Numerical control machining center tool setting gauge control system

Abstract

The invention discloses a numerical control machining center tool setting gauge control system, and the system comprises a data collection module which is used for collecting a multi-dimensional measurement data set of a numerical control machining center tool setting gauge through a sensor; the data preprocessing module is used for generating a standardized multi-dimensional measurement data set; the feature extraction and noise reduction module is used for constructing an improved variational automatic encoder model to obtain a low-dimensional potential feature representation data set and a multi-dimensional reconstruction measurement data set after noise reduction; the calibration optimization module is used for constructing a tool setting gauge calibration optimization objective function based on the low-dimensional potential feature representation data set and the multi-dimensional reconstruction measurement data set after noise reduction; the global optimization module is used for determining a global optimal calibration parameter set; and the calibration control module is used for enabling the cutter state to be optimally matched with the calibration parameters. The method has high adaptability under various tool types and machining conditions, and global optimality of calibration control parameters of the tool setting gauge is ensured.

Description

Technical Field

[0001] The invention relates to the technical field of numerical control, and in particular to a tool setting instrument control system for a numerical control machining center. Background Art

[0002] With the development of CNC machining technology, CNC machining centers are increasingly used in precision manufacturing, aerospace, and automobile manufacturing. Tool setting probes are important measuring tools for CNC machining centers. Their main function is to measure the position, shape, and wear of the tool to ensure machining accuracy and tool life.

[0003] At present, the calibration of tool setting instruments in CNC machining centers mainly relies on static measurement and simple deviation compensation technology. Traditional calibration methods are usually based on experience-based parameter setting and calibration control by manually adjusting tool compensation values. Traditional methods can achieve a certain degree of accuracy under ideal conditions, but under conditions of high-speed cutting, complex machining paths and dynamic environmental changes, it is often difficult to ensure a stable calibration effect. The main problems include:

[0004] During the CNC machining process, the tool will be affected by vibration, thermal deformation, and cutting force changes when rotating and feeding at high speed, resulting in a large amount of noise and irrelevant information in the data collected by the measurement sensor. Traditional data processing methods mainly use simple filtering or mean processing algorithms, but these methods are difficult to effectively remove nonlinear noise and may cause the loss of effective tool feature information, thereby affecting the calibration accuracy.

[0005] Existing tool setting calibration systems usually rely on preset compensation parameters and use local optimization strategies to adjust the calibration parameters. It is difficult to make fine adjustments under complex processing conditions. Due to changes in different types of tools, processing materials and processing environments, a single parameter setting is often not applicable to all processing tasks, resulting in accumulated calibration errors and affecting processing quality.

[0006] Traditional calibration methods usually perform static adjustments before processing. However, during the actual processing, tool wear and temperature changes will affect the tool state, causing the original calibration parameters to gradually become invalid. Due to the lack of real-time feedback and dynamic adjustment mechanisms, existing methods cannot perform online calibration during the processing process, resulting in continuous accumulation of errors and reduced processing accuracy.

[0007] At present, some optimization methods try to use genetic algorithms or particle swarm optimization algorithms to optimize the calibration parameters, but these algorithms are prone to fall into local optimality in high-dimensional optimization space, resulting in the calibration results not being able to achieve optimal accuracy. In addition, some methods have large computational complexity and slow optimization speed, making it difficult to meet the real-time requirements of CNC machining centers.

[0008] In view of the existing technical bottlenecks, there is an urgent need for a new tool setting instrument control method that can efficiently reduce noise and extract tool status characteristics, automatically optimize calibration parameters in combination with global optimization methods, and have dynamic adaptive adjustment capabilities to improve the machining accuracy and production efficiency of CNC machining centers. Summary of the invention

[0009] An object of the present invention is to provide a tool setting instrument control system for a CNC machining center, which exhibits high adaptability under various tool types and machining conditions, and ensures the global optimality of the tool setting instrument calibration control parameters.

[0010] A tool setting control system for a CNC machining center according to an embodiment of the present invention includes the following modules:

[0011] A data acquisition module, which is used to collect multi-dimensional measurement data sets of tool setting instruments in CNC machining centers through sensors;

[0012] A data preprocessing module, used for performing data cleaning, outlier removal, time series alignment and standardization on the multidimensional measurement data set to generate a standardized multidimensional measurement data set;

[0013] The feature extraction and denoising module is used to construct an improved variational autoencoder model, and perform denoising and feature extraction on the standardized multidimensional measurement data set based on the improved variational autoencoder model to obtain a low-dimensional potential feature representation data set and a denoised multidimensional reconstructed measurement data set;

[0014] A calibration optimization module is used to construct an optimization objective function for tool setting instrument calibration based on a low-dimensional latent feature representation dataset and a denoised multi-dimensional reconstructed measurement dataset;

[0015] A global optimization module is used to perform a global optimization search on the optimization objective function of the tool setting instrument calibration using a cuckoo search optimization algorithm to determine a global optimal calibration parameter set;

[0016] The calibration control module is used to generate the calibration control strategy of the tool setter of the CNC machining center based on the global optimal calibration parameter set, and to perform dynamic calibration adjustment according to the machining task of the tool setter and the real-time tool status, so as to keep the tool status and calibration parameters in the best match.

[0017] A tool setting instrument control method for a numerical control machining center is applied to a tool setting instrument control system for a numerical control machining center, and comprises the following steps:

[0018] S1. Use sensors to collect multi-dimensional measurement data sets of tool setting instruments in CNC machining centers;

[0019] S2. Perform data cleaning, outlier removal, time series alignment and standardization on the multidimensional measurement data set to form a standardized multidimensional measurement data set;

[0020] S3. Construct an improved variational autoencoder model, and perform denoising and feature extraction on the standardized multidimensional measurement data set through the improved variational autoencoder model to obtain a low-dimensional potential feature representation data set for tool setting instrument calibration of a CNC machining center and a denoised multidimensional reconstructed measurement data set;

[0021] S4. Constructing the optimization objective function of tool setting instrument calibration based on the low-dimensional latent feature representation dataset and the denoised multi-dimensional reconstructed measurement dataset;

[0022] S5. Use the cuckoo search optimization algorithm to perform a global search on the optimization objective function of the tool setting instrument calibration to determine the global optimal calibration parameter set;

[0023] S6. Generate a calibration control strategy for the tool setter of the CNC machining center based on the global optimal calibration parameter set, and dynamically calibrate and adjust the tool setter to achieve matching control between the tool state of the tool setter and the calibration parameters.

[0024] Optionally, the S1 includes the following steps:

[0025] S11. Use a sensor array to measure the tool state of the tool setting instrument in the CNC machining center. The sensor array includes a displacement sensor, a vibration sensor, a temperature sensor and an environmental monitoring sensor, which respectively collect the tool displacement signal, acceleration signal, temperature signal and environmental impact signal to construct an original multidimensional measurement data set. ;

[0026] S12. Divide the original multidimensional measurement data set into data acquisition windows, set a fixed time window, and group the measurement data in the original multidimensional measurement data set in chronological order so that each time window contains Group measurement data to form a windowed multi-dimensional measurement data set .

[0027] Optionally, S2 includes the following steps:

[0028] S21. The windowed multidimensional measurement data set Perform data cleaning, remove missing values, duplicate values ​​and abnormal measurement data, and construct a cleaned multidimensional measurement data set;

[0029] S22. According to the tool motion state and the environmental influence signal distribution characteristics, the mean and standard deviation of each measurement signal are calculated, and the abnormal judgment threshold is set to eliminate them, and a multidimensional measurement data set is obtained after the abnormal values ​​are eliminated;

[0030] S23. Performing time series alignment on the multidimensional measurement data set after removing outliers according to the timestamp, using interpolation to complete the missing data in the multidimensional measurement data set, and constructing a time series aligned multidimensional measurement data set;

[0031] S24. Normalize the multidimensional measurement data set aligned with the time series according to the standardization rules, and map each measurement signal to the interval , forming a standardized multidimensional measurement data set .

[0032] Optionally, S3 includes the following steps:

[0033] S31. Standardizing multidimensional measurement data sets An improved variational autoencoder model for deep noise reduction and feature extraction of tool setting instrument in CNC machining center is constructed using the : ;

[0034] in, To standardize the data points in a multidimensional measurement dataset, To improve the encoder network of the variational autoencoder model, is the encoder network parameter, which is used to adaptively capture the dynamic correlation characteristics between the tool's displacement signal, acceleration signal, temperature signal and environmental impact signal. To improve the decoder network of the variational autoencoder model, are the decoder network parameters used to reconstruct the noise-reduced measurement data, To standardize latent feature representations for multidimensional measurement datasets;

[0035] S32. Construct a latent feature mapping function that includes spatial attention mechanism and temporal attention mechanism, highlight the key signal dimensions and key moments of the tool state of the CNC machining center, and define the latent feature representation in the latent feature mapping function for: ;

[0036] in, , They represent the mean and variance vectors of the tool state features output by the encoder network, respectively, describing the potential feature distribution of the tool state. is a random variable with standard normal distribution, It is a spatial attention factor utilization function used to extract the spatial correlation characteristics between the tool's displacement signal, acceleration signal, temperature signal and environmental impact signal. is the spatial attention parameter, is the temporal attention factor utilization function, which is used to capture the sensitivity difference of the tool measurement signal at the preset time in the machining process. is the time attention parameter, l represents the index of the tool state feature in the spatial dimension, m represents the index of the tool state feature in the temporal dimension, and Represent the data points in the standardized multidimensional measurement data set in the spatial dimension and the temporal dimension, respectively. represents the Hadamard product;

[0037] S33. Construct a decoder reconstruction function that integrates the dynamic error feedback mechanism, and reconstruct the error of the previous step in real time during the calibration of the tool setting instrument in the CNC machining center As a dynamic adjustment factor, define the reconstruction data : ;

[0038] in, represents the initial reconstruction output of the decoder network based on the latent feature representation, is the error correction weight matrix, is a learnable parameter, is the reconstruction error of the current step, To dynamically adjust the intensity factor, Reconstruct the error for the previous step;

[0039] S34. Constructing a joint spatial-temporal attention-constrained deep denoising feature extraction loss function : ;

[0040] in, is the reconstruction error term, KL is the divergence loss function, is the KL divergence term, which is used to constrain the distribution of the potential features of the tool state. To standardize the total number of data in the multidimensional measurement dataset, is the standard normal distribution prior, and are spatial attention sparse constraint term and temporal attention sparse constraint term respectively, and is the weight coefficient;

[0041] S35. Back-propagate based on the loss function of the joint spatial-temporal attention constraint to optimize the encoder network parameters , decoder network parameters And the spatial attention parameter , temporal attention parameters , generating an improved variational autoencoder model for CNC machining center tool setting instrument measurement data:

[0042] S36. Using the improved variational autoencoder model to perform deep noise reduction and feature extraction on the standardized multidimensional measurement data set, a low-dimensional latent feature representation data set is obtained for tool setting calibration of CNC machining centers and the multi-dimensional reconstructed measurement data set after denoising .

[0043] Optionally, S4 includes the following steps:

[0044] S41. Define calibration parameter set , each calibration parameter in the calibration parameter set Used to describe the tool position compensation, dynamic deviation adjustment and signal compensation factors involved in the calibration process of the tool setting instrument of the CNC machining center;

[0045] S42. Construct tool setting probe calibration mapping function , the input of the knife calibration mapping function is a low-dimensional latent feature representation dataset and calibration parameter set The output is the predicted measurement data after the tool setting instrument is calibrated : ;

[0046] in, For the The low-dimensional potential feature representation of the data points obtained after dimensionality reduction by the improved variational autoencoder model, is the mapping weight matrix, is the bias vector, For depicting the The calibration parameters represent the low-dimensional latent features. The function of influence, is the total number of calibration parameters, represents the jth calibration parameter in the calibration parameter set;

[0047] S43. Constructing the tool setting calibration optimization objective function , tool setting calibration optimizes the objective function to minimize the predicted measurement data and the reconstructed multidimensional measurement data set after noise reduction The mean square error between is the target: ;

[0048] in, Indicates The calibration prediction error for data points, The multi-dimensional measurement data reconstructed after noise reduction is is the regularization coefficient, is the regularization term.

[0049] Optionally, S5 includes the following steps:

[0050] S51. Initialize the cuckoo population and define the initial position of the cuckoo population individuals as the calibration parameter set , the population size is set to , the cuckoo population initialization is defined as: ;

[0051] in, Indicates The initial calibration parameter set of cuckoo individuals, Indicates Among the cuckoo individuals The initial values ​​of the calibration parameters;

[0052] S52. Optimize the objective function based on tool setting calibration Evaluate the fitness of each individual in the cuckoo population and obtain the initial fitness value: ;

[0053] in, For the The initial fitness value of each cuckoo individual reflects the prediction error and stability of the corresponding calibration parameter set in the tool setting instrument calibration;

[0054] S53. Use the Levy flight mechanism to update the position of each individual in the cuckoo population and obtain an updated calibration parameter set : ;

[0055] in, For the After the update The set of calibration parameters for cuckoo individuals, For the After the update The calibration parameter set of each cuckoo individual represents the updated value of the tool setting calibration parameter of the CNC machining center. is the step size factor for position update, represents the random flight step vector following the Levy distribution, For the The set of calibration parameters of the individuals with the best fitness in the generation population;

[0056] S54. Using the updated calibration parameter set Calculate the fitness value again ;

[0057] S55. According to the optimal selection principle of the cuckoo search optimization algorithm, perform fitness comparison and replacement within the population and update the optimal position of the individuals in the population : ;

[0058] S56. Perform iterative search until the set termination condition is met to obtain the global optimal calibration parameter set : ;

[0059] in, is the iteration termination time, is the calibration optimization objective function value corresponding to the qth cuckoo individual in the Tth iteration.

[0060] Optionally, the S6 includes the following steps:

[0061] S61. Extract the calibration control strategy from the global optimal calibration parameter set, and set the calibration strategy category of the tool setting instrument of the CNC machining center according to the tool displacement compensation, tool wear correction, machining environment interference correction and dynamic error adjustment parameters:

[0062] Static calibration strategy, suitable for standard working conditions, tool displacement compensation parameters In the normal range, tool wear correction parameters ≤Preset threshold, dynamic error adjustment parameter In a stable state;

[0063] The adaptive calibration strategy is suitable for working conditions where the machining environment change is greater than the threshold or the tool wear rate is greater than the threshold. Out of normal range or tool wear correction parameters When it is higher than the preset threshold, the calibration control strategy is adjusted to the adaptive mode;

[0064] Emergency calibration strategy, suitable for abnormal tool vibration, machining accuracy reduction or error exceeding the limit, when the dynamic error adjustment parameters When it is higher than the set warning value, the calibration control strategy is adjusted to the emergency correction mode and the real-time compensation mechanism is triggered;

[0065] S62. Construct a calibration control matrix based on the tool status of the tool setting instrument of the CNC machining center and the type of machining task, classify the tool calibration accuracy requirements in combination with the tool calibration history data, and dynamically match the optimal calibration parameters when switching tasks;

[0066] S63. Real-time monitoring of tool displacement signals, acceleration signals, temperature signals and environmental impact signals. When the measured data deviates from the historical calibration mean and exceeds the preset error range, the calibration control strategy is adaptively adjusted to adapt the calibration parameters to the dynamic changes of the machining environment;

[0067] S64. A feedback optimization mechanism is adopted. After performing calibration adjustment, the calibration error is calculated based on the real-time measurement data, and the calibration control strategy is updated. When the calibration error is lower than the set threshold, the current calibration parameters are maintained. If the error exceeds the limit, the calibration parameters are dynamically corrected and the calibration control matrix is ​​optimized to realize the calibration control of the tool setting instrument of the CNC machining center.

[0068] The beneficial effects of the present invention are:

[0069] (1) In the preprocessing of the measurement data of the tool setting instrument in the CNC machining center, the present invention constructs an improved variational autoencoder model combined with the space-time attention mechanism, which is used to effectively remove the noise interference in the machining environment and extract the key features of the tool state. The improved variational autoencoder uses the spatial attention mechanism to assign feature weights to the tool displacement, vibration, temperature and environmental influence signals, thereby enhancing the ability to extract key tool state information. At the same time, combined with the temporal attention mechanism, the model's time series modeling ability in a complex machining environment is improved by dynamically adjusting the importance weights of data at different time points.

[0070] (2) The present invention introduces a cuckoo search optimization algorithm, enhances the global search capability through the Levy flight mechanism, and introduces an adaptive step size control strategy in the calibration parameter update process, so that the algorithm maintains a large search range in the early exploration stage to avoid falling into the local optimal solution. In the later stage, a high-precision local search mechanism is used to improve the optimization convergence speed. The calibration parameter convergence speed is improved, and it shows high adaptability under various tool types and processing conditions, ensuring the global optimality of the tool setting instrument calibration control parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0072] Figure 1 The present invention provides a flow chart of a method for controlling a tool setting instrument in a numerical control machining center. DETAILED DESCRIPTION

[0073] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0074] refer to Figure 1 , a tool setting control system for a CNC machining center, including the following modules:

[0075] A data acquisition module, which is used to collect multi-dimensional measurement data sets of tool setting instruments in CNC machining centers through sensors;

[0076] A data preprocessing module is used to perform data cleaning, outlier removal, time series alignment and standardization on the multidimensional measurement data set to generate a standardized multidimensional measurement data set;

[0077] The feature extraction and denoising module is used to construct an improved variational autoencoder model, and perform denoising and feature extraction on the standardized multidimensional measurement data set based on the improved variational autoencoder model to obtain a low-dimensional potential feature representation data set and a denoised multidimensional reconstructed measurement data set;

[0078] A calibration optimization module is used to construct an optimization objective function for tool setting instrument calibration based on a low-dimensional latent feature representation dataset and a denoised multi-dimensional reconstructed measurement dataset;

[0079] A global optimization module is used to perform a global optimization search on the optimization objective function of the tool setting instrument calibration using a cuckoo search optimization algorithm to determine a global optimal calibration parameter set;

[0080] The calibration control module is used to generate the calibration control strategy of the tool setter of the CNC machining center based on the global optimal calibration parameter set, and to perform dynamic calibration adjustment according to the machining task of the tool setter and the real-time tool status, so as to keep the tool status and calibration parameters in the best match.

[0081] A tool setting instrument control method for a numerical control machining center is applied to a tool setting instrument control system for a numerical control machining center, and comprises the following steps:

[0082] S1. Use sensors to collect multi-dimensional measurement data sets of tool setting instruments in CNC machining centers;

[0083] S2. Perform data cleaning, outlier removal, time series alignment and standardization on the multidimensional measurement data set to form a standardized multidimensional measurement data set;

[0084] S3. Construct an improved variational autoencoder model, and perform denoising and feature extraction on the standardized multidimensional measurement data set through the improved variational autoencoder model to obtain a low-dimensional latent feature representation data set for tool setting calibration of a CNC machining center and a denoised multidimensional reconstructed measurement data set;

[0085] S4. Constructing the optimization objective function of tool setting instrument calibration based on the low-dimensional latent feature representation dataset and the denoised multi-dimensional reconstructed measurement dataset;

[0086] S5. Use the cuckoo search optimization algorithm to perform a global search on the optimization objective function of the tool setting instrument calibration to determine the global optimal calibration parameter set;

[0087] S6. Generate a calibration control strategy for the tool setter of the CNC machining center based on the global optimal calibration parameter set, and dynamically calibrate and adjust the tool setter to achieve matching control between the tool state of the tool setter and the calibration parameters.

[0088] In this implementation, S1 includes the following steps:

[0089] S11. The sensor array is used to measure the tool status of the tool setting instrument in the CNC machining center. The sensor array includes displacement sensors, vibration sensors, temperature sensors and environmental monitoring sensors, which respectively collect the tool displacement signal, acceleration signal, temperature signal and environmental impact signal to construct the original multi-dimensional measurement data set. ;

[0090] S12. Divide the original multidimensional measurement data set into data collection windows, set a fixed time window, and group the measurement data in the original multidimensional measurement data set in chronological order so that each time window contains Group measurement data to form a windowed multi-dimensional measurement data set .

[0091] In this implementation, S2 includes the following steps:

[0092] S21. Windowed multidimensional measurement dataset Perform data cleaning, remove missing values, duplicate values ​​and abnormal measurement data, and construct a cleaned multidimensional measurement data set;

[0093] S22. According to the tool motion state and the environmental influence signal distribution characteristics, the mean and standard deviation of each measurement signal are calculated, and the abnormal judgment threshold is set to eliminate them, and a multidimensional measurement data set is obtained after the abnormal values ​​are eliminated;

[0094] S23. Performing time series alignment on the multidimensional measurement data set after removing outliers based on the timestamps, using interpolation to complete the missing data in the multidimensional measurement data set, and constructing a time series aligned multidimensional measurement data set;

[0095] S24. Normalize the multidimensional measurement data set aligned with the time series according to the standardization rules, and map each measurement signal to the interval , forming a standardized multidimensional measurement data set .

[0096] In this implementation, S3 includes the following steps:

[0097] S31. Standardizing multidimensional measurement data sets An improved variational autoencoder model for deep noise reduction and feature extraction of tool setting instrument in CNC machining center is constructed using the : ;

[0098] in, To standardize the data points in a multidimensional measurement dataset, To improve the encoder network of the variational autoencoder model, is the encoder network parameter, which is used to adaptively capture the dynamic correlation characteristics between the tool's displacement signal, acceleration signal, temperature signal and environmental impact signal. To improve the decoder network of the variational autoencoder model, are the decoder network parameters used to reconstruct the noise-reduced measurement data, To standardize latent feature representations for multidimensional measurement datasets;

[0099] S32. Construct a latent feature mapping function that includes spatial attention mechanism and temporal attention mechanism, highlight the key signal dimensions and key moments of the tool state of the CNC machining center, and define the latent feature representation in the latent feature mapping function for: ;

[0100] in, , They represent the mean and variance vectors of the tool state features output by the encoder network, respectively, describing the potential feature distribution of the tool state. is a random variable with standard normal distribution, It is a spatial attention factor utilization function used to extract the spatial correlation characteristics between the tool's displacement signal, acceleration signal, temperature signal and environmental impact signal. is the spatial attention parameter, is the temporal attention factor utilization function, which is used to capture the sensitivity difference of the tool measurement signal at the preset time in the machining process. is the time attention parameter, l represents the index of the tool state feature in the spatial dimension, m represents the index of the tool state feature in the temporal dimension, and Represent the data points in the standardized multidimensional measurement data set in the spatial dimension and the temporal dimension, respectively. represents the Hadamard product;

[0101] The formula is used to calculate the potential feature representation of the tool state. By introducing the spatial attention mechanism and the temporal attention mechanism, the feature extraction capability of the variational autoencoder in complex machining environments is improved. The spatial attention is used to highlight the influence of the key measurement signal dimensions, and the temporal attention is used to improve the feature expression capability at a specific moment (when the tool first contacts the workpiece and when cutting ends).

[0102] S33. Construct a decoder reconstruction function that integrates the dynamic error feedback mechanism, and reconstruct the error of the previous step in real time during the calibration of the tool setting instrument in the CNC machining center As a dynamic adjustment factor, define the reconstruction data : ;

[0103] in, represents the initial reconstruction output of the decoder network based on the latent feature representation, is the error correction weight matrix, is a learnable parameter, is the reconstruction error of the current step, To dynamically adjust the intensity factor, Reconstruct the error for the previous step;

[0104] The formula introduces a dynamic error feedback mechanism in the decoder stage, and dynamically adjusts the current reconstructed data through the reconstruction error of the previous step, so that the reconstruction error is optimized as the calibration parameters change, thereby improving the reconstruction accuracy of the tool measurement data.

[0105] S34. Constructing a joint spatial-temporal attention-constrained deep denoising feature extraction loss function : ;

[0106] in, is the reconstruction error term, KL is the divergence loss function, is the KL divergence term, which is used to constrain the distribution of the potential features of the tool state. To standardize the total number of data in the multidimensional measurement dataset, is the standard normal distribution prior, and are spatial attention sparse constraint term and temporal attention sparse constraint term respectively, and is the weight coefficient;

[0107] The formula is used to optimize the loss function of the VAE model. By combining the reconstruction error term, KL divergence loss and attention sparsity constraint term, it ensures that the model can efficiently denoise and extract the key features of the tool state.

[0108] S35. Back-propagate based on the loss function of the joint spatial-temporal attention constraint to optimize the encoder network parameters , decoder network parameters And the spatial attention parameter , temporal attention parameters , generating an improved variational autoencoder model for CNC machining center tool setting instrument measurement data:

[0109] S36. Using the improved variational autoencoder model to perform deep noise reduction and feature extraction on the standardized multidimensional measurement data set, a low-dimensional latent feature representation data set is obtained for tool setting calibration of CNC machining centers and the multi-dimensional reconstructed measurement data set after denoising .

[0110] In the preprocessing of the measurement data of the tool setting instrument in the CNC machining center, this embodiment constructs an improved variational autoencoder model combined with the space-time attention mechanism to effectively remove the noise interference in the machining environment and extract the key features of the tool state. The improved variational autoencoder uses the spatial attention mechanism to assign feature weights to the tool displacement, vibration, temperature and environmental influence signals, thereby enhancing the ability to extract key tool state information. At the same time, combined with the time attention mechanism, the model's time series modeling ability in complex machining environments is improved by dynamically adjusting the importance weights of data at different time points.

[0111] In this implementation, S4 includes the following steps:

[0112] S41. Define calibration parameter set , each calibration parameter in the calibration parameter set Used to describe the tool position compensation, dynamic deviation adjustment and signal compensation factors involved in the calibration process of the tool setting instrument of the CNC machining center;

[0113] S42. Construct tool setting probe calibration mapping function , the input of the knife calibration mapping function is a low-dimensional latent feature representation dataset and calibration parameter set The output is the predicted measurement data after the tool setting instrument is calibrated : ;

[0114] in, For the The low-dimensional potential feature representation of the data points obtained after dimensionality reduction by the improved variational autoencoder model, is the mapping weight matrix, is the bias vector, For depicting the The calibration parameters represent the low-dimensional latent features. The function of influence, is the total number of calibration parameters, represents the jth calibration parameter in the calibration parameter set;

[0115] The formula is used to construct the mapping relationship of tool calibration, mapping the low-dimensional feature representation data set and the calibration parameter set to the calibrated tool measurement data to ensure the accuracy of tool calibration.

[0116] S43. Constructing the tool setting calibration optimization objective function , tool setting calibration optimizes the objective function to minimize the predicted measurement data and the reconstructed multidimensional measurement data set after noise reduction The mean square error between is the target: ;

[0117] in, Indicates The calibration prediction error for data points, The multi-dimensional measurement data reconstructed after noise reduction is is the regularization coefficient, is the regularization term.

[0118] The formula is used to construct the calibration optimization objective function to minimize the error between the tool measurement data after calibration and the measurement data after noise reduction, ensuring the high accuracy of tool calibration.

[0119] This implementation method achieves accurate modeling of tool setter calibration parameters by constructing a tool setter calibration optimization objective function and combining a low-dimensional potential feature representation data set with a denoised multi-dimensional reconstructed measurement data set. This can effectively reduce calibration errors, improve the stability of calibration parameters, and avoid error accumulation caused by manual parameter setting in traditional methods. The optimization model can adapt to different tools and processing conditions, improve calibration accuracy, reduce calibration time, and enhance the intelligence and adaptability of tool setters in CNC machining centers.

[0120] In this implementation, S5 includes the following steps:

[0121] S51. Initialize the cuckoo population and define the initial position of the cuckoo population individuals as the calibration parameter set , the population size is set to , the cuckoo population initialization is defined as: ;

[0122] in, Indicates The initial calibration parameter set of cuckoo individuals, Indicates Among the cuckoo individuals The initial values ​​of the calibration parameters;

[0123] S52. Optimize the objective function based on tool setting calibration Evaluate the fitness of each individual in the cuckoo population and obtain the initial fitness value: ;

[0124] in, For the The initial fitness value of each cuckoo individual reflects the prediction error and stability of the corresponding calibration parameter set in the tool setting instrument calibration;

[0125] S53. Use the Levy flight mechanism to update the position of each individual in the cuckoo population and obtain an updated calibration parameter set : ;

[0126] in, For the After the update The set of calibration parameters for cuckoo individuals, For the After the update The calibration parameter set of each cuckoo individual represents the updated value of the tool setting calibration parameter of the CNC machining center. is the step size factor for position update, represents the random flight step vector following the Levy distribution, For the The set of calibration parameters of the individuals with the best fitness in the generation population;

[0127] The formula is used to update the position of individual cuckoos and guide the search through the Lévy flight mechanism to ensure the search ability of the global optimal solution and avoid falling into the local optimum.

[0128] S54. Using the updated calibration parameter set Calculate the fitness value again ;

[0129] S55. According to the optimal selection principle of the cuckoo search optimization algorithm, perform fitness comparison and replacement within the population and update the optimal position of the individuals in the population : ;

[0130] S56. Perform iterative search until the set termination condition is met to obtain the global optimal calibration parameter set : ;

[0131] in, is the iteration termination time, is the calibration optimization objective function value corresponding to the qth cuckoo individual in the Tth iteration.

[0132] This implementation method introduces a cuckoo search optimization algorithm, enhances the global search capability through the Levy flight mechanism, and introduces an adaptive step control strategy in the calibration parameter update process, so that the algorithm maintains a large search range in the early exploration stage to avoid falling into the local optimal solution. In the later stage, a high-precision local search mechanism is used to improve the optimization convergence speed. The calibration parameter convergence speed is improved, and it shows high adaptability under various tool types and processing conditions, ensuring the global optimality of the tool setting instrument calibration control parameters.

[0133] In this implementation, S6 includes the following steps:

[0134] S61. Extract the calibration control strategy from the global optimal calibration parameter set, and set the calibration strategy category of the tool setting instrument of the CNC machining center according to the tool displacement compensation, tool wear correction, machining environment interference correction and dynamic error adjustment parameters:

[0135] Static calibration strategy, suitable for standard working conditions, tool displacement compensation parameters In the normal range, tool wear correction parameters ≤Preset threshold, dynamic error adjustment parameter In a stable state;

[0136] The adaptive calibration strategy is suitable for working conditions where the machining environment change is greater than the threshold or the tool wear rate is greater than the threshold. Out of normal range or tool wear correction parameters When it is higher than the preset threshold, the calibration control strategy is adjusted to the adaptive mode;

[0137] Emergency calibration strategy, suitable for abnormal tool vibration, machining accuracy reduction or error exceeding the limit, when the dynamic error adjustment parameters When it is higher than the set warning value, the calibration control strategy is adjusted to the emergency correction mode and the real-time compensation mechanism is triggered;

[0138] S62. Construct a calibration control matrix based on the tool status of the tool setting instrument of the CNC machining center and the type of machining task, classify the tool calibration accuracy requirements in combination with the tool calibration history data, and dynamically match the optimal calibration parameters when switching tasks;

[0139] S63. Real-time monitoring of tool displacement signals, acceleration signals, temperature signals and environmental impact signals. When the measured data deviates from the historical calibration mean and exceeds the preset error range, the calibration control strategy is adaptively adjusted to adapt the calibration parameters to the dynamic changes of the machining environment;

[0140] S64. A feedback optimization mechanism is adopted. After performing calibration adjustment, the calibration error is calculated based on the real-time measurement data, and the calibration control strategy is updated. When the calibration error is lower than the set threshold, the current calibration parameters are maintained. If the error exceeds the limit, the calibration parameters are dynamically corrected and the calibration control matrix is ​​optimized to realize the calibration control of the tool setting instrument of the CNC machining center.

[0141] This implementation method realizes dynamic calibration adjustment of the tool setting instrument through a calibration control strategy based on a global optimal calibration parameter set, ensures real-time matching of the tool state and the calibration parameters, can adaptively adjust tool offset compensation and error correction, effectively reduce tool measurement errors, and improve the stability and accuracy of tool calibration, thereby improving the processing consistency and production efficiency of the CNC machining center, reducing human intervention, and improving the level of automation. It is suitable for high-precision manufacturing environments.

[0142] Embodiment 1:

[0143] In a CNC machining workshop of an aviation parts manufacturer, when a Haas UMC-750 five-axis machining center was batch processing aircraft engine blades, the operator found that the tool life was far lower than expected and the size deviation of the processed blades frequently exceeded the tolerance range. After preliminary inspection, it was found that the problem might be related to the insufficient calibration accuracy of the tool setting instrument. The traditional calibration method could not meet the dynamically changing machining environment, resulting in the continuous accumulation of machining errors. In order to optimize the calibration process, the company decided to apply the method of the present invention to improve the tool calibration accuracy, reduce errors, and improve production efficiency.

[0144] On March 15, 2024, workshop technician Mr. Li noticed during the production process that the dimensional deviations of a batch of blades processed with a φ12mm carbide ball end milling cutter all exceeded the tolerance range (within ±5μm). Further analysis revealed that the traditional tool setting instrument calibration method relies on fixed offset compensation parameters, which cannot be adjusted in real time during tool wear, resulting in continuous accumulation of calibration errors. Mr. Li decided to use the method of the present invention for experiments and enable the improved variational autoencoder model and cuckoo search optimization algorithm in the control system for tool calibration optimization.

[0145] First, in the control system of the CNC machining center, the data acquisition module is enabled to record the displacement, vibration, temperature and environmental impact signals of the tool during the calibration process in real time, sampling once every 100ms to form a measurement data set. From 10:00 to 12:00 on March 15, 2024, a total of 30,000 measurement data were collected, and the data are shown in the following example:

[0146] The collected data is input into the improved variational autoencoder model of the present invention for denoising to remove nonlinear noise interference. After the model runs for 2.5 seconds, a low-dimensional potential feature data set is generated, and the reconstructed measurement data after denoising is output. In the embodiment, the data collected by tool No. 04 at 10:00:04, the displacement signal in the original data is greatly affected by vibration noise. After denoising, the signal-to-noise ratio is improved by 18.6%, and the accuracy of tool state feature extraction is improved by 14.3%, providing high-quality input data for subsequent calibration optimization.

[0147] Subsequently, the system started the cuckoo search optimization algorithm, optimized the calibration parameters from 10:05 to 10:20, performed 100 rounds of search iterations, and finally obtained the optimal calibration parameter set. During the optimization process, the algorithm automatically adjusted the tool offset compensation, temperature compensation factor and dynamic error correction coefficient to reduce the final calibration error to within ±1.2μm. After optimization, the calibration error was reduced by 68.4%, and the parameter convergence time was reduced by 43.3%, significantly improving the calibration accuracy of the tool setting instrument.

[0148] After completing the parameter optimization, the system started the adaptive calibration adjustment function and entered the actual machining process at 10:30. According to the dynamic monitoring data, the system detected that the tool wear exceeded the preset threshold (tool offset increased by 2.4μm) after 2 hours, and automatically adjusted the compensation parameters to avoid further accumulation of calibration errors.

[0149] After 6 hours of continuous processing, the calibration accuracy changes of the method of the present invention and the traditional method are as follows:

[0150] It can be seen that during the long-term processing, the error of the traditional method gradually increases, while the method of the present invention can adjust the calibration parameters in real time so that the error is always controlled within ±2.1μm, thereby improving the long-term stability of the tool calibration.

[0151] Through 3 days of experiments, a total of 120 blade processing tasks were completed. The processing efficiency and qualification rate of the traditional method and the method of the present invention were compared. The experimental results showed that after adopting the method of the present invention, the tool calibration time was shortened by 43.0%, the processing qualification rate was increased by 6.2%, and the production efficiency of the CNC machining center was improved.

[0152] This experiment verifies the effectiveness of the method of the present invention in a real production environment. Compared with the traditional method, the present invention has significant advantages in the following aspects:

[0153] The data noise reduction capability is strong, which improves the signal-to-noise ratio by 18.6% and ensures the reliability of tool status characteristics;

[0154] By optimizing the calibration parameters, the calibration error was reduced by 68.4% and the convergence time was reduced by 43.3%;

[0155] Supports dynamic calibration adjustment. During 6 hours of processing, the error is always controlled within ±2.1μm.

[0156] Improve production efficiency, shorten tool calibration time by 43.0%, and increase machining qualification rate by 6.2%.

[0157] In summary, the method of the present invention can significantly improve the calibration accuracy and stability of the tool setting instrument of the CNC machining center in practical applications, is suitable for high-precision manufacturing scenarios, and has wide industrial application value.

[0158] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A tool setting control system for a CNC machining center, characterized in that: Includes the following modules: A data acquisition module, which is used to collect multi-dimensional measurement data sets of tool setting instruments in CNC machining centers through sensors; A data preprocessing module, used for performing data cleaning, outlier removal, time series alignment and standardization on the multidimensional measurement data set to generate a standardized multidimensional measurement data set; The feature extraction and denoising module is used to construct an improved variational autoencoder model, and perform denoising and feature extraction on the standardized multidimensional measurement data set based on the improved variational autoencoder model to obtain a low-dimensional potential feature representation data set and a denoised multidimensional reconstructed measurement data set; A calibration optimization module is used to construct an optimization objective function for tool setting instrument calibration based on a low-dimensional latent feature representation dataset and a denoised multi-dimensional reconstructed measurement dataset; A global optimization module is used to perform a global optimization search on the optimization objective function of the tool setting instrument calibration using a cuckoo search optimization algorithm to determine a global optimal calibration parameter set; The calibration control module is used to generate the calibration control strategy of the tool setter of the CNC machining center based on the global optimal calibration parameter set, and to perform dynamic calibration adjustment according to the machining task of the tool setter and the real-time tool status, so as to keep the tool status and calibration parameters in the best match.

2. A method for controlling a tool setting instrument of a CNC machining center, applied to a tool setting instrument control system of a CNC machining center according to claim 1, characterized in that: The steps include: S1. Use sensors to collect multi-dimensional measurement data sets of tool setting instruments in CNC machining centers; S2. Perform data cleaning, outlier removal, time series alignment and standardization on the multidimensional measurement data set to form a standardized multidimensional measurement data set; S3. Construct an improved variational autoencoder model, and perform denoising and feature extraction on the standardized multidimensional measurement data set through the improved variational autoencoder model to obtain a low-dimensional potential feature representation data set for tool setting instrument calibration of a CNC machining center and a denoised multidimensional reconstructed measurement data set; S4. Constructing the optimization objective function of tool setting instrument calibration based on the low-dimensional latent feature representation dataset and the denoised multi-dimensional reconstructed measurement dataset; S5. Use the cuckoo search optimization algorithm to perform a global search on the optimization objective function of the tool setting instrument calibration to determine the global optimal calibration parameter set; S6. Generate a calibration control strategy for the tool setter of the CNC machining center based on the global optimal calibration parameter set, and dynamically calibrate and adjust the tool setter to achieve matching control between the tool state of the tool setter and the calibration parameters.

3. A CNC machining center tool setting instrument control method according to claim 2, characterized in that: The S1 comprises the following steps: S11. Use a sensor array to measure the tool state of the tool setting instrument in the CNC machining center. The sensor array includes a displacement sensor, a vibration sensor, a temperature sensor and an environmental monitoring sensor, which respectively collect the tool displacement signal, acceleration signal, temperature signal and environmental impact signal to construct an original multidimensional measurement data set. ; S12. Divide the original multidimensional measurement data set into data acquisition windows, set a fixed time window, and group the measurement data in the original multidimensional measurement data set in chronological order so that each time window contains Group measurement data to form a windowed multi-dimensional measurement data set .

4. A method for controlling a tool setting instrument of a CNC machining center according to claim 3, characterized in that: The S2 comprises the following steps: S21. The windowed multidimensional measurement data set Perform data cleaning, remove missing values, duplicate values ​​and abnormal measurement data, and construct a cleaned multidimensional measurement data set; S22. According to the tool motion state and the environmental influence signal distribution characteristics, the mean and standard deviation of each measurement signal are calculated, and the abnormal judgment threshold is set to eliminate them, and a multidimensional measurement data set is obtained after the abnormal values ​​are eliminated; S23. Performing time series alignment on the multidimensional measurement data set after removing outliers according to the timestamp, using interpolation to complete the missing data in the multidimensional measurement data set, and constructing a time series aligned multidimensional measurement data set; S24. Normalize the multidimensional measurement data set aligned with the time series according to the standardization rules, and map each measurement signal to the interval , forming a standardized multidimensional measurement data set .

5. A method for controlling a tool setting instrument of a CNC machining center according to claim 4, characterized in that: The S3 comprises the following steps: S31. Standardizing multidimensional measurement data sets An improved variational autoencoder model for deep noise reduction and feature extraction of tool setting instrument in CNC machining center is constructed using the : ； in, To standardize the data points in a multidimensional measurement dataset, To improve the encoder network of the variational autoencoder model, is the encoder network parameter, which is used to adaptively capture the dynamic correlation characteristics between the tool's displacement signal, acceleration signal, temperature signal and environmental impact signal. To improve the decoder network of the variational autoencoder model, are the decoder network parameters used to reconstruct the noise-reduced measurement data, To standardize latent feature representations for multidimensional measurement datasets; S32. Construct a latent feature mapping function that includes spatial attention mechanism and temporal attention mechanism, highlight the key signal dimensions and key moments of the tool state of the CNC machining center, and define the latent feature representation in the latent feature mapping function for: ； in, , They represent the mean and variance vectors of the tool state features output by the encoder network, respectively, describing the potential feature distribution of the tool state. is a random variable with standard normal distribution, It is a spatial attention factor utilization function used to extract the spatial correlation characteristics between the tool's displacement signal, acceleration signal, temperature signal and environmental impact signal. is the spatial attention parameter, is the temporal attention factor utilization function, which is used to capture the sensitivity difference of the tool measurement signal at the preset time in the machining process. is the time attention parameter, l represents the index of the tool state feature in the spatial dimension, m represents the index of the tool state feature in the temporal dimension, and Represent the data points in the standardized multidimensional measurement data set in the spatial dimension and the temporal dimension, respectively. represents the Hadamard product; S33. Construct a decoder reconstruction function that integrates the dynamic error feedback mechanism, and reconstruct the error of the previous step in real time during the calibration of the tool setting instrument in the CNC machining center As a dynamic adjustment factor, define the reconstruction data : ； in, represents the initial reconstruction output of the decoder network based on the latent feature representation, is the error correction weight matrix, is a learnable parameter, is the reconstruction error of the current step, To dynamically adjust the intensity factor, Reconstruct the error for the previous step; S34. Constructing a joint spatial-temporal attention-constrained deep denoising feature extraction loss function : ； in, is the reconstruction error term, KL is the divergence loss function, is the KL divergence term, which is used to constrain the distribution of the potential features of the tool state. To standardize the total number of data in the multidimensional measurement dataset, is the standard normal distribution prior, and are spatial attention sparse constraint term and temporal attention sparse constraint term respectively, and is the weight coefficient; S35. Back-propagate based on the loss function of the joint spatial-temporal attention constraint to optimize the encoder network parameters , decoder network parameters And the spatial attention parameter , temporal attention parameters , generating an improved variational autoencoder model for CNC machining center tool setting instrument measurement data: S36. Using the improved variational autoencoder model to perform deep noise reduction and feature extraction on the standardized multidimensional measurement data set, a low-dimensional latent feature representation data set is obtained for tool setting calibration of CNC machining centers and the multi-dimensional reconstructed measurement data set after denoising .

6. A method for controlling a tool setting instrument of a CNC machining center according to claim 5, characterized in that: The S4 comprises the following steps: S41. Define calibration parameter set , each calibration parameter in the calibration parameter set Used to describe the tool position compensation, dynamic deviation adjustment and signal compensation factors involved in the calibration process of the tool setting instrument of the CNC machining center; S42. Construct tool setting probe calibration mapping function , the input of the knife calibration mapping function is a low-dimensional latent feature representation dataset and calibration parameter set The output is the predicted measurement data after the tool setting instrument is calibrated : ； in, For the The low-dimensional potential feature representation of the data points obtained after dimensionality reduction by the improved variational autoencoder model, is the mapping weight matrix, is the bias vector, For depicting the The calibration parameters represent the low-dimensional latent features. The function of influence, is the total number of calibration parameters, represents the jth calibration parameter in the calibration parameter set; S43. Constructing the tool setting calibration optimization objective function , tool setting calibration optimizes the objective function to minimize the predicted measurement data and the reconstructed multidimensional measurement data set after noise reduction The mean square error between is the target: ； in, Indicates The calibration prediction error for data points, The multi-dimensional measurement data reconstructed after noise reduction is is the regularization coefficient, is the regularization term.

7. A method for controlling a tool setting instrument of a CNC machining center according to claim 6, characterized in that: The S5 comprises the following steps: S51. Initialize the cuckoo population and define the initial position of the cuckoo population individuals as the calibration parameter set , the population size is set to , the cuckoo population initialization is defined as: ； in, Indicates The initial calibration parameter set of cuckoo individuals, Indicates Among the cuckoo individuals The initial values ​​of the calibration parameters; S52. Optimize the objective function based on tool setting calibration Evaluate the fitness of each individual in the cuckoo population and obtain the initial fitness value: ； in, For the The initial fitness value of each cuckoo individual reflects the prediction error and stability of the corresponding calibration parameter set in the tool setting instrument calibration; S53. Use the Levy flight mechanism to update the position of each individual in the cuckoo population and obtain an updated calibration parameter set : ； in, For the After the update The set of calibration parameters for each cuckoo individual, For the After the update The calibration parameter set of each cuckoo individual represents the updated value of the tool setting calibration parameter of the CNC machining center. is the step size factor for position update, represents the random flight step vector following the Levy distribution, For the The set of calibration parameters of the individuals with the best fitness in the generation population; S54. Using the updated calibration parameter set Calculate the fitness value again ; S55. According to the optimization principle of cuckoo search optimization algorithm, perform fitness comparison and replacement within the population and update the optimal position of the individuals in the population : ； S56. Perform iterative search until the set termination condition is met to obtain the global optimal calibration parameter set : ； in, is the iteration termination time, is the calibration optimization objective function value corresponding to the qth cuckoo individual in the Tth iteration.

8. A method for controlling a tool setting instrument of a CNC machining center according to claim 7, characterized in that: The S6 comprises the following steps: S61. Extract the calibration control strategy from the global optimal calibration parameter set, and set the calibration strategy category of the tool setting instrument of the CNC machining center according to the tool displacement compensation, tool wear correction, machining environment interference correction and dynamic error adjustment parameters: Static calibration strategy, suitable for standard working conditions, tool displacement compensation parameters In the normal range, tool wear correction parameters ≤Preset threshold, dynamic error adjustment parameter In a stable state; The adaptive calibration strategy is suitable for working conditions where the machining environment change is greater than the threshold or the tool wear rate is greater than the threshold. Out of normal range or tool wear correction parameters When it is higher than the preset threshold, the calibration control strategy is adjusted to the adaptive mode; Emergency calibration strategy, suitable for abnormal tool vibration, machining accuracy reduction or error exceeding the limit, when the dynamic error adjustment parameters When it is higher than the set warning value, the calibration control strategy is adjusted to the emergency correction mode and the real-time compensation mechanism is triggered; S62. Construct a calibration control matrix based on the tool status of the tool setting instrument of the CNC machining center and the type of machining task, classify the tool calibration accuracy requirements in combination with the tool calibration history data, and dynamically match the optimal calibration parameters when switching tasks; S63. Real-time monitoring of tool displacement signals, acceleration signals, temperature signals and environmental impact signals. When the measured data deviates from the historical calibration mean and exceeds the preset error range, the calibration control strategy is adaptively adjusted to adapt the calibration parameters to the dynamic changes of the machining environment; S64. A feedback optimization mechanism is adopted. After performing calibration adjustment, the calibration error is calculated based on the real-time measurement data, and the calibration control strategy is updated. When the calibration error is lower than the set threshold, the current calibration parameters are maintained. If the error exceeds the limit, the calibration parameters are dynamically corrected and the calibration control matrix is ​​optimized to realize the calibration control of the tool setting instrument of the CNC machining center.

Images