Tool condition monitoring method, system, device and storage medium

Abstract

The application discloses a tool state monitoring method, system, device and storage medium, and technical scheme points thereof are as follows: a dynamic equation of tool milling is established in advance, and experimental data of tool milling are obtained; simulation data are obtained by performing vibration simulation according to the dynamic equation, and a filter frequency band and a target feature are determined by analyzing the simulation data; experimental filtered data are obtained by filtering the experimental data according to the filter frequency band, experimental characteristic values of the target feature corresponding to the experimental data and the experimental filtered data are calculated, and the experimental characteristic values are taken as threshold values; vibration filtered data are obtained by filtering vibration data of a target tool according to the filter frequency band, and vibration characteristic values of the target feature corresponding to the vibration data and the vibration filtered data are calculated; a comparison result is obtained by comparing the vibration characteristic values with corresponding threshold values, and the state of the target tool is judged according to the comparison result. The scheme reduces the dependence on historical data of milling, and realizes real-time monitoring of the tool state.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of mechanical processing, and particularly relates to a tool state monitoring method, system, device and storage medium. BACKGROUND

[0002] In the mechanical manufacturing and processing industry, tool wear state is one of the key factors affecting processing efficiency, product quality and processing precision. Therefore, online monitoring of tool state is an important means to ensure the processing quality of parts and avoid economic losses.

[0003] With the development of sensing technology and algorithms, the tool wear state can be monitored by obtaining signals related to the tool state through indirect sensors and combining relevant algorithms. However, in existing technologies, most of the research on tool wear monitoring is based on artificial intelligence algorithms such as machine learning or deep learning. For example, the document with the publication number "CN 113780153 A" uses a deep learning algorithm to train the model, and weakens redundant information through an attention mechanism, thereby reducing the computational cost. The document with the publication number "CN 113997122 A" also continuously trains and corrects the tool model by combining historical command data, state data and wear data, thereby improving the accuracy of prediction.

[0004] These intelligent algorithms require a large amount of historical data for model training, and the more accurate the model requires, the more complete the experimental data under different parameters requires. At the same time, mechanical processing is affected by various factors, and it is difficult to obtain all data under various processing environments, so the application of tool monitoring algorithms based on machine learning or deep learning has been restricted. SUMMARY

[0005] The purpose of the present application is to provide a tool state monitoring method, system, device and storage medium, which reduces the dependence on milling processing historical data and realizes real-time monitoring of tool state.

[0006] The first aspect of the present application discloses a tool state monitoring method, comprising:

[0007] Step 1, a dynamic equation of tool milling processing is established in advance, and experimental data of tool milling processing is obtained in advance;

[0008] Step 2, vibration simulation is performed according to the dynamic equation to obtain simulation data, and the simulation data is analyzed to determine a filter frequency band and a target feature;

[0009] Step 3, experimental filtering data is obtained by filtering the experimental data according to the filter frequency band, the experimental feature value of the target feature corresponding to the experimental data and the experimental filtering data is calculated, and the experimental feature value is taken as a threshold value;

[0010] Step 4, collecting vibration data of the target tool;

[0011] Step 5, filtering the vibration data according to the filter frequency band to obtain vibration filtered data, and calculating vibration characteristic values of target characteristics corresponding to the vibration data and the vibration filtered data;

[0012] Step 6, comparing the vibration characteristic values with corresponding threshold values to obtain comparison results, and judging the state of the target tool according to the comparison results.

[0013] Optionally, the dynamic equation is:

[0014]

[0015] wherein, M is a two-dimensional mass matrix of the tool, C is a two-dimensional damping matrix of the tool, K is a two-dimensional stiffness matrix of the tool, is vibration acceleration, is vibration velocity, q(t) is vibration displacement, F s,st (t) is static shear force, F s,dy (t) is dynamic shear force, F p (t) is instantaneous impact force caused by the tool cutting into the workpiece, F e,st (t) is static plowing force, F e,da (t) is dynamic plowing force.

[0016] Optionally, the calculation formulas of the static shear force, the dynamic shear force, the instantaneous impact force, the static plowing force and the dynamic plowing force are respectively:

[0017]

[0018]

[0019]

[0020]

[0021]

[0022] wherein, N z represents the number of axial force units, N represents the number of tool teeth, i represents the i-th axial force unit, j represents the j-th tool tooth, g(Φ i,j (t)) represents a first window function, T(Φ i,j (t)) represents a coordinate conversion matrix, K ts represents a tangential shear force coefficient, K rs represents a radial shear force coefficient, K te represents a tangential plowing force coefficient, Kre represents a radial ploughing force coefficient, h s represents a static milling thickness, h r,j represents a radial dynamic milling thickness of the jth tooth, h r,j-1 represents a radial dynamic milling thickness of the j-1th tooth, S st represents a tool wear static area, S dy represents a tool wear dynamic area, w(Φ i,j (t)) represents a second window function, K a represents a stiffness coefficient, δ represents a deformation of the tool, e represents a penetration depth index, c p represents a damping coefficient of an instantaneous impact force, represents an angle of entry of the tool during milling.

[0023] Optionally, the analyzing the simulation data to determine a filtering frequency band and a target feature comprises:

[0024] selecting a frequency band with the highest energy in the simulation data as a first filtering frequency band, and selecting a frequency band with the lowest energy in the simulation data as a second filtering frequency band;

[0025] calculating a value of a simulation feature corresponding to the simulation data, the simulation feature comprising: an absolute average value, a root mean square value, a standard deviation, a pulse factor, a kurtosis, a skewness, and a Gini index;

[0026] selecting a first target feature from the simulation feature, the first target feature being any one of the simulation feature of the absolute average value, the root mean square value, and the variance related to the tool wear state in the simulation feature;

[0027] selecting a second target feature from the simulation feature, the second target feature being the Gini index in the simulation feature that can best express the impact signal in the milling process.

[0028] Optionally, the filtering experimental data according to the filtering frequency band to obtain experimental filtering data, calculating an experimental feature value of the target feature corresponding to the experimental data and the experimental filtering data, and taking the experimental feature value as a threshold value, comprises:

[0029] filtering the experimental data according to the first filtering frequency band to obtain first filtering data, and filtering the experimental data according to the second filtering frequency band to obtain second filtering data;

[0030] calculating an absolute average value of the experimental data of the tool in a non-milling state and a Gini index of the first filtering data to obtain a non-milling Gini index and a non-milling absolute average value, and determining a first threshold value according to the non-milling absolute average value and the non-milling Gini index;

[0031] calculating a Gini index of the first filtered data and an absolute average value of the second filtered data of the tool in a finish milling process and assuming a severe wear state, obtaining a finish milling Gini index and a finish milling absolute average value, and determining a second threshold value according to the finish milling Gini index and the finish milling absolute average value;

[0032] calculating a Gini index of the first filtered data and an absolute average value of the second filtered data of the tool in a finish milling process and assuming a severe wear state, obtaining a finish milling Gini index and a finish milling absolute average value, and determining a second threshold value according to the finish milling Gini index and the finish milling absolute average value;

[0033] Optionally, the vibration characteristic values include: an original absolute average value, a filtered Gini index and a filtered absolute average value, the original absolute value is an absolute average value of vibration data, the filtered Gini index is a Gini index of vibration filtered data filtered by a first filter frequency band, and the filtered absolute average value is an absolute average value of vibration filtered data filtered by a second filter frequency band.

[0034] The comparison of the vibration characteristic values with the corresponding threshold values obtains a comparison result, and the state of the target tool is determined according to the comparison result, including:

[0035] The original absolute average value and the filtered Gini index are compared with a first threshold value, and in a case where both the original absolute average value and the filtered Gini index are not less than the first threshold value, it is determined that the target tool is in a milling state, otherwise, it is determined that the target tool is in a non-milling state.

[0036] In a case where the target tool is in a finish milling process, the filtered Gini index and the filtered absolute average value are compared with a second threshold value, and in a case where both the filtered Gini index and the filtered absolute average value are not less than the second threshold value, it is determined that the target tool is in an assumed wear state, otherwise, it is determined that the target tool is in a normal wear state.

[0037] In a case where the target tool is in a finish milling process, the filtered Gini index and the filtered absolute average value are compared with a second threshold value, and in a case where both the filtered Gini index and the filtered absolute average value are not less than the second threshold value, it is determined that the target tool is in an assumed wear state, otherwise, it is determined that the target tool is in a normal wear state.

[0038] Optionally, the method further includes:

[0039] Step 7, determining whether the state of the target tool is an assumed wear state, if yes, returning to execute step 4 until the state of the target tool is determined as the assumed wear state for three times in succession, and then determining that the state of the target tool is a severe wear state.

[0040] A second aspect of the present application discloses a tool state monitoring system, including:

[0041] A pre-preparation module is configured to pre-establish a dynamic equation of the tool milling process and pre-acquire experimental data of the tool milling process.

[0042] A simulation analysis module is configured to perform vibration simulation according to the dynamic equation to obtain simulation data, analyze the simulation data to determine a filtering frequency band and a target feature.

[0043] A threshold determination module is configured to filter the experimental data according to the filtering frequency band to obtain experimental filtered data, calculate experimental feature values of the target feature corresponding to the experimental data and the experimental filtered data, and take the experimental feature values as threshold values.

[0044] A data acquisition module is configured to acquire vibration data of the target tool.

[0045] A data processing module is configured to filter the vibration data according to the filtering frequency band to obtain vibration filtered data, and calculate vibration feature values of the target feature corresponding to the vibration data and the vibration filtered data.

[0046] A comparison and determination module is configured to compare the vibration feature values with corresponding threshold values to obtain a comparison result, and determine the state of the target tool according to the comparison result.

[0047] The third aspect of the present application discloses a computer device, comprising a memory and a processor, the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0048] The fourth aspect of the present application discloses a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of the above method.

[0049] The technical solution provided by the present application has the following advantages and effects: by establishing a dynamic equation of the tool milling process, performing simulation according to the dynamic equation to obtain simulation data, determining a filtering frequency band and a target feature through the simulation data, the dependence on historical milling data is reduced, and a large amount of milling data is not required. Then, the experimental data of the tool milling process are filtered and calculated according to the filtering frequency band and the target feature to determine a threshold value for judging the tool state. After the vibration data of the target tool are acquired, the vibration data are filtered and calculated according to the filtering frequency band and the target feature to obtain vibration feature values corresponding to the target feature. The vibration feature values are compared with corresponding threshold values to obtain a comparison result, and the state of the target tool is determined according to the comparison result, thereby realizing real-time monitoring of the tool state. BRIEF DESCRIPTION OF DRAWINGS

[0050] Figure 1is a flowchart of a tool state monitoring method provided by the present application;

[0051] Figure 2 is a schematic diagram of a dynamic model of tool milling provided by an embodiment of the present application;

[0052] Figure 3 is a time-domain diagram of a simulated vibration response provided by an embodiment of the present application;

[0053] Figure 4 is a frequency-domain diagram of a simulated vibration response provided by an embodiment of the present application;

[0054] Figure 5 is a three-axis time-domain diagram of experimental data provided by an embodiment of the present application;

[0055] Figure 6 is a three-axis frequency-domain diagram of experimental data provided by an embodiment of the present application;

[0056] Figure 7 is a three-axis time-domain diagram of amplified experimental data provided by an embodiment of the present application;

[0057] Figure 8 is a three-axis frequency-domain diagram of amplified experimental data provided by an embodiment of the present application;

[0058] Figure 9 is a judgment result diagram of 7 groups of experimental groups provided by an embodiment of the present application;

[0059] Figure 10 is a structural block diagram of a tool state monitoring system provided by an embodiment of the present application;

[0060] Figure 11 is an internal structure diagram of a computer device provided by an embodiment of the present application.

[0061] Explanation of reference signs:

[0062] 100, workpiece; 200, tool tooth; 300a, ideal trajectory of the j-1th tool tooth; 300b, vibration displacement of the j-1th tool tooth; 400a, ideal trajectory of the jth tool tooth; 400b, vibration displacement of the jth tool tooth. DETAILED DESCRIPTION

[0063] In order to facilitate the understanding of the present application, specific embodiments of the present application will be described in more detail below with reference to the accompanying drawings.

[0064] Unless specifically stated or otherwise defined, "first, second, …" used herein is merely used for distinguishing names, and does not represent a specific quantity or order.

[0065] Unless otherwise specified or defined, the term “and / or” as used herein includes any and all combinations of one or more of the associated listed items.

[0066] It should be noted that in this article, "fixed to" or "connected to" can mean directly fixed to or connected to a component, or indirectly fixed to or connected to a component.

[0067] like Figure 1 As shown, the present invention provides a tool condition monitoring method, including:

[0068] Step 1: Establish the dynamic equations of the milling process in advance and obtain experimental data of the milling process in advance.

[0069] Specifically, such as Figure 2 As shown, in this embodiment of the invention, a dynamic model of milling machining is constructed in the XOY plane perpendicular to the milling machine spindle. This dynamic model is a two-degree-of-freedom model. In this dynamic model, 100 represents the workpiece to be cut, 200 represents the cutting tool teeth, 300a represents the ideal trajectory of the (j-1)th cutting tool tooth, 300b represents the vibration displacement of the (j-1)th cutting tool tooth, 400a represents the ideal trajectory of the jth cutting tool tooth, and 400b represents the vibration displacement of the jth cutting tool tooth. Based on this dynamic model, the dynamic equation of milling machining is as follows:

[0070]

[0071] Where M is the two-dimensional mass matrix of the cutting tool, C is the two-dimensional damping matrix of the cutting tool, and K is the two-dimensional stiffness matrix of the cutting tool. For vibration acceleration, Let q(t) be the vibration velocity, and F be the vibration displacement. s,st (t) represents the static shear force, F s,dy (t) represents the dynamic shear force, F p (t) represents the instantaneous impact force caused by the tool cutting into the workpiece, F. e,st (t) represents the static plowing force, F e,da (t) represents the dynamic plowing force. Figure 2 f in e Indicates the feed rate.

[0072] In the embodiment, the shearing force and ploughing force of the cutter are divided into static force and dynamic force, so that the simulation data based on the dynamic equation is more accurate. In the application, a sensor is arranged on a tool holder of the cutter, the sensor adopts a three-axis acceleration sensor, a milling experiment is performed on the cutter, vibration experimental data of the cutter in the milling experiment are collected by the sensor, static data in the vibration experimental data are removed to obtain experimental data, and the experimental data is convenient for determining a threshold value in combination with subsequent simulation data.

[0073] Further, the calculation formulas of the static shearing force, the dynamic shearing force, the instantaneous impact force, the static ploughing force and the dynamic ploughing force are respectively:

[0074]

[0075]

[0076]

[0077]

[0078]

[0079] wherein, N z represents the number of axial force units, N represents the number of cutter teeth, i represents the i-th axial force unit, j represents the j-th cutter tooth, g(Φ i,j (t)) represents a first window function, T(Φ i,j (t)) represents a coordinate conversion matrix, K ts represents a tangential shearing force coefficient, K rs represents a radial shearing force coefficient, K te represents a tangential ploughing force coefficient, K re represents a radial ploughing force coefficient, h s represents a static milling thickness, h r,j represents a radial dynamic milling thickness of the j-th cutter tooth, h r,j-1 represents a radial dynamic milling thickness of the j-1-th cutter tooth, S st represents a cutter wear static area, S dy represents a cutter wear dynamic area, w(Φ i,j (t)) represents a second window function, K a represents a stiffness coefficient, δ represents a deformation amount of the cutter, e represents a penetration depth index, c p represents a damping coefficient of the instantaneous impact force, represents a cutting-in angle of the cutter in the milling process.

[0080] Specifically, the calculation formulas of the first window function and the second window function are respectively:

[0081]

[0082]

[0083] wherein, the Φ i,j (t) represents the instantaneous milling angle of the jth tooth in the i axial force unit, represents the cutting-out angle of the tool in the milling process.

[0084] Step 2, vibration simulation is performed according to the dynamic equation to obtain simulation data, and the simulation data is analyzed to determine the filtering frequency band and the target feature.

[0085] In actual application, the vibration of tool milling can be simulated according to the dynamic equation, for example, the experimental parameters of the tool in the milling experiment are substituted into the dynamic equation, and the experimental parameters can include: tool mass, tool stiffness, number of axial force units, number of teeth, tangential shear force coefficient, radial shear force coefficient, tangential plowing force coefficient, radial plowing force coefficient, static milling thickness, tool wear static area, stiffness coefficient, damping coefficient of instantaneous impact force and tool cutting angle; in the simulation process, the damping is continuously increased to simulate the tool wear process, and the simulation result is obtained, for example, as shown in Figure 3 and Figure 4 When the radial dynamic milling thickness of the tooth decreases in the simulation process, it indicates that the tool is damaged, and when the dynamic area of tool wear increases sharply, it indicates that the tool wear is serious. As can be seen from Figure 3 , the milling vibration waveform is a plurality of pulse impact signals, in the case of tooth damage, the vibration amplitude of the tooth decreases, and the vibration amplitude of the other tooth does not change, however, with the aggravation of tool wear, the system damping increases, the decay of the pulse signal slows down, but the amplitude increases significantly. By analyzing the simulation data to determine the filtering frequency band and the target feature, subsequent data processing and calculation of experimental data and vibration data can be facilitated to better monitor the target tool state.

[0086] Further, the analyzing the simulation data to determine the filtering frequency band and the target feature comprises:

[0087] selecting the frequency band with the highest energy in the simulation data as the first filtering frequency band, and selecting the frequency band with the lowest energy in the simulation data as the second filtering frequency band;

[0088] calculating the value of the simulation feature corresponding to the simulation data, and the simulation feature includes: absolute average, root mean square, standard deviation, pulse factor, kurtosis, skewness and Gini index;

[0089] selecting a first target feature from the simulation features, the first target feature being any one of the absolute average value, the root mean square value and the variance of the simulation features related to the tool wear state;

[0090] selecting a second target feature from the simulation features, the second target feature being the Gini index of the simulation features most expressing the impact signal in the milling process.

[0091] Specifically, from the frequency domain graph of the simulation vibration response, Figure 4 ) it can be known that the discontinuous impact signal excites the resonance of the system, and the energy in the resonance frequency band is the highest, and the resonance frequency band is taken as the first filter frequency band, and the frequency band with the lowest energy is taken as the second filter frequency band, in the embodiment, the first filter frequency band is 800-1000 Hz, and the second filter frequency band is 1-200 Hz, and there is a surplus among the simulation features, and since the calculation results of the absolute average value, the root mean square value and the variance are consistent, therefore, any one of the absolute average value, the root mean square value and the variance can be selected as the first target feature, since the pulse factor, the skewness and the kurtosis are consistent and all irrelevant to the tool wear state, the pulse factor, the skewness or the kurtosis is not taken as the target feature, since the Gini index can better express the impact signal in the milling process, the Gini index is taken as the second target feature. In the embodiment, the absolute average value is selected as the first target feature, and the tool state is expressed by the absolute average value and the Gini index.

[0092] Step 3, filtering the experimental data according to the filter frequency band to obtain experimental filter data, calculating the experimental feature values of the target features corresponding to the experimental data and the experimental filter data, and taking the experimental feature values as the threshold values.

[0093] Further, the filtering of the experimental data according to the filter frequency band to obtain the experimental filter data, the calculation of the experimental feature values of the target features corresponding to the experimental data and the experimental filter data, and the taking of the experimental feature values as the threshold values comprise:

[0094] filtering the experimental data according to the first filter frequency band to obtain first filter data, and filtering the experimental data according to the second filter frequency band to obtain second filter data;

[0095] calculating the absolute average value of the experimental data of the tool in the non-milling state and the Gini index of the first filter data, obtaining the non-milling Gini index and the non-milling absolute average value, and determining the first threshold value according to the non-milling absolute average value and the non-milling Gini index;

[0096] calculating the Gini index of the first filtered data and the absolute average of the second filtered data of the cutting tool in the finish milling process and assuming the severe wear state, obtaining a finish milling Gini index and a finish milling absolute average, determining a second threshold according to the finish milling Gini index and the finish milling absolute average;

[0097] calculating the Gini index of the first filtered data and the absolute average of the second filtered data of the cutting tool in the finish milling process and assuming the severe wear state, obtaining a finish milling Gini index and a finish milling absolute average, determining a second threshold according to the finish milling Gini index and the finish milling absolute average;

[0098] Specifically, as shown in Figure 5 and Figure 6 , the time domain graph and the frequency domain graph of the experimental data are shown, as shown in Figure 7 and Figure 8 , the time domain graph and the frequency domain graph of the experimental data are shown, after obtaining the laboratory data, the experimental data is filtered through a resonance band-pass filter to obtain first filtered data in a first filter frequency band, and the experimental data is filtered through a low-pass filter to obtain second filtered data in a second filter frequency band; the calculation formulas of the absolute average, the root mean square, the standard deviation, the pulse factor, the skewness, the kurtosis, and the Gini index are as follows:

[0099]

[0100]

[0101]

[0102]

[0103]

[0104]

[0105]

[0106] wherein, μ x (t) represents the absolute average, RMS represents the root mean square, σ x (t) represents the standard deviation, P f represents the pulse factor, S k represents the skewness, K represents the kurtosis, GI represents the Gini index, N represents the signal length, t represents the time, x i (t) represents the time domain signal at t, max(x i (t)) represents the maximum value of the time series signal, min(x i (t)) represents the minimum value of the time series signal, w irepresents a weight factor. After obtaining the non-milling absolute average value and the non-milling Gini index, the non-milling absolute average value and the non-milling Gini index can both be taken as the first threshold value, after obtaining the absolute average value corresponding to the vibration data of the target tool, the absolute average value is compared with the non-milling absolute average value in the first threshold value, and after obtaining the Gini index corresponding to the vibration data filtered by the first filter frequency band, the Gini index is compared with the non-milling Gini index in the first threshold value; after obtaining the non-milling absolute average value and the non-milling Gini index, the non-milling absolute average value and the non-milling Gini index can also be weighted and calculated according to actual conditions to obtain a calculation result, which is taken as the first threshold value, after obtaining the absolute average value corresponding to the vibration data of the target tool and the Gini index corresponding to the vibration data filtered by the first filter frequency band, the absolute average value and the Gini index are weighted and calculated to obtain a calculation result, which is compared with the first threshold value; in actual application, the milling machining of the tool is divided into forward milling machining and reverse milling machining, and tool changing is required in the process of switching from forward milling machining to reverse milling machining or from reverse milling machining to forward milling machining, therefore, it is necessary to judge whether the tool is in the milling state during tool state monitoring, and the first threshold value can be used to judge whether the tool is in the milling state.

[0107] Specifically, in the state of forward milling machining of the tool, the second threshold value can be used to judge whether the tool is in the assumed wear state, after obtaining the forward milling Gini index and the forward milling absolute average value, the forward milling Gini index and the forward milling absolute average value can both be taken as the second threshold value, after obtaining the Gini index corresponding to the vibration data filtered by the first filter frequency band, the Gini index is compared with the forward milling Gini index in the second threshold value, and after obtaining the absolute average value corresponding to the vibration data filtered by the second filter frequency band, the absolute average value is compared with the forward milling absolute average value in the second threshold value; after obtaining the forward milling Gini index and the forward milling absolute average value, the forward milling Gini index and the forward milling absolute average value can also be weighted and calculated according to actual conditions to obtain a calculation result, which is taken as the second threshold value, after obtaining the Gini index corresponding to the vibration data filtered by the first filter frequency band and the absolute average value corresponding to the vibration data filtered by the second filter frequency band, the Gini index and the absolute average value are weighted and calculated to obtain a calculation result, which is compared with the second threshold value.

[0108] Specifically, in the state of tool reverse milling processing, whether the tool is in the assumed wear state can be judged by the third threshold value, after obtaining the reverse milling Gini index and the reverse milling absolute average value, the reverse milling Gini index and the reverse milling absolute average value can be used as the third threshold value, after obtaining the Gini index corresponding to the vibration data filtered by the first filter frequency band, the Gini index is compared with the reverse milling Gini index in the third threshold value, after obtaining the absolute average value corresponding to the vibration data filtered by the second filter frequency band, the absolute average value is compared with the reverse milling absolute average value in the third threshold value; after obtaining the reverse milling Gini index and the reverse milling absolute average value, the reverse milling Gini index and the reverse milling absolute average value can also be weighted and calculated according to the actual situation to obtain a calculation result, which is used as the third threshold value, after obtaining the Gini index corresponding to the vibration data filtered by the first filter frequency band and the absolute average value corresponding to the vibration data filtered by the second filter frequency band, the Gini index and the absolute average value are weighted and calculated to obtain a calculation result, which is compared with the third threshold value.

[0109] Step 4, collecting vibration data of the target tool; in actual application, vibration monitoring data is collected through a sensor on the handle of the target tool, and corresponding vibration data is obtained after static data removal of the vibration monitoring data. When data processing is performed each time, only a data packet of 0.2s is obtained, which is about 6 rotation periods for a spindle speed of 1800rpm. Feature calculation and extraction are performed on each data packet.

[0110] Step 5, filtering the vibration data according to the filter frequency band to obtain vibration filtered data, and calculating vibration characteristic values of target features corresponding to the vibration data and the vibration filtered data.

[0111] Specifically, the vibration data is filtered by a resonance band-pass filter to obtain first vibration filtered data in a first filter frequency band, and the vibration data is filtered by a low-pass filter to obtain second vibration filtered data in a second filter frequency band; the vibration characteristic values include an original absolute average value, a filtered Gini index and a filtered absolute average value, the original absolute value is an absolute average value of the vibration data, which is calculated according to an absolute average value calculation formula; the filtered Gini index is a Gini index of the vibration filtered data filtered by the first filter frequency band, that is, a Gini index of the first vibration filtered data, which is calculated according to a Gini index calculation formula; the filtered absolute average value is an absolute average value of the vibration filtered data filtered by the second filter frequency band, that is, an absolute average value of the second vibration filtered data, which is calculated according to the absolute average value calculation formula.

[0112] Step 6, comparing the vibration characteristic value with the corresponding threshold value to obtain a comparison result, and judging the state of the target tool according to the comparison result.

[0113] Further, the comparing the vibration characteristic value with the corresponding threshold value to obtain a comparison result, and judging the state of the target tool according to the comparison result, comprises:

[0114] Comparing the original absolute average value and the filtered Gini index with a first threshold value, and determining that the target tool is in a milling state if both the original absolute average value and the filtered Gini index are not less than the first threshold value, otherwise, determining that the target tool is in a non-milling state.

[0115] If the target tool is in a forward milling processing state, then comparing the filtered Gini index and the filtered absolute average value with a second threshold value, and determining that the target tool is in a presumed wear state if both the filtered Gini index and the filtered absolute average value are not less than the second threshold value, otherwise, determining that the target tool is in a normal wear state.

[0116] If the target tool is in a reverse milling processing state, then comparing the filtered Gini index and the filtered absolute average value with a third threshold value, and determining that the target tool is in a presumed wear state if both the filtered Gini index and the filtered absolute average value are not less than the third threshold value, otherwise, determining that the target tool is in a normal wear state.

[0117] Specifically, in the case that the non-milling absolute average value and the non-milling Gini index are both taken as the first threshold value, the original absolute average value and the filtered Gini index are both not less than the first threshold value means that the original absolute average value is not less than the non-milling absolute average value and the filtered Gini index is not less than the non-milling Gini index, in the case that the non-milling absolute average value and the non-milling Gini index are taken for weighted calculation, the original absolute average value and the filtered Gini index are both not less than the first threshold value means that the calculation result of the weighted calculation of the original absolute average value and the filtered Gini index is not less than the first threshold value; in the case that the forward milling Gini index and the forward milling absolute average value are both taken as the second threshold value, the filtered Gini index and the filtered absolute average value are both not less than the second threshold value means that the filtered Gini index is not less than the forward milling Gini index and the filtered absolute average value is not less than the forward milling absolute average value, in the case that the forward milling Gini index and the forward milling absolute average value are taken for weighted calculation, the filtered Gini index and the filtered absolute average value are both not less than the second threshold value means that the calculation result of the weighted calculation of the filtered Gini index and the filtered absolute average value is not less than the second threshold value; in the case that the reverse milling Gini index and the reverse milling absolute average value are both taken as the third threshold value, the filtered Gini index and the filtered absolute average value are both not less than the third threshold value means that the filtered Gini index is not less than the reverse milling Gini index and the filtered absolute average value is not less than the reverse milling absolute average value, in the case that the reverse milling Gini index and the reverse milling absolute average value are taken for weighted calculation, the filtered Gini index and the filtered absolute average value are both not less than the third threshold value means that the calculation result of the weighted calculation of the filtered Gini index and the filtered absolute average value is not less than the third threshold value.

[0118] In actual application, the time length of the forward milling and the reverse milling is the same, and the milling is processed in the order of forward milling-reverse milling, so that when the tool is determined to be in the milling state for the first time, the tool is in the forward milling state, when the tool is determined to be in the milling state for the second time, the tool is in the reverse milling state, when the tool is determined to be in the milling state for the third time, the tool is in the forward milling state, and when the tool is determined to be in the milling state for the fourth time, the tool is in the reverse milling state, so as to determine the forward milling or reverse milling state of the tool.

[0119] In the embodiment of the present application, the tool state monitoring method further comprises:

[0120] Step 7, determining whether the state of the target tool is the assumed wear state, if yes, returning to execute step 4 until the state of the target tool is determined to be the assumed wear state for three times in succession, then determining that the state of the target tool is the serious wear state.

[0121] In practical applications, when the filter Gini index and the absolute average value of the filter are both not less than the first threshold or the second threshold for the first time, in order to prevent false judgments, the target tool is only judged to be in a severely worn state after three consecutive tests in which the target tool is judged to be in a hypothetical wear state. If the target tool is not judged to be in a hypothetical wear state for three consecutive tests, it means that the previous one or two judgments of the wear state were false alarms, thus ensuring the accuracy of tool condition monitoring.

[0122] To verify the accuracy of the tool condition monitoring method, this embodiment of the invention also conducted a tool wear milling experiment. As shown in Table 1, seven experimental groups with different milling parameters were set up. The workpiece material was Cr12 with a hardness of 56-58 HRC and dimensions of 400mm x 400mm x 30mm. Each climb and conventional milling pass was recorded as one cut, and the length of one climb and conventional milling pass was 400mm. The tool holder model was EMR C20-5R20-150, and the radial milling depth was set to 80% of the tool holder diameter, i.e., 14mm. The tool shank model was BT40-ER32-100L. Two carbide cutting teeth were symmetrically mounted on the tool holder for milling. A sensor for collecting vibration signals was installed on the tool shank.

[0123] Table 1

[0124]

[0125] In the seven experimental groups, the axial cutting depth, spindle speed and feed rate were changed respectively. Vibration signals during the milling process were collected by sensors installed on the tool holder, and the wear VB value of the cutter tooth flank was measured after the experiment.

[0126] like Figure 9 As shown, the results of the 7 experimental groups are presented. In this embodiment, the tool status monitoring method was run on a laptop computer configured with an Intel(R) Core(TM) i7-9750H CPU@2.60GHz. The average processing time for each data packet was only 0.013s, which meets the requirements for online tool monitoring during the operation of the milling machine. Figure 8 The recall rate is calculated as Recall = TP / (TP+FN), and the precision rate as Precision = TP / (TP+FP), where TP represents actual severe wear that was classified as severe wear; FP represents actual severe wear that was classified as normal wear; and FN represents actual normal wear that was classified as severe wear. Figure 9 As can be seen, the recall rate of the embodiments of the present invention is above 74.9%, and the accuracy rate is above 81.6%, which can accurately monitor and identify the tool status.

[0127] The tool state monitoring method of the application, by establishing the dynamics equation of tool milling processing, and according to the dynamics equation, simulation data is obtained, the filtering frequency band and the target feature are determined through the simulation data, the dependence on historical data is reduced, without a large amount of milling processing data, then the filtering frequency band and the target feature are used to filter and calculate the experimental data of tool milling processing to determine the threshold value of judging the tool state, after the vibration data of the target tool is collected, the vibration data is filtered and calculated according to the filtering frequency band and the target feature to obtain the vibration characteristic value corresponding to the target feature, the vibration characteristic value is compared with the corresponding threshold value to obtain a comparison result, the state of the target tool is judged according to the comparison result, and the real-time monitoring of the tool state is realized.

[0128] As shown in Figure 10 The application also provides a tool state monitoring system, comprising:

[0129] The pre-preparation module 10 is used for pre-establishing the dynamics equation of tool milling processing, and pre-acquiring the experimental data of tool milling processing;

[0130] The simulation analysis module 20 is used for vibration simulation according to the dynamics equation, obtaining simulation data, and analyzing the simulation data to determine the filtering frequency band and the target feature;

[0131] The threshold value determination module 30 is used for filtering the experimental data according to the filtering frequency band to obtain experimental filtered data, calculating the experimental characteristic value of the target feature corresponding to the experimental data and the experimental filtered data, and taking the experimental characteristic value as the threshold value;

[0132] The data acquisition module 40 is used for collecting the vibration data of the target tool;

[0133] The data processing module 50 is used for filtering the vibration data according to the filtering frequency band to obtain vibration filtered data, and calculating the vibration characteristic value of the target feature corresponding to the vibration data and the vibration filtered data;

[0134] The comparison and judgment module 60 is used for comparing the vibration characteristic value with the corresponding threshold value to obtain a comparison result, and judging the state of the target tool according to the comparison result.

[0135] The specific structure of the tool state monitoring system can be referred to the structure of the tool state monitoring method in the above, which will not be repeated here. The modules of the tool state monitoring system can be realized by software, hardware and their combinations. The modules can be embedded in or independent of the processor in the computer device in hardware form, or stored in the memory in the computer device in software form, so that the processor can call and execute the operations of the modules.

[0136] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 11 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The network interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a tool condition monitoring method.

[0137] Those skilled in the art will understand that Figure 11 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a configuration of the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0138] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to perform the following steps:

[0139] Step 1: Establish the dynamic equations of the milling process in advance and obtain experimental data of the milling process in advance;

[0140] Step 2: Perform vibration simulation based on the dynamic equations to obtain simulation data, and analyze the simulation data to determine the filter frequency band and target characteristics;

[0141] Step 3: Filter the experimental data according to the filtering frequency band to obtain experimental filtered data, calculate the experimental feature value of the target feature corresponding to the experimental data and the experimental filtered data, and use the experimental feature value as the threshold.

[0142] Step 4: Collect vibration data of the target cutting tool;

[0143] Step 5: Filter the vibration data according to the filtering frequency band to obtain vibration filtered data, and calculate the vibration feature value of the target feature corresponding to the vibration data and vibration filtered data;

[0144] Step 6: Compare the vibration characteristic value with the corresponding threshold to obtain the comparison result, and determine the state of the target tool based on the comparison result.

[0145] In one embodiment, the kinetic equation is:

[0146]

[0147] wherein M is a two-dimensional mass matrix of the tool, C is a two-dimensional damping matrix of the tool, K is a two-dimensional stiffness matrix of the tool, is the vibration acceleration, is the vibration velocity, q(t) is the vibration displacement, F s,st (t) is the static shear force, F s,dy (t) is the dynamic shear force, F p (t) is the instantaneous impact force caused by the tool cutting into the workpiece, F e,st (t) is the static plowing force, F e,da (t) is the dynamic plowing force.

[0148] In one embodiment, the calculation formulas of the static shear force, the dynamic shear force, the instantaneous impact force, the static plowing force and the dynamic plowing force are respectively:

[0149]

[0150]

[0151]

[0152]

[0153]

[0154] wherein N z represents the number of axial force units, N represents the number of tool teeth, i represents the i-th axial force unit, j represents the j-th tool tooth, g(Φ i,j (t)) represents a first window function, T(Φ i,j (t)) represents a coordinate conversion matrix, K ts represents a tangential shear force coefficient, K rs represents a radial shear force coefficient, K te represents a tangential plowing force coefficient, K re represents a radial plowing force coefficient, h s represents a static milling thickness, h r,j represents a radial dynamic milling thickness of the j-th tool tooth, h r,j-1 represents a radial dynamic milling thickness of the j-1-th tool tooth, S st represents a tool wear static area, S dy represents a tool wear dynamic area, w(Φ i,j (t)) represents a second window function, K a represents a stiffness coefficient, δ represents a deformation of the tool, e represents a penetration depth index, c pa damping coefficient representing an instantaneous impact force, an angle of attack representing an angle of a tool in a milling process.

[0155] In one embodiment, the analyzing the simulation data to determine a filtering frequency band and a target feature comprises:

[0156] selecting a frequency band with the highest energy in the simulation data as a first filtering frequency band, and selecting a frequency band with the lowest energy in the simulation data as a second filtering frequency band;

[0157] calculating a value of a simulation feature corresponding to the simulation data, the simulation feature comprising: an absolute average value, a root mean square value, a standard deviation, a pulse factor, a kurtosis, a skewness, and a Gini index;

[0158] selecting a first target feature from the simulation feature, the first target feature being any one of the absolute average value, the root mean square value, and the variance of the simulation feature related to a tool wear state;

[0159] selecting a second target feature from the simulation feature, the second target feature being the Gini index of the simulation feature that can best express an impact signal in a milling process.

[0160] In one embodiment, the filtering experimental data according to the filtering frequency band to obtain experimental filtering data, and calculating an experimental feature value of the target feature corresponding to the experimental data and the experimental filtering data, and taking the experimental feature value as a threshold value, comprises:

[0161] filtering experimental data according to the first filtering frequency band to obtain first filtering data, and filtering experimental data according to the second filtering frequency band to obtain second filtering data;

[0162] calculating an absolute average value of the experimental data of the tool in a non-milling state and a Gini index of the first filtering data, to obtain a non-milling Gini index and a non-milling absolute average value, and determining a first threshold value according to the non-milling absolute average value and the non-milling Gini index;

[0163] calculating a Gini index of the first filtering data and an absolute average value of the second filtering data of the tool in a down-milling process and a hypothetical severe wear state, to obtain a down-milling Gini index and a down-milling absolute average value, and determining a second threshold value according to the down-milling Gini index and the down-milling absolute average value;

[0164] calculating a Gini index of the first filtering data and an absolute average value of the second filtering data of the tool in an up-milling process and a hypothetical severe wear state, to obtain an up-milling Gini index and an up-milling absolute average value, and determining a third threshold value according to the up-milling Gini index and the up-milling absolute average value.

[0165] In one embodiment, the vibration feature values include: an original absolute average value, a filtered Gini index and a filtered absolute average value, the original absolute average value being an absolute average value of vibration data, the filtered Gini index being a Gini index of vibration filtered data filtered by a first filter frequency band, and the filtered absolute average value being an absolute average value of vibration filtered data filtered by a second filter frequency band.

[0166] The comparison of the vibration feature values with the corresponding threshold values obtains a comparison result, and the state of the target tool is determined according to the comparison result, including:

[0167] The original absolute average value and the filtered Gini index are compared with a first threshold value, and in a case where both the original absolute average value and the filtered Gini index are not less than the first threshold value, it is determined that the target tool is in a milling state, otherwise, it is determined that the target tool is in a non-milling state.

[0168] In a case where the target tool is in a forward milling processing state, the filtered Gini index and the filtered absolute average value are compared with a second threshold value, and in a case where both the filtered Gini index and the filtered absolute average value are not less than the second threshold value, it is determined that the target tool is in a supposed wear state, otherwise, it is determined that the target tool is in a normal wear state.

[0169] In a case where the target tool is in a reverse milling processing state, the filtered Gini index and the filtered absolute average value are compared with a third threshold value, and in a case where both the filtered Gini index and the filtered absolute average value are not less than the third threshold value, it is determined that the target tool is in a supposed wear state, otherwise, it is determined that the target tool is in a normal wear state.

[0170] In one embodiment, further comprising:

[0171] Step 7, determining whether the state of the target tool is a supposed wear state, if yes, returning to execute step 4 until the state of the target tool is determined to be a supposed wear state for three times in succession, and then determining that the state of the target tool is a serious wear state.

[0172] In one embodiment, a computer readable storage medium is provided, and a computer program is stored on the computer readable storage medium, and the computer program is executed by a processor to implement the following steps:

[0173] Step 1, a dynamic equation of tool milling processing is established in advance, and experimental data of tool milling processing is obtained in advance;

[0174] Step 2, vibration simulation is performed according to the dynamic equation to obtain simulation data, and the simulation data is analyzed to determine a filter frequency band and a target feature;

[0175] Step 3, filtering the experimental data according to the filter frequency band to obtain filtered experimental data, calculating the experimental characteristic value of the target feature corresponding to the experimental data and the filtered experimental data, and taking the experimental characteristic value as a threshold value;

[0176] Step 4, collecting vibration data of the target tool;

[0177] Step 5, filtering the vibration data according to the filter frequency band to obtain vibration filtered data, and calculating the vibration characteristic value of the target feature corresponding to the vibration data and the vibration filtered data;

[0178] Step 6, comparing the vibration characteristic value with the corresponding threshold value to obtain a comparison result, and judging the state of the target tool according to the comparison result.

[0179] In one embodiment, the dynamic equation is:

[0180]

[0181] Wherein, M is a two-dimensional mass matrix of the tool, C is a two-dimensional damping matrix of the tool, K is a two-dimensional stiffness matrix of the tool, is a vibration acceleration, is a vibration velocity, q(t) is a vibration displacement, F s,st (t) is a static shear force, F s,dy (t) is a dynamic shear force, F p (t) is an instantaneous impact force caused by the tool cutting into the workpiece, F e,st (t) is a static plowing force, F e,da (t) is a dynamic plowing force.

[0182] In one embodiment, the calculation formulas of the static shear force, the dynamic shear force, the instantaneous impact force, the static plowing force and the dynamic plowing force are respectively:

[0183]

[0184]

[0185]

[0186]

[0187]

[0188] Wherein, N z represents the number of axial force units, N represents the number of tool teeth, i represents the i-th axial force unit, j represents the j-th tool tooth, g(Φ i,j (t)) represents a first window function, T(Φ i,j(t)) represents a coordinate transformation matrix, K ts represents a tangential shear force coefficient, K rs represents a radial shear force coefficient, K te represents a tangential ploughing force coefficient, K re represents a radial ploughing force coefficient, h s represents a static milling thickness, h r,j represents a radial dynamic milling thickness of the jth cutter tooth, h r,j-1 represents a radial dynamic milling thickness of the j-1th cutter tooth, S st represents a cutter wear static area, S dy represents a cutter wear dynamic area, w(Φ i,j (t)) represents a second window function, K a represents a stiffness coefficient, δ represents a deformation of the cutter, e represents a penetration depth index, c p represents a damping coefficient of an instantaneous impact force, represents an angle of entry of the cutter during milling.

[0189] In one embodiment, the analyzing the simulation data determines a filtering frequency band and a target feature, including:

[0190] selecting a frequency band with the highest energy in the simulation data as a first filtering frequency band, and selecting a frequency band with the lowest energy in the simulation data as a second filtering frequency band;

[0191] calculating a value of a simulation feature corresponding to the simulation data, the simulation feature including: an absolute average value, a root mean square value, a standard deviation, a pulse factor, a kurtosis, a skewness, and a Gini index;

[0192] selecting a first target feature from the simulation feature, the first target feature being any one of the simulation feature of the absolute average value, the root mean square value, and the variance related to the state of the cutter wear in the simulation feature;

[0193] selecting a second target feature from the simulation feature, the second target feature being the Gini index in the simulation feature that can best express the impact signal in the milling process.

[0194] In one embodiment, the filtering experimental data according to the filtering frequency band to obtain experimental filtering data, calculating an experimental feature value of the target feature corresponding to the experimental data and the experimental filtering data, taking the experimental feature value as a threshold, including:

[0195] filtering experimental data according to the first filtering frequency band to obtain first filtering data, and filtering experimental data according to the second filtering frequency band to obtain second filtering data;

[0196] calculating a Gini index of the first filtered data and an absolute average of the second filtered data of the tool in the non-milling state to obtain a non-milling Gini index and a non-milling absolute average, and determining the first threshold value according to the non-milling absolute average and the non-milling Gini index;

[0197] calculating a Gini index of the first filtered data and an absolute average of the second filtered data of the tool in the non-milling state to obtain a non-milling Gini index and a non-milling absolute average, and determining the first threshold value according to the non-milling absolute average and the non-milling Gini index;

[0198] calculating a Gini index of the first filtered data and an absolute average of the second filtered data of the tool in the non-milling state to obtain a non-milling Gini index and a non-milling absolute average, and determining the first threshold value according to the non-milling absolute average and the non-milling Gini index.

[0199] In one embodiment, the vibration feature values include: an original absolute average, a filtered Gini index and a filtered absolute average, the original absolute average being an absolute average of vibration data, the filtered Gini index being a Gini index of vibration filtered data filtered by a first filter frequency band, and the filtered absolute average being an absolute average of vibration filtered data filtered by a second filter frequency band.

[0200] The comparison of the vibration feature values with the corresponding threshold values obtains a comparison result, and the state of the target tool is determined according to the comparison result, including:

[0201] The original absolute average and the filtered Gini index are compared with the first threshold value, and in a case where both the original absolute average and the filtered Gini index are not less than the first threshold value, it is determined that the target tool is in a milling state, otherwise, it is determined that the target tool is in a non-milling state.

[0202] In a case where the target tool is in a forward milling state, the filtered Gini index and the filtered absolute average are compared with the second threshold value, and in a case where both the filtered Gini index and the filtered absolute average are not less than the second threshold value, it is determined that the target tool is in a presumed wear state, otherwise, it is determined that the target tool is in a normal wear state.

[0203] In a case where the target tool is in a reverse milling state, the filtered Gini index and the filtered absolute average are compared with the third threshold value, and in a case where both the filtered Gini index and the filtered absolute average are not less than the third threshold value, it is determined that the target tool is in a presumed wear state, otherwise, it is determined that the target tool is in a normal wear state.

[0204] In one embodiment, further comprising:

[0205] Step 7, judging whether the state of the target tool is the assumed wear state, if yes, returning to execute step 4 until the state of the target tool is judged as the assumed wear state for three times in succession, then judging that the state of the target tool is the serious wear state.

[0206] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiment methods can be included. Any reference to memory, storage, database or other medium used in the embodiments provided in the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0207] The technical features of the above embodiments can be combined in any manner. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combinations of the technical features do not contradict, they should be considered as the scope of the present disclosure.

Claims

1. A tool condition monitoring method, characterized by, The method comprises the following steps: Step 1, pre-establishing a dynamic equation of tool milling, and pre-acquiring experimental data of tool milling; Step 2, performing vibration simulation according to the dynamic equation to obtain simulation data, and analyzing the simulation data to determine a filtering frequency band and a target feature; Step 3, filtering experimental data according to the filtering frequency band to obtain experimental filtered data, calculating experimental characteristic values of the target feature corresponding to the experimental data and the experimental filtered data, and taking the experimental characteristic values as threshold values; Step 4, collecting vibration data of a target tool; Step 5, filtering the vibration data according to the filtering frequency band to obtain vibration filtered data, and calculating vibration characteristic values of the target feature corresponding to the vibration data and the vibration filtered data; Step 6, comparing the vibration characteristic values with corresponding threshold values to obtain a comparison result, and judging a state of the target tool according to the comparison result. The analyzing the simulation data to determine the filtering frequency band and the target feature comprises the following steps: selecting a frequency band with the highest energy in the simulation data as a first filtering frequency band, and selecting a frequency band with the lowest energy in the simulation data as a second filtering frequency band; calculating values of simulation features corresponding to the simulation data, wherein the simulation features comprise absolute average value, root mean square value, standard deviation, pulse factor, kurtosis, skewness and Gini index; selecting a first target feature from the simulation features, wherein the first target feature is any one of the simulation features of the absolute average value, the root mean square value and the variance in the simulation features related to the tool wear state; selecting a second target feature from the simulation features, wherein the second target feature is the Gini index in the simulation features that can best express impact signals in the milling process.

2. The tool condition monitoring method according to claim 1, characterized in that The dynamic equation is as follows: Wherein, the M is two-dimensional mass matrix of the tool, C is two-dimensional damping matrix of the tool, K is two-dimensional stiffness matrix of the tool, is the vibration acceleration, is the vibration velocity, q(t) is the vibration displacement, F s,st (t) is the static shear force, F s,dy (t) is the dynamic shear force, F p (t) is the instantaneous impact force caused by the tool cutting into the workpiece, F e,st (t) is the static plowing force, F e,da (t) is the dynamic plowing force.

3. The tool condition monitoring method according to claim 2, characterized in that The calculation formulas of the static shear force, the dynamic shear force, the instantaneous impact force, the static plowing force and the dynamic plowing force are as follows: where N z represents the number of axial force units, N represents the number of cutter teeth, i represents the i-th axial force unit, j represents the j-th cutter tooth, g(Φ i,j (t)) represents a first window function, T(Φ i,j (t)) represents a coordinate conversion matrix, K ts represents a tangential shear force coefficient, K rs represents a radial shear force coefficient, K te represents a tangential plowing force coefficient, K re represents a radial plowing force coefficient, h s represents a static milling thickness, h r,j represents a radial dynamic milling thickness of the j-th cutter tooth, h r,j-1 represents a radial dynamic milling thickness of the j-1-th cutter tooth, S st represents a cutter wear static area, S dy represents a cutter wear dynamic area, w(Φ i,j (t)) represents a second window function, K a represents a stiffness coefficient, δ represents a deformation of the cutter, e represents a penetration depth index, c p represents a damping coefficient of an instantaneous impact force, represents an angle of entry of the cutter during milling.

4. The tool condition monitoring method according to claim 1, characterized in that, The filtering experimental data according to the filtering frequency band, the calculating the experimental characteristic values of the target feature corresponding to the experimental data and the experimental filtered data, and the taking the experimental characteristic values as threshold values comprise the following steps: filtering the experimental data according to the first filtering frequency band to obtain first filtered data, and filtering the experimental data according to the second filtering frequency band to obtain second filtered data; calculating the absolute average value of the experimental data and the Gini index of the first filtered data of the tool in a non-milling state to obtain a non-milling Gini index and a non-milling absolute average value, and determining a first threshold value according to the non-milling absolute average value and the non-milling Gini index; calculating the Gini index of the first filtered data and the absolute average value of the second filtered data of the tool in a down-milling process and in a hypothetical severe wear state to obtain a down-milling Gini index and a down-milling absolute average value, and determining a second threshold value according to the down-milling Gini index and the down-milling absolute average value; calculating the Gini index of the first filtered data and the absolute average value of the second filtered data of the tool in an up-milling process and in a hypothetical severe wear state to obtain an up-milling Gini index and an up-milling absolute average value, and determining a third threshold value according to the up-milling Gini index and the up-milling absolute average value.

5. A tool condition monitoring method according to claim 4, characterised in that, The vibration characteristic values include: an original absolute average value, a filtered Gini index, and a filtered absolute average value, the original absolute value is an absolute average value of vibration data, the filtered Gini index is a Gini index of vibration filtered data filtered through a first filter frequency band, and the filtered absolute average value is an absolute average value of vibration filtered data filtered through a second filter frequency band; The comparison of the vibration characteristic values with corresponding threshold values obtains a comparison result, and the state of the target tool is determined according to the comparison result, including: The original absolute average value and the filtered Gini index are compared with a first threshold value, and in the case that the original absolute average value and the filtered Gini index are both not less than the first threshold value, it is determined that the target tool is in a milling state, otherwise, it is determined that the target tool is in a non-milling state; In the case that the target tool is in a forward milling state, the filtered Gini index and the filtered absolute average value are compared with a second threshold value, and in the case that the filtered Gini index and the filtered absolute average value are both not less than the second threshold value, it is determined that the target tool is in a supposed wear state, otherwise, it is determined that the target tool is in a normal wear state; In the case that the target tool is in a reverse milling state, the filtered Gini index and the filtered absolute average value are compared with a third threshold value, and in the case that the filtered Gini index and the filtered absolute average value are both not less than the third threshold value, it is determined that the target tool is in a supposed wear state, otherwise, it is determined that the target tool is in a normal wear state.

6. The tool condition monitoring method according to claim 1, characterized in that, Further comprising: Step 7, determining whether the state of the target tool is a supposed wear state, if yes, returning to execute step 4 until the state of the target tool is determined as a supposed wear state for three times in succession, and then determining that the state of the target tool is a serious wear state.

7. A tool condition monitoring system characterised in that, Including: A pre-preparation module is configured to pre-establish a dynamic equation of tool milling and pre-acquire experimental data of tool milling; A simulation analysis module is configured to perform vibration simulation according to the dynamic equation to obtain simulation data, and analyze the simulation data to determine a filter frequency band and a target characteristic; A threshold value determination module is configured to filter experimental data according to the filter frequency band to obtain experimental filtered data, calculate experimental characteristic values of the target characteristic corresponding to the experimental data and the experimental filtered data, and take the experimental characteristic values as threshold values; A data acquisition module is configured to acquire vibration data of a target tool; A data processing module is configured to filter vibration data according to the filter frequency band to obtain vibration filtered data, and calculate vibration characteristic values of the target characteristic corresponding to the vibration data and the vibration filtered data; A comparison and determination module is configured to compare the vibration characteristic values with corresponding threshold values to obtain a comparison result, and determine the state of the target tool according to the comparison result; The analysis of the simulation data to determine a filter frequency band and a target characteristic includes: Selecting a frequency band with the highest energy in the simulation data as a first filter frequency band, and selecting a frequency band with the lowest energy in the simulation data as a second filter frequency band; calculating values of simulation features corresponding to the simulation data, the simulation features comprising: absolute average value, root mean square value, standard deviation, pulse factor, kurtosis, skewness and Gini index; selecting a first target feature from the simulation features, the first target feature being any one of the absolute average value, the root mean square value and the variance of the simulation features related to the tool wear state; selecting a second target feature from the simulation features, the second target feature being the Gini index of the simulation features most capable of expressing the impact signal in the milling process.

8. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor implements the steps of the method of any one of claims 1 to 6 when executing the computer program.

9. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program implements the steps of the method of any one of claims 1 to 6 when executed by the processor.

Images