CN117260386A - Intelligent production line mechanical type numerical control cnc engraving and milling machine cutter wear monitoring system - Google Patents
Intelligent production line mechanical type numerical control cnc engraving and milling machine cutter wear monitoring system Download PDFInfo
- Publication number
- CN117260386A CN117260386A CN202311186910.9A CN202311186910A CN117260386A CN 117260386 A CN117260386 A CN 117260386A CN 202311186910 A CN202311186910 A CN 202311186910A CN 117260386 A CN117260386 A CN 117260386A
- Authority
- CN
- China
- Prior art keywords
- cutter
- wear
- historical
- value
- abnormal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003801 milling Methods 0.000 title claims abstract description 37
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 30
- 238000012544 monitoring process Methods 0.000 title claims abstract description 29
- 238000005299 abrasion Methods 0.000 claims abstract description 91
- 230000002159 abnormal effect Effects 0.000 claims abstract description 75
- 238000010801 machine learning Methods 0.000 claims abstract description 61
- 238000011156 evaluation Methods 0.000 claims abstract description 50
- 238000012549 training Methods 0.000 claims abstract description 39
- 238000013480 data collection Methods 0.000 claims abstract description 9
- 230000006870 function Effects 0.000 claims description 28
- 238000000034 method Methods 0.000 claims description 24
- 239000013598 vector Substances 0.000 claims description 22
- 239000000463 material Substances 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 9
- 238000013145 classification model Methods 0.000 claims description 3
- 238000005520 cutting process Methods 0.000 description 11
- 238000004590 computer program Methods 0.000 description 8
- 238000005457 optimization Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 238000003860 storage Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000003754 machining Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000004065 semiconductor Substances 0.000 description 2
- 238000012795 verification Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
- B23Q—DETAILS, COMPONENTS, OR ACCESSORIES FOR MACHINE TOOLS, e.g. ARRANGEMENTS FOR COPYING OR CONTROLLING; MACHINE TOOLS IN GENERAL CHARACTERISED BY THE CONSTRUCTION OF PARTICULAR DETAILS OR COMPONENTS; COMBINATIONS OR ASSOCIATIONS OF METAL-WORKING MACHINES, NOT DIRECTED TO A PARTICULAR RESULT
- B23Q17/00—Arrangements for observing, indicating or measuring on machine tools
- B23Q17/09—Arrangements for observing, indicating or measuring on machine tools for indicating or measuring cutting pressure or for determining cutting-tool condition, e.g. cutting ability, load on tool
- B23Q17/0952—Arrangements for observing, indicating or measuring on machine tools for indicating or measuring cutting pressure or for determining cutting-tool condition, e.g. cutting ability, load on tool during machining
- B23Q17/0957—Detection of tool breakage
Landscapes
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- Numerical Control (AREA)
Abstract
The invention belongs to the technical field of cutter wear monitoring, and discloses a cutter wear monitoring system of a mechanical numerical control engraving and milling machine of an intelligent production line; comprising the following steps: the first data collection module is used for collecting first historical data of the cutter; the first model training module is used for training a first machine learning model for predicting the cutter abrasion degree value based on the first historical data; the wear evaluation module is used for analyzing the cutter wear degree value predicted by the first machine learning model and a preset cutter wear threshold value and generating a moderate wear mark or a severe wear mark for the cutter; the second data collection module is used for collecting second historical data of the tool corresponding to the heavy wear mark, wherein the second historical data comprises historical second characteristic data; analyzing the historical second characteristic data to generate a cutter wear evaluation coefficient, and generating a normal wear mark or an abnormal wear mark according to the cutter wear evaluation coefficient and a preset cutter wear evaluation coefficient threshold; saving the use cost of the cutter.
Description
Technical Field
The invention relates to the technical field of cutter wear monitoring, in particular to a cutter wear monitoring system of a mechanical numerical control engraving and milling machine of an intelligent production line.
Background
In the prior art, the invention publication number CN113997122A discloses a tool wear monitoring method and system, wherein an initial tool virtual model corresponding to a target tool system is constructed, and test data is acquired through a cutting machining test of the target tool system, so that the initial tool virtual model is corrected, a target tool virtual model is obtained, the tool wear condition is monitored in real time, and the accuracy and efficiency of tool wear monitoring are improved.
However, the above-mentioned invention only detects the cutter wear in real time, but does not monitor the cause of cutter wear, but does not fundamentally solve the problem of cutter abnormal wear if the cause of cutter abnormal wear is not monitored.
In view of the above, the present invention provides a tool wear monitoring system for a mechanical numerical control engraving and milling machine of an intelligent production line to solve the above problems.
Disclosure of Invention
In order to overcome the defects in the prior art, the embodiment of the invention provides a cutter wear monitoring system of a mechanical numerical control engraving and milling machine of an intelligent production line.
In order to achieve the above purpose, the present invention provides the following technical solutions:
in a preferred embodiment, an intelligent production line mechanical numerical control engraving and milling machine cutter wear monitoring system is characterized by comprising: the first data collection module is used for collecting first historical data of the cutter;
the first model training module is used for training a first machine learning model for predicting the cutter abrasion degree value based on the first historical data;
the wear evaluation module is used for analyzing the cutter wear degree predicted by the first machine learning model and a preset cutter wear threshold value and generating a moderate wear mark or a severe wear mark for the cutter;
a second data collection module for collecting second historical data of the tool corresponding to the heavy wear mark,
analyzing the historical second characteristic data to generate a cutter wear evaluation coefficient, and generating a normal wear mark or an abnormal wear mark according to the cutter wear evaluation coefficient and a preset cutter wear evaluation coefficient threshold;
and the second model training module is used for training a second machine learning model for identifying the abnormal abrasion fault reason of the cutter corresponding to the abnormal abrasion mark based on the second historical data.
Further, the first historical data includes historical first feature data and a tool wear level value corresponding to the historical first feature data, the tool wear level value is i=1, 2, 3..i, I being a number greater than 1; the first characteristic data includes a temperature value, a pressure value, an acoustic frequency value, and an operating power value of the tool.
Further, the temperature value is used for collecting the temperature of the cutter during operation through a temperature sensor arranged near the cutter;
the pressure value is obtained by a pressure sensor arranged on a workbench below the cutter engraving member, and the pressure exerted by the own weight of the engraving member on the pressure sensor is marked as P 1 The pressure detected by the pressure sensor during the operation of the cutter is marked as P 2 Will P 2 Subtracting P 1 The obtained difference is marked as a pressure value;
the acoustic audio frequency values collect low frequency sound V by being arranged near the sound level respectively l And high-frequency sound V h ；
The operating power information W is acquired by a multifunctional power tester.
Further, a first machine learning model for predicting the tool wear degree value is trained based on the historical first feature data, and each group of historical first feature data is converted into a corresponding group of first feature vectors;
taking each group of first feature vectors as input of a first machine learning model, wherein the first machine learning model takes a group of cutter wear degree value predictions corresponding to each group of historical first feature data as output, takes the wear degree value of a cutter corresponding to each group of real-time first feature data as a prediction target, and takes a minimized first machine learning model loss function as a training target; and stopping training when the first machine learning model loss function is smaller than or equal to the first target loss value.
Further, the first machine learning model is one of a two-classification model.
Further, the preset cutter abrasion threshold value comprises a second-level cutter abrasion threshold value and a first-level cutter abrasion threshold value, wherein the second-level cutter abrasion threshold value is smaller than the first-level cutter abrasion threshold value, and when the cutter abrasion degree is smaller than the second-level threshold value, the cutter is not marked; when the cutter wear degree is greater than or equal to the second-level cutter wear threshold and less than the first-level cutter wear threshold, marking the cutter wear condition as moderate wear, and generating a moderate wear mark; when the tool wear level is greater than or equal to the primary tool wear threshold, the marked tool wear condition is a heavy wear mark.
Further, the second historical data includes historical second feature data and numbers of abnormal wear and tear fault causes of the cutter corresponding to the historical second feature data, the numbers are n=1, 2, 3..n, N is an integer greater than or equal to 1, and each number corresponds to one abnormal wear and tear fault cause of the cutter;
the historical second characteristic data includes abnormal drop depth duration, abnormal operating temperature duration, and hardness of the work material;
the abnormal descending depth duration h is obtained through a distance sensor and a timer which are arranged at the cutter clamping arm;
the abnormal operating temperature duration s is that the operating temperature T of the tool is greater than the safe operating temperature T of the tool 1 Obtained by means of a thermometer and a timer, the safe operating temperature T of the tool 1 The tool is obtained through a use instruction of the tool;
the hardness of the processing material is set correspondingly according to different processing materials.
Further, tool wear evaluation coefficients are generated for the abnormal descent depth duration, the abnormal operating temperature duration, and the hardness normalization of the work material.
Further, when the cutter abrasion evaluation coefficient is smaller than or equal to the cutter abrasion evaluation coefficient threshold value, generating a normal abrasion mark for the cutter corresponding to the corresponding heavy abrasion mark; and when the cutter abrasion evaluation coefficient is larger than the cutter abrasion evaluation coefficient threshold value, generating an abnormal abrasion mark for the cutter corresponding to the corresponding heavy abrasion mark.
Further, a second machine learning model for identifying the cause of the abnormal wear fault of the cutter corresponding to the abnormal wear mark is trained based on the historical second feature data, and each group of the historical second feature data is converted into a corresponding group of second feature vectors; taking each group of second feature vectors as input of a second machine learning model, wherein the second machine learning model takes a group of abnormal cutter wear fault reason numbers corresponding to each group of historical second feature data as output, takes a group of abnormal cutter wear fault reason numbers corresponding to each group of real-time second feature data as a prediction target, and takes a minimized second machine learning model loss function as a training target; and stopping training when the second machine learning model loss function is smaller than or equal to the first target loss value.
The second machine learning model sends the outputted tool abnormal wear failure cause number to the wear evaluation module.
The wear evaluation module numbers the reasons of abnormal wear of the cutter, for example, the cutter carving speed is abnormally accelerated due to the fact that an electronic element for controlling the cutter carving speed is failed, and the abnormal wear of the cutter is marked as number 1 due to the fact that the temperature is too high; the failure of the clamp arm to control the movement of the tool results in excessive cutting depth of the tool, exceeding a predetermined depth, and excessive wear in the middle of the tool, causing vibration and tool breakage marked number 2.
The invention discloses a technical effect and advantages of a cutter abrasion monitoring system of an intelligent production line mechanical numerical control engraving and milling machine, which are as follows:
1. the wear condition of the cutter can be monitored in real time, different cutter wear marks are marked according to different wear conditions of the cutter, and the cutter marked with the different cutter wear marks is processed differently, so that the condition that the quality of a cutter processed product is too poor due to too serious cutter wear can be effectively prevented, the cutting quality is improved, and the processing precision is improved.
2. The method is used for analyzing the reasons of abnormal cutter wear, avoiding the problem that the reasons of abnormal cutter wear are monitored only, but not the reasons of abnormal cutter wear, eliminating the normal wear caused by normal cutter wear, detecting the abnormal cutter wear caused by other reasons, effectively prolonging the normal service life of the cutter, reducing unnecessary cutter wear and saving the use cost of the cutter.
Drawings
FIG. 1 is a schematic diagram of a tool wear monitoring system of an intelligent production line mechanical numerical control engraving and milling machine;
FIG. 2 is a schematic diagram of a method for monitoring the wear of a mechanical numerical control engraving and milling machine cutter of an intelligent production line;
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Example 1
Referring to fig. 1, the tool wear monitoring system of the intelligent production line mechanical numerical control engraving and milling machine according to the embodiment includes a first data collection module, a first model training module, a wear evaluation module, a second data collection module and a second model training module, wherein the modules are connected by a wired and/or wireless network.
The first data acquisition module is used for collecting first historical data of the cutter and sending the first historical data to the first model training module.
The first historical data includes historical first feature data and a tool wear level value corresponding to the historical first feature data, the tool wear level value being i=1, 2, 3..i, I being a number greater than 1; the first characteristic data includes a temperature value, a pressure value, an acoustic frequency value, and an operating power value of the tool.
The numerical control engraving and milling machine is mainly used for manufacturing precise parts, and the cutter is used for engraving the precise parts. The temperature value, the pressure value, the acoustic frequency value and the running power value of the cutter are relatively stable without great fluctuation under the condition that the cutter is not worn. When the cutter is worn, the numerical value is changed, and the larger the change range is, the more the cutter is worn. By monitoring the information, the cutter abrasion condition can be predicted.
The historical first characteristic data comprises a temperature value, a pressure value sound value and an operation power value of the cutter;
the temperature value is acquired by a temperature sensor arranged near the cutter and used for collecting the temperature of the cutter during operation, the temperature sensor can be a laser temperature sensor, interference on the operation of the cutter is avoided, and the condition that the cutter is worn due to overhigh temperature can occur;
the pressure value is obtained through a pressure sensor arranged on a workbench below the cutter engraving member, after the cutter is worn, the cutting edge of the cutter becomes blunt, the cutting force becomes large, namely the contact area of the part of the cutter responsible for engraving is increased, and the pressure applied to the engraving member by the corresponding cutter is also increased when the operation is performed; the pressure exerted by the own weight of the engraving member on the pressure sensor is marked as P 1 The pressure detected by the pressure sensor during the operation of the cutter is marked as P 2 Will P 2 Subtracting P 1 The obtained difference is marked as a pressure value;
the sound value is obtained by a sound level meter, the sound level score is divided into an A-type sound level meter and a B-type sound level meter, and the low-frequency sound V is respectively collected l And high-frequency sound V h In the running process of the equipment, the generated sound is the sound of the running machine and the rotating sound of the motor, and after the cutter is worn, the cutting edge of the cutter becomes blunt, the cutting force becomes large, and the cutter has an obvious high-frequency sound value V in the engraving process h And the bottom frequency sound value V l ；
The operation power value is obtained by a multifunctional electric power tester, the actual power of the machine is relatively stable in the long-term stable operation process, and after the cutter is worn, the cutter is required to consume more energy to complete carving in the process of carving operation due to the reduction of sharpness, so that the power of equipment is increased, and the operation power W of the intelligent production line mechanical numerical control engraving and milling machine is firstly collected in the state that the cutter is not worn 1 Will W 1 As intelligent production lineWhen the power threshold value of the mechanical numerical control engraving and milling machine is larger than W 1 When the tool is in a worn state.
The first model training module trains a first machine learning model for calculating the prediction of the abrasion degree of the cutter based on the historical first characteristic data of the numerical control engraving and milling machine;
the method for training the first machine learning model for calculating the cutter wear degree prediction based on the historical first characteristic data of the numerical control engraving and milling machine comprises the following steps:
converting each group of cutter characteristic data in the historical first characteristic data into a corresponding group of first characteristic vectors, wherein the cutter characteristic data of the mechanical numerical control engraving and milling machine of the production line comprises a temperature value, a pressure value, an operation power value, a low-frequency sound value and a high-frequency sound value at the same moment;
taking each group of first feature vectors as input of a first machine learning model, wherein the first machine learning model takes a group of cutter wear degree predictions corresponding to each group of historical feature data as output, takes the wear degree of cutters corresponding to each group of real-time cutter feature data as a prediction target, and takes a minimized first machine learning model loss function as a training target; and stopping training when the first machine learning model loss function is smaller than or equal to the first target loss value.
The first machine learning model loss function is mean square error or cross entropy; MSE (mean square error) 1
The mean square error is one of the commonly used loss functions by formulating the loss function:
the model is trained for the purpose of minimization, so that the machine learning model is better fitted with data, and the performance and accuracy of the model are improved;
MSE in loss function 1 P is the first eigenvector lease number for the loss function value; w is the first feature vector group number; y is p The wear degree value of the tool corresponding to the p-th group of first feature vectors,the wear degree value of the tool actually corresponding to the p-th group of the first feature vector is, for example, the wear degree value of the tool is set to be a value between 1 and 10, and the larger the value is, the more serious the wear is.
Preferably, the first machine learning model is one of a classification model;
and other model parameters of the first machine learning model, a first target loss value, an optimization algorithm, a training set test set verification set proportion, loss function optimization and the like are realized through actual engineering, and are obtained after experimental optimization is continuously carried out.
After the first machine learning model is trained, the first model training module can calculate the predicted wear degree value of the cutter in real time and send the predicted wear degree value to the wear evaluation module.
Further, a cutter abrasion threshold is set, the predicted abrasion degree is compared with the cutter abrasion threshold, the cutter abrasion threshold is divided into a second-level cutter abrasion threshold and a first-level cutter abrasion threshold, the second-level cutter abrasion threshold is smaller than the first-level cutter abrasion threshold, and when the cutter abrasion degree is smaller than the second-level threshold, the cutter is not marked and does not need to be processed; when the cutter wear degree is greater than or equal to the second-level cutter wear threshold and less than the first-level cutter wear threshold, marking the cutter wear condition as moderate wear, and generating a moderate wear mark; when the cutter wear degree is greater than or equal to the first-stage cutter wear threshold, the cutter wear condition is marked as heavy wear, and a heavy wear mark is generated.
The cutter abrasion threshold is set by a worker according to the historical service life of the cutter, and the higher the historical service life is, the lower the cutter abrasion threshold is, and the service lives of the cutters are different in different areas due to different consideration. If the cutter is slightly worn in the area due to the consideration of the quality of the processed product of the cutter, the cutter can be replaced, and the service life of the cutter is shorter; while the other areas are not replaced and continue to be used even if the cutter is worn out in some ways due to cost saving, so that the service life of the cutter is longer. The longer the service life of the cutter, the worse the abrasion condition of the cutter, and the longer the historical service life of the numerical control engraving and milling machine for a long time in a certain area, and the lower the corresponding cutter abrasion threshold value, the second-level cutter abrasion threshold value is set to be 0.5, the first-level cutter abrasion threshold value is set to be 0.7, if the historical service life of the numerical control engraving and milling machine for a long time in a certain area is shorter, the corresponding cutter abrasion threshold value is set to be higher, and if the second-level cutter abrasion threshold value is set to be 0.7, and the first-level cutter abrasion threshold value is set to be 0.9. The setting of the wear threshold of the primary cutter with the specific service life is judged by local staff according to actual conditions.
Different processing strategies are implemented on the generated moderate wear marks and heavy wear marks: for equipment marked as a moderate wear mark, the cutter of the equipment is worn to a certain extent, and when the production task is not busy, technicians can be arranged to overhaul and replace the equipment; the two pairs of equipment marked as heavy wear marks, the cutter of which is worn seriously, also affect the efficiency and speed of producing products, and the technicians need to be arranged immediately for replacement so as not to affect the quality of the products.
And the second data collection module is used for collecting second historical data of the cutter corresponding to the heavy wear mark.
The second historical data of the cutter corresponding to the heavy wear mark comprises historical second characteristic data and a number of the abnormal wear fault reason of the cutter corresponding to the historical second characteristic data, wherein the number is n=1, 2,3.
The historical second characteristic data includes abnormal drop depth duration, abnormal operating temperature duration s, and hardness of the work material.
The abnormal descent depth duration is obtained by a distance sensor and a timer arranged at the cutter clamping arm, and the specific method is as follows:
setting the threshold value of the descending depth of the cutter as h 1 ，h 1 Obtained from the corresponding tool instruction, the actual height of each tool running is marked as h, and the statistics is carried out initiallyStarting the cutter to the cutter within a time interval with heavy wear marks, wherein each time h is greater than h 1 Obtaining the duration time of the abnormal descending depth by the sum of the duration time, wherein h is obtained by a distance sensor at the cutter clamping arm; when h is greater than h 1 When the cutting depth of the cutter is too large, the contact area between the cutter and the processed material is increased, the friction force is increased, and the possibility of abnormal abrasion of the cutter is increased;
the abnormal operating temperature duration obtaining method comprises the following steps:
setting a cutter working temperature threshold T 1 Marking the actual temperature of each cutter running as T, and counting the time interval from the initial cutter to the cutter with heavy wear mark, wherein each T is larger than T 1 Obtaining abnormal working temperature duration time by the sum of the duration time, wherein the working temperature threshold value of the cutter is the safe working temperature of the cutter and is obtained through the use instruction of the corresponding cutter; the higher the abnormal working temperature, the greater the likelihood of abnormal wear of the tool; the initial cutter is brand new cutter starting use time;
the hardness of the processing material is set correspondingly according to different processing materials.
Normalizing the historical second characteristic data, and passing through the formula: f=γ 1*h +γ 2*s +γ 3*m, generating a tool wear evaluation coefficient f, γ1+γ2+γ3=1; the values of gamma 1, gamma 2 and gamma 3 are all larger than 0; γ1 is a weight coefficient of the tool descent depth, γ2 is a weight coefficient of the abnormal working temperature duration, and γ3 is a weight coefficient of the hardness of the processed material; the weight coefficient reflects the importance of the working characteristic data of each cutter on the cutter abrasion evaluation coefficient, and the larger the importance is, the larger the numerical value of the corresponding weight coefficient is; the opposite is true;
generating a cutter abrasion evaluation coefficient threshold value, comparing the cutter abrasion evaluation coefficient threshold value with the cutter abrasion evaluation coefficient, and marking the cutter marked with heavy abrasion as a normal abrasion mark if the cutter abrasion evaluation coefficient is smaller than or equal to the cutter abrasion evaluation coefficient threshold value; if the tool wear evaluation coefficient is greater than the tool wear evaluation coefficient threshold, the tool marked with the heavy wear mark is marked with an abnormal wear mark.
The generation mode of the cutter wear evaluation coefficient threshold value is as follows: under an experimental environment, running the numerical control engraving and milling machine, continuously changing historical second characteristic data of the cutter on the premise of conforming to the use specification of the cutter, recording the historical second characteristic data each time until the cutter conforms to the condition of generating a heavy wear mark, generating a plurality of groups of cutter wear evaluation coefficients for the collected plurality of groups of historical second characteristic data through a formula, removing the maximum value and the minimum value of the plurality of groups of cutter wear evaluation coefficients, and taking the average value of the rest cutter wear evaluation coefficients as a cutter wear evaluation coefficient threshold value.
The second model training module trains a second machine learning model for calculating the abnormal abrasion fault reason number of the cutter based on the historical second characteristic data of the cutter;
the method for training the second machine learning model for calculating the abnormal abrasion fault reason number of the cutter based on the historical second characteristic data of the cutter includes the following steps:
converting historical second characteristic data of the tool into a corresponding set of second characteristic vectors, wherein the historical second characteristic data of the tool set comprises abnormal descent depth duration of the tool, dangerous tool operating temperature duration and tool machining material hardness;
taking each group of second feature vectors as input of a second machine learning model, wherein the second machine learning model takes a group of abnormal wear fault reason numbers of the cutters corresponding to the historical second feature data of each group of cutters as output, takes the abnormal wear fault reason numbers of the cutters corresponding to each group of real-time historical second feature data as a prediction target, and takes a minimized second machine learning model loss function as a training target; and stopping training when the second machine learning model loss function is smaller than or equal to the second target loss value.
The second machine learning model loss function is mean square error or cross entropy;
the mean square error is one of the usual loss functions by formulating the loss functionTraining for minimisationThe model enables the machine learning model to better fit data, thereby improving the performance and accuracy of the model;
MSE in loss function 2 For the loss function value, x is the second eigenvector set number; m is the number of the second feature vector group; y is x Numbering the abnormal abrasion fault reasons of the cutter corresponding to the x-th group of second feature vectors,numbering the abnormal abrasion fault reasons of the cutter actually corresponding to the x-th group of second feature vectors;
and other model parameters of the second machine learning model, a second target loss value, an optimization algorithm, a training set test set verification set proportion, loss function optimization and the like are realized through actual engineering, and are obtained after experimental optimization is continuously carried out.
The second machine learning model sends the outputted tool abnormal wear failure cause number to the wear evaluation module.
The wear evaluation module numbers the reasons of abnormal wear of the cutter, for example, the cutter carving speed is abnormally accelerated due to the fact that an electronic element for controlling the cutter carving speed is failed, and the abnormal wear of the cutter is marked as number 1 due to the fact that the temperature is too high; the failure of the clamp arm to control the movement of the tool results in excessive cutting depth of the tool, exceeding a predetermined depth, and excessive wear in the middle of the tool, causing vibration and tool breakage marked number 2.
According to the embodiment, the abrasion of the cutters is monitored through the machine learning model, so that the abrasion condition of each cutter can be timely and accurately detected, the cutting quality is improved, and the machining precision is improved; the method can accurately grasp the fault reason of abnormal abrasion of the cutter, avoid abnormal damage of the cutter, prolong the service life of the cutter and reduce the cost.
Example 2
Referring to fig. 2, the embodiment is not described in detail, but in part, is described in embodiment one, and a method for monitoring tool wear of a mechanical numerical control engraving and milling machine of an intelligent production line is provided, which includes:
collecting first historical data of a cutter, wherein the first historical data comprises historical first characteristic data and cutter wear degree values corresponding to the historical first characteristic data, the cutter wear degree values are i=1, 2,3. The first characteristic data includes a temperature value, a pressure value, an acoustic frequency value, and an operating power value of the tool.
Training a first machine learning model for calculating the prediction of the cutter wear degree based on the historical first characteristic data of the numerical control engraving and milling machine; converting each group of cutter characteristic data in the historical first characteristic data into a corresponding group of first characteristic vectors, wherein the cutter characteristic data of the mechanical numerical control engraving and milling machine of the production line comprises temperature information, pressure value, running power and low-frequency sound V at the same moment l And high-frequency sound V h ；
Taking each group of first feature vectors as input of a first machine learning model, wherein the first machine learning model takes a group of cutter wear degree predictions corresponding to each group of historical feature data as output, takes the wear degree of cutters corresponding to each group of real-time cutter feature data as a prediction target, and takes a minimized first machine learning model loss function as a training target; and stopping training when the first machine learning model loss function is smaller than or equal to the first target loss value.
And setting a cutter abrasion threshold value, and comparing the predicted abrasion degree with the cutter abrasion threshold value. The cutter abrasion threshold is divided into a second-level cutter abrasion threshold and a first-level cutter abrasion threshold, wherein the second-level cutter abrasion threshold is smaller than the first-level cutter abrasion threshold, and when the cutter abrasion degree is smaller than the second-level threshold, the cutter is not marked and does not need to be processed; when the cutter wear degree is greater than or equal to the second-level cutter wear threshold and less than the first-level cutter wear threshold, marking the cutter wear condition as moderate wear, and generating a moderate wear mark; when the cutter wear degree is greater than or equal to the first-stage cutter wear threshold, the cutter wear condition is marked as heavy wear, and a heavy wear mark is generated.
Collecting second historical data of the tool corresponding to the heavy wear mark;
the second historical data of the cutter corresponding to the heavy wear mark comprises historical second characteristic data and a number of the abnormal wear fault reason of the cutter corresponding to the historical second characteristic data, wherein the number is n=1, 2,3.
The historical second characteristic data includes abnormal drop depth duration h, abnormal operating temperature duration s, and hardness of the work material;
normalizing the historical second characteristic data, generating a cutter abrasion evaluation coefficient through analysis, generating a cutter abrasion evaluation coefficient threshold value, comparing the cutter abrasion evaluation coefficient threshold value with the cutter abrasion evaluation coefficient, and marking the cutter with the heavy abrasion mark as a normal abrasion mark if the cutter abrasion evaluation coefficient is smaller than or equal to the cutter abrasion evaluation coefficient threshold value; if the cutter abrasion evaluation coefficient is larger than the cutter abrasion evaluation coefficient threshold value, marking the cutter marked with the heavy abrasion as an abnormal abrasion mark;
training a second machine learning model for calculating the abnormal abrasion fault reason number of the cutter based on the historical second characteristic data of the cutter;
converting historical second characteristic data of the cutters into a corresponding set of second characteristic vectors, taking each set of second characteristic vectors as input of a second machine learning model, taking a set of abnormal wear fault reason numbers of the cutters corresponding to the historical second characteristic data of each set of cutters as output, taking the abnormal wear fault reason numbers of the cutters corresponding to each set of real-time historical second characteristic data as a prediction target, and taking a minimized second machine learning model loss function as a training target; and stopping training when the second machine learning model loss function is smaller than or equal to the second target loss value.
Numbering the reasons of abnormal wear of the cutter, for example, the cutter engraving speed is abnormally accelerated due to the fact that an electronic element for controlling the cutter engraving speed is failed, and the abnormal wear of the cutter is marked as number 1 due to the fact that the temperature is too high; the failure of the clamp arm to control the movement of the tool results in excessive cutting depth of the tool, exceeding a predetermined depth, and excessive wear in the middle of the tool, causing vibration and tool breakage marked number 2.
Example 3
An electronic device is shown according to an exemplary embodiment, comprising: a processor and a memory, wherein the memory stores a computer program for the processor to call;
the processor executes the method for monitoring the cutter abrasion of the mechanical numerical control engraving and milling machine of the intelligent production line by calling the computer program stored in the memory.
Example 4
A computer readable storage medium having stored thereon a computer program that is erasable according to an exemplary embodiment is shown;
when the computer program runs on the computer equipment, the computer equipment is enabled to execute the intelligent production line mechanical numerical control engraving and milling machine cutter abrasion monitoring method.
The above embodiments may be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented in software, the above-described embodiments may be implemented in whole or in part in the form of a computer program product. The computer program product comprises one or more computer instructions or computer programs. When the computer instructions or computer program are loaded or executed on a computer, the processes or functions described in accordance with embodiments of the present invention are produced in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center over a wired network or a wireless network. The computer readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. that contains one or more sets of available media. The usable medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium may be a solid state disk.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, and are not repeated herein.
In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely one, and there may be additional divisions in actual implementation, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.
The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Finally: the foregoing description of the preferred embodiments of the invention is not intended to limit the invention to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (10)
1. The utility model provides an intelligent production line mechanical type numerical control cnc engraving and milling machine cutter wear monitoring system which characterized in that includes:
the first data collection module is used for collecting first historical data of the cutter;
the first model training module is used for training a first machine learning model for predicting the cutter abrasion degree value based on the first historical data;
the wear evaluation module is used for analyzing the cutter wear degree value predicted by the first machine learning model and a preset cutter wear threshold value and generating a moderate wear mark or a severe wear mark for the cutter;
the second data collection module is used for collecting second historical data of the tool corresponding to the heavy wear mark, wherein the second historical data comprises historical second characteristic data;
analyzing the historical second characteristic data to generate a cutter wear evaluation coefficient, and generating a normal wear mark or an abnormal wear mark according to the cutter wear evaluation coefficient and a preset cutter wear evaluation coefficient threshold;
and the second model training module is used for training a second machine learning model for identifying the abnormal wear fault reason of the cutter corresponding to the abnormal wear mark based on the historical second characteristic data.
2. The intelligent production line mechanical numerical control engraving and milling machine tool wear monitoring system of claim 1, wherein the first historical data includes historical first characteristic data and a tool wear value corresponding to the historical first characteristic data, the tool wear value is i=1, 2, 3..i, I is a value greater than 1; the historical first characteristic data comprises a cutter temperature value, a cutter pressure value, a cutter sound frequency value and a numerical control engraving and milling machine operation power value.
3. The intelligent production line mechanical numerical control engraving and milling machine cutter abrasion monitoring system according to claim 2, wherein the cutter temperature value is used for collecting the temperature of the cutter during operation through a temperature sensor arranged near the cutter;
the cutter pressure value is obtained by a pressure sensor arranged on a workbench below the cutter carving part, and the pressure exerted by the self weight of the carving part on the pressure sensor is marked as P 1 The pressure detected by the pressure sensor during the operation of the cutter is marked as P 2 Will P 2 Subtracting P 1 The obtained difference is marked as a cutter pressure value;
the tool sound frequency value collects low frequency sound V through sound level respectively arranged near the tool l And high-frequency sound V h ；
The operation power value of the numerical control engraving and milling machine is obtained through a multifunctional electric power tester.
4. The intelligent production line mechanical numerical control engraving and milling machine tool wear monitoring system according to claim 3, wherein a first machine learning model for predicting a tool wear degree value is trained based on historical first characteristic data, and each set of historical first characteristic data is converted into a corresponding set of first characteristic vectors;
taking each group of first feature vectors as input of a first machine learning model, wherein the first machine learning model takes a group of cutter wear degree value predictions corresponding to each group of historical first feature data as output, takes the wear degree value of a cutter corresponding to each group of real-time first feature data as a prediction target, and takes a minimized first machine learning model loss function as a training target; and stopping training when the first machine learning model loss function is smaller than or equal to the first target loss value.
5. The method for monitoring tool wear of an intelligent production line mechanical numerical control engraving and milling machine of claim 4, wherein the first machine learning model is one of two classification models.
6. The method for monitoring cutter wear of the intelligent production line mechanical numerical control engraving and milling machine according to claim 5, wherein the cutter wear threshold comprises a second cutter wear threshold and a first cutter wear threshold, the second cutter wear threshold is smaller than the first cutter wear threshold, and when the cutter wear level value is smaller than the second threshold, the cutter is not marked; when the cutter abrasion degree value is larger than or equal to the second-level cutter abrasion threshold value and smaller than the first-level cutter abrasion threshold value, marking the cutter abrasion condition as moderate abrasion, and generating a moderate abrasion mark; when the tool wear level value is greater than or equal to the primary tool wear threshold, the marked tool wear condition is a heavy wear mark.
7. The intelligent production line mechanical numerical control engraving and milling machine tool wear monitoring system of claim 6, wherein the second historical data further comprises a number of abnormal tool wear failure causes corresponding to the second historical characteristic data, the number is n=1, 2, 3..n, N is an integer greater than or equal to 1, and each number corresponds to one abnormal tool wear failure cause;
the historical second characteristic data includes abnormal drop depth duration, abnormal operating temperature duration, and hardness of the work material;
the abnormal descent depth duration obtaining method includes:
setting the threshold value of the descending depth of the cutter as h 1 The actual height of each cutter running is marked as h, and the initial cutter is countedWithin the time interval that the cutter has heavy wear marks, each time h is greater than h 1 Obtaining the duration time of the abnormal descending depth by the sum of the duration time, wherein h is obtained by a distance sensor at the cutter clamping arm;
the abnormal operating temperature duration obtaining method comprises the following steps:
setting a cutter working temperature threshold T 1 Marking the actual temperature of each cutter running as T, and counting the time interval from the initial cutter to the cutter with heavy wear mark, wherein each T is larger than T 1 The sum of the durations to obtain the duration of the abnormal working temperature; the initial cutter is brand new cutter starting use time;
the hardness of the processing material is set correspondingly according to different processing materials.
8. The intelligent production line mechanical numerical control engraving and milling machine tool wear monitoring system of claim 7, wherein the tool wear evaluation coefficients are generated by normalizing the abnormal descent depth duration, the abnormal operating temperature duration and the hardness of the processed material.
9. The intelligent production line mechanical numerical control engraving and milling machine cutter wear monitoring system according to claim 8 is characterized in that when the cutter wear evaluation coefficient is smaller than or equal to the cutter wear evaluation coefficient threshold value, a normal wear mark is generated for the cutter corresponding to the corresponding heavy wear mark; and when the cutter abrasion evaluation coefficient is larger than the cutter abrasion evaluation coefficient threshold value, generating an abnormal abrasion mark for the cutter corresponding to the corresponding heavy abrasion mark.
10. The intelligent production line mechanical numerical control engraving and milling machine tool wear monitoring system of claim 9, wherein each set of historical second feature data is converted into a corresponding set of second feature vectors; taking each group of second feature vectors as input of a second machine learning model, wherein the second machine learning model takes a group of abnormal cutter wear fault reason numbers corresponding to each group of historical second feature data as output, takes a group of abnormal cutter wear fault reason numbers corresponding to each group of real-time second feature data as a prediction target, and takes a minimized second machine learning model loss function as a training target; and stopping training when the second machine learning model loss function is smaller than or equal to the second target loss value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311186910.9A CN117260386A (en) | 2023-09-14 | 2023-09-14 | Intelligent production line mechanical type numerical control cnc engraving and milling machine cutter wear monitoring system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311186910.9A CN117260386A (en) | 2023-09-14 | 2023-09-14 | Intelligent production line mechanical type numerical control cnc engraving and milling machine cutter wear monitoring system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117260386A true CN117260386A (en) | 2023-12-22 |
Family
ID=89215259
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311186910.9A Pending CN117260386A (en) | 2023-09-14 | 2023-09-14 | Intelligent production line mechanical type numerical control cnc engraving and milling machine cutter wear monitoring system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117260386A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118071334A (en) * | 2024-04-22 | 2024-05-24 | 靖边县天润农业科技有限公司 | Licorice tablet preparation equipment operation and maintenance method and system based on Internet of things |
-
2023
- 2023-09-14 CN CN202311186910.9A patent/CN117260386A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118071334A (en) * | 2024-04-22 | 2024-05-24 | 靖边县天润农业科技有限公司 | Licorice tablet preparation equipment operation and maintenance method and system based on Internet of things |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109909804B (en) | Tool wear damage online monitoring method based on spindle driving current and process steps | |
| WO2021043192A1 (en) | Method for online detection of milling blade damage | |
| CN109766334B (en) | Processing method and system for online monitoring abnormal data of power equipment | |
| TWI422460B (en) | Tool nose detection method for cutting machine tool | |
| CN106514434B (en) | A kind of milling cutter wear monitoring method based on data | |
| CN103760820A (en) | Evaluation device of state information of machining process of numerical control milling machine | |
| TWI472402B (en) | Tool flutter monitoring method | |
| CN107186547A (en) | Numerical control turning batch machining tool wear on-line monitoring method based on cutting power | |
| CN117260386A (en) | Intelligent production line mechanical type numerical control cnc engraving and milling machine cutter wear monitoring system | |
| CN111113150B (en) | Method for monitoring state of machine tool cutter | |
| CN205362716U (en) | Intelligent cutter | |
| CN110334728A (en) | A kind of fault early warning method and device towards industry internet | |
| CN116596322B (en) | Property equipment management method and system based on big data visualization | |
| CN113894617A (en) | Cutter state monitoring system and method based on machine tool vibration signals | |
| CN101870076A (en) | Method for predicting service life of guide pair of numerical control machine on basis of performance degradation model | |
| JP4182399B2 (en) | Machine tool operation information collection system | |
| CN112571149A (en) | Cooling, heat dissipation monitoring and alarming system of large numerical control machine tool | |
| CN113126563A (en) | Numerical control machine tool data management system and method | |
| CN109725599A (en) | A kind of numerically-controlled machine tool efficiency on-line monitoring method | |
| CN114135477A (en) | Pump equipment state monitoring dynamic threshold early warning method | |
| CN114326593B (en) | Cutter life prediction system and method | |
| CN109240253B (en) | Online equipment diagnosis and preventive maintenance method and system | |
| CN104503361A (en) | Multimodal fusion based gear machining process tool change decision method | |
| CN116910680B (en) | Remote fault detection method and system for fitness equipment | |
| CN116307938B (en) | Health state assessment method for feeding system of machining center |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |