Search

CN-122007931-A - Method, device and computer program product for adaptively clamping tool holder of machine tool

CN122007931ACN 122007931 ACN122007931 ACN 122007931ACN-122007931-A

Abstract

The invention relates to the technical field of clamping of machine tools and discloses a method, a device and a computer program product for self-adaptive clamping of a machine tool holder, wherein the method comprises the steps of solving a modified kinetic equation by a fourth-order Dragon-Gregorian tower method, generating simulation data under different clamping states and constructing a source domain data set; the method comprises the steps of collecting clamping data of a cutter handle in different clamping states, constructing an actual measurement data set, adopting a CS-GWO optimized VMD algorithm to reduce noise of the actual measurement data set to obtain a target domain data set, training an improved alternate migration learning model based on a training set consisting of a source domain data set and the target domain data set, and self-adaptively adjusting the clamping force of a cutter holder according to the material, specification and clamping state of the cutter handle by utilizing the trained improved alternate migration learning model. The invention realizes closed-loop self-adaptive regulation and control of the clamping force, can dynamically adjust parameters according to the material, specification and processing working condition of the cutter handle, avoids the problem of over-tightening or over-loosening, and improves the processing precision and the service life of the cutter.

Inventors

  • WEI CHENGHONG
  • FENG LEI
  • WEI SHIYONG
  • GAO XIAOYANG
  • WU JINGLIANG

Assignees

  • 巨冈精工(广东)股份有限公司

Dates

Publication Date
20260512
Application Date
20260129

Claims (10)

  1. 1. A method for adaptively clamping a tool holder of a machine tool based on improved alternate transfer learning, comprising: Based on Hertz contact theory, constructing a cutter holder-cutter holder nonlinear dynamic model according to the elastic contact characteristics of the cutter holder and the cutter holder, correcting a radial displacement excitation function according to the real-time position of the cutter holder, and correcting contact stiffness and impact force parameters according to the specification and the clamping state of the cutter holder to obtain a corrected dynamic equation; Solving a modified kinetic equation by a fourth-order Dragon-Gregory tower method, generating simulation data containing clamping force, radial displacement and vibration signals under different clamping states, and constructing a source domain data set; Collecting clamping data of tool shanks of different specifications and different materials in different clamping states to construct an actual measurement data set, wherein the clamping data comprises clamping force, radial displacement and vibration signals; Adopting CS-GWO to optimize VMD algorithm to make noise reduction on the actual measurement data set so as to obtain target domain data set; training a pre-constructed improved alternate migration learning model based on a training set consisting of a source domain data set and a target domain data set; and the trained improved alternate transfer learning model is utilized to adaptively adjust the clamping force of the tool holder according to the material, specification and clamping state of the tool holder.
  2. 2. The method for adaptively clamping a tool holder of a machine tool based on improved alternate mobility learning according to claim 1, wherein the method is based on the Hertz contact theory and constructs a tool holder-tool holder nonlinear dynamics model according to the elastic contact characteristics of a tool holder and the tool holder, and specifically comprises the following steps: based on Hertz contact theory, the contact force and the contact deformation of the tool handle and the tool holder show a nonlinear relation: , , ; Wherein, the Is the normal pressing force generated when the tool holder contacts the tool handle, For the equivalent modulus of elasticity, the elastic modulus, For equivalent radii of curvature, E 1 and E 2 are the elastic moduli of the tool holder and shank materials respectively, And Poisson ratio of the cutter holder and the cutter handle respectively, R 1 is the curvature radius of the cutter handle conical surface, R 2 is the curvature radius of the cutter holder conical surface, Is the normal deformation vector of the contact point; The kinetic equation before correction is: , ; Wherein M is, 、C、 、K I 、 、 And The mass distribution matrix, the acceleration, the damping distribution matrix, the speed, the linear stiffness matrix, the nonlinear contact stiffness, the displacement and the external excitation of the tool holder-tool handle system are sequentially adopted; The method comprises the steps of correcting a radial displacement excitation function according to the real-time position of a cutter handle, correcting parameters of contact stiffness and impact force according to the specification and the clamping state of the cutter handle, and obtaining a corrected kinetic equation, and specifically comprises the following steps: introducing real-time position coordinates of the cutter handle into a radial displacement excitation function: u Correction (t)=ecos(ωt)+u 0 (t), wherein ecos (ωt) is a periodic radial displacement due to rotational eccentricity, ω is a rotational angular velocity, e is an eccentricity, and u 0 (t) is an additional displacement due to cutting force fluctuation; introducing a deformation correction coefficient of the contact area, and correcting nonlinear contact stiffness as follows: Wherein eta is less than or equal to 1, and eta is smaller as the abrasion is more serious; based on the strong correlation of the impact force and the specification of the tool handle and the clamping state, the momentum theorem is utilized to realize the following The correction is as follows: In the formula (I), in the formula (II), In order to correct the coefficient of the coefficient, In order to change the amount of speed before and after impact, For the duration of the impact.
  3. 3. The method for adaptively clamping a tool holder of a machine tool based on improved alternate mobility learning according to claim 2, wherein the method is characterized in that a modified kinetic equation is solved by a fourth-order longgrid-base tower method, simulation data containing clamping force, radial displacement and vibration signals under different clamping states are generated, and a source domain data set is constructed, specifically: And solving a modified dynamic equation by a fourth-order Dragon-Gregory tower method, generating simulation data of the tool holder in five clamping states of normal clamping, overtightening, contact abrasion and drive failure, and constructing a source domain data set with a clamping state label.
  4. 4. The method for adaptively clamping a tool holder of a machine tool based on improved alternate mobility learning of claim 3, wherein the adopting CS-GWO to optimize the VMD algorithm to reduce the noise of the measured dataset to obtain the target domain dataset specifically comprises: selecting an amplitude spectrum of clamping data in the measured data set as a fitness function of a CS-GWO algorithm, using an extremely small value of amplitude spectrum entropy as the fitness function, and searching for an optimal parameter combination in a VMD algorithm Wherein, K is the mode number K, Is a secondary penalty factor; decomposing clamping data in the actual measurement data set by utilizing a VMD algorithm of the optimal parameter combination to obtain K modal components; and calculating the correlation between each modal component and the measured sample, reserving the modal components with high correlation, and summing and reconstructing to obtain the target domain data for constructing the target domain data set.
  5. 5. The method for adaptively clamping a tool holder of a machine tool based on improved alternate shift learning as set forth in claim 4, wherein said selecting an amplitude spectrum of clamping data in the measured dataset as an fitness function of the CS-GWO algorithm, and searching for an optimal combination of parameters in the VMD algorithm with a minimum value of the entropy of the amplitude spectrum as the fitness function The method specifically comprises the following steps: Initializing parameters of a CS-GWO algorithm and a VMD; the minimum value of the amplitude spectrum entropy is taken as the fitness function, and the CS-GWO algorithm is used for searching the optimal parameter combination in the VMD algorithm Wherein, the calculation formula of the amplitude spectrum entropy is as follows: ; In the above formula, L i is the amplitude spectrum of the modal component u i , H i is the amplitude spectrum entropy of the modal component u i , and N is the length of the modal component.
  6. 6. The method for adaptively clamping a tool holder of a machine tool based on improved alternate shift learning according to claim 5, wherein the method for constructing the improved alternate shift learning model is as follows: Constructing a CNN model comprising five convolution layers, two pooling layers and two full connection layers, adding a batch of standardization layers after each convolution layer, adopting a ReLU function as a hidden layer activation function and a Softmax function as an output layer activation function, wherein a first pooling layer is positioned between a second convolution layer and a third convolution layer, and a second pooling layer is positioned before the first full connection layer after the fifth convolution layer; Calculating a CORAL loss function after a first convolution layer of the CNN model, and reducing the second-order statistic difference of a source domain data set and a target domain data set; the sum of the MMD penalty function and the classification penalty is calculated at the full connectivity layer, updating the network weights and bias parameters by alternating back propagation.
  7. 7. The method for adaptively clamping a tool holder of a machine tool based on improved alternate shift learning according to claim 6, wherein the method for setting a loss function of the improved alternate shift learning model is as follows: The distance of the second order statistic of the data features in the source domain dataset from the target domain dataset is defined as the CORAL loss function L CORAL : , the Frobenius norm of the mean square matrix; Wherein D is the feature dimension output by the first convolution layer, and the covariance matrices of the features of the source domain dataset D s and the target domain dataset D t are C s and C t respectively; MMD is used to measure the difference in distribution of feature sets D s and D t in the regenerated kernel hilbert space: , a feature mapping function for mapping the output of the full connection layer to a high dimensional space; Wherein n s 、n t is the number of samples of D s and D t , respectively, 、 Features of the ith sample of D s and the jth sample of D t , respectively; For the ith target domain data sample, predicting the probability of being class c : ; Wherein, the For the original output of the full connection layer on class c for the ith target domain data sample, Is a natural constant; The total loss of the full connection layer is L total =L MMD +λL CE , wherein lambda is a preset weight coefficient; ; where N is the total number of samples, c is the type of clamped state, A real label of the ith target domain data sample; the weight and bias of the full connection layer are updated firstly through back propagation, and then the parameters of the convolution layer are updated by combining with the CORAL loss of the first convolution layer, and the repeated alternate iteration is carried out.
  8. 8. The device for adaptively clamping the tool holder of the machine tool based on the improved alternate transfer learning comprises a tool holder, an elastic tool holder for clamping the tool holder, a piezoelectric ceramic driver and a wedge block transmission mechanism, wherein the piezoelectric ceramic driver is in driving connection with an elastic clamping jaw through the wedge block transmission mechanism, and the device is characterized in that the piezoelectric ceramic driver realizes recognition of a clamping state and adaptive adjustment of the clamping force of the tool holder through the method for adaptively clamping the tool holder of the machine tool based on the improved alternate transfer learning according to any one of claims 1-7.
  9. 9. The device for adaptively clamping a tool holder of a machine tool based on improved alternate shift learning according to claim 8, wherein when the recognition result of the clamping state is "overtightening", the piezoelectric ceramic driver reduces the output voltage by 5% -8%; when the recognition result of the clamping state is "over loose", controlling the piezoelectric ceramic driver to increase the output voltage by 8% -12%; When the recognition result of the clamping state is 'overtightening', controlling the piezoelectric ceramic driver to reduce the output voltage by 5% -8%; When the recognition result of the clamping state is contact abrasion, dynamically adjusting the compensation coefficient of the contact stiffness, wherein the adjustment range of the compensation coefficient is 1.0-1.15.
  10. 10. A computer program product comprising computer programs/instructions which, when executed by a processor, implement a method of improved adaptive clamping of machine tool holders according to any one of claims 1-7.

Description

Method, device and computer program product for adaptively clamping tool holder of machine tool Technical Field The invention relates to the technical field of machine tool cutter clamping, in particular to a method, a device and a computer program product for self-adaptive clamping of a machine tool holder. Background The clamping reliability of the tool holder of the machine tool, which is used as a core component for connecting the main shaft and the tool holder, directly influences the machining precision, the service life of the tool and the production safety. The current mainstream tool holder clamping modes comprise hydraulic clamping, pneumatic clamping, mechanical locking and the like, and the tool holder has the following technical defects: (1) The clamping force control depends on preset parameters, so that the control is difficult to adapt to the different requirements of tool shanks with different materials and specifications, and the problems that the tool shanks deform due to too tight clamping or cutting vibration is caused due to too loose clamping are easy to occur; (2) The lack of accurate modeling of dynamic characteristics in the clamping process can not sense dynamic changes of key parameters such as contact stiffness, radial displacement and the like in real time, so that the clamping stability is insufficient; (3) The fault diagnosis relies on manual experience or single sensor data, so that potential problems such as clamping failure, contact abrasion and the like are difficult to quickly identify, and adaptability of the cross-working condition and cross-model tool holder is poor. Therefore, a method for adaptively clamping a tool holder is needed in the art to solve the problems of poor adaptability, low diagnosis precision and weak anti-interference capability of the traditional method. Disclosure of Invention The present invention aims to provide a method, a device and a computer program product for adaptively clamping a tool holder of a machine tool, which solve or at least partially solve the technical problems mentioned in the background art. To achieve the purpose, the invention adopts the following technical scheme: In a first aspect, the present invention provides a method for adaptive clamping of a tool holder of a machine tool based on improved alternate transfer learning, comprising: Based on Hertz contact theory, constructing a cutter holder-cutter holder nonlinear dynamic model according to the elastic contact characteristics of the cutter holder and the cutter holder, correcting a radial displacement excitation function according to the real-time position of the cutter holder, and correcting contact stiffness and impact force parameters according to the specification and the clamping state of the cutter holder to obtain a corrected dynamic equation; Solving a modified kinetic equation by a fourth-order Dragon-Gregory tower method, generating simulation data containing clamping force, radial displacement and vibration signals under different clamping states, and constructing a source domain data set; Collecting clamping data of tool shanks of different specifications and different materials in different clamping states to construct an actual measurement data set, wherein the clamping data comprises clamping force, radial displacement and vibration signals; Adopting CS-GWO to optimize VMD algorithm to make noise reduction on the actual measurement data set so as to obtain target domain data set; training a pre-constructed improved alternate migration learning model based on a training set consisting of a source domain data set and a target domain data set; and the trained improved alternate transfer learning model is utilized to adaptively adjust the clamping force of the tool holder according to the material, specification and clamping state of the tool holder. Optionally, based on the Hertz contact theory, the tool holder-tool holder nonlinear dynamics model is constructed according to the elastic contact characteristic of the tool holder and the tool holder, and specifically comprises the following steps: based on Hertz contact theory, the contact force and the contact deformation of the tool handle and the tool holder show a nonlinear relation: ,,; Wherein, the Is the normal pressing force generated when the tool holder contacts the tool handle,For the equivalent modulus of elasticity, the elastic modulus,For equivalent radii of curvature, E 1 and E 2 are the elastic moduli of the tool holder and shank materials respectively,AndPoisson ratio of the cutter holder and the cutter handle respectively, R 1 is the curvature radius of the cutter handle conical surface, R 2 is the curvature radius of the cutter holder conical surface,Is the normal deformation vector of the contact point; The kinetic equation before correction is: , ; Wherein M is, 、C、、KI、、AndThe mass distribution matrix, the acceleration, the damping distribution matrix, the speed, the linear stiffness ma