Ball screw feeding system thermal error prediction method, device, equipment and medium

Abstract

The application provides a kind of thermal error prediction method, device, equipment and medium of ball screw feed system, involve thermal error prediction technical field, the method comprises: obtaining the thermal error data, temperature data and position data of ball screw feed system in preset time period under different conditions;Thermal error data, temperature data and position data are input into thermal error prediction model, and the thermal error prediction result of ball screw feed system output by thermal error prediction model is obtained.The technical scheme of the application is based on the dynamic thermal error prediction model trained based on thermal error sample data, temperature sample data and position sample data, based on the dynamic thermal error prediction model, the feature relationship between time series information such as thermal error data, temperature data and position data in the preset time period is extracted, and the influence of temperature data and position data on thermal error data of ball screw feed system in a period of time is considered according to the feature relationship to obtain the thermal error prediction result, which improves the prediction accuracy.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of thermal error prediction, in particular to a thermal error prediction method, device and equipment of ball screw feeding system and a medium. BACKGROUND

[0002] Ball screw is the core component of most numerical control machine tool feeding system, the change of environmental temperature and the friction between bearing and nut and other parts during the working of the component can cause the change of screw temperature, produce thermal error, and finally affect the positioning accuracy of the feeding system. Therefore, the thermal error research of ball screw feeding system has become an important research direction to improve the positioning accuracy of machine tool.

[0003] At present, in the prior art, the prediction of error usually adopts multiple linear regression theory to establish a ball screw prediction model, which is a static model, only considers the relationship between current temperature and current thermal deformation, and ignores the influence of previous time temperature, therefore, there is the problem of inaccurate thermal error prediction caused by temperature change. SUMMARY

[0004] The present application provides a thermal error prediction method, device and equipment of ball screw feeding system, to solve the defect of inaccurate thermal error prediction caused by temperature change in the prior art by using static model to predict thermal error, only considering the relationship between current temperature and current thermal deformation, and ignoring the influence of previous time temperature, realizing the dynamic thermal error prediction model trained based on the thermal error sample data, temperature sample data and position sample data obtained based on time sequence, extracting the feature relationship between thermal error data, temperature data and position data based on the thermal error prediction model, considering the influence of temperature data and position data on thermal error data of ball screw feeding system in a period of time according to the feature relationship, so as to obtain thermal error prediction result and improve prediction accuracy.

[0005] The present application provides a thermal error prediction method of ball screw feeding system, comprising the following steps.

[0006] Thermal error data, temperature data and position data of ball screw feeding system in a preset time period under different working conditions are obtained; wherein the thermal error data represents the error condition between screw, front bearing and rear bearing of ball screw feeding system during the working process, the temperature data is used to represent the heating condition of ball screw feeding system during the working process, and the position data is used to represent the position condition of workbench of ball screw feeding system during the working process.

[0007] Input the thermal error data, the temperature data and the position data into the thermal error prediction model to obtain a thermal error prediction result of the ball screw feeding system output by the thermal error prediction model; wherein the thermal error prediction model is obtained based on the thermal error sample data, the temperature sample data and the position sample data, and the thermal error prediction model is a model for extracting the feature relationship between the thermal error data, the temperature data and the position data to determine the thermal error prediction result.

[0008] According to the thermal error prediction method of the ball screw feeding system provided by the application, the thermal error data, the temperature data and the position data of the ball screw feeding system in a preset time period under different working conditions are obtained, including: obtaining candidate thermal error data of the ball screw feeding system under different working conditions; obtaining candidate temperature data of the ball screw feeding system under different working conditions; obtaining candidate position data of the ball screw feeding system under different working conditions; and performing normalization processing on the candidate thermal error data, the candidate temperature data and the candidate position data respectively to obtain the thermal error data, the temperature data and the position data.

[0009] According to the thermal error prediction method of the ball screw feeding system provided by the application, the candidate thermal error data of the ball screw feeding system under different working conditions are obtained, including: obtaining a geometric error result of the ball screw feeding system under different working conditions; wherein the geometric error result is obtained by measuring the error of the ball screw feeding system according to a standard; obtaining a positioning error result of the ball screw feeding system; wherein the positioning error result is obtained by measuring the error of the measurement points at different times by segmenting the screw stroke in the ball screw feeding system through a laser interferometer; and determining the candidate thermal error data according to the difference between the positioning error result and the geometric error result.

[0010] According to the thermal error prediction method of the ball screw feeding system provided by the application, the candidate temperature data of the ball screw feeding system under different working conditions are obtained, including: obtaining an initial temperature data set of the ball screw feeding system under different working conditions; wherein the initial temperature data set includes initial temperature data of at least two measurement points in the ball screw feeding system; performing clustering on the initial temperature data to obtain a temperature clustering data set; wherein the temperature clustering data set includes at least two types of temperature clustering data; obtaining a thermal deformation parameter; wherein the thermal deformation parameter is used to represent the thermal deformation of the front bearing, the rear bearing and the screw of the ball screw feeding system; determining a correlation coefficient between the thermal deformation parameter and the initial temperature data of any measurement point; and selecting a temperature clustering data from each type of temperature clustering data in the temperature clustering data set according to the correlation coefficient to obtain the candidate temperature data.

[0011] According to the heat error prediction method of the ball screw feeding system provided by the application, the heat error data, the temperature data and the position data are input into a heat error prediction model to obtain a heat error prediction result of the ball screw feeding system output by the heat error prediction model, which comprises: inputting the heat error data, the temperature data and the position data into a dimension increasing module in the heat error prediction model to obtain a three-dimensional tensor output by the dimension increasing module; inputting the three-dimensional tensor into a hidden feature extraction module in the heat error prediction model to obtain a special tensor output by the hidden feature extraction module; inputting the special tensor into a position encoding module in the heat error prediction model to obtain an encoding result output by the position encoding module; inputting the encoding result into a feature relationship extraction module in the heat error prediction model to obtain the heat error prediction result of the ball screw feeding system output by the feature relationship extraction module; and the feature relationship extraction module is used to extract the feature relationship between the heat error data, the temperature data and the position data according to the encoding result.

[0012] According to the heat error prediction method of the ball screw feeding system provided by the application, the training step of the heat error prediction model comprises the following steps: obtaining heat error sample data, temperature sample data and position sample data; inputting the heat error sample data, the temperature sample data and the position sample data into a heat error basic prediction model to obtain an initial heat error prediction result output by the heat error basic prediction model; and repeatedly iteratively optimizing the heat error basic prediction model according to the initial heat error prediction result to obtain the heat error prediction model.

[0013] According to the heat error prediction method of the ball screw feeding system provided by the application, the heat error prediction model is obtained by repeatedly iteratively optimizing the heat error basic prediction model according to the initial heat error prediction result, which comprises the following steps: after obtaining the initial heat error prediction result, determining whether the model iteration number is equal to a number threshold value; in the case that the model iteration number is equal to the number threshold value, obtaining an optimization hyperparameter; obtaining the heat error prediction model according to the optimization hyperparameter and the heat error basic prediction model; in the case that the model iteration number is less than the number threshold value, determining whether the initial heat error prediction result is less than a result threshold value; in the case that the initial heat error prediction result is less than the result threshold value, updating the optimization hyperparameter and obtaining the heat error prediction model according to the updated optimization hyperparameter and the heat error basic prediction model; in the case that the initial heat error prediction result is greater than or equal to the result threshold value, changing the optimization hyperparameter according to a Bayesian algorithm, updating the heat error basic prediction model, and continuing to iteratively optimize the heat error sample data, the temperature sample data and the position sample data according to the updated heat error basic prediction model to obtain an iteratively optimized initial heat error prediction result, and continuing to return to execute the step of determining whether the model iteration number is equal to the number threshold value after obtaining the initial heat error prediction result until the case that the model iteration number is equal to the number threshold value or the iteratively optimized initial heat error prediction result is less than the result threshold value, and stopping optimization to obtain the heat error prediction model.

[0014] The application further provides a thermal error prediction device of a ball screw feeding system, comprising the following modules.

[0015] The data acquisition module is configured to acquire thermal error data, temperature data and position data of the ball screw feeding system in a preset time period under different working conditions, wherein the thermal error data represent error conditions among the screw, the front bearing and the rear bearing of the ball screw feeding system in a working process, the temperature data are used to represent heat generation conditions of the ball screw feeding system in the working process, and the position data are used to represent position conditions of a worktable of the ball screw feeding system in the working process.

[0016] The error prediction module is configured to input the thermal error data, the temperature data and the position data into a thermal error prediction model to obtain a thermal error prediction result of the ball screw feeding system output by the thermal error prediction model, wherein the thermal error prediction model is obtained based on thermal error sample data, temperature sample data and position sample data, and the thermal error prediction model is a model used to extract a feature relationship among the thermal error data, the temperature data and the position data to determine the thermal error prediction result.

[0017] The application further provides an electronic device comprising a memory, a processor and a computer program stored in the memory and capable of running on the processor, and the processor implements the thermal error prediction method of the ball screw feeding system according to any one of the above when executing the computer program.

[0018] The application further provides a non-transitory computer readable storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the thermal error prediction method of the ball screw feeding system according to any one of the above.

[0019] The application further provides a computer program product comprising a computer program, and the computer program is executed by a processor to implement the thermal error prediction method of the ball screw feeding system according to any one of the above.

[0020] This invention provides a method, apparatus, device, and medium for predicting the thermal error of a ball screw feed system. It acquires thermal error data, temperature data, and position data of the ball screw feed system under different operating conditions within a preset time period. The thermal error data represents the error between the screw, front bearing, and rear bearing during operation; the temperature data represents the heat generation during operation; and the position data represents the position of the worktable during operation. The thermal error data, temperature data, and position data are input into a thermal error prediction model to obtain the thermal error prediction result of the ball screw feed system. The thermal error prediction model is trained based on thermal error sample data, temperature sample data, and position sample data, and is used to extract the feature relationships between the thermal error data, temperature data, and position data to determine the thermal error prediction result. The technical solution of this invention addresses the shortcomings of existing technologies that predict thermal errors using static models. These models only consider the relationship between the current temperature and the current thermal deformation, neglecting the influence of temperatures at previous times. Consequently, they suffer from inaccurate thermal error predictions due to temperature changes. This invention implements a dynamic thermal error prediction model trained using thermal error sample data, temperature sample data, and position sample data obtained from time series data. Based on this model, feature relationships between thermal error data, temperature data, and position data are extracted. The influence of temperature and position data on the thermal error data of the ball screw feed system over a period of time is considered according to these feature relationships, thereby obtaining the thermal error prediction result and improving prediction accuracy. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 This is a flowchart illustrating the thermal error prediction method for the ball screw feed system provided by the present invention.

[0023] Figure 2 This is a schematic diagram of the data processing flow provided by the present invention.

[0024] Figure 3 This is a schematic diagram of the ball screw feed system provided by the present invention.

[0025] Figure 4 This is a schematic diagram of the model training process provided by the present invention.

[0026] Figure 5This is a schematic diagram of the model testing process provided by the present invention.

[0027] Figure 6 This is a flowchart of the model optimization provided by the present invention.

[0028] Figure 7 This is a schematic diagram of the thermal error prediction device for the ball screw feed system provided by the present invention.

[0029] Figure 8 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0031] The following is combined with Figure 1 The present invention describes a method for predicting the thermal error of a ball screw feed system. This method is applicable to predicting the thermal error of a ball screw feed system based on a thermal error prediction model. The execution subject of this method can be an electronic device or a thermal error prediction device for a ball screw feed system installed in the electronic device. The configuration device of the control interface can be implemented by software, hardware, or a combination of both. Figure 1 This is a flowchart illustrating the thermal error prediction method for the ball screw feed system provided by the present invention, as shown below. Figure 1 As shown, the method includes the following steps 101 and 102.

[0032] Step 101: Obtain thermal error data, temperature data, and position data of the ball screw feed system under different working conditions within a preset time period.

[0033] In this step, thermal error data represents the error between the screw, front bearing, and rear bearing of the ball screw feed system during operation. That is, thermal error data is the result of the combined effects of the screw, front bearing, and rear bearing. Temperature data is used to represent the heat generation of the ball screw feed system during operation. Position data is used to represent the position of the worktable of the ball screw feed system during operation, with the worktable moving back and forth on the screw.

[0034] Wherein, different working conditions are the state and performance parameters of the ball screw feeding system under specific working conditions, which may be, for example, working load, temperature and other factors that affect the running state and performance of the ball screw feeding system, which are not limited in the embodiment.

[0035] Different working conditions may be, for example, working conditions with different feeding speeds, such as 3 meters per minute, 6 meters per minute, etc., which are not limited in the embodiment.

[0036] Specifically, the thermal error data of the error conditions between the screw, the front bearing and the rear bearing of the ball screw feeding system in the working process under different working conditions are obtained, the temperature data of the heating conditions of the ball screw feeding system in the working process under different working conditions are obtained, and the position conditions of the workbench of the ball screw feeding system in the working process under different working conditions are obtained.

[0037] In a specific embodiment, obtaining the thermal error data, temperature data and position data of the ball screw feeding system within a preset time period under different working conditions comprises: obtaining candidate thermal error data of the ball screw feeding system under different working conditions; obtaining candidate temperature data of the ball screw feeding system under different working conditions; obtaining candidate position data of the ball screw feeding system under different working conditions; and performing normalization processing on the candidate thermal error data, candidate temperature data and candidate position data respectively to obtain the thermal error data, temperature data and position data.

[0038] Specifically, Figure 2 is a flowchart of data processing provided by the present application, as Figure 2 shown, the steps of obtaining the thermal error data, temperature data and position data of the ball screw feeding system within a preset time period under different working conditions specifically comprise steps 201, 202, 203 and 204.

[0039] Step 201, obtaining candidate thermal error data of the ball screw feeding system under different working conditions.

[0040] In a specific embodiment, obtaining the candidate thermal error data of the ball screw feeding system under different working conditions comprises: obtaining a geometric error result of the ball screw feeding system under different working conditions; wherein the geometric error result is a result obtained by measuring the error of the ball screw feeding system according to a standard; obtaining a positioning error result of the ball screw feeding system; wherein the positioning error result is a result obtained by measuring the error of the measurement points at different times by segmenting the screw stroke in the ball screw feeding system with a laser interferometer; and determining the candidate thermal error data according to the difference between the positioning error result and the geometric error result.

[0041] In this step, the laser interferometer is a high-precision measuring instrument based on the principle of laser interference. The laser interferometer segments the screw stroke in the ball screw feeding system, for example, it can be segmented into 10 segments, that is, 11 measurement point positioning error results are obtained for each round trip. The present embodiment does not limit this.

[0042] According to the standard, the geometric error result obtained by measuring the error of the ball screw feeding system can be, for example, the average of the first five results of GB / T 17421.2-2000 as the geometric error result. The present embodiment does not limit this.

[0043] The result of the geometric error can change slightly according to different working conditions, or it can not change according to different working conditions. The present embodiment does not limit this.

[0044] Specifically, according to the geometric error result obtained by measuring the error of the ball screw feeding system according to the standard, the positioning error result of the error measurement of the measurement point at different times is determined by segmenting the screw stroke in the ball screw feeding system through the laser interferometer, and finally the candidate thermal error data is determined according to the difference between the positioning error result and the geometric error result.

[0045] For example, the average of the first five results obtained by measuring the error of the ball screw feeding system according to the standard can be used as the geometric error result, and 11 measurement points at different times are determined for error measurement by segmenting the screw stroke in the ball screw feeding system through the laser interferometer, and the positioning error result is determined. Finally, the difference between the determined positioning error result and the determined geometric error result is used to determine multiple candidate data, which is determined as candidate thermal error data. For example, for 11 measurement points, the first 5 measurement results of the 11 measurement points are obtained, after the first 5 measurements, the workbench reciprocates, and one movement will obtain 11 forward measurement errors and 11 reverse measurement errors. Usually, one direction is taken, so 11 forward measurement errors of 11 measurement points are obtained. After 200 rounds of movement, the first 5 forward measurement errors of 11 measurement points are used as the geometric error result, and the last 195 forward measurement errors of 11 measurement points are used as the positioning error result. The positioning error result minus the geometric error result obtains the candidate thermal error data.

[0046] Step 202, obtaining candidate temperature data of the ball screw feeding system under different working conditions.

[0047] In this step, the candidate temperature data is obtained by the temperature sensor of the ball screw feeding system under different working conditions. The temperature sensor may be arranged in the front bearing, nut seat, rear bearing, workbench, motor, motor seat, working environment, etc. of the ball screw feeding system, which is not limited in the embodiment.

[0048] In the process of collecting the candidate temperature data under different working conditions, how to determine to stop collecting is that when a certain working condition reaches thermal equilibrium, stop measuring, and consider that the measurement of the working condition is completed (i.e. it usually takes about 2 hours for a working condition to reach thermal equilibrium), which is not limited in the embodiment.

[0049] Specifically, the candidate temperature data of the ball screw feeding system under different working conditions is obtained.

[0050] In a specific embodiment, the candidate temperature data of the ball screw feeding system under different working conditions is obtained, including: obtaining the initial temperature of the ball screw feeding system under different working conditions, performing window sliding filtering on the initial temperature, and obtaining the current environment temperature of the ball screw feeding system under different working conditions. The current environment temperature is used as the basis (because the environment temperature changes, the environment temperature is measured, and the current environment temperature is used as the basis to obtain the current temperature difference), and the initial temperature after window sliding filtering is subtracted from the current environment temperature, thereby obtaining the candidate temperature data.

[0051] The initial temperature is fluctuating data, which may affect the accuracy of the subsequent model, so window sliding filtering is performed, and the window size may be 20, for example. The advantage of this setting is to improve the processing accuracy of the subsequent model.

[0052] In a specific embodiment, the candidate temperature data of the ball screw feeding system under different working conditions is obtained, including: obtaining an initial temperature data set of the ball screw feeding system under different working conditions; wherein the initial temperature data set includes initial temperature data of at least two measurement points in the ball screw feeding system; clustering the initial temperature data to obtain a temperature clustering data set; wherein the temperature clustering data set includes at least two types of temperature clustering data; obtaining a thermal deformation parameter; wherein the thermal deformation parameter is used to represent the thermal deformation of the ball screw feeding system front bearing, rear bearing and screw; determining the correlation coefficient of the thermal deformation parameter and the initial temperature data of any measurement point; selecting a temperature clustering data from each type of temperature clustering data in the temperature clustering data set according to the correlation coefficient to obtain the candidate temperature data.

[0053] In this step, the initial temperature data is clustered, and a Gaussian Mixture Model (GMM) is usually used for clustering analysis and probability density estimation. The GMM is a statistical model for clustering analysis and probability density estimation, which is not limited in the embodiment.

[0054] correlation coefficient The Pearson calculation of the correlation coefficient, for example, can be a Pearson correlation coefficient, also known as a Pearson product-moment correlation coefficient (PPMCC or PCCs), which is used to measure the correlation (linear correlation) between the thermal deformation parameter X and the initial temperature data of the measurement points Y, and the value is between -1 and 1, which is not limited in the embodiment.

[0055] correlation coefficient The calculation of the correlation coefficient is shown in formula (1).

[0056] (1)

[0057] In formula (1), represents the thermal deformation coefficient, represents the initial temperature data of any measurement point, represents the total number of measurement points. represents and the covariance of represents the standard deviation of represents the standard deviation of

[0058] Specifically, a set of initial temperature data of the ball screw feeding system under different working conditions is obtained; wherein the set of initial temperature data includes initial temperature data of at least two measurement points in the ball screw feeding system; the initial temperature data is clustered to obtain a set of temperature clustering data; wherein the set of temperature clustering data includes at least two types of temperature clustering data; a thermal deformation parameter is obtained; wherein the thermal deformation parameter is used to represent the thermal deformation of the front bearing, the rear bearing and the screw of the ball screw feeding system; the correlation coefficient of the thermal deformation parameter and the initial temperature data of any measurement point is determined; according to the correlation coefficient, one temperature clustering data is selected from each type of temperature clustering data in the set of temperature clustering data to obtain a candidate temperature data.

[0059] For example, a set of initial temperature data of the ball screw feeding system under different working conditions is obtained, it is assumed that the set of initial temperature data includes 11 initial temperature data, then the 11 initial temperature data is clustered to obtain a set of temperature clustering data, it is assumed that the set of temperature clustering data includes 3 types of temperature aggregation data, then the correlation coefficient of the initial temperature of any measurement point in the initial temperature data of the 11 measurement points and the thermal deformation parameter is determined, and according to the correlation coefficient, one temperature clustering data is selected from each of the 3 types of temperature aggregation data, so that the 3 selected temperature clustering data are determined as candidate temperature data.

[0060] The advantage of such setting is that selecting part of the temperature clustering data as candidate temperature data is beneficial to reducing redundancy and improving accuracy in subsequent model training and application process.

[0061] Step 203, obtaining candidate position data of the ball screw feeding system under different working conditions.

[0062] Specifically, the measurement points of the screw feeding process in the ball screw feeding system under different working conditions are determined, and the candidate position data corresponding to the measurement points are determined.

[0063] In a specific embodiment, Figure 3 is a schematic diagram of the ball screw feeding system provided by the application, as Figure 3 shown, the ball screw feeding system includes a motor, a front bearing, a rear bearing, and a workbench, the assembly between the front bearing and the rear bearing is a screw, the workbench moves on the screw, 11 black origin points represent measurement points for dividing the screw feeding process into 10 segments, and 0, 50, 100, 150, 200, 250, 300, 350, 400, 450, and 500 under each measurement point represent candidate position data, which is not limited in this embodiment.

[0064] Further, in Figure 3 , the fixed mode of the screw is one end fixed and one end hinged. The front bearing is an angular contact bearing with axial force and radial force to form a fixed end, and the rear bearing is a deep groove ball bearing only bearing radial force to form a hinged end, so that the screw will elongate from the fixed end to the free end after thermal deformation. The screw stroke is divided into 10 segments, and the laser interferometer can obtain the positioning error results of the 11 points at different times.

[0065] The advantage of such setting is that the thermal error of the ball screw feeding system mainly comes from the thermal deformation of the screw, but the heating of different positions of the screw is inconsistent, so it cannot be completely considered that the screw full stroke is linear elongation. In this scheme, the screw is divided into 10 segments, the relationship between the thermal deformation of the 11 position points and the temperature is predicted by the model, and the thermal deformation prediction of the full stroke of the screw is completed by linear interpolation between adjacent positions, which better solves the problem of uneven deformation caused by uneven heating of the screw.

[0066] Step 204, normalize the candidate thermal error data, candidate temperature data and candidate position data respectively to obtain thermal error data, temperature data and position data.

[0067] In this step, the value range and unit of the candidate thermal error data, candidate temperature data and candidate position data are different, and the difference between the data can be large, which can cause some large variables to occupy a large proportion in the subsequent prediction process of the thermal error prediction model, and thus affect the prediction accuracy of the thermal error prediction model. The normalization processing is to standardize the candidate thermal error data, candidate temperature data and candidate position data, and to reduce the difference between the data by normalization processing with the maximum value and the minimum value.

[0068] The formula of the normalization processing is shown in the following formula (2).

[0069] (2)

[0070] In formula (2), represents the value of the original sequence that needs to be normalized, which can be the candidate thermal error data, candidate temperature data or candidate position data; represents the minimum value of the original sequence (candidate thermal error data, candidate temperature data or candidate position data) that needs to be normalized; represents the maximum value of the original sequence (candidate thermal error data, candidate temperature data or candidate position data) that needs to be normalized, represents the thermal error data, temperature data or position data after normalization.

[0071] Specifically, for the candidate thermal error data, the current candidate data in the candidate thermal error data, the minimum candidate data in the candidate thermal error data and the maximum candidate data in the candidate thermal error data are input into formula (2) to obtain the thermal error data; for the candidate temperature data, the current temperature clustering data in the candidate temperature data, the minimum temperature clustering data in the candidate temperature data and the maximum temperature clustering data in the candidate temperature data are input into formula (2) to obtain the normalized temperature data; the current position data, the minimum position data and the maximum position data of the position data of each measurement point are input into formula (2) to obtain the normalized position data.

[0072] The advantage of such setting is that by normalization processing, each data is scaled to a dimensionless pure value.

[0073] In a specific embodiment, before the candidate thermal error data, the candidate temperature data and the candidate position data are normalized to obtain the thermal error data, the temperature data and the position data, the candidate thermal error data, the candidate temperature data and the candidate position data can also be subjected to time alignment processing, which is specifically used to eliminate redundant temperature data, and the present embodiment does not limit this.

[0074] Step 102, input the thermal error data, the temperature data and the position data into the thermal error prediction model to obtain a thermal error prediction result of the ball screw feeding system output by the thermal error prediction model.

[0075] In the present step, the thermal error prediction model is obtained based on thermal error sample data, temperature sample data and position sample data (training set), and the thermal error prediction model is a model for extracting the feature relationship between the thermal error data, the temperature data and the position data to determine the thermal error prediction result. The thermal error prediction model mainly considers the hysteresis of the thermal error and establishes a dynamic thermal error prediction model considering the time sequence. The thermal error prediction model mainly trains the training set by constructing a temporal convolutional network (TCN) convolutional neural network model, optimizes the TCN hyperparameters through Bayesian optimization (BO), processes the comprehensive features learned by the TCN using a multi-head attention mechanism (MHA), and finally establishes a BO-TCN-MHA model (thermal error prediction model).

[0076] Specifically, after obtaining the thermal error data, the temperature data, the position data and the trained thermal error prediction model, the thermal error data, the temperature data and the position data are input into the thermal error prediction model, the feature relationship between the thermal error data, the temperature data and the position data is extracted based on the thermal error prediction model, and thus the thermal error prediction result of the ball screw feeding system output by the thermal error prediction model is obtained based on the feature relationship.

[0077] The advantage of such arrangement is that the BO algorithm is used to optimize the hyperparameters in the TCN to find the optimal network parameters of the model, thereby obtaining the thermal error prediction model and improving the prediction accuracy of the thermal error prediction model in predicting the thermal error prediction result.

[0078] In one specific embodiment, thermal error data, temperature data, and position data are input into a thermal error prediction model to obtain the thermal error prediction result of the ball screw feed system output by the thermal error prediction model. This includes: inputting the thermal error data, temperature data, and position data into a dimensionality-upgrading module in the thermal error prediction model to obtain a three-dimensional tensor output by the dimensionality-upgrading module; inputting the three-dimensional tensor into a hidden feature extraction module in the thermal error prediction model to obtain a special tensor output by the hidden feature extraction module; inputting the special tensor into a position encoding module in the thermal error prediction model to obtain an encoding result output by the position encoding module; and inputting the encoding result into a feature relationship extraction module in the thermal error prediction model to obtain the thermal error prediction result of the ball screw feed system output by the feature relationship extraction module. The feature relationship extraction module is used to extract the feature relationships between the thermal error data, temperature data, and position data based on the encoding results.

[0079] Specifically, after obtaining thermal error data, temperature data, and location data, these data are input into the dimensionality-upgrading module of the thermal error prediction model to obtain a three-dimensional tensor output by the module. This three-dimensional tensor is then input into the hidden feature extraction module of the thermal error prediction model to obtain a special tensor output by the hidden feature extraction module. This special tensor is then input into the location encoding module of the thermal error prediction model to obtain the encoding result output by the location encoding module. The encoding result is then input into the feature relationship extraction module of the thermal error prediction model. The feature relationship extraction module extracts the feature relationships between the thermal error data, temperature data, and location data from the encoding result. Based on these feature relationships, iteratively calculates weighted attention and then passes it through a fully connected layer to obtain the thermal error prediction result of the ball screw feed system output by the feature relationship extraction module.

[0080] The location coding module performs location coding calculations as shown in formulas (3) and (4).

[0081] (3)

[0082] (4)

[0083] In formulas (3) and (4), This indicates the position within the input special tensor (i.e., the position of thermal error data, temperature data, and position data within the special tensor). This represents the dimension of the input special tensor. This represents the dimension of the positional data in the input special tensor. The dimension of a special tensor is Time position The encoding results of the data, The dimension of a special tensor is Time position the encoding result of the data.

[0084] In a specific embodiment, the training step of the thermal error prediction model comprises the following: obtaining thermal error sample data, temperature sample data and position sample data; inputting the thermal error sample data, the temperature sample data and the position sample data into the thermal error base prediction model to obtain an initial thermal error prediction result output by the thermal error base prediction model; and repeatedly iteratively optimizing the thermal error base prediction model according to the initial thermal error prediction result to obtain the thermal error prediction model.

[0085] In this step, initial data is obtained before the thermal error model is trained, the initial data including a test set and a training set, the ratio of the test set to the training set being 1:1, the test set including thermal error data, temperature data and position data; the training set including thermal error sample data, temperature sample data and position sample data, the thermal error sample data, the temperature sample data and the position sample data being obtained based on the same processing mode as steps 201-204, which will not be described here.

[0086] Specifically, in the process of training the thermal error prediction model, the obtained thermal error sample data, temperature sample data and position sample data are input into the thermal error base prediction model to obtain an initial thermal error prediction result output by the thermal error base prediction model; the thermal error base prediction model is repeatedly iteratively optimized according to the initial thermal error prediction result to obtain the thermal error prediction model.

[0087] In this step, the thermal error base prediction model includes a TCN convolutional neural network base model, a base multi-head attention mechanism and a base position encoding.

[0088] In a specific embodiment, Figure 4 is a flowchart of model training provided by the present application, as Figure 4 shown, the training process of the thermal error prediction model specifically comprises steps 401, 402, 403, 404, 405, 406 and 407.

[0089] Step 401, obtaining thermal error sample data, temperature sample data and position sample data.

[0090] Specifically, the thermal error sample data, the temperature sample data and the position sample data are obtained, and the thermal error sample data, the temperature sample data and the position sample data are taken as a training set, so as to train the thermal error prediction model based on subsequent steps.

[0091] Step 402, optimizing the TCN convolutional neural network base model.

[0092] Specifically, the training set of the thermal error sample data, the temperature sample data and the position sample data is input into the TCN convolutional neural network base model, and the TCN convolutional neural network base model is optimized based on BO to obtain a first result output by the TCN convolutional neural network base model.

[0093] Step 403, result encoding.

[0094] Specifically, after obtaining the first result output by the TCN convolutional neural network base model, the first result is position-encoded by the base position encoding to obtain a second result.

[0095] Step 404, weight calculation.

[0096] Specifically, after obtaining the second result output by the base position encoding, the second result is calculated by the base multi-head attention mechanism to obtain a third result.

[0097] Step 405, model training.

[0098] Specifically, after obtaining the third result output by the base attention mechanism, the thermal error base prediction model is optimized and trained according to the third result to obtain an initial thermal error prediction result.

[0099] Step 406, determining whether the training is completed.

[0100] Specifically, after obtaining the initial thermal error prediction result, it is determined whether the training is completed. If the initial thermal error prediction result does not meet the result threshold, it is determined that the training is not completed, and step 405 is returned to continue to be executed by repeated iteration and optimization, until the initial thermal error prediction result meets the result threshold, and step 407 is executed.

[0101] Step 407, obtaining a thermal error prediction model.

[0102] Specifically, if the initial thermal error prediction result meets the result threshold, it is determined that the model training is completed, and a thermal error prediction model is obtained.

[0103] The result threshold is a threshold value for model training judgment, which is not limited in the embodiment.

[0104] The advantages of such setting are that the TCN has the characteristics of high calculation efficiency, good stability and high model precision, and supports parallel feature processing. The hyperparameters are optimized by BO, and the attention mechanism is combined with the deep neural network to avoid problems such as gradient explosion, training redundancy, etc., thereby improving the model training efficiency and training precision.

[0105] In a specific embodiment, after obtaining the trained thermal error prediction model, a test set (thermal error test data, temperature test data and position test data) is further obtained, the thermal error prediction model is tested by the test set, the thermal error test data, the temperature test data and the position test data (test set) are input into the thermal error prediction model, the test result of the ball screw feeding system output by the thermal error prediction model is obtained, and after obtaining the test result, the prediction accuracy of the thermal error prediction model is evaluated by the test result and the evaluation index, Figure 5 is a flowchart of model testing provided by the present application, as shown in Figure 5 , specifically comprising steps 501, 502 and 503.

[0106] Step 501, obtaining a test set.

[0107] Step 502, inputting the test set into the thermal error prediction model to obtain a test result.

[0108] Step 503, after obtaining the test result, evaluating the prediction accuracy of the thermal error prediction model according to the test result and the evaluation index.

[0109] In this step, the evaluation index may include mean absolute error (MAE), root mean squared error (RMSE) and determination coefficient (R-Square, ), which are not limited in the present embodiment.

[0110] The calculation of RMSE is shown in formula (6), and the calculation of MAE is shown in formula (7).

[0111] (5)

[0112] In formula (5), represents the sum of squared residuals, represents the total variance of the test result.

[0113] (6)

[0114] In formula (6), represents the total number of samples of the input test set, represents the true error result of the sample , and represents the error test result.

[0115] (7)

[0116] In formula (7), represents the total number of samples of the input test set, represents the true error result of the sample . represents the test result of the error.

[0117] In a specific embodiment, the thermal error prediction model is obtained by repeatedly iterating and optimizing the thermal error basic prediction model according to the initial thermal error prediction result, including: after obtaining the initial thermal error prediction result, determining whether the model iteration number is equal to the number threshold; in the case that the model iteration number is equal to the number threshold, obtaining the optimization hyperparameter; obtaining the thermal error prediction model according to the optimization hyperparameter and the thermal error basic prediction model; in the case that the model iteration number is less than the number threshold, determining whether the initial thermal error prediction result is less than the result threshold; in the case that the initial thermal error prediction result is less than the result threshold, updating the optimization hyperparameter, and obtaining the thermal error prediction model according to the updated optimization hyperparameter and the thermal error basic prediction model; in the case that the initial thermal error prediction result is greater than or equal to the result threshold, changing the optimization hyperparameter according to the Bayesian algorithm, and updating the thermal error basic prediction model, continuing to iterate and optimize the thermal error sample data, the temperature sample data and the position sample data according to the updated thermal error basic prediction model, obtaining the iteratively optimized initial thermal error prediction result, and continuing to return to execute the step of determining whether the model iteration number is equal to the number threshold after obtaining the initial thermal error prediction result, until the case that the model iteration number is equal to the number threshold or the iteratively optimized initial thermal error prediction result is less than the result threshold, stopping optimization, and obtaining the thermal error prediction model.

[0118] Specifically, Figure 6 is the flowchart of model optimization provided by the present application, as Figure 6 , specifically including steps 601, 602, 603, 604, 605, 606 and 607.

[0119] Step 601, determining the initial thermal error prediction result, and determining the evaluation index according to the initial thermal error prediction result.

[0120] Step 602, determining whether the model iteration number is equal to the number threshold.

[0121] Specifically, after obtaining the initial thermal error prediction result, it is determined whether the model iteration number is equal to the number threshold. In the case that the model iteration number is equal to the number threshold, step 603 is continued to be executed. It is further determined whether the model iteration number is less than the number threshold, and in the case that the model iteration number is less than the number threshold, step 605 is continued to be executed.

[0122] Step 603, obtaining the optimization hyperparameter.

[0123] Specifically, in a case where the number of iterations of the model is equal to the number threshold, the optimized hyperparameters are obtained.

[0124] At step 604, the thermal error prediction model is obtained according to the optimized hyperparameters and the thermal error base prediction model.

[0125] At step 605, it is determined whether the evaluation index is less than the result threshold.

[0126] Specifically, in a case where the number of iterations of the model is less than the number threshold, it is determined whether the evaluation index is less than the result threshold, and in a case where the evaluation index is less than the result threshold, step 606 is performed; in a case where the evaluation index is greater than or equal to the result threshold, step 607 is performed.

[0127] At step 606, the optimized hyperparameters are updated.

[0128] Specifically, in a case where the evaluation index is less than the result threshold, the optimized hyperparameters are updated and the evaluation index is updated, and step 604 is performed again according to the updated optimized hyperparameters and the thermal error base prediction model to obtain the thermal error prediction model.

[0129] At step 607, the optimized hyperparameters are changed according to the Bayesian algorithm, the thermal error base prediction model is iteratively optimized according to the changed optimized hyperparameters, and the training set is processed according to the iteratively optimized thermal error base prediction model.

[0130] Specifically, in a case where the evaluation index is greater than or equal to the result threshold, the optimized hyperparameters are changed according to the Bayesian algorithm, the thermal error base prediction model is iteratively optimized according to the changed optimized hyperparameters, and the training set such as the thermal error sample data, the temperature sample data and the position sample data is processed according to the iteratively optimized thermal error base prediction model, the initial thermal error prediction result is determined, and step 601 is performed again until the number of iterations of the model is equal to the number threshold, or the evaluation index is determined according to the initial thermal error prediction result, the evaluation index is less than the result threshold, the optimized hyperparameters are updated and the evaluation index is updated, and the thermal error prediction model is obtained according to the updated optimized hyperparameters and the thermal error base prediction model, thereby obtaining the trained thermal error prediction model.

[0131] In an embodiment, the thermal error prediction model BO-TCN-MHA is compared with an existing Multilayer Perceptron (MLP) model and a Support Vector Regression (SVR) model.

[0132] Table 1 shows the position data of the lead screw at 0 millimeters (mm), 210 mm, and 420 mm, respectively.

[0133] As shown in Table 1, the units of MAE and RMSE are microns (μm), and the units of position data are millimeters (mm).

[0134] The test set data is imported into the thermal error prediction model BO-TCN-MHA to verify the prediction accuracy of the thermal error prediction model BO-TCN-MHA, and the prediction results of the two models MLP and SVR are compared. The results show that the thermal error prediction model BO-TCN-MHA has high accuracy. As shown in Table 1, it can be seen from Table 1 that the thermal error prediction model BO-TCN-MHA has high accuracy.

[0135] Table 1

[0136]

[0137] In summary, the thermal error prediction model BO-TCN-MHA of the present scheme is different from the conventional static model which only considers the temperature and thermal deformation relationship at the current time. The present thermal error prediction model BO-TCN-MHA is a dynamic model which effectively extracts the time series information features of the thermal error of the feeding system by TCN and considers the influence of temperature on thermal deformation for a period of time.

[0138] This invention provides a method for predicting the thermal error of a ball screw feed system. The method involves acquiring thermal error data, temperature data, and position data of the ball screw feed system under different operating conditions within a preset time period. The thermal error data represents the error between the screw, front bearing, and rear bearing during operation; the temperature data represents the heat generation during operation; and the position data represents the position of the worktable during operation. The thermal error data, temperature data, and position data are input into a thermal error prediction model to obtain the predicted thermal error result of the ball screw feed system. The thermal error prediction model is trained based on thermal error sample data, temperature sample data, and position sample data, and is used to extract the feature relationships between the thermal error data, temperature data, and position data to determine the predicted thermal error result. The technical solution of this invention addresses the shortcomings of existing technologies that predict thermal errors using static models. These models only consider the relationship between the current temperature and the current thermal deformation, neglecting the influence of temperatures at previous times. Consequently, they suffer from inaccurate thermal error predictions due to temperature changes. This invention implements a dynamic thermal error prediction model trained using thermal error sample data, temperature sample data, and position sample data obtained from time series data. Based on this model, feature relationships between thermal error data, temperature data, and position data are extracted. The influence of temperature and position data on the thermal error data of the ball screw feed system over a period of time is considered according to these feature relationships, thereby obtaining the thermal error prediction result and improving prediction accuracy.

[0139] The thermal error prediction device for the ball screw feed system provided by the present invention is described below. The thermal error prediction device for the ball screw feed system described below can be referred to in correspondence with the thermal error prediction method for the ball screw feed system described above.

[0140] Figure 7 This is a schematic diagram of the thermal error prediction device for the ball screw feed system provided by the present invention, with reference to... Figure 7 As shown, the thermal error prediction device 700 of the ball screw feed system includes: a data acquisition module 701 and an error prediction module 702.

[0141] The data acquisition module 701 is used to acquire thermal error data, temperature data, and position data of the ball screw feed system under different working conditions within a preset time period. Among them, the thermal error data represents the error between the screw, front bearing, and rear bearing of the ball screw feed system during operation; the temperature data represents the heat generation of the ball screw feed system during operation; and the position data represents the position of the worktable of the ball screw feed system during operation.

[0142] The error prediction module 702 is configured to input the thermal error data, the temperature data and the position data into a thermal error prediction model to obtain a thermal error prediction result of the ball screw feeding system output by the thermal error prediction model, wherein the thermal error prediction model is trained based on thermal error sample data, temperature sample data and position sample data, and the thermal error prediction model is a model for extracting a feature relationship between the thermal error data, the temperature data and the position data to determine the thermal error prediction result.

[0143] In an example embodiment, the data acquisition module 701 is specifically configured to: acquire candidate thermal error data of the ball screw feeding system under different working conditions; acquire candidate temperature data of the ball screw feeding system under different working conditions; acquire candidate position data of the ball screw feeding system under different working conditions; and perform normalization processing on the candidate thermal error data, the candidate temperature data and the candidate position data respectively to obtain the thermal error data, the temperature data and the position data.

[0144] In an example embodiment, the data acquisition module 701 acquires the candidate thermal error data of the ball screw feeding system under different working conditions, and is specifically configured to: acquire a geometric error result of the ball screw feeding system under different working conditions, wherein the geometric error result is a result obtained by performing error measurement on the ball screw feeding system according to a standard; acquire a positioning error result of the ball screw feeding system, wherein the positioning error result is a result obtained by performing error measurement on measurement points at different times by segmenting a screw stroke in the ball screw feeding system through a laser interferometer; and determine the candidate thermal error data according to a difference between the positioning error result and the geometric error result.

[0145] In an example embodiment, the data acquisition module 701 acquires the candidate temperature data of the ball screw feeding system under different working conditions, and is specifically configured to: acquire an initial temperature data set of the ball screw feeding system under different working conditions, wherein the initial temperature data set includes initial temperature data of at least two measurement points in the ball screw feeding system; perform clustering on the initial temperature data to obtain a temperature clustering data set, wherein the temperature clustering data set includes at least two types of temperature clustering data; acquire a thermal deformation parameter, wherein the thermal deformation parameter is used to represent a thermal deformation condition of a front bearing, a rear bearing and a screw of the ball screw feeding system; determine a correlation coefficient between the thermal deformation parameter and initial temperature data of any measurement point; and select a temperature clustering data from each type of temperature clustering data in the temperature clustering data set according to the correlation coefficient to obtain the candidate temperature data.

[0146] In an example embodiment, the error prediction module 702 is specifically configured to: input the thermal error data, the temperature data, and the position data into a dimension increasing module in the thermal error prediction model to obtain a three-dimensional tensor output by the dimension increasing module; input the three-dimensional tensor into a hidden feature extraction module in the thermal error prediction model to obtain a special tensor output by the hidden feature extraction module; input the special tensor into a position encoding module in the thermal error prediction model to obtain an encoding result output by the position encoding module; and input the encoding result into a feature relationship extraction module in the thermal error prediction model to obtain a thermal error prediction result of the ball screw feeding system output by the feature relationship extraction module; wherein the feature relationship extraction module is configured to extract a feature relationship between the thermal error data, the temperature data, and the position data according to the encoding result.

[0147] In an example embodiment, the apparatus further includes a model training module. The model training module is configured to: obtain thermal error sample data, temperature sample data, and position sample data; input the thermal error sample data, the temperature sample data, and the position sample data into a thermal error base prediction model to obtain an initial thermal error prediction result output by the thermal error base prediction model; and repeatedly iteratively optimize the thermal error base prediction model according to the initial thermal error prediction result to obtain the thermal error prediction model.

[0148] In an example embodiment, the model training module repeatedly iteratively optimizes the thermal error base prediction model according to the initial thermal error prediction result to obtain the thermal error prediction model, and is specifically configured to: after obtaining the initial thermal error prediction result, determine whether a model iteration number is equal to a number threshold; in a case where the model iteration number is equal to the number threshold, obtain an optimization hyperparameter; obtain the thermal error prediction model according to the optimization hyperparameter and the thermal error base prediction model; in a case where the model iteration number is less than the number threshold, determine whether the initial thermal error prediction result is less than a result threshold; in a case where the initial thermal error prediction result is less than the result threshold, update the optimization hyperparameter, and obtain the thermal error prediction model according to the updated optimization hyperparameter and the thermal error base prediction model; in a case where the initial thermal error prediction result is greater than or equal to the result threshold, change the optimization hyperparameter according to a Bayesian algorithm, update the thermal error base prediction model, continue to iteratively optimize the thermal error sample data, the temperature sample data, and the position sample data according to the updated thermal error base prediction model to obtain an iteratively optimized initial thermal error prediction result, and continue to return to execute, after obtaining the initial thermal error prediction result, the determination of whether the model iteration number is equal to the number threshold, until the case where the model iteration number is equal to the number threshold or the iteratively optimized initial thermal error prediction result is less than the result threshold, and stop optimization to obtain the thermal error prediction model.

[0149] The device of the embodiment can be used to execute the method of any one of the embodiments of the thermal error prediction method of the ball screw feeding system, and the specific implementation process and technical effects are similar to those of the thermal error prediction method of the ball screw feeding system. For details, refer to the detailed description in the thermal error prediction method of the ball screw feeding system, which will not be repeated here.

[0150] Figure 8 is a structural schematic diagram of an electronic device provided by the application, as Figure 8 shown, the electronic device can include a processor 810, a communications interface 820, a memory 830, and a communications bus 840, wherein the processor 810, the communications interface 820, and the memory 830 complete mutual communication through the communications bus 840. The processor 810 can invoke the logic instructions in the memory 830 to execute the thermal error prediction method of the ball screw feeding system, which includes: acquiring thermal error data, temperature data, and position data of the ball screw feeding system within a preset time period under different working conditions; wherein the thermal error data represents the error condition between the screw, the front bearing, and the rear bearing of the ball screw feeding system during the working process, the temperature data is used to represent the heating condition of the ball screw feeding system during the working process, and the position data is used to represent the position condition of the worktable of the ball screw feeding system during the working process; inputting the thermal error data, the temperature data, and the position data into a thermal error prediction model to obtain a thermal error prediction result of the ball screw feeding system output by the thermal error prediction model; wherein the thermal error prediction model is obtained by training based on thermal error sample data, temperature sample data, and position sample data, and the thermal error prediction model is a model for extracting the feature relationship between the thermal error data, the temperature data, and the position data to determine the thermal error prediction result.

[0151] In addition, the logic instructions in the memory 830 described above can be implemented in the form of a software functional unit and sold or used as an independent product, which can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the application essentially or the part that contributes to the prior art or part of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the embodiments of the application. The foregoing storage medium includes a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and various media that can store program codes.

[0152] In another aspect, the present application also provides a computer program product, which comprises a computer program, the computer program being stored in a non-transitory computer-readable storage medium, and the computer program being capable of executing the thermal error prediction method of the ball screw feeding system provided by the above-mentioned methods when executed by a processor, the method comprising: obtaining thermal error data, temperature data and position data of the ball screw feeding system in a preset time period under different working conditions; wherein the thermal error data represents error conditions among the screw, the front bearing and the rear bearing of the ball screw feeding system in the working process, the temperature data is used to represent heat generation conditions of the ball screw feeding system in the working process, and the position data is used to represent position conditions of the worktable of the ball screw feeding system in the working process; inputting the thermal error data, the temperature data and the position data into a thermal error prediction model to obtain a thermal error prediction result of the ball screw feeding system output by the thermal error prediction model; wherein the thermal error prediction model is obtained based on thermal error sample data, temperature sample data and position sample data, and the thermal error prediction model is a model used to extract a feature relationship among the thermal error data, the temperature data and the position data to determine the thermal error prediction result.

[0153] In another aspect, the present application also provides a non-transitory computer-readable storage medium, which stores a computer program, the computer program being capable of implementing the thermal error prediction method of the ball screw feeding system provided by the above-mentioned methods when executed by a processor, the method comprising: obtaining thermal error data, temperature data and position data of the ball screw feeding system in a preset time period under different working conditions; wherein the thermal error data represents error conditions among the screw, the front bearing and the rear bearing of the ball screw feeding system in the working process, the temperature data is used to represent heat generation conditions of the ball screw feeding system in the working process, and the position data is used to represent position conditions of the worktable of the ball screw feeding system in the working process; inputting the thermal error data, the temperature data and the position data into a thermal error prediction model to obtain a thermal error prediction result of the ball screw feeding system output by the thermal error prediction model; wherein the thermal error prediction model is obtained based on thermal error sample data, temperature sample data and position sample data, and the thermal error prediction model is a model used to extract a feature relationship among the thermal error data, the temperature data and the position data to determine the thermal error prediction result.

[0154] The device embodiments described above are only schematic, wherein the units shown as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e., can be located in one place or distributed on a plurality of network units. Part or all of the modules can be selected according to actual needs to achieve the purpose of the present embodiment scheme. Those skilled in the art can understand and implement it without creative labor.

[0155] Those skilled in the art can clearly understand the implementation of the embodiments by the description of the above embodiments, and the embodiments can be implemented by means of software and necessary universal hardware platforms, and of course, can also be implemented by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, and the computer software product can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the method described in each embodiment or some parts of the embodiment.

[0156] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for some technical features thereof; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A thermal error prediction method of a ball screw feed system, characterized by, The method comprises the steps of: acquiring thermal error data, temperature data and position data of a ball screw feeding system under different working conditions within a preset time period; wherein the thermal error data represents error conditions between a screw, a front bearing and a rear bearing of the ball screw feeding system during the working process, the temperature data is used to represent heat generation conditions of the ball screw feeding system during the working process, and the position data is used to represent position conditions of a workbench of the ball screw feeding system during the working process; inputting the thermal error data, the temperature data and the position data into a thermal error prediction model to obtain a thermal error prediction result of the ball screw feeding system output by the thermal error prediction model; wherein the thermal error prediction model is trained based on thermal error sample data, temperature sample data and position sample data, and the thermal error prediction model is a model used to extract a feature relationship between the thermal error data, the temperature data and the position data to determine the thermal error prediction result; the inputting of the thermal error data, the temperature data and the position data into the thermal error prediction model to obtain the thermal error prediction result of the ball screw feeding system output by the thermal error prediction model comprises: inputting the thermal error data, the temperature data and the position data into a dimension increasing module in the thermal error prediction model to obtain a three-dimensional tensor output by the dimension increasing module; inputting the three-dimensional tensor into a hidden feature extraction module in the thermal error prediction model to obtain a special tensor output by the hidden feature extraction module; inputting the special tensor into a position encoding module in the thermal error prediction model to obtain an encoding result output by the position encoding module; and inputting the encoding result into a feature relationship extraction module in the thermal error prediction model to obtain the thermal error prediction result of the ball screw feeding system output by the feature relationship extraction module; wherein the feature relationship extraction module is used to extract a feature relationship between the thermal error data, the temperature data and the position data according to the encoding result.

2. The thermal error prediction method of a ball screw feed system according to claim 1, characterized by, The acquiring of the thermal error data, the temperature data and the position data of the ball screw feeding system under different working conditions within a preset time period comprises: acquiring candidate thermal error data of the ball screw feeding system under different working conditions; acquiring candidate temperature data of the ball screw feeding system under different working conditions; acquiring candidate position data of the ball screw feeding system under different working conditions; respectively performing normalization processing on the candidate thermal error data, the candidate temperature data and the candidate position data to obtain the thermal error data, the temperature data and the position data.

3. The thermal error prediction method of a ball screw feed system according to claim 2, characterized by, The acquiring of the candidate thermal error data of the ball screw feeding system under different working conditions comprises: acquiring a geometric error result of the ball screw feeding system under different working conditions; wherein the geometric error result is obtained by performing error measurement on the ball screw feeding system according to a standard specification; Obtaining a positioning error result of the ball screw feeding system; wherein the positioning error result is an error measurement result of a laser interferometer on a screw stroke in the ball screw feeding system by segmenting to determine measurement points at different times; Determining the candidate thermal error data according to a difference between the positioning error result and the geometric error result.

4. The thermal error prediction method of a ball screw feed system according to claim 2, characterized by, The candidate temperature data of the ball screw feeding system under different working conditions includes: Obtaining an initial temperature data set of the ball screw feeding system under different working conditions; wherein the initial temperature data set includes initial temperature data of at least two measurement points in the ball screw feeding system; Clustering the initial temperature data to obtain a temperature clustering data set; wherein the temperature clustering data set includes at least two types of temperature clustering data; Obtaining a thermal deformation parameter; wherein the thermal deformation parameter is used to represent the thermal deformation of the front bearing, the rear bearing and the screw of the ball screw feeding system; Determining a correlation coefficient of the thermal deformation parameter and the initial temperature data of any measurement point; According to the correlation coefficient, selecting one temperature clustering data from each type of temperature clustering data in the temperature clustering data set to obtain the candidate temperature data.

5. The thermal error prediction method of a ball screw feed system according to claim 1, characterized by, The training steps of the thermal error prediction model include the following: Obtaining the thermal error sample data, the temperature sample data and the position sample data; Inputting the thermal error sample data, the temperature sample data and the position sample data into a thermal error basic prediction model to obtain an initial thermal error prediction result output by the thermal error basic prediction model; According to the initial thermal error prediction result, repeatedly iterating and optimizing the thermal error basic prediction model to obtain the thermal error prediction model.

6. The thermal error prediction method of a ball screw feed system according to claim 5, wherein According to the initial thermal error prediction result, repeatedly iterating and optimizing the thermal error basic prediction model to obtain the thermal error prediction model, includes: After obtaining the initial thermal error prediction result, determining whether the model iteration number is equal to the number threshold; In the case that the model iteration number is equal to the number threshold, obtaining an optimization hyperparameter; According to the optimization hyperparameter and the thermal error basic prediction model, obtaining the thermal error prediction model; In the case that the model iteration number is less than the number threshold, determining whether the initial thermal error prediction result is less than a result threshold; In the case that the initial thermal error prediction result is less than the result threshold, updating the optimization hyperparameter, and according to the updated optimization hyperparameter and the thermal error basic prediction model, obtaining the thermal error prediction model; In the case that the initial thermal error prediction result is greater than or equal to the result threshold, the optimization hyperparameters are changed according to a Bayesian algorithm, the thermal error base prediction model is updated, the thermal error sample data, the temperature sample data and the position sample data are continuously iteratively optimized according to the updated thermal error base prediction model, the initial thermal error prediction result after the iterative optimization is obtained, and the process of determining whether the model iteration number is equal to the number threshold after the initial thermal error prediction result is obtained is continuously executed until the model iteration number is equal to the number threshold or the initial thermal error prediction result after the iterative optimization is less than the result threshold, and the optimization is stopped, and the thermal error prediction model is obtained.

7. A thermal error prediction device of a ball screw feed system, characterized by, Comprise: A data acquisition module is configured to acquire thermal error data, temperature data and position data of a ball screw feeding system under different working conditions within a preset time period; wherein the thermal error data represents the error between the screw, the front bearing and the rear bearing of the ball screw feeding system during the working process, the temperature data is used to represent the heating condition of the ball screw feeding system during the working process, and the position data is used to represent the position condition of the workbench of the ball screw feeding system during the working process; An error prediction module is configured to input the thermal error data, the temperature data and the position data into a thermal error prediction model to obtain a thermal error prediction result of the ball screw feeding system output by the thermal error prediction model; wherein the thermal error prediction model is trained based on thermal error sample data, temperature sample data and position sample data, and the thermal error prediction model is a model for extracting the feature relationship between the thermal error data, the temperature data and the position data to determine the thermal error prediction result; the inputting of the thermal error data, the temperature data and the position data into the thermal error prediction model to obtain the thermal error prediction result of the ball screw feeding system output by the thermal error prediction model comprises: inputting the thermal error data, the temperature data and the position data into a dimension increasing module in the thermal error prediction model to obtain a three-dimensional tensor output by the dimension increasing module; inputting the three-dimensional tensor into a hidden feature extraction module in the thermal error prediction model to obtain a special tensor output by the hidden feature extraction module; inputting the special tensor into a position encoding module in the thermal error prediction model to obtain an encoding result output by the position encoding module; and inputting the encoding result into a feature relationship extraction module in the thermal error prediction model to obtain the thermal error prediction result of the ball screw feeding system output by the feature relationship extraction module; wherein the feature relationship extraction module is configured to extract the feature relationship between the thermal error data, the temperature data and the position data according to the encoding result.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and running on the processor, characterized in that, The processor executes the computer program to implement the thermal error prediction method of the ball screw feeding system according to any one of claims 1 to 6. 9.A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program, when executed by a processor, implements the thermal error prediction method of the ball screw feed system according to any one of claims 1 to 6.

Images