Machine learning-based intelligent cutting tool life cycle management method and system

Abstract

The present invention relates to the technical field of intelligent cutting tool management, and relates to a machine learning-based intelligent cutting tool life cycle management method and system. A specific method comprises: establishing a database of quartz machining cutting tools; after operation in a factory is finished every day, recovering and recording cutting tool use and circulation information of the quartz machining cutting tools; establishing a machine learning model, to evaluate and calculate daily wear amounts of the cutting tools, and a daily cutting tool use influence coefficient; evaluating remaining service lives of quartz machining cutting tools having different abnormality grades; arranging remaining service life data of quartz machining cutting tools of a same machining category in order, to form a cutting tool life sequence, and updating the cutting tool life sequence to the database of quartz machining cutting tools in real time; dynamically formulating a factory cutting tool purchase and maintenance plan. The present invention resolves the problems in the prior art of lack of a unified data management platform for quartz machining cutting tools, chaotic tool management at a production site, and inaccurate calculation of lifespans of the cutting tools.

Description

A Machine Learning-Based Intelligent Tool Lifecycle Management Method and System Technical Field

[0001] This invention relates to the field of intelligent tool management technology, and is a tool lifecycle intelligent management method and system based on machine learning. Background Technology

[0002] Quartz machining tools are typically diamond tools, mainly diamond grinding heads or diamond grinding wheels. When machining quartz components, it is necessary to select appropriate diamond tools based on the machining requirements and install them on the machining equipment. Machining parameters such as rotational speed and feed rate must be adjusted. During machining, the cutting effect and machining quality must be ensured by controlling machining parameters and tool pressure. Currently, there is a lack of a unified data management platform for tool parameters, resulting in low tool selection rates among process engineers. Tool usage data is not correlated with the machining process, and tool management departments cannot objectively evaluate the lifespan of tools supplied by vendors. Furthermore, arbitrary tool use is common on the production floor, leading to a large backlog of tools and difficulty in finding replacement tools when needed. Improving tool management and maintenance efficiency is a significant challenge for quartz machining plants.

[0003] Among the existing publicly disclosed inventions, such as Chinese Patent Publication No. CN115421455A, a production management system, supply chain management method, and electronic equipment for photomasks are disclosed. The system includes: a material feeding management unit that compares quartz substrate parameter information with substrate standard values ​​to obtain a first detection result; a photomask management unit, including a thin film detection module corresponding to each thin film layer, which obtains thin film parameter information and compares it with thin film standard parameter values ​​to obtain a second detection result; a photomask management unit that obtains etching process parameters and compares them with etching process standard values ​​to obtain a third detection result; and a decision-making unit that determines whether to shut down for maintenance based on the first, second, and third detection results.

[0004] The aforementioned patent relies on operators inputting parameter information and making comparisons. If the operators make mistakes or are negligent, it may lead to inaccurate test results. Summary of the Invention

[0005] The technical problem to be solved by this invention is that in the prior art, there is a lack of a unified data management platform for quartz machining tools, chaotic tool management on the production site, and inaccurate calculation of the life of each tool. This invention proposes a tool life cycle intelligent management method and system based on machine learning.

[0006] To achieve the above objectives, this invention discloses a machine learning-based intelligent tool lifecycle management method, comprising the following steps:

[0007] S1: Classify the quartz machining tools based on the factory's historical purchase information and establish a quartz machining tool database;

[0008] Specifically, step S1 further includes collecting the initial state of the quartz machining tool that is entered into the database for the first time. The initial state of the tool includes: tool parameter information, specifications, model, manufacturer, initial high-resolution image data of the tool surface, and the factory wear amount R0 obtained by comparing the initial high-resolution image of the tool surface with the factory surface image of the tool.

[0009] S2: Before the start of each day's factory operations, the quartz machining tools required for the pre-processed quartz devices of that day are distributed and issued, and after the end of each day's factory operations, the tool usage and circulation information of the quartz machining tools are collected and recorded.

[0010] Specifically, in step S2, the tool usage and circulation information of the quartz machining tool includes: the ambient temperature and humidity of the tool usage environment on the day, the number of quartz devices processed, the cumulative processing time on the day, and the high-resolution image data of the tool surface on the day.

[0011] S3: Establish a machine learning model to extract features of abnormal tool wear from the tool usage and circulation information, evaluate and calculate the daily wear amount and daily tool usage impact coefficient of the tool, and perform recycling anomaly classification and marking processing on the daily recycled quartz machining tools.

[0012] Specifically, step S3 includes the following steps:

[0013] S311: Establish the machine learning model, wherein the neural network of the machine learning model includes: an input layer, four convolutional layers, four sampling layers, a fully connected layer, and an output layer, wherein each of the convolutional layers contains 10 convolutional kernels of size 8×8;

[0014] S312: Perform edge detection and mean filtering noise reduction preprocessing on the high-resolution image data of the tool surface obtained in step S2;

[0015] S313: Convert the preprocessed high-resolution image of the tool surface to grayscale, and input the grayscale high-resolution image of the tool surface into the machine learning model to perform grayscale threshold segmentation on the defect region. The formula for performing the grayscale threshold segmentation in the convolutional layer is:

[0016] g i,n (x n,y n )∈[g i-1 ,g i ], i∈{1,2...4};

[0017] Among them, g i,n (x n ,y n (x) is the nth coordinate on the edge line of the i-th target region. n ,y n The gradient value of the pixel;

[0018] g i The maximum value among all the gradient values ​​of all pixels in the i-th target region;

[0019] g i-1 It is the maximum value among all the gradient values ​​of all pixels in the (i-1)th target region;

[0020] x n The x-coordinate of the nth pixel in the two-dimensional plane coordinate system of the high-resolution image of the tool surface on that day;

[0021] y n The ordinate of the nth pixel in the two-dimensional plane coordinate system of the high-resolution image of the tool surface on that day;

[0022] [g i-1 ,g i ] represents the gradient value threshold range of the pixel in the i-th target region; n is the subscript representing the pixel number; i is the subscript representing the target region number; the criterion for dividing the target region is: when the gradient value of a pixel in the high-resolution image of the tool surface on the same day belongs to the threshold range, then the pixel belongs to the i-th target region;

[0023] S314: Separate the tool environment background from the tool abnormal wear area, perform region identification and manual annotation of the tool abnormal wear area on the high-resolution image of the tool surface on the same day, extract the feature information of the tool abnormal wear area, and evaluate and calculate the tool wear amount on the same day.

[0024] Specifically, in step S314, the strategy for evaluating the daily wear of the cutting tool is as follows:

[0025] Where M is the total number of abnormal wear areas of the tool; m is a subscript indicating the m-th abnormal wear area of ​​the tool; m≤M;

[0026] R x In the subscript, x represents the cumulative usage time of the cutting tool after it was purchased and entered the quartz processing plant, which is x days; R xThis indicates the daily wear amount of the cutting tool on the xth cumulative day of use;

[0027] H m The total number of pixels in the m-th abnormal wear region of the tool;

[0028] Let be the average pixel value of all pixels in the m regions of abnormal tool wear.

[0029] Specifically, in step S3, the calculation strategy for the daily tool usage impact coefficient is as follows:

[0030] S321: Collect the characteristic information of the abnormal wear area of ​​the tool and the curve of the daily wear amount of the tool as a function of the tool during the use of the tool on that day; the wear factor variables include: cutting speed and cutting pressure;

[0031] S322: Set up a control group by using multiple tools with a similarity of more than 95% in the size of the quartz devices cut by the tools. Extract the daily wear amount of each abnormal wear area of ​​the tools in the control group as a function of the wear factor variable. Perform curve separation processing on the change curve according to the wear factor variable level.

[0032] S323: Calculate the daily tool usage impact coefficients for different wear factor variables.

[0033] When the wear factor variable is the cutting speed, the calculation strategy is as follows:

[0034] Extract the curves showing the variation of abnormal tool wear with cutting speed under the same cutting pressure in the control group. There are a total of G curves. Simultaneously extract the cutting speed dataset {v1, v2...v...} g ...v G} and the daily wear data of the tool in the feature information of the abnormal wear area of ​​the tool {b 1 ,b 2 ...b g ...b G}, where v g b is the cutting speed in the g-th variation curve; g Let g be the daily wear amount of the tool in the g-th variation curve; g∈{1,2,...,G};

[0035] Calculate the daily tool usage influence coefficient S1 when the wear factor variable is the cutting speed:

[0036] in, The average cutting speed is found in the G variation curves. The average daily wear of the cutting tool in the G variation curves;

[0037] When the wear factor variable is the cutting pressure, the specific calculation strategy is as follows:

[0038] J curves representing the same cutting speed were extracted from the control group. Simultaneously, the cutting pressure dataset {p1, p2...p} was extracted. j ...p J} and the daily wear data of the tool in the feature information of the abnormal wear area of ​​the tool {c 1 ,c 2 ...c j ...c J}, where p j c represents the cutting pressure in the j-th variation curve. j Let J represent the daily wear of the tool in the j-th variation curve; j∈{1,2,...,J};

[0039] Calculate the daily tool usage influence coefficient S2 when the wear factor variable is the cutting pressure:

[0040] in, The average cutting speed is found in J variation curves. Let J be the average daily wear of the cutting tools in the J variation curves.

[0041] S4: Evaluate the remaining service life of the quartz machining tools for different anomaly levels;

[0042] Specifically, in step S4, the assessment of the remaining service life of the quartz machining tools with different anomaly levels includes:

[0043] S41: Extract the cutting speed and cutting pressure data required for the processing of the quartz device in the next working day at the quartz processing plant.

[0044] S42: Predict and calculate the expected wear of the tool in the next working day;

[0045] Where R0 is the tool's factory wear amount;

[0046] R x+1 The predicted wear amount of the tool for the next working day;

[0047] v x+1 The cutting speed during the tool's use in the next working day;

[0048] px+1 The cutting pressure during tool use in the next working day;

[0049] S43: Evaluate the remaining service life of the quartz machining tools for different anomaly levels;

[0050] On the (x+1)th working day, the remaining service life T of the quartz machining tool x+1 for:

[0051] S5: Arrange the remaining service life data of the quartz machining tools of the same processing category in an orderly manner to form a tool life sequence, and update the tool life sequence to the quartz machining tool database in real time;

[0052] Specifically, S51: The remaining service life data of the quartz machining tools of the same processing category are arranged in an orderly manner to form the tool life sequence, and the tool life sequence is updated to the quartz machining tool database in real time;

[0053] S52: Provide distribution and requisition suggestions for the subsequent distribution and requisition of quartz machining tools based on the tool life sequence, wherein the distribution and requisition suggestions include: prioritizing the requisition of the quartz machining tool with the longest remaining service life among the quartz machining tools of the same machining category;

[0054] S6: Statistically analyze the remaining service life of all quartz machining tools in the factory, and dynamically formulate factory tool purchase and maintenance plans.

[0055] Specifically, in step S6, the factory tool purchase and maintenance plan includes:

[0056] When T x+1 ≥T w In such cases, contact the tool manufacturer to have the tool maintained and serviced in advance based on the type of abnormal wear area.

[0057] when At that time, the tool management department checks the corresponding tool inventory and supplier information to determine whether it is necessary to replace or procure the tool of that model.

[0058] Among them, T w and T c T represents the maintenance threshold and replacement threshold corresponding to the abnormal wear area of ​​the cutting tool. x The remaining service life of the quartz machining tool on the xth working day.

[0059] In addition, this invention discloses a machine learning-based intelligent tool lifecycle management system, which is implemented based on the aforementioned machine learning-based intelligent tool lifecycle management method and includes the following modules:

[0060] The tool initial information acquisition module is used to classify tools based on the information of the quartz machining tools purchased by the factory in the past and to establish the quartz machining tool database;

[0061] The tool circulation information recording module is used to distribute and issue the quartz machining tools needed for pre-processing quartz devices before the start of each day's factory operation, and to collect and record the tool usage and circulation information of the quartz machining tools after the end of each day's factory operation.

[0062] The wear feature analysis module is used to establish the machine learning model, extract features of abnormal tool wear from the tool usage and circulation information, evaluate the daily wear amount of the calculated tool and the daily tool usage impact coefficient, and perform recycling anomaly classification and marking processing on the daily recycled quartz machining tools.

[0063] The remaining service life assessment module is used to assess the remaining service life of the quartz machining tools for different anomaly levels.

[0064] The lifespan sequence update module arranges the remaining lifespan data of the quartz machining tools of the same processing category in an orderly manner to form the tool lifespan sequence, and updates the tool lifespan sequence to the quartz machining tool database in real time.

[0065] The purchase and maintenance plan output module is used to statistically analyze the remaining service life of all types of quartz machining tools in the factory and dynamically formulate the factory's tool purchase and maintenance plan.

[0066] The present invention also discloses a storage medium storing instructions, which, when read by a computer, cause the computer to execute the aforementioned machine learning-based intelligent tool lifecycle management method.

[0067] The present invention also discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the above-described machine learning-based intelligent tool lifecycle management method.

[0068] Compared with the prior art, the technical effects of the present invention are as follows:

[0069] 1. This invention manages and monitors quartz machining tools, enabling timely distribution and recycling of tools, thus avoiding production line downtime due to insufficient or damaged tools and improving production efficiency.

[0070] 2. This invention uses a machine learning model to evaluate abnormal tool wear, which can detect and handle abnormalities, extend tool life, reduce tool replacement frequency, and lower production costs.

[0071] 3. This invention records and updates the usage of cutting tools to the database in real time, allowing for the viewing and analysis of tool conditions at any time, and helping managers to formulate timely tool purchase and maintenance plans. Attached Figure Description

[0072] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0073] Figure 1 is a flowchart illustrating a machine learning-based intelligent tool lifecycle management method according to the present invention.

[0074] Figure 2 is a schematic diagram of the structure of a machine learning-based intelligent management system for the life cycle of cutting tools according to the present invention. Detailed Implementation

[0075] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0076] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0077] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0078] Example 1

[0079] As shown in Figure 1, an embodiment of the present invention provides a machine learning-based intelligent tool lifecycle management method, which includes the following specific steps:

[0080] S1: Classify the quartz machining tools based on the factory's historical purchase information and establish a quartz machining tool database.

[0081] S1 also includes the acquisition of the initial state of quartz machining tools that are entered into the database for the first time, including: tool parameter information, specifications, model, manufacturer, initial high-resolution image data of the tool surface, and the factory wear amount R0 obtained by comparing the initial high-resolution image of the tool surface with the factory surface image of the tool.

[0082] S2: Before the start of each day's factory operations, distribute and issue the quartz machining tools needed for the pre-processed quartz components of that day, and collect and record the tool usage and circulation information of the quartz machining tools after the end of each day's factory operations.

[0083] In S2, the tool usage and circulation information of the quartz machining tool includes: the ambient temperature and humidity of the tool usage environment on the day, the number of quartz devices processed, the cumulative processing time on the day, and the high-resolution image data of the tool surface on the day.

[0084] S3: Establish a machine learning model to extract features of abnormal tool wear from tool usage and circulation information, evaluate and calculate the daily wear amount and daily tool usage impact coefficient, and perform recycling anomaly classification and marking processing on the daily recycled quartz machining tools.

[0085] S311: Establish a machine learning model, wherein the neural network of the machine learning model includes: an input layer, four convolutional layers, four sampling layers, a fully connected layer, and an output layer, wherein each convolutional layer contains 10 convolutional kernels of size 8×8.

[0086] S312: Perform edge detection and mean filtering denoising preprocessing on the high-resolution image data of the tool surface obtained in S2 for the same day.

[0087] S313: Convert the preprocessed high-resolution image of the tool surface to grayscale, and input the grayscale high-resolution image of the tool surface into the machine learning model to perform grayscale thresholding on the defect area. The formula for grayscale thresholding in the convolutional layer is as follows:

[0088] g i,n (x n ,y n )∈[g i-1 ,g i ], i∈{1,2...4};

[0089] Among them, g i,n (x n ,y n (x) is the nth coordinate on the edge line of the i-th target region.n ,y n The gradient value of each pixel.

[0090] g i It is the maximum gradient value among all pixels in the i-th target region.

[0091] g i-1 It is the maximum gradient value among all pixels in the (i-1)th target region.

[0092] x n The x-coordinate of the nth pixel in the two-dimensional plane coordinate system of the high-resolution image of the tool surface on that day.

[0093] y n The ordinate of the nth pixel in the two-dimensional plane coordinate system of the high-resolution image of the tool surface on that day is given.

[0094] [g i-1 ,g i ] represents the gradient value threshold range of a pixel in the i-th target region; n is a subscript representing the pixel number; i is a subscript representing the target region number; the criterion for dividing the target region is: when the gradient value of a pixel in the high-resolution image of the tool surface on the same day belongs to the threshold range, then the pixel belongs to the i-th target region.

[0095] S314: Separate the tool environment background from the tool abnormal wear area, identify and manually annotate the abnormal wear area in the high-resolution image of the tool surface on the same day, extract the feature information of the abnormal wear area, and evaluate and calculate the tool wear amount on the same day.

[0096] In S314, the specific strategy for evaluating the daily wear of the cutting tool is as follows:

[0097] Where M is the total number of abnormal tool wear areas; m is a subscript indicating the m-th abnormal tool wear area; m≤M;

[0098] R x In the subscript, x represents the cumulative usage time of the cutting tool after it was purchased and entered the quartz processing plant, which is x days; R x This indicates the daily wear amount of the cutting tool on the xth cumulative day of use;

[0099] H m The total number of pixels in the m-th abnormal wear region;

[0100] Let be the average pixel value of all pixels in the m regions of abnormal tool wear.

[0101] Let be the average pixel value of all pixels in the I regions of abnormal tool wear.

[0102] In S3, the calculation strategy for the daily tool usage impact coefficient is as follows:

[0103] S321: Collect characteristic information of the abnormal wear area of ​​the tool and the curve of the daily wear amount of the tool with the wear factor variables during the tool's use on that day; the wear factor variables include: cutting speed and cutting pressure.

[0104] S322: Set up a control group for the use of multiple cutting tools with a similarity of more than 95% in the size of the quartz devices cut by the cutting tools. Extract the daily wear amount of each abnormal wear area of ​​the cutting tools in the control group as a function of wear factor variables. Perform curve separation processing on the change curves according to the wear factor variable level.

[0105] S323: Calculate the daily tool usage impact coefficients for different wear factor variables.

[0106] When the wear factor variable is cutting speed, the calculation strategy is as follows:

[0107] Extract the curves of abnormal tool wear as a function of cutting speed under the same cutting pressure in the control group. There are a total of G curves. Simultaneously extract the cutting speed dataset {v1, v2...v...} g ...v G The daily wear data of the tool in the feature information of the abnormal wear area of ​​the tool. 1 ,b 2 ...b g ...b G}, where v g b is the cutting speed in the g-th variation curve; g Let g represent the daily wear of the tool in the g-th variation curve; g∈{1,2,...,G}.

[0108] Calculate the daily tool usage influence coefficient S1 when the wear factor variable is cutting speed:

[0109] in, The average cutting speed is found in the G variation curves. Let G be the average daily wear of the cutting tools among the G variation curves.

[0110] When the wear factor variable is cutting pressure, the specific calculation strategy is as follows:

[0111] Extract the cutting speed variation curves from the control group, resulting in J variation curves. Simultaneously extract the cutting pressure dataset {p1, p2...p...}. j ...pJ The daily wear data of the tool in the feature information of the abnormal wear area of ​​the tool. 1 ,c 2 ...c j ...c J}, where p j c represents the cutting pressure in the j-th variation curve. j Let J represent the daily wear of the tool in the j-th variation curve; j∈{1,2,...,J}.

[0112] Calculate the daily tool usage influence coefficient S2 when the wear factor variable is cutting pressure:

[0113] in, The average cutting speed is found in J variation curves. Let J be the average daily wear of the cutting tools in the J variation curves.

[0114] S4: Assess the remaining service life of quartz machining tools with different anomaly levels.

[0115] In S4, the assessment of the remaining service life of quartz machining tools with different anomaly levels includes:

[0116] S41: Extract the cutting speed and cutting pressure data required for the processing of the quartz device in the next working day at the quartz processing plant.

[0117] S42: Predicts the expected wear of the tool during the next working day.

[0118] Where R0 is the tool's wear amount at the time of manufacture.

[0119] R x+1 This is the predicted wear amount of the tool for the next working day.

[0120] v x+1 This refers to the cutting speed during the tool's use in the next working day.

[0121] p x+1 This refers to the cutting pressure during the tool's use in the next working day.

[0122] S43: Assess the remaining service life of quartz machining tools with different anomaly levels.

[0123] On the (x+1)th working day, the remaining service life T of the quartz machining tool x+1 for:

[0124] S5: Arrange the remaining service life data of quartz machining tools of the same machining category in an orderly manner to form a tool life sequence, and update the tool life sequence to the quartz machining tool database in real time.

[0125] For example, in this embodiment, the processing categories of quartz processing tools include: cutting tools, grinding tools, and polishing tools for different quartz device sizes and different quartz device surface hardness.

[0126] S5 includes the following specific steps:

[0127] S51: Arrange the remaining service life data of quartz machining tools of the same machining category in an orderly manner to form a tool life sequence, and update the tool life sequence to the quartz machining tool database in real time.

[0128] S52: Based on the quartz machining tool database and the tool life sequence, provide distribution and requisition suggestions for subsequent quartz machining tool distribution and requisition processing, wherein the distribution and requisition suggestions include: prioritizing the requisition of the quartz machining tools with the longest remaining service life among the quartz machining tools of the same machining category.

[0129] S6: Statistically analyze the remaining service life of all quartz machining tools in the factory, and dynamically formulate factory tool purchase and maintenance plans.

[0130] In S6, the factory tool purchase and maintenance plan includes:

[0131] When T x+1 ≥T w In such cases, contact the tool manufacturer to have the tools maintained and serviced in advance based on the type of abnormal wear area.

[0132] when At that time, the tool management department checks the corresponding tool inventory and supplier information to determine whether it is necessary to replace or procure the tool of that model.

[0133] Among them, T w and T c T represents the maintenance and replacement thresholds corresponding to areas of abnormal tool wear. x The remaining service life of the quartz machining tool on the xth working day.

[0134] Example 2

[0135] As shown in Figure 2, an embodiment of the present invention provides a machine learning-based intelligent tool lifecycle management system. The system is implemented based on the machine learning-based intelligent tool lifecycle management method and includes the following modules:

[0136] The module includes: initial tool information acquisition module, tool usage flow information recording module, wear characteristic analysis module, remaining service life assessment module, service life sequence update module, and purchase and maintenance plan output module.

[0137] The tool initial information acquisition module classifies tools based on the information of quartz machining tools purchased by the factory in the past, establishes a quartz machining tool database, and collects the initial tool status of quartz machining tools that are entered into the database for the first time.

[0138] The tool usage flow information recording module is used to distribute and issue quartz machining tools needed for pre-processing quartz devices before the start of each day's factory operations, and to collect and record the tool usage and flow information of the quartz machining tools after the end of each day's factory operations.

[0139] The wear feature analysis module is used to establish a machine learning model, extract features of abnormal tool wear from tool usage and circulation information, evaluate and calculate the daily wear amount and daily tool usage impact coefficient, and perform recycling anomaly classification and marking processing on the daily recycled quartz machining tools.

[0140] The remaining service life assessment module is used to assess the remaining service life of quartz machining tools with different anomaly levels.

[0141] The lifespan sequence update module arranges the remaining lifespan data of quartz machining tools of the same machining category into a tool lifespan sequence, and updates the tool lifespan sequence to the quartz machining tool database in real time.

[0142] The purchase and maintenance plan output module is used to statistically analyze the remaining service life of all types of quartz machining tools in the factory and dynamically formulate the factory's tool purchase and maintenance plan.

[0143] Example 3

[0144] This embodiment provides an electronic device, including: a processor and a memory, wherein the memory stores a computer program that can be called by the processor;

[0145] The processor executes the aforementioned machine learning-based intelligent tool lifecycle management method by calling computer programs stored in memory.

[0146] The electronic device can vary considerably depending on its configuration or performance. It may include one or more Central Processing Units (CPUs) and one or more memories, wherein the memory stores at least one computer program, which is loaded and executed by the processor to implement the machine learning-based intelligent tool lifecycle management method provided in the above-described embodiment. The electronic device may also include other components for implementing its functions; for example, it may have wired or wireless network interfaces and input / output interfaces for data input and output. Further details are omitted in this embodiment.

[0147] Example 4

[0148] This embodiment proposes a computer-readable storage medium on which an erasable and rewritable computer program is stored.

[0149] When a computer program runs on a computer device, it causes the computer device to execute the aforementioned machine learning-based intelligent tool lifecycle management method.

[0150] For example, computer-readable storage media can be read-only memory (ROM), random access memory (RAM), compact disc read-only memory (CD-ROM), magnetic tape, floppy disk, and optical data storage devices.

[0151] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0152] It should be understood that determining B based on A does not mean determining B solely based on A; it also means determining B based on A and / or other information.

[0153] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the flow or function according to the embodiments of the present invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via a wired network and / or wireless network. A computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0154] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0155] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0156] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only one method, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0157] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0158] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0159] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0160] In summary, compared with the prior art, the technical effects of the present invention are as follows:

[0161] 1. This invention manages and monitors quartz machining tools, enabling timely distribution and recycling of tools, thus avoiding production line downtime due to insufficient or damaged tools and improving production efficiency.

[0162] 2. This invention uses a machine learning model to evaluate abnormal tool wear, which can detect and handle abnormalities, extend tool life, reduce tool replacement frequency, and lower production costs.

[0163] 3. This invention records and updates the usage of cutting tools to the database in real time, allowing for the viewing and analysis of tool conditions at any time, and helping managers to formulate timely tool purchase and maintenance plans.

[0164] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A machine learning based intelligent management of tool life cycle method, characterized by: The method comprises the following steps: S1: classifying tools according to quartz processing tool information purchased historically by the factory, and establishing a quartz processing tool database; S2: before the factory starts operation each day, processing tool distribution and use of the quartz processing tools required for pre-processing quartz devices on the day, and after the factory ends operation each day, recycling and recording tool use and turnover information of the quartz processing tools; S3: establishing a machine learning model, extracting features of tool abnormal wear from the tool use and turnover information, evaluating and calculating daily wear of the tool and a daily tool use influence coefficient, and processing abnormal classification and marking of the quartz processing tools recycled each day; S4: evaluating the remaining service life of the quartz processing tools of different abnormal levels; S5: sequentially arranging the remaining service life data of the quartz processing tools of the same processing category to form a tool life sequence, and updating the tool life sequence to the quartz processing tool database in real time; S6: statistically analyzing the remaining service life of all the quartz processing tools in the factory, and dynamically formulating a factory tool purchase and maintenance plan.

2. The machine learning based intelligent management of tool life cycle method according to claim 1, wherein, The step S1 further comprises: collecting an initial state of the quartz processing tools first entered into the database; the tool initial state collection comprises: collecting parameter information, specifications, models, manufacturers of the tools, tool surface initial high-resolution image data, and factory wear R0 obtained by comparing and evaluating the tool surface initial high-resolution image with a factory surface image of the tool; In the step S2, the tool use and turnover information of the quartz processing tools comprises: daily tool use environment temperature, environment humidity, number of processed quartz devices, daily cumulative processing use time length, and tool surface daily high-resolution image data.

3. The machine learning based intelligent tool life cycle management method according to claim 2, wherein, The step S3 comprises the following specific steps: S311: establishing the machine learning model, a neural network of the machine learning model comprising: an input layer, 4 convolution layers, 4 sampling layers, a full connection layer, and an output layer, wherein each convolution layer contains 10 convolution kernels with a size of 8x8; S312: performing edge detection and mean filter denoising preprocessing on the tool surface daily high-resolution image data obtained in the step S2; S313: performing grayscale processing on the preprocessed tool surface daily high-resolution image, inputting the grayscale tool surface daily high-resolution image into the machine learning model, performing grayscale threshold segmentation processing on the defect area, and performing the grayscale threshold segmentation processing in the convolution layer according to the formula: g i,n (x n ,y n )∈[g i-1 ,g i ],i∈{1,2...4}; wherein g i,n (x n ,y n ) is the gradient value of the pixel point with the coordinate of (x n ,y n ) on the i-th target region edge line; g i the maximum value among the gradient values of all the pixel points of the i-th target region; g i-1 the maximum value among the gradient values of all the pixel points of the i-1th target region; x n is the horizontal coordinate of the nth pixel point in the two-dimensional plane coordinate system of the high-resolution image of the tool surface on the day; y n is the vertical coordinate of the n-th pixel point in the two-dimensional plane coordinate system of the high-resolution image of the tool surface on the day [g i-1 ,g i ] is a threshold interval of the gradient value of the pixel point in the ith target region; n is a subscript, indicating the number of the pixel point; i is a subscript, indicating the number of the target region; the division criterion of the target region is that when the gradient value of a certain pixel point in the high-resolution image of the tool surface on the day belongs to the threshold interval, the pixel point belongs to the ith target region. S314: separating tool environment background and tool abnormal wear area, performing area identification and manual labeling of the tool abnormal wear area on the tool surface daily high-resolution image, extracting feature information of the tool abnormal wear area, and evaluating and calculating the daily wear of the tool.

4. The machine learning based intelligent tool life cycle management method according to claim 3, wherein, In the step S314, the evaluation strategy of the tool's wear amount of the day is specifically as follows: Wherein, M is the total number of the tool abnormal wear areas; m is a subscript, representing the mth tool abnormal wear area; m≤M; R x The subscript x indicates the cumulative use day length of the tool after being purchased into the quartz processing factory is x days; R x represents the tool wear amount on the xth cumulative use day. H m is the total number of pixel points of the mth abnormal wear area of the tool. is a pixel mean value of all pixel points in the m tool abnormal wear areas.

5. The machine learning based intelligent tool life cycle management method according to claim 4, wherein, In the step S3, the calculation strategy of the tool use influence coefficient in the day is as follows: S321: Collect the feature information of the abnormal wear area of the tool and the change curve of the tool wear amount in the day with the wear factor variable data during the tool use in the day; The wear factor variable includes cutting speed and cutting pressure; S322: Set multiple tool use processes of tools with a size similarity of quartz devices cut by the tools of more than 95% as a control group, extract the change curve of the tool wear amount in the day of each abnormal wear area of the tool with the wear factor variable in the control group, and perform curve separation processing on the change curve according to the wear factor variable level; S323: Calculate the tool use influence coefficient in the day corresponding to different wear factor variables respectively, When the wear factor variable is the cutting speed, the specific calculation strategy is as follows: extracting a variation curve of an abnormal tool wear amount of the tool with respect to a change of the cutting speed at the same cutting pressure in the control group, there are G variation curves, and synchronously extracting a cutting speed data set {v1, v2...vg...v G} and a characteristic information of the abnormal tool wear area of the tool, a tool wear data set {b 1 ,b 2 ...b g ...b G} of the tool on the day, wherein v g is the cutting speed in the gth variation curve; b g is the tool wear on the day in the gth variation curve; g ∈ {1, 2,..., G}. calculating the tool use influence coefficient S1 at the cutting speed when the wear factor variable is calculated: wherein, G is the average cutting speed in the G number of varying profiles; is the average tool wear amount in the day in the G change curves; When the wear factor variable is the cutting pressure, the specific calculation strategy is as follows: extracting a variation curve of the same cutting speed in the control group, a total of J variation curves, synchronously extracting a cutting pressure data set {p1, p2...p j ...p J} and a tool abnormal wear area feature information tool daily wear data set {c 1 ,c 2 ...c j ...c J} of the day, wherein p j is the cutting pressure in the jth variation curve; c j is the tool daily wear in the jth variation curve; j∈{1, 2,..., J}. calculating the wear factor variable as the cutting pressure S2: wherein the average cutting speed in the J varying profiles; is the average tool wear amount in the day in the J change curves.

6. The machine learning based intelligent tool life cycle management method according to claim 5, wherein, In the step S4, the evaluation of the remaining use life of the quartz processing tool with different abnormal levels includes: S41: Extract the cutting speed and cutting pressure data required by the quartz devices to be processed in the processing process in the next working day of the quartz processing factory; S42: predictively calculating a predicted wear amount of the tool in the next workday; Wherein, R0 is the factory wear amount of the tool; R x+1 is the predicted wear amount of the tool for the next workday; v x+1 for the cutting speed during tool use in the next workday; p x+1 is the cutting pressure during tool use on the next workday; S43: Evaluate the remaining use life of the quartz processing tool with different abnormal levels; On the x+1th working day, the remaining useful life T of the quartz processing tool x+1 is:

7. The machine learning based intelligent tool life cycle management method according to claim 6, wherein, The step S5 includes the following specific steps: S51: Orderly arrange the remaining use life data of the quartz processing tools of the same processing category to form the tool life sequence, and update the tool life sequence to the quartz processing tool database in real time; S52: Provide a distribution and use suggestion for the subsequent quartz processing tool distribution and use processing according to the tool life sequence, wherein the distribution and use suggestion includes preferentially using the quartz processing tool with the longest remaining use life among the quartz processing tools of the same processing category; In the step S6, the factory tool purchase and maintenance plan includes: When T x+1 ≥ T w , contact the tool manufacturer to maintain and repair the tool in advance according to the type of the abnormal wear area of the tool. When When the tool management department finds that the tool needs to be replaced and purchased, the tool management department finds the tool inventory and supplier information corresponding to the tool, and confirms whether the tool needs to be replaced and purchased; Wherein, T w and T c are the repair threshold and replacement threshold corresponding to the abnormal wear area of the tool, T x is the remaining service life of the quartz processing tool on the xth working day.

8. A machine learning based tool life cycle intelligent management system, which is implemented based on a machine learning based tool life cycle intelligent management method according to any one of claims 1-7, characterized in that, The system includes the following modules: A tool initial information collection module is configured to classify tools according to the quartz processing tool information purchased by the factory in the past, and establish a quartz processing tool database; A tool use flow information recording module is configured to distribute and use the quartz processing tools required by the quartz devices to be processed in the factory before the daily work starts, and recycle and record the tool use and flow information of the quartz processing tools after the daily work of the factory ends; a wear characteristic analysis module, configured to establish the machine learning model, perform feature extraction of tool abnormal wear on the tool use and turnover information, evaluate and calculate the daily wear of the tool and the daily tool use influence coefficient, and perform recovery abnormal classification and marking processing on the quartz processing tool recovered daily; a remaining service life evaluation module, configured to evaluate the remaining service life of the quartz processing tool of different abnormal grades; a life sequence updating module, configured to sequentially arrange the remaining service life data of the quartz processing tool of the same processing category to form the tool life sequence, and update the tool life sequence to the quartz processing tool database in real time; a purchase and maintenance plan output module, configured to statistically analyze the remaining service life of all categories of the quartz processing tool in the factory, and dynamically formulate a tool purchase and maintenance plan for the factory.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by a processor to implement a machine learning-based tool life cycle intelligent management method according to any one of claims 1-7.

10. An electronic device, comprising: comprising: a memory for storing instructions; a processor for executing the instructions, so that the device implements a machine learning-based tool life cycle intelligent management method according to any one of claims 1-7.

Images