibly due to uniform wear or external light interference), such as changing from a concentrated distribution of 230-255 to a concentrated distribution of 130-155, this is reflected in the image entropy, which is large and there is a defect.

[0054] The above image entropy calculation formula can add a grayscale value factor in the calculation of edge image entropy to overcome the above defects. In one embodiment, the image entropy calculation formula of the pixel point on the first edge can also be: ,in, represents the image entropy of the pixel on the first edge, represents the gray level, Indicates grayscale The probability of appearing on the first edge; the image entropy of the pixel points on the second edge is calculated in the same way as the image entropy calculation method of the pixel points on the first edge, Represents the accumulated value of the gray level present on the first edge.

[0055] S22: The thickness wear is obtained based on the ratio of the average Euclidean distance to the image entropy.

[0056] Specifically, the formula for thickness wear is: ,in, Indicates thickness wear, represents the hyperbolic tangent function; represents the average value of the Euclidean distances between the pixels on the first edge that are closest to the pixels on the second edge. represents the image entropy of the pixel on the first edge, Represents the image entropy of the pixels on the second edge. Indicates An exponential function with base .

[0057] For the factor , The purpose is to Normalization, The larger the value is, the greater the change in the position of the second edge compared to the first edge. The larger the value, the greater the thickness wear, and vice versa, the smaller the thickness wear. Other normalization functions that can be applied to the above increase and decrease relationship, such as sigmoid function, etc., can also be used, which will not be described in detail here.

[0058] for Factor, The closer it is to 1, the higher the similarity of the grayscale distribution between the first edge and the second edge. As the tool gradually wears, the image entropy of the projected edge continues to increase, so Greater than or equal to , The smaller, The larger the value, the greater the thickness wear, and vice versa, the smaller the thickness wear.

[0059] In other embodiments, the factor Alternative Calculate the thickness wear. The smaller it is, the higher the similarity of the grayscale distribution between the first edge and the second edge is. The larger the value, the greater the thickness wear, and vice versa, the smaller the thickness wear.

[0060] S3: Calculate the edge wear of the tool based on the torque of the tool axis.

[0061] Tool wear will lead to an increase in cutting force, because the contact area between the worn tool and the workpiece increases, and the friction also increases. Torque, as the embodiment of cutting force on the rotating axis, will also change with tool wear, specifically manifested as an increase in torque. This is because as the tool wears, the resistance of the tool in the cutting process increases, resulting in an increase in the required rotational torque. Therefore, the degree of tool wear can be reflected based on the torque of the tool axis.

[0062] In one embodiment, the edge wear calculation formula is:

[0063] ,in, express The edge wear of the moment, Indicates from time 1, time 2 to The torque sequence of the force on the tool axis at each moment, Indicates the minimum value of the force torque in the torque sequence, Indicates the maximum value of the applied torque in the torque series.

[0064] Calculate from the first moment to the The minimum and maximum values ​​of all torque values ​​at the moment are obtained by dividing the minimum value by the maximum value. The closer the ratio is to 1, the more stable the force on the tool during cutting and the lower the wear. If the ratio is far from 1, it means that the force on the tool fluctuates greatly during cutting and the wear is higher.

[0065] represents the median in the torque series, Indicates the number of torque values ​​between the median and the minimum value, Indicates the number of torque values ​​between the maximum value and the median. If most of the torque values ​​are concentrated on one side of the median, it means that the force on the tool during cutting is unstable, and the tool wear is high at this time; if the torque values ​​are evenly distributed on both sides of the median, then this ratio may be close to 1, indicating that the force on the tool during cutting is relatively stable, and the tool wear is low at this time.

[0066] This embodiment combines the ratio of the maximum and minimum torque of the tool and the distribution of the torque to comprehensively judge the stability of the force applied to the tool during operation, thereby quantifying the degree of wear of the tool. Considering multiple factors, the quantification result of the degree of wear is more accurate.

[0067] In one embodiment, the edge wear calculation formula is: ,in, express The edge wear of the moment, Indicates from time 1, time 2 to The torque sequence of the force on the tool axis at each moment, Indicates the minimum value of the force torque in the torque sequence, Indicates the maximum value of the force torque in the sequence. Calculate from the first moment to the The minimum and maximum values ​​of all torque values ​​at the moment are obtained by dividing the minimum value by the maximum value. The closer the ratio is to 1, the more stable the force on the tool during cutting and the lower the wear. If the ratio is far from 1, it means that the force on the tool during cutting fluctuates greatly and the wear is higher. This embodiment is suitable for application scenarios with low computational requirements and low precision requirements.

[0068] In the above process, in order to reduce abnormal vibration caused by non-wear reasons (such as the fixture is not clamped tightly when holding the workpiece, the tool is loosely installed, etc.), the data needs to be denoised. The specific denoising process is: obtain the initial torque sequence and initial amplitude sequence of the tool; calculate the synchronization of the initial torque sequence and the amplitude sequence.

[0069] The calculation method of synchronization is: supplement the shorter sequence of the torque sequence and the amplitude sequence to make the torque sequence and the amplitude sequence equal in length; and use the Pearson correlation coefficient of the torque sequence and the amplitude sequence as synchronization.

[0070] Specifically, if one sequence is shorter than the other, the shorter sequence can be supplemented by filling in missing values. The filling method can be: calculate the mean or median of the existing data and use the mean or median to fill in the missing part.

[0071] The Pearson correlation coefficient is a statistical indicator used to measure the strength of the linear correlation between two variables. Between 0 and 1, where 1 indicates a perfect positive correlation, that is, when one sequence increases, the other sequence also increases accordingly. Indicates a perfect negative correlation, that is, when one series increases, the other decreases.

[0072] When the synchronization is less than or equal to the synchronization threshold, the corresponding force torque and vibration amplitude are deleted to obtain the torque sequence and amplitude sequence. For example, the synchronization threshold can be set to 0.4. When the synchronization of the torque sequence and the amplitude sequence is greater than 0.4, it means that the synchronization of the torque sequence and the amplitude sequence is strong, that is, synchronous rise, synchronous fall or synchronous gentle fluctuation. On the contrary, the synchronization is poor. Deleting this part of the data can improve the accuracy of the calculation.

[0073] If the noise continues, an alarm signal is generated to prompt relevant personnel to check and adjust. The method for judging whether the noise continues is to set the noise threshold and time window. In the time window, if the vibration amplitude of the set ratio exceeds the noise threshold, the noise is considered to be continuing. The time window and noise threshold can be set by the user according to the actual application scenario. For example, the time window can be set to 5 minutes, the set ratio can be 20%, and the noise threshold can be set to twice the average vibration amplitude in the previous time window.

[0074] S4: Construct a data set based on the thickness wear and edge wear in the historical data, label the data set with the standard cutting speed, train the neural network model based on the labeled data set to obtain a matching model; input the thickness wear and edge wear of the real-time data into the matching model, and output the corresponding standard cutting speed.

[0075] The tool is subjected to a wear test in advance. During the test, different cutting speeds are adopted for different degrees of wear to find the optimal cutting speed corresponding to different degrees of wear. For example, at the optimal cutting speed, the wear change of the tool is small while ensuring that the workpiece is processed to be qualified.

[0076] The thickness wear and edge wear are taken as a two-dimensional data group, and each two-dimensional data group is labeled, and the label is the standard cutting speed corresponding to the two-dimensional data group.

[0077] The neural network model may be a BP (Back Propagation) neural network model, and the loss function is a mean square error loss function. When the model training reaches a preset training stop condition (the number of training times is 1000 or the loss function value is 0.01), the training is stopped to obtain a matching model.

[0078] It should be noted that the present application is applicable to a workpiece processed by a fixed tool, and is not applicable to application scenarios where the tool or the workpiece is frequently changed, and the tool is a rotary cutting tool.

[0079] An embodiment of the present application also discloses a CNC machine tool tool wear identification system, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a CNC machine tool tool wear identification method according to the present application is implemented.

[0080] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0081] In the present application, the aforementioned memory may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high bandwidth memory HBM (High Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module, or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0082] Although this specification has shown and described a plurality of embodiments of the present application, it is obvious to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will think of many changes, modifications and alternatives without departing from the thought and spirit of the present application. It should be understood that in the process of practicing the present application, various alternatives to the embodiments of the present application described herein may be adopted.

[0083] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A method for identifying tool wear of a CNC machine tool, characterized in that: Includes steps: The image projected by the tool before processing is taken as the first image, and the image projected by the tool after processing is taken as the second image, wherein the light source in the first image and the second image is the same, the position and placement angle of the tool are the same, and the light source is tilted; The thickness wear of the tool is calculated according to the first image and the second image; the thickness wear is calculated by: extracting the projection of the edge of the tool blade in the first image as the first edge, extracting the projection of the edge of the tool blade in the second image as the second edge, calculating the average value of the Euclidean distance between the pixels closest to each pixel of the second edge in the first edge; calculating the image entropy ratio of the image entropy of the pixel on the first edge to the image entropy of the pixel on the second edge; and obtaining the thickness wear according to the average value of the Euclidean distance and the image entropy ratio; The edge wear of the tool is calculated based on the torque of the tool axis; a data set is constructed based on the thickness wear and edge wear in the historical data, and the data set is labeled with the standard cutting speed. A neural network model is trained based on the labeled data set to obtain a matching model; the thickness wear and edge wear of the real-time data are input into the matching model, and the corresponding standard cutting speed is output.

2. A method for identifying tool wear of a CNC machine tool according to claim 1, characterized in that: The calculation formula for the image entropy of the pixel on the first edge is: ,in, represents the image entropy of the pixel on the first edge, represents the gray level, Indicates grayscale The probability of appearing on the first edge; the image entropy of the pixel points on the second edge is calculated in the same way as the image entropy calculation method of the pixel points on the first edge.

3. The method for identifying tool wear of a CNC machine tool according to claim 1, characterized in that: The calculation formula for the image entropy of the pixel on the first edge is: ,in, represents the image entropy of the pixel on the first edge, represents the gray level, Indicates grayscale The probability of appearing on the first edge; the image entropy of the pixel points on the second edge is calculated in the same way as the image entropy calculation method of the pixel points on the first edge, Represents the accumulated value of the gray level present on the first edge.

4. A method for identifying tool wear of a CNC machine tool according to any one of claims 1 to 3, characterized in that: Calculate the edge wear of the tool using the following formula: ,in, express The edge wear of the moment, Indicates from time 1, time 2 to The torque sequence of the force on the tool axis at each moment, Indicates the minimum value of the force torque in the torque sequence, Indicates the maximum value of the force torque in the sequence.

5. A method for identifying tool wear of a CNC machine tool according to any one of claims 1 to 3, characterized in that: Calculate the edge wear of the tool using the following formula: ,in, express The edge wear of the moment, Indicates from time 1, time 2 to The torque sequence of the force on the tool axis at each moment, Indicates the minimum value of the force torque in the torque sequence, Indicates the maximum value of the force torque in the torque sequence, represents the median in the torque series, Indicates the number of torque values ​​between the median and the minimum value, Indicates the number of torque values ​​between the maximum value and the median value.

6. A method for identifying tool wear of a CNC machine tool according to claim 5, characterized in that: The method also includes the steps of: denoising the collected initial torque sequence to obtain a torque sequence, and the denoising step includes: obtaining the initial torque sequence and initial amplitude sequence of the tool; calculating the synchronization of the initial torque sequence and the amplitude sequence, and when the synchronization is less than or equal to the synchronization threshold, deleting the corresponding force torque and vibration amplitude to obtain the torque sequence and amplitude sequence.

7. A method for identifying tool wear of a CNC machine tool according to claim 6, characterized in that: The synchronization is calculated by supplementing the shorter sequence of the torque sequence and the amplitude sequence to make the torque sequence and the amplitude sequence equal in length; The Pearson correlation coefficient between the torque series and the amplitude series was taken as the synchronicity.

8. A CNC machine tool tool wear identification system, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for identifying tool wear of a CNC machine tool according to any one of claims 1 to 7 is implemented.

Images