A numerical control tool full life cycle quality monitoring system

Abstract

The application discloses a kind of full life cycle quality monitoring systems of numerical control cutter, including cutter performance evaluation module, cutter factory quality detection module and cutter state monitoring module;The cutter performance evaluation module includes cutting force evaluation module, machining quality evaluation module, life evaluation module and stability evaluation module;The cutter state monitoring module includes data acquisition and transmission module, data preprocessing module, wear prediction model training module and state monitoring module.The application can find problems in time in each link of cutter production and use, and can find cutter with manufacturing defects in time in the process of cutter manufacturing.In the use process of cutter, dullness, chipping and other conditions of cutter can be found in time, to eliminate the security risks caused by it, and improve the quality of product.The application integrates the data generated in each link into database, maximizes the utilization rate of data in each link, and reduces the waste of useful data.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing technology, and in particular relates to a quality monitoring system for the entire life cycle of CNC cutting tools. Background Technology

[0002] With the continuous development of the manufacturing industry, the complexity of cutting tools has greatly increased, becoming one of the most significant factors affecting processing quality and efficiency in the manufacturing process. Currently, there are many problems and challenges throughout the entire lifecycle of a cutting tool, from manufacturing to use.

[0003] Due to machining errors, a large number of non-standard scrap tools are generated during the tool manufacturing process. Therefore, to ensure product quality, scrapped tools need to be identified in advance. Currently, most companies use manual inspection for outgoing quality checks, which is costly and inefficient. During tool use, the timing of tool replacement mainly relies on manual experience, leading to tool replacement too early or too late. Replacing a tool too early fails to utilize its value, resulting in waste, while replacing it too late may affect the machining quality of the product and even the precision of the machine tool.

[0004] In conclusion, there is an urgent need for an efficient tool lifecycle quality control system to promptly identify problems at each stage of the tool lifecycle, ensure product quality, and thereby enhance the core competitiveness of enterprises. Summary of the Invention

[0005] To address the aforementioned issues, this invention proposes a full lifecycle quality monitoring system for CNC cutting tools, enabling quality monitoring throughout the entire lifecycle of CNC cutting tools and reducing problems during tool design, manufacturing, and application.

[0006] To achieve the above objectives, the technical solution of the present invention is as follows: a full life cycle quality monitoring system for CNC cutting tools, comprising a tool performance evaluation module, a tool factory quality inspection module, and a tool condition monitoring module;

[0007] The tool performance evaluation module includes a cutting force evaluation module, a machining quality evaluation module, a life evaluation module, and a stability evaluation module.

[0008] The function of the cutting force evaluation module is to collect the cutting force signal of the tool during the cutting process by a force measuring instrument, and evaluate the cutting force performance of the tool by the magnitude, direction and stability of the cutting force.

[0009] The function of the machining quality evaluation module is to conduct cutting experiments and evaluate the machining quality of the tool by measuring the machining accuracy, surface roughness, residual stress layer, and cutting pattern of the machined workpiece.

[0010] The life evaluation module is used to conduct cutting experiments with fixed cutting parameters, record the maximum value of tool flank wear once every cutting period, and use the total cutting distance before the maximum value of tool flank wear exceeds 300μm as the evaluation standard for tool life.

[0011] The stability evaluation module evaluates the stability of the cutting tool by measuring the changes in cutting force during the cutting process and the accuracy stability of the machined workpiece.

[0012] The function of the tool performance evaluation module is to determine the importance of the evaluation indicators obtained from the four evaluation modules based on the specific application of the tool, and to assign weights to these four indicators. Based on the assigned weights, a comprehensive evaluation value for the tool performance is calculated, and the comprehensive evaluation value is used to determine whether the tool meets the machining requirements.

[0013] The tool manufacturing quality inspection module includes a tool image acquisition module, a tool image database, a defect detection model training module, and a manufacturing quality inspection module.

[0014] The function of the tool image acquisition module is to obtain a digital image of the tool to be detected through a vision platform. The digital image of the tool includes the image of the front face and the image of the back face.

[0015] The tool image database annotates the digital images acquired by the tool image acquisition module with defects and classifies the images according to different defect types to form the tool image database.

[0016] The defect detection model training module first expands the tool image database using image enhancement methods, then divides the images in the tool image database into training and testing sets to train the tool defect classification model.

[0017] The tool manufacturing quality inspection module sends the tool to be inspected to the vision platform of the tool image acquisition module to acquire images of the tool's rake face and flank face. The images are then input into a trained tool defect classification model to detect the presence of defects and mark defective tools.

[0018] The tool condition monitoring module includes a data acquisition and transmission module, a data preprocessing module, a wear prediction model training module, and a condition monitoring module.

[0019] The data acquisition and transmission module utilizes sensors to collect static and dynamic data information during machine tool operation and transmits the collected static and dynamic signals to the data preprocessing module. The static data information includes tool grade, tool material, workpiece size, workpiece material, spindle speed, cutting speed, depth of cut, and feed rate. The dynamic data information includes force signals in the x, y, and z directions, acceleration signals in the x, y, and z directions, and acoustic emission signals.

[0020] The data preprocessing module includes data truncation, noise reduction, and average segment aggregation. First, the stable part of the entire data segment is extracted. Then, the mean filtering method is used to reduce the noise of the data. Finally, the average segment aggregation method is used to reduce the number of sample points while preserving the trend of the original data.

[0021] The wear prediction model training module uses the collected data to train the tool wear prediction model offline.

[0022] The status monitoring module inputs the information obtained from the data acquisition and transmission module into the trained tool wear prediction model to obtain the tool wear status and the predicted tool wear value.

[0023] Furthermore, the method for evaluating the direction of cutting force involves calculating the resultant cutting force by measuring the cutting forces in the x, y, and z directions using a force gauge, and then evaluating the resultant cutting force by comparing its direction with the ideal direction. The formula for calculating the direction of the resultant cutting force is as follows:

[0024]

[0025] Where α, β, and γ are the angles between the resultant cutting force and the x, y, and z coordinate axes, respectively, and F x F y F z These represent the measured cutting force values ​​in the x, y, and z directions, respectively.

[0026] Furthermore, the workpiece machining accuracy evaluation includes dimensional accuracy evaluation and shape accuracy evaluation, which is performed by comparing the dimensions and shape of the machined workpiece with those in the design drawings. The evaluation of the residual stress layer is based on the magnitude and distribution of residual stress on the surface of the machined workpiece.

[0027] Furthermore, the formula for calculating the comprehensive evaluation value of the tool performance is as follows:

[0028] R = IW T

[0029] Where R is the comprehensive evaluation value of tool performance, I represents the score vector composed of the scores of the cutting force evaluation module, machining quality evaluation module, life evaluation module and stability evaluation module, W represents the weight vector composed of the assigned weights, and T is the transpose symbol.

[0030] Furthermore, the vision platform consists of an industrial camera, a lens, and a light source.

[0031] Furthermore, the types of tool defects include edge chipping, delamination, cracks, porosity, and vibration.

[0032] Furthermore, the tool wear prediction model consists of the following components: First, a one-dimensional convolutional neural network containing six convolutional layers, three pooling layers, and one fully connected layer is used to further extract features from the obtained data. Then, deep generalized canonical correlation analysis is used to achieve feature fusion. Finally, a softmax classifier is used based on the fused feature values ​​to determine the tool wear state. The tool wear state includes initial wear state, stable wear state, wear sag, and chipping.

[0033] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0034] 1. This invention can promptly identify problems at every stage of tool production and use. During the tool manufacturing process, tools with manufacturing defects can be detected in a timely manner. During the use of the tools, dulling, chipping, and other issues can be detected promptly, eliminating the safety hazards they pose, improving product quality, and ultimately enhancing the company's core competitiveness.

[0035] 2. The system proposed in this invention integrates the data generated in each stage into the database, maximizing the utilization rate of data in each stage and reducing the waste of useful data. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the components of the present invention.

[0037] Figure 2 This is a flowchart of the tool performance evaluation method of the present invention. Detailed Implementation

[0038] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0039] The specific embodiments described herein are particular embodiments of the present invention, used to illustrate the concept of the invention, and are illustrative and exemplary, and should not be construed as limiting the embodiments or scope of the present invention. In addition to the embodiments described herein, those skilled in the art can employ other obvious technical solutions based on the content disclosed in the claims and specification of this application. These technical solutions include those that make any obvious substitutions and modifications to the embodiments described herein, all of which are within the protection scope of the present invention.

[0040] like Figure 1 As shown, a CNC tool lifecycle quality monitoring system includes a tool performance evaluation module, a tool factory quality inspection module, and a tool condition monitoring module.

[0041] (1) The tool performance evaluation module specifically includes:

[0042] The cutting force evaluation module collects cutting force signals of the tool during the cutting process using a force measuring instrument, and evaluates the cutting force performance of the tool by measuring the magnitude, direction, and stability of the cutting force.

[0043] Furthermore, the method for evaluating the direction of cutting force is to calculate the resultant cutting force by measuring the cutting force components with a force gauge, and to evaluate the resultant cutting force by comparing the deviation of its direction from the ideal direction.

[0044] The machining quality evaluation module conducts cutting experiments and evaluates the machining quality of the tool by measuring the machining accuracy, surface roughness, residual stress layer, and cutting pattern of the machined workpiece.

[0045] Furthermore, the machining accuracy evaluation mainly includes dimensional accuracy evaluation and shape accuracy evaluation, which is conducted by comparing the dimensions and shape of the machined workpiece with those in the design drawings. The evaluation of the residual stress layer is mainly based on the magnitude and distribution of residual stress on the surface of the machined workpiece.

[0046] The tool life evaluation module uses fixed cutting parameters to conduct cutting experiments. The maximum value of tool flank wear is recorded once every cutting period. The total cutting distance before the maximum value of tool flank wear exceeds 300μm is used as the evaluation standard for tool life.

[0047] The stability evaluation module evaluates the stability of the cutting tool based on the changes in cutting force during the cutting process and the accuracy stability of the machined workpiece.

[0048] The overall tool performance evaluation module determines the importance of the four evaluation indicators based on the specific application of the tool and assigns weights to these indicators. It then calculates a comprehensive evaluation value for the tool performance based on the assigned weights and determines whether the tool meets the machining requirements based on this comprehensive evaluation value.

[0049] In the embodiments, the evaluation process for tool performance is as follows: Figure 2 As shown, in this embodiment, firstly, through a cutting experiment, the magnitude and direction of the three-dimensional cutting force during the cutting process are measured. After each cutting operation, the machining quality of the workpiece is measured using an optical microscope and a coordinate measuring machine, while the maximum wear of the tool's flank face is measured using an optical microscope. Next, each specific performance indicator of the tool is scored and assigned a corresponding weight, as shown in the specific weight allocation diagram. Figure 2 As shown, (a1+a2+a3)+(b1+b2+b3+b4)+(c1)+(d1+d2)=1. Finally, the score of each evaluation index is multiplied by its weight, and the final results are added together to obtain the comprehensive value of the tool performance.

[0050] (2) The tool factory quality inspection module specifically includes:

[0051] The tool image acquisition module obtains the digitization of the tool to be inspected through a vision platform.

[0052] Furthermore, the vision platform mainly consists of an industrial camera, a lens, and a light source. The digital image of the cutting tool mainly includes the image of the tool's rake face and the image of the tool's flank face.

[0053] The tool image database is formed by annotating defects in the digital images acquired by the tool image acquisition module and classifying the images according to different defect types.

[0054] Furthermore, the types of tool defects mainly include edge chipping, delamination, cracks, porosity, and vibration.

[0055] The model training module first expands the image database using image enhancement methods, and then divides the images in the database into training and testing sets to train the tool defect classification model.

[0056] The factory quality inspection module inputs the tool to be inspected into the tool image acquisition module, uses a trained tool defect classification model to detect whether the tool has defects, and marks the defective tool.

[0057] (3) The tool condition monitoring module specifically includes:

[0058] The data acquisition and transmission module uses sensors to collect static and dynamic data information during the machine tool's operation and transmits the collected static and dynamic signals to the data preprocessing module.

[0059] The data preprocessing module includes data truncation, noise reduction, and average segmentation aggregation. First, it extracts the stable portion of the entire data segment. Then, it uses mean filtering to reduce noise. Finally, it uses average segmentation aggregation to reduce the number of sample points while preserving the trend of the original data.

[0060] The model training module uses the collected data to develop an offline tool wear prediction model.

[0061] The condition monitoring module inputs the information obtained from the data acquisition module into the trained tool wear prediction model to obtain the tool wear status and the predicted tool wear value.

[0062] Furthermore, the tool wear prediction model specifically comprises the following components: First, a one-dimensional convolutional neural network containing six convolutional layers, three pooling layers, and one fully connected layer performs further feature extraction on the obtained data. Then, deep generalized canonical correlation analysis is used to achieve feature fusion. Finally, a softmax classifier is used based on the fused feature values ​​to determine the tool wear state. The tool wear state includes initial wear state, stable wear state, wear sag, and chipping.

[0063] This invention is not limited to this embodiment. Any equivalent concept or modification within the technical scope disclosed in this invention shall be included within the protection scope of this invention.

Claims

1. A full life cycle quality monitoring system for a CNC tool, characterized by: The tool performance evaluation module includes a cutting force evaluation module, a machining quality evaluation module, a life evaluation module and a stability evaluation module. The cutting force evaluation module collects cutting force signals of the tool in the cutting process through a dynamometer, and evaluates the cutting force performance of the tool according to the cutting force size, the cutting force direction and the cutting force stability. The machining quality evaluation module performs cutting experiments, and evaluates the machining quality of the tool according to the machining precision, the surface roughness, the residual stress layer and the cutting form of the machined workpiece. The life evaluation module performs cutting experiments with fixed cutting parameters, and records the maximum wear value of the tool flank surface every certain time interval. The stability evaluation module evaluates the stability of the tool according to the change of the cutting force and the precision stability of the machined workpiece. The tool performance evaluation module evaluates the importance of the evaluation indexes obtained by the four evaluation modules according to the specific use of the tool, and performs weight distribution on the four indexes. The tool image acquisition module obtains digital images of the tool to be detected through a visual platform. The tool image database labels the digital images collected by the tool image acquisition module, and classifies the images according to different defect types to form a tool image database. The defect detection model training module first expands the tool image database by image enhancement, then divides the images in the tool image database into a training set and a test set, and trains a tool defect classification model. The tool image acquisition module obtains digital images of the tool to be detected through a visual platform. The data acquisition and transmission module collects static data information and dynamic data information of the machine tool running process, and transmits the collected static and dynamic signals to the data preprocessing module. ​ ​ ​ The data preprocessing module includes data interception, noise reduction and average segmentation aggregation; first, the stable part in the whole data is intercepted, then the mean filtering method is used to reduce the noise of the data, finally, the average segmentation aggregation method is used to reduce the number of sample points while retaining the trend of the original data; The wear prediction model training module trains the tool wear prediction model offline using the collected data; The state monitoring module inputs the information obtained from the data acquisition and transmission module into the trained tool wear prediction model to obtain the tool wear state and tool wear prediction value; The tool wear prediction model is composed as follows: first, a one-dimensional convolutional neural network containing six convolutional layers, three pooling layers and one fully connected layer is used to further extract features from the obtained data, then a deep generalized canonical correlation analysis method is used to realize feature fusion, and finally a softmax classifier is used to obtain the tool wear state according to the fused feature values; the tool wear state includes initial wear state, stable wear state, wear and collapse.

2. The full life cycle quality monitoring system of a CNC tool according to claim 1, characterized in that: The cutting force direction evaluation method is to calculate the cutting resultant force by measuring the cutting forces in x, y and z directions by the dynamometer, and to evaluate by comparing the deviation degree of the direction of the cutting resultant force and the ideal direction; wherein the direction calculation formula of the cutting resultant force is: wherein, respectively, the cutting resultant force and the included angle of the three coordinate axes, respectively, the measured cutting force size values in the three directions.

3. The full life cycle quality monitoring system of a CNC tool according to claim 1, wherein: The workpiece machining precision evaluation includes size precision evaluation and shape precision evaluation, and the evaluation is performed by comparing the size and shape of the machined workpiece with the design drawing; the evaluation of the residual stress layer takes the residual stress size and distribution of the machined workpiece surface as the evaluation basis.

4. The full life cycle quality monitoring system of a CNC tool according to claim 1, wherein: The calculation formula of the tool performance comprehensive evaluation value is as follows: wherein, is a comprehensive evaluation value of the tool performance, denotes a score vector composed of scores from the cutting force evaluation module, the machining quality evaluation module, the lifetime evaluation module and the stability evaluation module, denotes a weight vector composed of the assigned weights, is a transpose symbol.

5. The full life cycle quality monitoring system of a CNC tool according to claim 1, wherein: The visual platform is composed of an industrial camera, a lens and a light source.

6. The full life cycle quality monitoring system of a CNC tool according to claim 1, wherein: The tool defect types include edge collapse, layering, cracks, pores and vibration marks.

Images