Cutter wear state determination method and device, medium and terminal

Abstract

The invention discloses a method and a device for determining a tool wear state, a medium and a terminal, relates to the technical field of machine tool state monitoring, and mainly aims to solve the problem that the accuracy of a classification result of the tool wear state is reduced due to the change of machining parameters in the prior art. Comprising the following steps: acquiring a current effective value data set corresponding to a target cutter within a preset duration, predicting a three-way milling force value data set of the target cutter according to the current effective value data set based on a milling force prediction model of which model training is completed, and respectively calculating three-way milling force mean values; according to the three-way milling force mean value, based on a cutting angle change matrix corresponding to the target tool and current machining parameters, calculating a three-way cutting force coefficient mean value of the target tool within the preset duration; and determining the wear state of the target tool according to the mean value of the three-way cutting force coefficients.

Description

Technical Field

[0001] The present application relates to the technical field of machine tool status monitoring, and particularly to a method and device, medium, and terminal for determining the tool wear status. Background Art

[0002] Milling is a highly efficient machining method, often used for machining complex spatial planes, and is widely applied in industries such as aerospace, automotive manufacturing, mold engraving, and fluid machinery. During the milling process, due to the contact between the milling cutter and the workpiece, the extrusion and friction between the cutting edge and the machining plane will cause wear of the cutting edge, resulting in an increase in the interaction force between the milling cutter and the workpiece, that is, an increase in the cutting force. Furthermore, a large elastic deflection will occur at the cutting end, leading to a decrease in the geometric accuracy of the workpiece and an increase in the surface roughness of the machining plane. Therefore, in order to ensure that the geometric accuracy of the workpiece and the surface roughness of the machining plane meet the requirements, it is necessary to monitor the wear status of the milling cutter. In modern numerically controlled machine tools, there is a non-linear mapping relationship between the spindle motor current and the milling force. When the mechanical load on the motor changes, the servo controller can be used to change the motor current to overcome the load change. Therefore, the change in the milling force can be mapped through the spindle current signal, thereby realizing the monitoring of the tool wear status.

[0003] Currently, to monitor the tool wear status through the spindle current signal, first, a classification model for the tool wear status needs to be constructed, and the model is trained using historical current signals and the corresponding tool wear status. During actual monitoring, the current signal is input into the trained classification model to determine the tool wear status, thereby realizing the monitoring of the tool wear status.

[0004] However, when the machining parameters during milling change, the above classification model will no longer be applicable to the updated machining parameters. Even if the model is trained using training data under different machining parameters during model training, all machining parameters cannot be covered. Therefore, the accuracy of the classification result of the tool wear status will be reduced, and further, the accuracy of the monitoring will be reduced. Summary of the Invention

[0005] In view of this, the present application provides a method and device, medium, and terminal for determining the tool wear status, mainly aiming at the problem of the reduced accuracy of the classification result of the tool wear status caused by the change of machining parameters.

[0006] According to one aspect of the present application, a method for determining the tool wear status is provided, including:

[0007] Obtain the effective current value data set corresponding to the target tool within a preset time period. Based on the milling force prediction model with completed model training, predict the three-direction milling force value data set of the target tool according to the effective current value data set, and calculate the average value of the three-direction milling force respectively. The milling force prediction model is obtained by training with the training effective current values under different processing parameters and the corresponding training three-direction milling force values;

[0008] According to the average value of the three-direction milling force, based on the cutting angle change matrix corresponding to the target tool and the current processing parameters, calculate the average value of the three-direction cutting force coefficients of the target tool within the preset time period;

[0009] Determine the wear state of the target tool according to the average value of the three-direction cutting force coefficients.

[0010] Preferably, the determining the wear state of the target tool according to the average value of the three-direction cutting force coefficients includes:

[0011] If any one of the average values of the three-direction cutting force coefficients exceeds the preset cutting force coefficient threshold, it is determined that the wear state of the target tool is severe wear.

[0012] Preferably, the calculating the average value of the three-direction cutting force coefficients of the target tool within the preset time period according to the average value of the three-direction milling force, based on the cutting angle change matrix corresponding to the target tool and the current processing parameters, includes:

[0013] Substitute the average value of the three-direction milling force into the following formula to calculate the average value of the three-direction cutting force coefficients,

[0014]

[0015] Wherein, represents the average value of the tangential cutting force coefficient, represents the average value of the radial cutting force coefficient, represents the average value of the axial cutting force coefficient, N represents the number of tool teeth, a p represents the cutting depth in the processing parameters, f e represents the equivalent feed rate in the processing parameters, θ represents the tool cutting angle, θ en represents the angle when the tool participates in cutting, θ ex represents the angle when the tool leaves cutting, represents the average value of the milling force in the X direction, represents the average value of the milling force in the Y direction, represents the average value of the milling force in the Z direction.

[0016] Preferably, before the milling force prediction model based on the completed model training predicts the three-direction milling force value data set of the target tool according to the effective current value data set, the method further includes:

[0017] Construct an initial milling force prediction model;

[0018] Obtain a training data set, where the training data set includes training effective current values under different machining parameters and corresponding training three-direction milling force values;

[0019] Based on the initial milling force prediction model, predict the corresponding predicted three-direction milling force value according to each training effective current value;

[0020] Calculate the loss function between the predicted three-direction milling force value and the training three-direction milling force value, and perform a primary optimization on the model parameters of the initial milling force prediction model based on the loss function to obtain an intermediate milling force prediction model;

[0021] Obtain a validation data set, and perform a secondary optimization on the model parameters of the intermediate milling force prediction model based on the validation data set to obtain a milling force prediction model with completed model training.

[0022] Preferably, before obtaining the training data set, the method further includes:

[0023] Use a current sensor to collect three-phase spindle current signals of a training tool under different machining parameters, and use a force sensor to synchronously collect the corresponding training three-direction milling force values;

[0024] Perform current signal fusion on each of the three-phase spindle current signals to obtain multiple training effective current values;

[0025] Map each training effective current value to the synchronously collected training three-direction milling force value to obtain multiple sets of mapped training data;

[0026] Divide the multiple sets of mapped training data into a training data set and a validation data set according to a preset ratio, so as to perform a primary optimization on the initial milling force prediction model based on the training data set and perform a secondary optimization on the intermediate milling force prediction model based on the validation data set.

[0027] Preferably, before performing current signal fusion on each of the three-phase spindle current signals, the method further includes:

[0028] Perform first data normalization processing on each of the three-phase spindle current signals to obtain the three-phase spindle current signals after the first normalization processing, so as to perform current signal fusion based on the three-phase spindle current signals after the first normalization processing;

[0029] Perform a second data normalization process on each of the training three-way milling force values to obtain the training three-way milling force values after the second normalization process, and generate mapping training data based on the training three-way milling force values after the second normalization process.

[0030] Preferably, the current machining parameters include cutting depth and equivalent feed rate.

[0031] According to another aspect of the present application, a device for determining the tool wear state is provided, including:

[0032] A milling force prediction module, configured to obtain a group of effective current value data corresponding to a target tool within a preset time period, predict a group of three-way milling force value data of the target tool based on a trained milling force prediction model, and calculate the average value of the three-way milling force respectively. The milling force prediction model is obtained by training using the training effective current values under different machining parameters and the corresponding training three-way milling force values;

[0033] A cutting force coefficient calculation module, configured to calculate the average value of the three-way cutting force coefficients of the target tool within the preset time period based on the average value of the three-way milling force, the cutting angle change matrix corresponding to the target tool, and the current machining parameters;

[0034] A tool wear state determination module, configured to determine the wear state of the target tool according to the average value of the three-way cutting force coefficients.

[0035] Preferably, the tool wear state determination module is configured to:

[0036] If any one of the average values of the three-way cutting force coefficients exceeds a preset cutting force coefficient threshold, it is determined that the wear state of the target tool is severe wear.

[0037] Preferably, the cutting force coefficient calculation module is configured to:

[0038] Substitute the average value of the three-way milling force into the following formula to calculate the average value of the three-way cutting force coefficients,

[0039]

[0040] where, represents the average value of the tangential cutting force coefficient, represents the average value of the radial cutting force coefficient, represents the average value of the axial cutting force coefficient, N represents the number of tool teeth, a p represents the cutting depth in the machining parameters, f e represents the equivalent feed rate in the machining parameters, θ represents the tool cutting angle, θ en represents the angle when the tool participates in cutting, θex Indicates the angle when the cutting tool leaves the cutting. Indicates the average milling force in the X direction. Indicates the average milling force in the Y direction. Indicates the average milling force in the Z direction.

[0041] Preferably, before the milling force prediction module, the device further includes a model training module for:

[0042] Construct an initial milling force prediction model;

[0043] Obtain a training data set, which includes the effective value of the training current under different processing parameters and the corresponding training three-way milling force values;

[0044] Based on the initial milling force prediction model, predict the corresponding predicted three-way milling force values according to each effective value of the training current;

[0045] Calculate the loss function between the predicted three-way milling force values and the training three-way milling force values, and perform a primary optimization on the model parameters of the initial milling force prediction model based on the loss function to obtain an intermediate milling force prediction model;

[0046] Obtain a validation data set, and perform a secondary optimization on the model parameters of the intermediate milling force prediction model based on the validation data set to obtain a milling force prediction model with completed model training.

[0047] Preferably, the model training module is further used for:

[0048] Use a current sensor to collect three-phase spindle current signals of the training tool under different processing parameters, and use a force sensor to synchronously collect the corresponding training three-way milling force values;

[0049] Fuse the three-phase spindle current signals respectively to obtain multiple effective values of the training current;

[0050] Map each effective value of the training current to the synchronized training three-way milling force values to obtain multiple groups of mapped training data;

[0051] Divide the multiple groups of mapped training data into a training data set and a validation data set according to a preset ratio, so as to perform a primary optimization on the initial milling force prediction model based on the training data set, and perform a secondary optimization on the intermediate milling force prediction model based on the validation data set.

[0052] Preferably, the model training module is further used for:

[0053] Perform first data normalization processing on each of the three-phase spindle current signals to obtain the three-phase spindle current signals after the first normalization processing, so as to perform current signal fusion based on the three-phase spindle current signals after the first normalization processing;

[0054] Perform second data normalization processing on each of the training three-way milling force values to obtain the training three-way milling force values after the second normalization processing, so as to generate mapped training data based on the training three-way milling force values after the second normalization processing.

[0055] Preferably, the current machining parameters include cutting depth and equivalent feed rate.

[0056] According to another aspect of the present application, there is provided a storage medium storing at least one executable instruction, and the executable instruction causes a processor to perform operations corresponding to the method for determining the tool wear state as described above.

[0057] According to still another aspect of the present application, there is provided a terminal, including: a processor, a memory, a communication interface, and a communication bus, and the processor, the memory, and the communication interface complete communication with each other through the communication bus;

[0058] The memory is used to store at least one executable instruction, and the executable instruction causes the processor to perform operations corresponding to the method for determining the tool wear state as described above.

[0059] By means of the above technical solutions, the technical solutions provided by the embodiments of the present application have at least the following advantages:

[0060] The present application provides a method and device for determining the tool wear state, a medium, and a terminal. First, an effective current value data set corresponding to a target tool within a preset time period is obtained. Based on a trained milling force prediction model, the three-direction milling force value data set of the target tool is predicted according to the effective current value data set, and the average values of the three-direction milling forces are calculated respectively. The milling force prediction model is trained using training effective current values under different machining parameters and corresponding training three-direction milling force values. Secondly, according to the average values of the three-direction milling forces, based on the cutting angle change matrix corresponding to the target tool and the current machining parameters, the average values of the three-direction cutting force coefficients of the target tool within the preset time period are calculated. Finally, the wear state of the target tool is determined according to the average values of the three-direction cutting force coefficients. Compared with the prior art, in the embodiment of the present application, by using the milling force prediction model to predict the three-direction milling force value of the tool according to the effective current value, and then combining the cutting angle change matrix and the current machining parameters, the three-direction cutting force coefficients are calculated, and the wear state of the tool is determined according to the three-direction cutting force coefficients. Since the increase in the cutting force after tool wear is caused by the increase in the milling force coefficient, determining the wear state of the tool according to the cutting force coefficient can avoid the problem of reduced accuracy of the tool wear state classification result caused by changes in the milling machining parameters, effectively improving the accuracy of the tool wear state determination result. At the same time, by using the average value of the three-direction cutting force coefficients to determine the tool wear state, the possible errors in calculating the cutting force coefficients are eliminated, further improving the accuracy of the tool wear state determination result.

[0061] The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features, and advantages of the present application more obvious and understandable, the following specifically describes the embodiments of the present application. Brief Description of the Drawings

[0062] By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present application. And throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0063] Figure 1 Shows a flowchart of a method for determining the tool wear state provided by an embodiment of the present application;

[0064] Figure 2 Shows a flowchart of a training method for a milling force prediction model provided by an embodiment of the present application;

[0065] Figure 3The block diagram of a device for determining the tool wear state provided by an embodiment of the present application is shown;

[0066] Figure 4 The structural schematic diagram of a terminal provided by an embodiment of the present application is shown. Detailed implementation manners

[0067] Hereinafter, the exemplary embodiments of the present disclosure will be described in more detail with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art.

[0068] At the same time, it should be understood that, for the sake of description, the dimensions of the various parts shown in the drawings are not drawn in actual proportional relationship.

[0069] The following description of at least one exemplary embodiment is merely illustrative in nature and in no way limits the present application and its application or use.

[0070] Techniques, methods, and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the techniques, methods, and devices should be regarded as part of the specification.

[0071] It should be noted that like reference numerals and letters denote like items in the following drawings, and thus, once an item is defined in one drawing, it need not be further discussed in subsequent drawings.

[0072] The embodiments of the present application can be applied to a computer system / server, which can operate together with many other general-purpose or special-purpose computing system environments or configurations. Examples of well-known computing systems, environments, and / or configurations suitable for use with a computer system / server include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, minicomputer systems, mainframe computer systems, and distributed cloud computing technology environments including any of the above systems, and so on.

[0073] A computer system / server may be described in the general context of computer system-executable instructions, such as program modules, executed by a computer system. Generally, program modules may include routines, programs, object programs, components, logics, data structures, etc., which perform specific tasks or implement specific abstract data types. The computer system / server may be implemented in a distributed cloud computing environment where tasks are executed by remote processing devices linked through a communication network. In a distributed cloud computing environment, program modules may be located on local or remote computing system storage media including storage devices.

[0074] An embodiment of the present application provides a method for determining the tool wear state, as Figure 1 shown, the method includes:

[0075] 101. Obtain a group of effective current value data corresponding to a target tool within a preset time period, and based on a trained milling force prediction model, predict a group of three-direction milling force value data of the target tool according to the group of effective current value data, and calculate the average value of the three-direction milling force respectively.

[0076] Among them, the milling force prediction model is trained using the training effective current value under different machining parameters and the corresponding training three-direction milling force values. It should be noted that in the existing method, the classification model classifies the spindle current signal to determine the tool wear state. When the spindle current signal is a new type of signal (i.e., the machining parameters change), it is easy to cause inaccurate classification results. Even when the model is trained using training data under different machining parameters during model training, all machining parameters cannot be covered. However, the milling force prediction model in the embodiment of the present application learns the non-linear relationship between the effective current value and the milling force value during model training. Using data under different machining parameters for model training can increase the diversity of training data, thereby improving the prediction accuracy of the milling force value. Then, the milling force value is predicted using the model, and the milling force coefficient is calculated. The wear state is determined according to the milling force coefficient, eliminating the influence of machining parameters and improving the accuracy of the tool wear state determination result; the target tool may be a tool of a numerical control machine tool, such as a milling cutter, etc.; the effective current value can be obtained by collecting the three-phase spindle current signal using a current sensor and then performing current fusion; the average value of the three-direction milling force includes the average value of the tangential milling force, the average value of the radial milling force, and the average value of the axial milling force. It should be noted that the error that may exist when calculating the cutting force coefficient is eliminated. In the embodiment of the present application, the average value of the three-direction cutting force coefficient is used to determine the tool wear state. Therefore, the three-direction milling force also takes the average value. In the embodiment of the present application, the current execution end may be a tool wear state monitoring unit of a numerical control machine tool.

[0077] 102. Calculate the average value of the three-direction cutting force coefficients of the target tool within a preset time period based on the average value of the three-direction milling forces, the cutting angle change matrix corresponding to the target tool, and the current machining parameters.

[0078] Among them, the current machining parameters can include the cutting depth and the equivalent feed rate, and can be obtained through the machine tool interface using PLC network communication. In the embodiments of the present application, specifically, the average value of the three-direction milling forces calculated in step 101 of the embodiment can be substituted into the following formula to calculate the average value of the three-direction cutting force coefficients.

[0079]

[0080] Among them, represents the average value of the tangential cutting force coefficient. represents the average value of the radial cutting force coefficient. represents the average value of the axial cutting force coefficient, N represents the number of tool teeth, a p represents the cutting depth in the machining parameters, f e represents the equivalent feed rate in the machining parameters, θ represents the tool cutting angle, θ en represents the angle when the tool participates in cutting, θ ex represents the angle when the tool leaves the cutting. represents the average value of the milling force in the X direction. represents the average value of the milling force in the Y direction. represents the average value of the milling force in the Z direction.

[0081] It should be noted that first, establish the conversion formulas between the tangential force and the milling force in the X direction, the radial force and the milling force in the Y direction, and the axial force and the milling force in the Z direction of the tool.

[0082]

[0083] Among them, T represents the cutting angle change matrix, which can be expressed as the following formula.

[0084]

[0085] Among them, θ j,z represents the tool cutting angle, k f represents the instantaneous feed direction of the tool, which can be expressed as the following formula.

[0086]

[0087] Among them, f represents the instantaneous feed vector, and i represents the unit vector in the X direction.

[0088] The tool cutting angle θ j,z , which can be expressed as the following formula.

[0089] θ j,z = φ + k f - jΨ p - Ψ j,z

[0090] where φ represents the rotational angle of the cutting tool, Ψ p represents the tooth space angle, and j represents the height from the tool tip to the tooth edge.

[0091] Combined with the milling force expression and the conversion formulas between the tangential force, radial force, axial force and the milling forces in the X, Y, and Z directions, the conversion formula between the three-direction cutting force coefficients and the three-direction milling force values can be derived. It should be noted that in the formula derivation, since the machining method is straight-line machining, the instantaneous feed direction k of the cutting tool f takes the value of 0.

[0092] 103. Determine the wear state of the target tool according to the average value of the three-direction cutting force coefficients.

[0093] It should be noted that in milling, the wear amount of the cutting tool directly affects the cutting force. When the cutting tool is severely worn, the contact form between the cutting tool and the workpiece changes from point contact to surface contact. At the same time, the wear also causes an increase in the radius of the tool tip arc, increasing the frictional force between the cutting tool and the workpiece, and thus increasing the milling power. The additional cutting force caused by tool wear consists of two parts, namely the radial force F caused by the flank wear of the cutting tool NW and the frictional force F between the cutting tool and the workpiece caused by wear FW , which can be expressed by the following formula

[0094]

[0095] where H represents the Brinell hardness of the workpiece material, VB represents the wear amount of the flank of the cutting tool, s represents the length of the flank wear band of the cutting tool, and δ represents the sliding friction coefficient between the cutting tool and the workpiece.

[0096] When the cutting tool is worn, the tangential force F of the milling cutter t ′ can be expressed by the following formula

[0097] F t ′ = F t + F FW ,

[0098] where F t represents the tangential force when the cutting tool is not worn.

[0099] For face milling, cutting parameters such as the feed per tooth, milling width, and spindle speed will affect the change of milling force. The j-th edge of the tool is subjected to tangential force, axial force, and radial force during cutting. The angle between the instantaneous feed direction and the X-axis is k, and the milling force can be expressed by the following formula.

[0100]

[0101] Among them,

[0102]

[0103] h(θ j,z ) = f e sinθ j,z

[0104] Among them, dF t,j,z (θ j,z ) represents the tangential force of the tool at the cutting angle of θ j,z . dF r,j,z (θ j,z ) represents the radial force of the tool at the cutting angle of θ j,z . dF a,j,z (θ j,z ) represents the axial force of the tool at the cutting angle of θ j,z . K t represents the tangential milling force coefficient, K r represents the radial milling force coefficient, K a represents the axial milling force coefficient, μ(θ j,z ) represents the step function, h(θ j,z ) represents the cutting thickness of the tool at the cutting angle of θ j,z . f e represents the equivalent feed in the machining parameters.

[0105] According to the expression of the tangential force F t ′ of the milling cutter when the tool wears, F t ′ = F t + F FW , the expression of the milling force coefficient after tool wear can be deduced.

[0106]

[0107] Among them, K t ′ represents the tangential milling force coefficient after tool wear, K r ′ represents the radial milling force coefficient after tool wear, K a ′ represents the axial milling force coefficient after tool wear, ΔK t represents the increment of the tangential milling force coefficient caused by tool wear, ΔK rDenote the increment of the radial milling force coefficient caused by tool wear, ΔK a Denote the increment of the axial milling force coefficient caused by tool wear.

[0108] According to the tangential force F t ′ expression of the milling cutter during tool wear, the milling force expression, and the milling force coefficient expression after tool wear, the following formula can be derived,

[0109]

[0110] Based on this formula, it can be known that the increase in the cutting force after tool wear is caused by the increase in the milling force coefficient. Therefore, when the milling force coefficient increases to a certain threshold, it indicates that the target tool has suffered severe wear. Based on this, in the embodiments of the present application, a cutting force coefficient threshold can be preset in advance. When this threshold is exceeded, it is determined that the target tool has worn. The threshold can be the same threshold for three directions or different thresholds can be used respectively. In the embodiments of the present application, no specific limitation is made.

[0111] Compared with the prior art, in the embodiments of the present application, by using the milling force prediction model to predict the three-direction milling force values of the tool according to the effective current value, and then combining the cutting angle change matrix and the current machining parameters, the three-direction cutting force coefficients are calculated, and the wear state of the tool is determined according to the three-direction cutting force coefficients. Since the increase in the cutting force after tool wear is caused by the increase in the milling force coefficient, therefore, determining the wear state of the tool according to the cutting force coefficient can avoid the problem of reduced accuracy of the tool wear state classification result caused by the change of the milling machining parameters, effectively improving the accuracy rate of the tool wear state determination result; at the same time, by using the average value of the three-direction cutting force coefficients to determine the tool wear state, the possible errors in calculating the cutting force coefficients are eliminated, further improving the accuracy rate of the tool wear state determination result.

[0112] In an embodiment of the present application, for further limitation and explanation, step 103 of the embodiment determines the wear state of the target tool according to the average value of the three-direction cutting force coefficients, including: if any one of the average values of the three-direction cutting force coefficients exceeds the preset cutting force coefficient threshold, it is determined that the wear state of the target tool is severe wear.

[0113] Exemplarily, the three-direction cutting force coefficient thresholds are all set to the same threshold, such as 5000. When any one of the average values of the three-direction cutting force coefficients exceeds 5000, it is determined that the wear state of the target tool is severe wear.

[0114] In an embodiment of the present application, for further limitation and explanation, such as Figure 2As shown, before predicting the three - dimensional milling force value data set of the target tool according to the effective current value data set based on the completed milling force prediction model in step 101 of the embodiment, the method of the embodiment further includes:

[0115] 201. Construct an initial milling force prediction model.

[0116] Preferably, an initial milling force prediction model can be constructed based on a CNN - LSTM network model, which can include an encoding layer and a decoding layer. In the encoding layer, the CNN layer is used to extract the features of the effective current value, and the multi - faceted features of the effective current value are mixed into the convolutional body. Relying on the memory function of LSTM to bridge the time lag of the mapping relationship, the problem of unknown duration in the prediction relationship between the effective current value and the three - dimensional milling force value is solved; in the decoding layer,

[0117] a non - linear relationship between the effective current value and the three - dimensional milling force value is established through multiple fully - connected neural layers, and the three - dimensional milling force value is reconstructed through the characteristics of the effective current value. Exemplarily, the detailed settings of each layer of the initial milling force prediction model are shown in Table 1. The two convolutional layers and pooling layers respectively adopt the same input parameter settings, and the last layer is the predicted three - dimensional milling force value.

[0118] Table 1 Initial milling force prediction model parameters

[0119]

[0120] 202. Use a current sensor to collect the three - phase spindle current signals of the training tool under different machining parameters, and use a force sensor to synchronously collect the corresponding training three - dimensional milling force values; perform first - data normalization processing on each three - phase spindle current signal to obtain the first - normalized three - phase spindle current signal; perform second - data normalization processing on each training three - dimensional milling force value to obtain the second - normalized training three - dimensional milling force value.

[0121] Among them, the training tool is used to represent the tool used to collect the training data required for model training; the first - data normalization processing is used to represent the data normalization processing performed on all three - phase spindle current signals, which can include removing singular - point data and regularizing the data; the second - data normalization processing is used to represent the data normalization processing performed on all training three - dimensional milling force values, which can include removing singular - point data and regularizing the data;

[0122] In a specific application scenario, a CMV-850A vertical machining center can be used, in combination with a two-flute carbide flat-end end mill with a diameter of 12 mm and a helix angle of 30°, to perform machining operations on a workpiece made of 40Cr with dimensions of 90×120×70 mm. Its numerical control system is FANUC0I-MC, the spindle is driven by a synchronous belt with a transmission ratio of 1:1, and the spindle motor is a β12 / 8000i type three-phase four-pole asynchronous motor. An LT 108-S7 type closed-loop Hall effect current sensor is used to collect the three-phase spindle current signals of the training tool under different machining parameters, and a YDX-III9702 type three-component dynamometer is used to synchronously collect the corresponding training three-component milling force values. Different machining parameters are shown in Table 2,

[0123] Table 2 Machining Parameters

[0124]

[0125] Furthermore, the normalized data is processed by a sliding window, that is, the step size of the sliding window is set to 1 to establish the input data of the model.

[0126] 203. The three-phase spindle current signals after the first normalization processing are respectively subjected to current signal fusion to obtain multiple training current effective values; the respective training current effective values are mapped with the training three-component milling force values after the second normalization processing in synchronization to obtain multiple sets of mapped training data; according to a preset ratio, the multiple sets of mapped training data are divided into a training data set and a validation data set.

[0127] Furthermore, the initial milling force prediction model is optimized once based on the training data set, and the intermediate milling force prediction model is optimized twice based on the validation data set.

[0128] In the embodiment of the present application, the training current effective value and the training three-component milling force value after the second normalization processing in synchronization can form a set of mapped training data, and then all the mapped training data is divided into a training data set and a validation data set. Preferably, the preset ratio can be 60% training data set, 20% validation data set, and 20% test data set.

[0129] 204. Obtain the training data set; and optimize the model parameters of the initial milling force prediction model based on the training data set to obtain an intermediate milling force prediction model.

[0130] Among them, the training data set includes the training current effective values under different machining parameters and the corresponding training three-component milling force values; the first optimization is used to represent the model parameter optimization operation when the initial milling force prediction model is trained based on the training data set.

[0131] Specifically, based on the initial milling force prediction model, the corresponding predicted three-direction milling force values are predicted according to the effective values of each training current; the loss function between the predicted three-direction milling force values and the training three-direction milling force values is calculated, and the model parameters of the initial milling force prediction model are optimized once based on the loss function to obtain an intermediate milling force prediction model.

[0132] 205. Obtain a validation data set, and perform a secondary optimization on the model parameters of the intermediate milling force prediction model based on the validation data set to obtain a milling force prediction model with completed model training.

[0133] Among them, the validation data set includes the effective values of the validation current under different machining parameters and the corresponding validation three-direction milling force values; the secondary optimization is used to characterize the model parameter optimization operation when the intermediate milling force prediction model is trained based on the validation data set.

[0134] Specifically, based on the intermediate milling force prediction model, the corresponding predicted three-direction milling force values are predicted according to the effective values of each validation current; the loss function between the predicted three-direction milling force values and the validation three-direction milling force values is calculated, and the model parameters of the intermediate milling force prediction model are optimized twice based on the loss function to obtain a milling force prediction model with completed model training.

[0135] The present application provides a method for determining the tool wear state. First, a data set of the effective values of the current corresponding to the target tool within a preset time period is obtained. Based on the milling force prediction model with completed model training, the three-direction milling force value data set of the target tool is predicted according to the data set of the effective values of the current, and the average values of the three-direction milling forces are calculated respectively. The milling force prediction model is obtained by training the model using the effective values of the training current under different machining parameters and the corresponding training three-direction milling force values. Secondly, according to the average values of the three-direction milling forces, based on the cutting angle change matrix corresponding to the target tool and the current machining parameters, the average value of the three-direction cutting force coefficients of the target tool within the preset time period is calculated. Finally, the wear state of the target tool is determined according to the average value of the three-direction cutting force coefficients. Compared with the prior art, in the embodiment of the present application, by using the milling force prediction model to predict the three-direction milling force value of the tool according to the effective value of the current, and then combining the cutting angle change matrix and the current machining parameters, the three-direction cutting force coefficient is calculated, and the wear state of the tool is determined according to the three-direction cutting force coefficient. Since the increase in the cutting force after tool wear is caused by the increase in the milling force coefficient, therefore, determining the wear state of the tool according to the cutting force coefficient can avoid the problem of reduced accuracy of the tool wear state classification result caused by changes in the milling machining parameters, and effectively improve the accuracy rate of the tool wear state determination result; at the same time, by using the average value of the three-direction cutting force coefficient to determine the tool wear state, the error that may exist when calculating the cutting force coefficient is eliminated, and the accuracy rate of the tool wear state determination result is further improved.

[0136] Further, as an implementation of the method described above Figure 1 shown, an embodiment of the present application provides a device for determining the tool wear state, as Figure 3 shown, the device includes:

[0137] A milling force prediction module 31, a cutting force coefficient calculation module 32, and a tool wear state determination module 33;

[0138] The milling force prediction module 31 is configured to obtain a group of effective current value data corresponding to a target tool within a preset time period, predict a three-way milling force value data group of the target tool based on a trained milling force prediction model, and calculate the average value of the three-way milling force respectively. The milling force prediction model is trained using training effective current values under different processing parameters and corresponding training three-way milling force values;

[0139] The cutting force coefficient calculation module 32 is configured to calculate the average value of the three-way cutting force coefficients of the target tool within the preset time period according to the average value of the three-way milling force, based on the cutting angle change matrix corresponding to the target tool and the current processing parameters;

[0140] The tool wear state determination module 33 is configured to determine the wear state of the target tool according to the average value of the three-way cutting force coefficients.

[0141] In a specific application scenario, the tool wear state determination module is configured to:

[0142] If any one of the average values of the three-way cutting force coefficients exceeds a preset cutting force coefficient threshold, it is determined that the wear state of the target tool is severe wear.

[0143] In a specific application scenario, the cutting force coefficient calculation module is configured to:

[0144] Substitute the average value of the three-way milling force into the following formula to calculate the average value of the three-way cutting force coefficients,

[0145]

[0146] where, represents the average value of the tangential cutting force coefficient, represents the average value of the radial cutting force coefficient, represents the average value of the axial cutting force coefficient, N represents the number of tool teeth, a p represents the cutting depth in the processing parameters, f e represents the equivalent feed rate in the processing parameters, θ represents the tool cutting angle, θ en represents the angle when the tool participates in cutting, θ exRepresents the angle when the cutting tool leaves the cutting. Represents the average milling force in the X direction. Represents the average milling force in the Y direction. Represents the average milling force in the Z direction.

[0147] In a specific application scenario, before the milling force prediction module, the device further includes a model training module for:

[0148] Construct an initial milling force prediction model;

[0149] Obtain a training data set, which includes the effective value of the training current and the corresponding training three-way milling force values under different processing parameters;

[0150] Based on the initial milling force prediction model, predict the corresponding predicted three-way milling force values according to each effective value of the training current;

[0151] Calculate the loss function between the predicted three-way milling force values and the training three-way milling force values, and optimize the model parameters of the initial milling force prediction model once based on the loss function to obtain an intermediate milling force prediction model;

[0152] Obtain a validation data set, and perform a secondary optimization on the model parameters of the intermediate milling force prediction model based on the validation data set to obtain a milling force prediction model with completed model training.

[0153] In a specific application scenario, the model training module is further used for:

[0154] Use a current sensor to collect three-phase spindle current signals of the training tool under different processing parameters, and use a force sensor to synchronously collect the corresponding training three-way milling force values;

[0155] Fuse the three-phase spindle current signals respectively to obtain multiple effective values of the training current;

[0156] Map each effective value of the training current to the synchronized training three-way milling force values to obtain multiple groups of mapped training data;

[0157] Divide the multiple groups of mapped training data into a training data set and a validation data set according to a preset ratio, so as to perform a primary optimization on the initial milling force prediction model based on the training data set and a secondary optimization on the intermediate milling force prediction model based on the validation data set.

[0158] In a specific application scenario, the model training module is further used for:

[0159] Perform first data normalization processing on each of the three-phase spindle current signals to obtain the three-phase spindle current signals after the first normalization processing, so as to perform current signal fusion based on the three-phase spindle current signals after the first normalization processing;

[0160] Perform second data normalization processing on each of the training three-way milling force values to obtain the training three-way milling force values after the second normalization processing, so as to generate mapping training data based on the training three-way milling force values after the second normalization processing.

[0161] In a specific application scenario, the current machining parameters include cutting depth and equivalent feed rate.

[0162] This application provides a device for determining the tool wear state. First, obtain the effective current data group corresponding to the target tool within a preset time period. Based on the milling force prediction model that has completed model training, predict the three-way milling force value data group of the target tool according to the effective current data group, and calculate the average value of the three-way milling force respectively. The milling force prediction model is obtained by training the model using the training effective current under different machining parameters and the corresponding training three-way milling force values. Secondly, according to the average value of the three-way milling force, based on the cutting angle change matrix corresponding to the target tool and the current machining parameters, calculate the average value of the three-way cutting force coefficients of the target tool within the preset time period. Finally, determine the wear state of the target tool according to the average value of the three-way cutting force coefficients. Compared with the prior art, in the embodiment of this application, by using the milling force prediction model to predict the three-way milling force value of the tool according to the effective current, and then combining the cutting angle change matrix and the current machining parameters, calculate the three-way cutting force coefficient, and determine the wear state of the tool according to the three-way cutting force coefficient. Since the increase in cutting force after tool wear is caused by the increase in the milling force coefficient, therefore, determining the wear state of the tool according to the cutting force coefficient can avoid the problem of reducing the accuracy of the tool wear state classification result caused by the change of milling machining parameters, and effectively improve the accuracy rate of the tool wear state determination result; at the same time, by using the average value of the three-way cutting force coefficient to determine the tool wear state, the error that may exist when calculating the cutting force coefficient is eliminated, and the accuracy rate of the tool wear state determination result is further improved.

[0163] According to an embodiment of the present application, a storage medium is provided. The storage medium stores at least one executable instruction, and the computer executable instruction can execute the method for determining the tool wear state in any of the above method embodiments.

[0164] Based on such an understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including several instructions for causing a computer device (such as a personal computer, a server, or a network device, etc.) to execute the methods described in various implementation scenarios of the present application.

[0165] Figure 4 FIG. 4 shows a schematic structural diagram of a terminal provided according to an embodiment of the present application. The specific implementation of the terminal is not limited in the specific embodiments of the present application.

[0166] As Figure 4 shown, the terminal may include: a processor 402, a communication interface 404, a memory 406, and a communication bus 408.

[0167] Among them: the processor 402, the communication interface 404, and the memory 406 communicate with each other through the communication bus 408.

[0168] The communication interface 404 is used to communicate with network elements of other devices such as clients or other servers, etc.

[0169] The processor 402 is used to execute the program 410, and specifically can execute the relevant steps in the above-mentioned embodiment of the method for determining the tool wear state.

[0170] Specifically, the program 410 may include program code, and the program code includes computer operation instructions.

[0171] The processor 402 may be a central processing unit CPU, or a specific integrated circuit ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application. One or more processors included in the computer device may be of the same type of processor, such as one or more CPUs; or may be of different types of processors, such as one or more CPUs and one or more ASICs.

[0172] The memory 406 is used to store the program 410. The memory 406 may include a high-speed RAM memory, and may also include a non-volatile memory, such as at least one disk memory.

[0173] The program 410 is specifically used to cause the processor 402 to perform the following operations:

[0174] Obtain the effective current value data set corresponding to the target tool within a preset time period. Based on the milling force prediction model that has completed model training, predict the three-direction milling force value data set of the target tool according to the effective current value data set, and calculate the average value of the three-direction milling forces respectively. The milling force prediction model is obtained by using the training effective current values under different processing parameters and the corresponding training three-direction milling force values for model training;

[0175] According to the average value of the three-direction milling forces, based on the cutting angle change matrix corresponding to the target tool and the current processing parameters, calculate the average value of the three-direction cutting force coefficients of the target tool within the preset time period;

[0176] Determine the wear state of the target tool according to the average value of the three-direction cutting force coefficients.

[0177] The storage medium may further include an operating system and a network communication module. The operating system is a program for managing the entity device hardware and software resources of the above method for determining the tool wear state, and supports the operation of the information processing program and other software and / or programs. The network communication module is used to implement the communication between the components inside the storage medium, and the communication between other hardware and software in the information processing entity device.

[0178] Each embodiment in this specification is described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. The same or similar parts among the embodiments can be referred to each other. For the system embodiment, since it basically corresponds to the method embodiment, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment.

[0179] The methods and systems of the present application can be implemented in many ways. For example, the methods and systems of the present application can be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of the steps for the method is only for illustration, and the steps of the method of the present application are not limited to the specific order described above, unless otherwise specifically stated. In addition, in some embodiments, the present application can also be implemented as a program recorded in a recording medium, and these programs include machine-readable instructions for implementing the method according to the present application. Therefore, the present application also covers the recording medium storing the program for executing the method according to the present application.

[0180] Obviously, those skilled in the art should understand that the various modules or steps of the present application described above can be implemented by a general-purpose computing device. They can be concentrated on a single computing device or distributed on a network composed of multiple computing devices. Optionally, they can be implemented by program codes executable by the computing device. Thus, they can be stored in a storage device and executed by the computing device. And in some cases, the steps shown or described can be executed in a sequence different from that here, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module for implementation. In this way, the present application is not limited to any specific combination of hardware and software.

[0181] The foregoing is only a preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

Claims

1. A method for determining a tool wear state, characterized in that: include: Obtain a current effective value data group corresponding to a target tool within a preset time period, and based on a milling force prediction model that has completed model training, predict a three-dimensional milling force value data group of the target tool according to the current effective value data group, and calculate the mean of the three-dimensional milling force respectively, wherein the milling force prediction model is obtained by model training using the training current effective value under different processing parameters and the corresponding training three-dimensional milling force value; According to the three-dimensional milling force average value, based on the cutting angle change matrix corresponding to the target tool and the current machining parameters, the three-dimensional cutting force coefficient average value of the target tool within the preset time length is calculated; The wear state of the target tool is determined according to the mean values ​​of the three-way cutting force coefficients.

2. The method according to claim 1, characterized in that Determining the wear state of the target tool according to the mean value of the three-way cutting force coefficients includes: If any one of the three-way cutting force coefficient mean values ​​exceeds a preset cutting force coefficient threshold, the wear state of the target tool is determined to be severe wear.

3. The method according to claim 1, characterized in that The calculating, according to the three-way milling force mean value, based on the cutting angle change matrix corresponding to the target tool and the current machining parameters, the three-way cutting force coefficient mean value of the target tool within the preset time period includes: Substitute the three-dimensional milling force mean into the following formula to calculate the three-dimensional cutting force coefficient mean: in, represents the mean value of the tangential milling force coefficient, represents the mean value of radial milling force coefficient, represents the mean value of the axial milling force coefficient, N represents the number of tool teeth, a p Indicates the cutting depth in the machining parameters, f e represents the equivalent feed amount in the machining parameters, θ represents the tool cutting angle, θ en Indicates the angle at which the tool is involved in cutting, θ ex Indicates the angle at which the tool leaves the cutting surface. Indicates the mean value of milling force in X direction, Indicates the mean Y-axis milling force, F Z Indicates the mean value of the milling force in the Z direction.

4. The method according to claim 1, characterized in that Before the milling force prediction model based on the completed model training predicts the three-dimensional milling force value data group of the target tool according to the current effective value data group, the method further includes: Constructing the initial milling force prediction model; Acquire a training data set, wherein the training data set includes training current effective values ​​and corresponding training three-dimensional milling force values ​​under different machining parameters; Based on the initial milling force prediction model, predict corresponding three-dimensional milling force values ​​according to each training current effective value; Calculating a loss function between the predicted three-dimensional milling force value and the trained three-dimensional milling force value, and optimizing the model parameters of the initial milling force prediction model based on the loss function to obtain an intermediate milling force prediction model; A validation data set is obtained, and model parameters of the intermediate milling force prediction model are optimized secondary based on the validation data set to obtain a milling force prediction model for which model training has been completed.

5. The method according to claim 4, characterized in that Before obtaining the training data set, the method further includes: The current sensor is used to collect the three-phase spindle current signal of the training tool under different processing parameters, and the force sensor is used to synchronously collect the corresponding training three-dimensional milling force value; Performing current signal fusion on each of the three-phase main shaft current signals respectively to obtain a plurality of training current effective values; Mapping each training current effective value with the synchronized training three-dimensional milling force value to obtain multiple sets of mapping training data; According to a preset ratio, multiple groups of mapping training data are divided into a training data set and a verification data set, so as to optimize the initial milling force prediction model once based on the training data set and to optimize the intermediate milling force prediction model twice based on the verification data set.

6. The method according to claim 5, characterized in that Before respectively fusing the three-phase main shaft current signals, the method further includes: Performing a first data normalization process on each of the three-phase main shaft current signals to obtain the three-phase main shaft current signals after the first normalization process, and performing current signal fusion based on the three-phase main shaft current signals after the first normalization process; A second data normalization process is performed on each of the training three-dimensional milling force values ​​to obtain the second normalized training three-dimensional milling force values, so as to generate mapping training data based on the second normalized training three-dimensional milling force values.

7. The method according to claim 1, characterized in that The current machining parameters include cutting depth and equivalent feed rate.

8. A device for determining the wear state of a tool, characterized in that: include: A milling force prediction module is used to obtain a current effective value data group corresponding to a target tool within a preset time length, and based on a milling force prediction model that has completed model training, predict a three-dimensional milling force value data group of the target tool according to the current effective value data group, and calculate the three-dimensional milling force mean values ​​respectively, wherein the milling force prediction model is obtained by model training using the training current effective value under different processing parameters and the corresponding training three-dimensional milling force values; A cutting force coefficient calculation module, configured to calculate the mean value of the three-directional cutting force coefficient of the target tool within the preset time period according to the mean value of the three-directional milling force, based on the cutting angle change matrix corresponding to the target tool and the current machining parameters; The tool wear state determination module is used to determine the wear state of the target tool according to the mean value of the three-way cutting force coefficient.

9. A storage medium, wherein at least one executable instruction is stored in the storage medium, characterized in that: The executable instructions enable the processor to execute operations corresponding to the method for determining the tool wear state according to any one of claims 1 to 7.

10. A terminal, comprising: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, wherein the executable instruction enables the processor to execute an operation corresponding to the method for determining the tool wear state according to any one of claims 1-7.

Images