Intelligent milling tool wear state monitoring method

Abstract

This invention relates to an intelligent milling tool wear state monitoring method. It pre-constructs a milling tool type sub-matching model and a set of milling signal curves generated during the milling of workpieces of different materials. This allows for the acquisition of benchmark values ​​for the wear state of the milling tool when milling workpieces of different materials under different wear conditions. Furthermore, it pre-constructs a set of workpiece surface feature parameters and a set of milling waste chip feature parameters when milling workpieces of different materials under different wear conditions. Then, it constructs separate milling tool wear state judgment sub-models for milling workpieces of different materials using milling tools of the same material. The workpiece material, the current milling tool material type, the current set of milling signal curves, the current workpiece surface feature parameters, and the current milling waste chip feature parameters are input into the milling tool wear state judgment sub-models to obtain the monitored wear state of the current milling tool.

Description

Technical Field

[0001] This invention relates to the field of intelligent milling, and more particularly to a method for monitoring the wear condition of intelligent milling tools. Background Technology

[0002] As a key tool in CNC milling, milling typically includes planar milling or curved surface milling. The milling structure consists of a cutter head mounted on a tool holder, and then milling tools (inserts) mounted on the cutter head for machining. During the machining process using milling tools, the tools wear due to friction with the workpiece surface, negatively impacting the dimensional accuracy, surface roughness, and surface texture of the machined workpiece. Furthermore, severe tool wear significantly affects the quality of the machined workpiece and severely reduces milling efficiency.

[0003] Therefore, timely and accurate detection of the wear condition of milling tools is crucial for understanding the wear status of milling tools in CNC milling operations. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an intelligent milling tool wear condition monitoring method in light of the above-mentioned prior art.

[0005] The technical solution adopted by this invention to solve the above-mentioned technical problems is: a method for monitoring the wear condition of intelligent milling tools, characterized by comprising the following steps:

[0006] Step 1: Pre-construct a milling tool type sub-matching model that represents a milling tool of a certain material type when it is not performing milling operations; wherein, the milling tool has a milling tool type sub-matching model that corresponds one-to-one with its material type, and the milling tool has its own cutting edge curve;

[0007] Step 2: Pre-set multiple key milling points for each milling tool corresponding to the milling tool type sub-matching model, and construct a three-dimensional model of the milling tool corresponding to the milling tool type sub-matching model in a preset rectangular coordinate system; wherein, each milling tool has a three-dimensional model of the milling tool that corresponds to it one by one.

[0008] Step 3: Obtain the set of three-dimensional coordinates of the three-dimensional model of the milling tool and the coordinates of the key points of each milling key point on the three-dimensional model of the milling tool, forming a set of key point coordinates of the milling tool corresponding to the milling tool; wherein, the set of key point coordinates of the milling tool corresponds one-to-one with the milling tool.

[0009] Step 4: In the process of milling workpieces of different materials with milling cutters of the same material, vibration signal information and pressure signal information of each key point of milling on the milling cutter are collected in advance, forming a set of milling signal curves when the milling cutter of the same material is milling the corresponding workpiece. The set of milling signal curves corresponds one-to-one with the milling cutter and the workpiece material milled by the milling cutter. The set of milling signal curves includes a set of vibration signal curves and a set of pressure signal curves. The set of vibration signal curves is formed by the vibration signal curves of all key points of milling on the milling cutter, and the set of pressure signal curves is formed by the pressure signal curves of all key points of milling on the milling cutter.

[0010] Step 5: Determine the abnormal wear times when the milling cutter exhibits different wear states during the milling of workpieces of different materials, and the amplitudes of the milling cutter vibration signal and pressure signal on the milling signal curves of each key milling point corresponding to each abnormal wear time. Establish a list of abnormal wear amplitude relationships when the milling cutter mills workpieces of different materials. The list of abnormal wear amplitude relationships includes a one-to-one correspondence between the milling cutter wear state and the amplitudes of the milling cutter vibration signal and pressure signal at each key milling point.

[0011] Step 6: Based on the list of abnormal wear amplitudes of various milling tools, calculate the baseline value of the wear state of the milling tools when milling workpieces of different materials under different wear states.

[0012] Step 7: Pre-construct sets of milling waste chip characteristic parameters for milling waste chips obtained from workpieces of different materials when milling with milling cutters under different wear conditions; wherein, the milling cutter material, the workpiece material being milled, the wear condition of the milling cutter, and the milling waste chip characteristic parameters obtained from the workpiece are all in one-to-one correspondence.

[0013] Step 8: Based on the obtained set of benchmark values ​​for the wear state of each milling tool and the set of characteristic parameters of milling waste, construct sub-models for judging the wear state of milling tools of different materials when milling workpieces with milling tools of this material.

[0014] Step 9: Determine the workpiece material in the actual milling process and identify the current milling tool material type in the actual milling process based on the milling tool type matching model. Obtain the current milling signal curve set and current milling waste characteristic parameters of the milling tool in the milling process. Among them, the milling tool in the actual milling process and the milling tool that forms the corresponding milling tool type sub-matching model are tools of the same material and specifications.

[0015] Step 10: Input the current milling tool material type, workpiece material, current milling signal curve set, and current milling chip characteristic parameters into the milling tool wear state judgment sub-model, and output the current milling tool wear state.

[0016] Improved in this invention, the intelligent milling tool wear condition monitoring method further includes:

[0017] Steps 1 to 8 are executed sequentially for milling cutters of different materials to construct sub-models for judging the wear state of milling cutters when milling workpieces of different materials.

[0018] The overall model for judging the wear state of milling tools is formed by combining all the sub-models for judging the wear state of milling tools.

[0019] Furthermore, the current milling tool material type, workpiece material, current milling signal curve set, and current milling waste chip characteristic parameters are input as input parameters into the milling tool wear state judgment sub-model located within the overall milling tool wear state judgment model and corresponding to the milling tool, and the current milling tool wear state is output.

[0020] Furthermore, in the intelligent milling tool wear condition monitoring method, the construction process of the milling tool type sub-matching model is as follows: steps a1 to a5:

[0021] Step a1: Simultaneously apply load voltage to milling tools of each material type during the first time period to obtain the current and resistance values ​​of each material type milling tool; wherein, the total number of material types of milling tools is denoted as I, the total number of resistance values ​​of the i-th material milling tool during the first time period is denoted as M, and the m-th resistance value of the i-th material milling tool during the first time period is denoted as r. i,m , 1≤i≤I, 1≤m≤M;

[0022] Step a2: Parallel light rays are simultaneously emitted from the same light source onto milling tools of different material types during the second time period, obtaining the refractive index set of each material type milling tool; wherein, the total number of refractive indices of the i-th material milling tool during the second time period is denoted as K, and the k-th refractive index value of the i-th material milling tool during the second time period is denoted as n. i,k , 1≤k≤K;

[0023] Step a3: Based on the obtained sets of resistance and refractive index of milling tools for each material type, obtain the characteristic values ​​of the milling tool type matching parameters for each material type; wherein, the characteristic values ​​of the milling tool type matching parameters are as follows:

[0024]

[0025] Where, χ i,T To characterize the feature value of the milling tool type matching parameter for milling tools of the i-th material, α i These are intermediate parameters obtained based on the resistance value of the milling tool for the i-th type of material. It is the average resistance of the milling tool for the i-th type of material; β i These are intermediate parameters obtained based on the optical characteristics of milling tools for the i-th type of material. It is the average refractive index of the milling tool for the i-th type of material;

[0026] Step a4: Pre-set the actual values ​​of the milling tool resistance and the set of refractive indices of each material type of milling tool before milling, using a labeling method, to obtain the actual values ​​of the milling tool type matching parameters for each material type of milling tool; wherein, the actual value of the milling tool type matching parameter for the i-th material type milling tool is labeled as χ. i,s :

[0027]

[0028] Step a5: Based on the obtained feature values ​​of the milling tool type matching parameters for various materials and the actual values ​​of the milling tool type matching parameters, a milling tool type matching model is formed; the milling tool type matching model is as follows:

[0029]

[0030] Furthermore, in the intelligent milling tool wear condition monitoring method, the process of setting the key points of the milling process in step 2 is as follows:

[0031] Obtain the cutting edge curve equations for each cutting edge of the milling tool; wherein, the cutting edge of the milling tool includes the main cutting edge of the main cutting edge and the secondary cutting edge of the secondary cutting edge;

[0032] Based on the obtained equations of each cutting edge curve, determine all singular points on each cutting edge; where singular points are the inflection points on the corresponding cutting edge curves.

[0033] All the identified singular points are taken as the critical points for milling operations using this milling tool.

[0034] Furthermore, in the intelligent milling tool wear condition monitoring method, in step 4, the vibration signal curve is the curve of the vibration signal amplitude changing with time at the key points of milling, and the pressure signal curve set is the curve of the pressure signal amplitude changing with time at the key points of milling.

[0035] Furthermore, in the intelligent milling tool wear condition monitoring method, in step 6, the calculation method for the benchmark value of the milling tool wear condition when the milling tool is milling a workpiece of any material is as follows:

[0036]

[0037]

[0038] Where, ψ i,j For a milling tool made of material i, the vibration signal amplitude of its vibration signal curve is A. j The reference value for the wear condition of the milling tool at that time, ψ i,j The wear condition of the milling tool corresponding to the i-th type of material is marked as C. i,j A q B represents the amplitude of the qth vibration signal of the vibration signal curve corresponding to the milling tool of the i-th material. q μ represents the q-th pressure signal amplitude of the pressure signal curve corresponding to the i-th type of material milling tool; A Let μ be the average amplitude of the first j vibration signals of the milling tool of material i in its vibration signal curve. B Let A be the average value of the first j pressure signal amplitudes of the milling tool of material i in its pressure signal curve; where the vibration signal amplitude A is... j Each wear anomaly corresponds one-to-one with the moment when a milling tool of material i is used to mill a workpiece of any material.

[0039] Improved, in the intelligent milling tool wear condition monitoring method, the construction process of the milling tool wear condition judgment sub-model is as follows:

[0040] Step b1: Pre-acquire images of workpiece milling waste chips and infrared images of workpiece milling waste chips obtained when milling workpieces of different materials under different wear conditions. Pre-process the milling waste chip images of each workpiece to obtain pre-processed workpiece milling waste chip images corresponding to different wear conditions of the milling cutter. The pre-processed workpiece milling waste chip images correspond one-to-one with the type of milling cutter used, the wear condition of the milling cutter, and the workpiece of the milled material.

[0041] Step b2: Process the milling waste images of each preprocessed workpiece to obtain the milling waste histogram of each preprocessed workpiece milling waste image.

[0042] Step b3: Process the histogram of milling waste chips for each workpiece to obtain the grayscale mean value of the histogram of milling waste chips for each material workpiece.

[0043] Step b4: Based on the obtained infrared images of milling waste from each workpiece, process them to obtain the average temperature of the milling waste for each material workpiece.

[0044] Step b5: Based on the obtained average gray value and average temperature value, obtain the workpiece milling chip wear characterization factor that represents the wear state of milling chips for each material workpiece.

[0045] Step b6: Use the obtained wear characterization factor values ​​of milling waste chips for each workpiece as a sub-model for judging the wear state of milling tools during the milling process of milling workpieces of different materials.

[0046] Furthermore, in the intelligent milling tool wear condition monitoring method, in step b5 above, the workpiece milling chip wear characterization factor is calculated as follows:

[0047]

[0048] Where, λ i,t,j G represents the wear factor of milling chips when a milling tool of material i is milling a workpiece of material j at time t. i,t,j T is the grayscale mean of the workpiece milling chip histogram corresponding to the milling tool of material i when milling a workpiece of material j at time t. i,t,j It is the average temperature and grayscale value G of the milling chips corresponding to the workpiece of material i when the milling tool of material i is milling the workpiece of material j at time t. i,t,j Average temperature T i,t,j And the wear condition of the milling cutter when milling the workpiece corresponds one-to-one.

[0049] Compared with the prior art, the advantages of this invention are as follows: The intelligent milling tool wear condition monitoring method of this invention pre-constructs a milling tool type matching model, a set of milling signal curves formed during the milling of workpieces of different materials by the milling tool, and obtains the benchmark value of the wear condition of the milling tool when milling workpieces of different materials under different wear conditions. It also pre-constructs a set of workpiece surface feature parameters and a set of milling waste feature parameters of the milling waste obtained from the workpiece when milling workpieces of different materials under different wear conditions. Finally, based on the obtained wear condition of each milling tool... The following sets of reference values, workpiece surface feature parameters, and milling waste chip feature parameters are used to construct sub-models for judging the wear state of milling tools made of different materials. Then, the workpiece material in the actual milling process, the current milling tool material type, the current milling signal curve set, the current workpiece surface feature parameters, and the current milling waste chip feature parameters are identified and executed based on the milling tool type matching model and input into the milling tool wear state judgment sub-model to obtain the monitored wear state of the current milling tool.

[0050] Because this invention considers and integrates the characteristics of the milling tool during the milling process, the characteristics of the workpiece itself, and the characteristics of the milling chips as judgment factors when judging the wear state of the milling tool, it achieves a comprehensive judgment from both the characteristics of the tool performing the milling action and the characteristics of the workpiece being milled, thereby improving the accuracy of the judgment and avoiding the one-sidedness of judging the tool wear state solely based on the characteristics of the milling tool. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of the intelligent milling tool wear condition monitoring method in an embodiment of the present invention;

[0052] Figure 2 This invention relates to the accuracy of the intelligent milling tool wear monitoring method in this embodiment for monitoring the wear condition (wear depth) of milling tools made of different materials. Detailed Implementation

[0053] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0054] This embodiment provides a method for monitoring the wear condition of intelligent milling tools. Specifically, see [link to documentation]. Figure 1 As shown, the intelligent milling tool wear condition monitoring method of this embodiment includes the following steps:

[0055] Step 1: Pre-construct a milling tool type sub-matching model representing a milling tool of a certain material type when not performing milling operations; wherein, the milling tool has a milling tool type sub-matching model that corresponds one-to-one with its material type, and the milling tool has its own cutting edge curve; for example, in this embodiment, the milling tool used here is a milling tool of material X1, and after the construction process, the resulting milling tool type sub-matching model matching the milling tool of material X1 is labeled as MODEL(X1);

[0056] Specifically, in this embodiment, the construction process of the milling tool type sub-matching model is as follows: steps a1 to a5:

[0057] Step a1: Simultaneously apply load voltage to milling tools of each material type during the first time period to obtain the current and resistance values ​​of each material type milling tool; wherein, the total number of material types of milling tools is denoted as I, the total number of resistance values ​​of the i-th material milling tool during the first time period is denoted as M, and the m-th resistance value of the i-th material milling tool during the first time period is denoted as r. i,m , 1≤i≤I, 1≤m≤M;

[0058] Step a2: Parallel light rays are simultaneously emitted from the same light source onto milling tools of different material types during the second time period, obtaining the refractive index set of each material type milling tool; wherein, the total number of refractive indices of the i-th material milling tool during the second time period is denoted as K, and the k-th refractive index value of the i-th material milling tool during the second time period is denoted as n. i,k , 1≤k≤K;

[0059] Step a3: Based on the obtained sets of resistance and refractive index of milling tools for each material type, obtain the characteristic values ​​of the milling tool type matching parameters for each material type; wherein, the characteristic values ​​of the milling tool type matching parameters are as follows:

[0060]

[0061] Where, χ i,T To characterize the feature value of the milling tool type matching parameter for milling tools of the i-th material, α i These are intermediate parameters obtained based on the resistance value of the milling tool for the i-th type of material. It is the average resistance of the milling tool for the i-th type of material; β i These are intermediate parameters obtained based on the optical characteristics of milling tools for the i-th type of material. It is the average refractive index of the milling tool for the i-th type of material;

[0062] Step a4: Pre-set the actual values ​​of the milling tool resistance and the set of refractive indices of each material type of milling tool before milling, using a labeling method, to obtain the actual values ​​of the milling tool type matching parameters for each material type of milling tool; wherein, the actual value of the milling tool type matching parameter for the i-th material type milling tool is labeled as χ. i,s :

[0063]

[0064] Step a5: Based on the obtained feature values ​​of the milling tool type matching parameters for various materials and the actual values ​​of the milling tool type matching parameters, a milling tool type matching model is formed; the milling tool type matching model is as follows:

[0065]

[0066] Step 2: Pre-set multiple key milling points for each milling tool corresponding to the milling tool type sub-matching model, and construct a three-dimensional model of the milling tool corresponding to the milling tool type sub-matching model in a preset rectangular coordinate system; wherein, each milling tool has a three-dimensional model of the milling tool that corresponds to it one by one.

[0067] Specifically, for the milling tool type sub-matching model MODEL(X1) constructed in this embodiment, after processing, multiple key milling points on the milling tool type sub-matching model MODEL(X1) are marked as D1, D2, ..., D... n-1 and D n The three-dimensional model of the milling tool (X1) corresponding to the sub-matching model MODEL(X1) for this milling tool type is labeled as V(X1);

[0068] Step 3: Obtain the set of three-dimensional coordinates of the three-dimensional model of the milling tool and the coordinates of the key points of each milling key point on the three-dimensional model of the milling tool, forming a set of key point coordinates of the milling tool corresponding to the milling tool; wherein, the set of key point coordinates of the milling tool corresponds one-to-one with the milling tool.

[0069] In this embodiment, after processing, the key point coordinates of each milling machining key point on the three-dimensional model V(X1) of the milling tool are marked as D1(x1,y1,z1), D2(x2,y2,z2), ..., D... n-1 (x n-1 ,y n-1 ,z n-1 ) and D n (x n ,y n ,z nCorrespondingly, the set of coordinates of the key milling points corresponding to the milling tool X1 is labeled D(X1), where D(X1) = {D1(x1,y1,z1),D2(x2,y2,z2),…,D…} n-1 (x n-1 ,y n-1 ,z n-1 ),D n (x n ,y n ,z n )};

[0070] Step 4: In the process of milling workpieces of different materials with milling cutters of the same material, vibration signal information and pressure signal information of each key point of milling on the milling cutter are collected in advance, forming a set of milling signal curves when the milling cutter of the same material is milling the corresponding workpiece. The set of milling signal curves corresponds one-to-one with the milling cutter and the workpiece material milled by the milling cutter. The set of milling signal curves includes a set of vibration signal curves and a set of pressure signal curves. The set of vibration signal curves is formed by the vibration signal curves of all key points of milling on the milling cutter, and the set of pressure signal curves is formed by the pressure signal curves of all key points of milling on the milling cutter.

[0071] It should be noted that in step 4, the vibration signal curve is the curve of the amplitude of the vibration signal at the key points of the milling process changing with time, and the pressure signal curve set is the curve of the amplitude of the pressure signal at the key points of the milling process changing with time.

[0072] For example, in this embodiment, for a milling cutter made of material X1, the set of milling signal curves during the milling process of the milling cutter made of material Y1 is denoted as S. X1-Y1 Milling signal curve set S X1-Y1 The set of vibration signal curves within is labeled S. A,X1-Y1 Milling signal curve set S X1-Y1 The set of pressure signal curves within is labeled S. B,X1-Y1 Vibration signal curve set S A,X1-Y1 Key point D in internal milling n The vibration signal curve is marked as s A,Dn,X1-Y1 Pressure signal curve set S B,X1-Y1 Key point D in internal milling n The pressure signal curve is marked as s B,Dn,X1-Y1 ;

[0073] Step 5: Determine the abnormal wear times when the milling cutter exhibits different wear states during the milling of workpieces of different materials, and the amplitudes of the milling cutter vibration signal and pressure signal on the milling signal curves of each key milling point corresponding to each abnormal wear time. Establish a list of abnormal wear amplitude relationships when the milling cutter mills workpieces of different materials. The list of abnormal wear amplitude relationships includes a one-to-one correspondence between the milling cutter wear state and the amplitudes of the milling cutter vibration signal and pressure signal at each key milling point.

[0074] Step 6: Based on the list of abnormal wear amplitudes of various milling tools, calculate the baseline value of the milling tool wear state when milling workpieces of different materials under different wear conditions; for example, in this embodiment, the calculation method of the baseline value of the milling tool wear state is as follows:

[0075]

[0076]

[0077] Where, η i,c Let ψ be the baseline value for the wear state of a milling tool of material i when its wear degree is c. i,p,m The vibration signal amplitude of the vibration signal curve of the milling tool of material i at the u-th critical point of its milling process is A. p The reference value for the wear condition of the milling tool corresponding to the time, A q Let B be the amplitude of the qth vibration signal on the vibration signal curve of the i-th material milling tool at its u-th critical milling point. q μ represents the amplitude of the qth pressure signal on the pressure signal curve of the i-th material milling tool at its u-th critical milling point; A,u Let μ be the average amplitude of the first p vibration signals of the vibration signal curve at the u-th critical point of the milling process for the i-th type of material. B,u The average value of the first p pressure signal amplitudes of the pressure signal curve at the u-th critical point of the milling process for the i-th type of material milling tool;

[0078] Step 7: Pre-construct sets of milling waste chip characteristic parameters for milling waste chips obtained from workpieces of different materials under different wear conditions. Among them, the milling tool material, the workpiece material being milled, the wear state of the milling tool, and the milling waste chip characteristic parameters obtained from the workpiece are all in one-to-one correspondence. The milling waste chip characteristic parameters are the gray-scale mean of the histogram corresponding to the milling waste chip and the temperature mean of the infrared image corresponding to the milling waste chip.

[0079] Step 8: Based on the obtained set of benchmark values ​​for the wear state of each milling tool and the set of characteristic parameters of milling waste, construct sub-models for judging the wear state of milling tools of different materials when milling workpieces with milling tools of this material.

[0080] Step 9: Determine the workpiece material in the actual milling process and identify the current milling tool material type in the actual milling process based on the milling tool type matching model. Obtain the current milling signal curve set and current milling waste characteristic parameters of the milling tool in the milling process. Among them, the milling tool in the actual milling process and the milling tool that forms the corresponding milling tool type sub-matching model are tools of the same material and specifications.

[0081] Step 10: Input the current milling tool material type, workpiece material, current milling signal curve set, and current milling chip characteristic parameters into the milling tool wear state judgment sub-model, and output the current milling tool wear state.

[0082] To improve the accuracy of monitoring the wear condition of milling tools, in step 8 of this embodiment, the constructed milling tool wear condition judgment sub-model is optimized using a convolutional neural network to form an optimized milling tool wear condition judgment sub-model. Specifically, the formation process of the optimized milling tool wear condition judgment sub-model includes the following steps:

[0083] Step S1: Based on the different wear states of the milling cutter, the tool wear state benchmark value and milling chip characteristic parameters corresponding to one or more wear states of the milling cutter are used as training samples, while the tool wear state benchmark value and milling chip characteristic parameters corresponding to other wear states of the milling cutter are used as test samples.

[0084] Step S2: Input the training samples into a convolutional autoencoder for model pre-training to obtain the model pre-training results;

[0085] Step S3: Continue model training through a convolutional neural network to obtain an initial model for judging the wear state of milling tools for this material.

[0086] Step S4: Use the obtained initial model for judging the wear state of milling tools to judge the wear state of the test sample and output the milling tool wear state judgment result;

[0087] Step S5: Based on the error between the actual tool wear state of the tool and the obtained milling tool wear state judgment result, adjust the parameters of the initial model for judging the milling tool wear state, and perform iterative update processing on the initial model for judging the milling tool wear state after adjusting the model parameters to obtain the final optimized milling tool wear state judgment sub-model.

[0088] Specifically, in step 2 of this embodiment, the process of setting the key points for milling is as follows:

[0089] Obtain the cutting edge curve equations for each cutting edge of the milling tool; wherein, the cutting edge of the milling tool includes the main cutting edge of the main cutting edge and the secondary cutting edge of the secondary cutting edge;

[0090] Based on the obtained equations of each cutting edge curve, determine all singular points on each cutting edge; where singular points are the inflection points on the corresponding cutting edge curves.

[0091] All the identified singular points are taken as the critical points for milling operations using this milling tool.

[0092] More specifically, regarding the milling tool wear state judgment sub-model in step 8 above, in this embodiment, the construction process of the milling tool wear state judgment sub-model is as follows:

[0093] Step b1: Pre-acquire images of workpiece milling waste chips and infrared images of workpiece milling waste chips obtained when milling workpieces of different materials under different wear conditions. Pre-process the milling waste chip images of each workpiece to obtain pre-processed workpiece milling waste chip images corresponding to different wear conditions of the milling cutter. The pre-processed workpiece milling waste chip images correspond one-to-one with the type of milling cutter used, the wear condition of the milling cutter, and the workpiece of the milled material.

[0094] Step b2: Process the milling waste images of each preprocessed workpiece to obtain the milling waste histogram of each preprocessed workpiece milling waste image.

[0095] Step b3: Process the histogram of milling waste chips for each workpiece to obtain the grayscale mean value of the histogram of milling waste chips for each material workpiece.

[0096] Step b4: Based on the obtained infrared images of milling waste from each workpiece, process them to obtain the average temperature of the milling waste for each material workpiece.

[0097] Step b5: Based on the obtained average gray value and average temperature value, obtain the workpiece milling chip wear characterization factor that represents the wear state of milling chips for each material workpiece.

[0098] Step b6: Use the obtained wear characterization factor values ​​of milling waste chips for each workpiece as a sub-model for judging the wear state of milling tools during the milling process of milling workpieces of different materials.

[0099] In step b5 above, the workpiece milling chip wear characterization factor is calculated as follows:

[0100]

[0101] Where, λ i,t,j G represents the wear factor of milling chips when a milling tool of material i is milling a workpiece of material j at time t. i,t,j T is the grayscale mean of the workpiece milling chip histogram corresponding to the milling tool of material i when milling a workpiece of material j at time t. i,t,j It is the average temperature and grayscale value G of the milling chips corresponding to the workpiece of material i when the milling tool of material i is milling the workpiece of material j at time t. i,t,j Average temperature T i,t,j And the wear condition of the milling cutter when milling the workpiece corresponds one-to-one.

[0102] To test the accuracy of the intelligent milling tool wear monitoring method in this embodiment in monitoring the wear of milling tools, this embodiment used the intelligent milling tool wear monitoring method to conduct an experimental simulation of the tool wear (characterized by tool wear depth) of a milling tool made of a certain material during the cutting process of a workpiece. For details of the experimental simulation, please refer to [link to simulation details]. Figure 2 As shown. From Figure 2 As can be seen, the accuracy (marked as Pd) of the intelligent milling tool wear monitoring method in this embodiment increases as the wear of the tool deepens.

[0103] Although preferred embodiments of the present invention have been described in detail above, it should be clearly understood that various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for monitoring the wear condition of intelligent milling tools, characterized in that, Includes the following steps: Step 1: Pre-construct a milling tool type sub-matching model that represents a milling tool of a certain material type when it is not performing milling operations; wherein, the milling tool has a milling tool type sub-matching model that corresponds one-to-one with its material type, and the milling tool has its own cutting edge curve; Step 2: Pre-set multiple key milling points for each milling tool corresponding to the milling tool type sub-matching model, and construct a three-dimensional model of the milling tool corresponding to the milling tool type sub-matching model in a preset rectangular coordinate system; wherein, each milling tool has a three-dimensional model of the milling tool that corresponds to it one by one. Step 3: Obtain the set of three-dimensional coordinates of the three-dimensional model of the milling tool and the coordinates of the key points of each milling key point on the three-dimensional model of the milling tool, forming a set of key point coordinates of the milling tool corresponding to the milling tool; wherein, the set of key point coordinates of the milling tool corresponds one-to-one with the milling tool. Step 4: In the process of milling workpieces of different materials with milling cutters of the same material, vibration signal information and pressure signal information of each key point of milling on the milling cutter are collected in advance, forming a set of milling signal curves when the milling cutter of the same material is milling the corresponding workpiece. The set of milling signal curves corresponds one-to-one with the milling cutter and the workpiece material milled by the milling cutter. The set of milling signal curves includes a set of vibration signal curves and a set of pressure signal curves. The set of vibration signal curves is formed by the vibration signal curves of all key points of milling on the milling cutter, and the set of pressure signal curves is formed by the pressure signal curves of all key points of milling on the milling cutter. Step 5: Determine the abnormal wear times when the milling cutter exhibits different wear states during the milling of workpieces of different materials, and the amplitudes of the milling cutter vibration signal and pressure signal on the milling signal curves of each key milling point corresponding to each abnormal wear time. Establish a list of abnormal wear amplitude relationships when the milling cutter mills workpieces of different materials. The list of abnormal wear amplitude relationships includes a one-to-one correspondence between the milling cutter wear state and the amplitudes of the milling cutter vibration signal and pressure signal at each key milling point. Step 6: Based on the list of abnormal wear amplitudes of various milling tools, calculate the baseline value of the wear state of the milling tools when milling workpieces of different materials under different wear states. Step 7: Pre-construct sets of milling waste chip characteristic parameters for milling waste chips obtained from workpieces of different materials when milling with milling cutters under different wear conditions; wherein, the milling cutter material, the workpiece material being milled, the wear condition of the milling cutter, and the milling waste chip characteristic parameters obtained from the workpiece are all in one-to-one correspondence. Step 8: Based on the obtained set of benchmark values ​​for the wear state of each milling tool and the set of characteristic parameters of milling waste, construct sub-models for judging the wear state of milling tools of different materials when milling workpieces with milling tools of this material. Step 9: Determine the workpiece material in the actual milling process and identify the current milling tool material type in the actual milling process based on the milling tool type matching model. Obtain the current milling signal curve set and current milling waste characteristic parameters of the milling tool in the milling process. Among them, the milling tool in the actual milling process and the milling tool that forms the corresponding milling tool type sub-matching model are tools of the same material and specifications. Step 10: Input the current milling tool material type, workpiece material, current milling signal curve set, and current milling chip characteristic parameters into the milling tool wear state judgment sub-model, and output the current milling tool wear state.

2. The intelligent milling tool wear condition monitoring method according to claim 1, characterized in that, Also includes: Steps 1 to 8 are executed sequentially for milling cutters of different materials to construct sub-models for judging the wear state of milling cutters when milling workpieces of different materials. The overall model for judging the wear state of milling tools is formed by combining all the sub-models for judging the wear state of milling tools. Furthermore, the current milling tool material type, workpiece material, current milling signal curve set, and current milling waste chip characteristic parameters are input as input parameters into the milling tool wear state judgment sub-model located within the overall milling tool wear state judgment model and corresponding to the milling tool, and the current milling tool wear state is output.

3. The intelligent milling tool wear condition monitoring method according to claim 2, characterized in that, The process of constructing the milling tool type sub-matching model is as follows: steps a1 to a5: Step a1: Simultaneously apply load voltage to milling tools of each material type during the first time period to obtain the current and resistance values ​​of each material type milling tool; wherein, the total number of material types of milling tools is denoted as I, the total number of resistance values ​​of the i-th material milling tool during the first time period is denoted as M, and the m-th resistance value of the i-th material milling tool during the first time period is denoted as r. i,m , 1≤i≤I, 1≤m≤M; Step a2: Parallel light rays are simultaneously emitted from the same light source onto milling tools of different material types during the second time period, obtaining the refractive index set of each material type milling tool; wherein, the total number of refractive indices of the i-th material milling tool during the second time period is denoted as K, and the k-th refractive index value of the i-th material milling tool during the second time period is denoted as n. i,k , 1≤k≤K; Step a3: Based on the obtained sets of resistance and refractive index of milling tools for each material type, obtain the characteristic values ​​of the milling tool type matching parameters for each material type; wherein, the characteristic values ​​of the milling tool type matching parameters are as follows: Where, χ i,T To characterize the feature value of the milling tool type matching parameter for milling tools of the i-th material, α i These are intermediate parameters obtained based on the resistance value of the milling tool for the i-th type of material. It is the average resistance of the milling tool for the i-th type of material; β i These are intermediate parameters obtained based on the optical characteristics of milling tools for the i-th type of material. It is the average refractive index of the milling tool for the i-th type of material; Step a4: Pre-set the actual values ​​of the milling tool resistance and the set of refractive indices of each material type of milling tool before milling, using a labeling method, to obtain the actual values ​​of the milling tool type matching parameters for each material type of milling tool; wherein, the actual value of the milling tool type matching parameter for the i-th material type milling tool is labeled as χ. i,s : Step a5: Based on the obtained feature values ​​of the milling tool type matching parameters for various materials and the actual values ​​of the milling tool type matching parameters, a milling tool type matching model is formed; the milling tool type matching model is as follows:

4. The intelligent milling tool wear condition monitoring method according to claim 3, characterized in that, In step 2, the process of setting the key points for the milling process is as follows: Obtain the cutting edge curve equations for each cutting edge of the milling tool; wherein, the cutting edge of the milling tool includes the main cutting edge of the main cutting edge and the secondary cutting edge of the secondary cutting edge; Based on the obtained equations of each cutting edge curve, determine all singular points on each cutting edge; where singular points are the inflection points on the corresponding cutting edge curves. All the identified singular points are taken as the critical points for milling operations using this milling tool.

5. The intelligent milling tool wear condition monitoring method according to claim 4, characterized in that, In step 4, the vibration signal curve is the curve showing the change of vibration signal amplitude over time at key points of milling, and the pressure signal curve set is the curve showing the change of pressure signal amplitude over time at key points of milling.

Images