An optimization method and system for ball screw pairs driven by digital twins

Abstract

The present invention provides an optimization method and system for ball screw pairs driven by digital twins, including: obtaining the original data and scene perception data of the ball screw pair; establishing a digital twin three-dimensional digital model including a ball screw pair model based on the original data and the scene perception data; obtaining an accuracy prediction index based on the digital twin three-dimensional digital model; obtaining an accuracy loss prediction result based on the obtained scene perception data and a preset accuracy loss prediction model; obtaining an operation accuracy optimization strategy for the ball screw pair based on the accuracy loss prediction result and a preset accuracy optimization model; the present invention realizes the modeling, accuracy loss prediction and accuracy optimization of the ball screw pair, provides a theoretical support and research method for improving the transmission accuracy of the ball screw pair; and provides an optimization idea for obtaining a ball screw pair that meets the production requirements of high-end CNC equipment.

Description

Technical Field

[0001] The present invention belongs to the technical field of ball screw pairs, and particularly relates to an optimization method and system for ball screw pairs driven by digital twin. Background Technique

[0002] As an important precision transmission functional component in intelligent numerical control equipment and industrial fields such as aerospace, the transmission accuracy of the ball screw pair is directly related to the use accuracy of high-end equipment; in recent years, the ball screw pair products have made obvious progress in transmission accuracy, but there is still room for improvement

[0003] The inventor found that for the existing ball screw pairs, their transmission accuracy still cannot meet the production requirements of high-grade numerical control equipment, seriously restricting the production of high-grade numerical control equipment and affecting the production of high-precision products Summary of the Invention

[0004] In order to solve the above problems, the present invention proposes an optimization method and system for ball screw pairs driven by digital twin. The present invention is based on the accuracy loss prediction and its optimization design method of ball screw pairs driven by digital twin. Through the application of digital twin technology, the modeling, accuracy loss prediction and accuracy optimization of ball screw pairs are realized, providing theoretical support and research methods for improving the transmission accuracy of ball screw pairs

[0005] In order to achieve the above object, the present invention is realized through the following technical solutions:

[0006] In the first aspect, the present invention provides an optimization method for ball screw pairs driven by digital twin, including:

[0007] Obtain the original data and scenario perception data of the ball screw pair;

[0008] According to the original data and the scenario perception data, establish a digital twin three-dimensional digital model including a ball screw pair model; according to the digital twin three-dimensional digital model, obtain an accuracy prediction index;

[0009] According to the obtained scenario perception data and a preset accuracy loss prediction model, obtain an accuracy loss prediction result; wherein, the accuracy loss prediction model is trained according to the scenario perception data, the accuracy prediction index and a fuzzy neural network

[0010] Based on the precision loss prediction result and a preset precision optimization model, an operation precision optimization strategy for the ball screw pair is obtained. Specifically, in the operation scenario, the precision loss result in this scenario is obtained through the precision loss prediction model, and then the operation parameters and structural parameters of the ball screw pair are coupled and adjusted through the precision optimization model. Furthermore, the optimal value of the operation precision is obtained based on the precision loss prediction model. Among them, the precision optimization model is constructed using an artificial neural network.

[0011] Furthermore, based on the scenario perception data, the service life self-update of the ball screw pair model is realized.

[0012] Furthermore, according to different application working conditions, the ball screw pair model is self-configured and self-optimized.

[0013] Furthermore, a ball screw pair scenario perception model is constructed. Using historical perception data, the ball screw pair scenario perception model is trained and data mining and processing are realized to achieve the intelligent perception of the application scenario of the physical ball screw pair by the digital twin three-dimensional digital model.

[0014] Furthermore, when constructing the precision optimization model, based on the context decoupled data for multi-parameter optimization, the backpropagation algorithm is used to train the multi-parameter neural network model to obtain a multi-parameter function model. Based on the precision loss law of the ball screw pair reflected by the digital twin three-dimensional digital model, the dynamic update of the constraint conditions is realized. The multi-objective particle swarm algorithm is used to solve the optimal solution set of the model.

[0015] Furthermore, the multi-parameter optimal solution set is analyzed to output the optimal parameters. Online simulation analysis of the optimal parameters is carried out on the digital twin three-dimensional digital model, and the simulation results are fed back to the scenario perception model.

[0016] Furthermore, when analyzing the multi-parameter optimal solution, it includes normalizing the decision matrix, solving the weighted normalized matrix, determining the interval positive ideal solution and negative ideal solution, calculating the proximity of each solution to the ideal solution, and determining the pros and cons of the solutions according to the proximity to the positive ideal solution.

[0017] In the second aspect, the present invention also provides an optimization system for a ball screw pair driven by digital twin, including:

[0018] A data acquisition module, configured to: acquire the original data and scenario perception data of the ball screw pair;

[0019] A digital twin three-dimensional digital model establishment module, configured to: establish a digital twin three-dimensional digital model including the ball screw pair model according to the original data and the scenario perception data; obtain a precision prediction index according to the digital twin three-dimensional digital model;

[0020] A prediction module, configured to: obtain a precision loss prediction result based on the acquired scene perception data and a preset precision loss prediction model; wherein, the precision loss prediction model is trained based on the scene perception data, the precision prediction index, and a fuzzy neural network.

[0021] An optimization module, configured to: obtain an operation precision optimization strategy for the ball screw pair based on the precision loss prediction result and a preset precision optimization model; specifically, in an operation scenario, obtain the precision loss result in this scenario through the precision loss prediction model, and then couple and adjust the operation parameters and structural parameters of the ball screw pair through the precision optimization model; and then obtain the optimal value of the operation precision through the precision loss prediction model; wherein, the precision optimization model is constructed using an artificial neural network.

[0022] In a third aspect, the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the steps of the ball screw pair optimization method based on digital twin driving described in the first aspect are implemented.

[0023] In a fourth aspect, the present invention also provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, and when the processor executes the program, the steps of the ball screw pair optimization method based on digital twin driving described in the first aspect are implemented.

[0024] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0025] The precision loss prediction and its optimization design method for the ball screw pair based on digital twin driving in the present invention realizes the modeling, precision loss prediction, and precision optimization of the ball screw pair through the application of digital twin technology, providing a theoretical support and research method for improving the transmission precision of the ball screw pair; and providing an optimization idea for obtaining a ball screw pair that meets the production requirements of high-end numerical control equipment. Description of the Drawings

[0026] The specification drawings constituting a part of this embodiment are used to provide a further understanding of this embodiment. The schematic embodiments and descriptions of this embodiment are used to explain this embodiment and do not constitute an improper limitation on this embodiment.

[0027] Figure 1 It is the overall technical roadmap of Embodiment 1 of the present invention;

[0028] Figure 2 It is the Digital Twin model and intelligent perception technology framework diagram of Embodiment 1 of the present invention;

[0029] Figure 3The accuracy loss prediction process of the ball screw pair in Embodiment 1 of the present invention;

[0030] Figure 4 The accuracy optimization design process of the ball screw pair in Embodiment 1 of the present invention. Detailed implementation manners

[0031] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0032] It should be noted that the following detailed descriptions are all illustrative and are intended to provide further descriptions of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present application belongs.

[0033] As an important precision transmission functional component in intelligent numerical control equipment and industrial fields such as aerospace, the transmission accuracy of the ball screw pair is directly related to the use accuracy of high-end equipment; in recent years, domestic ball screw by-products have made obvious progress in transmission accuracy, but there is still a gap compared with foreign high-end products; according to statistics, about half of the demand for high-precision rolling functional components from domestic numerical control equipment manufacturers needs to be imported from foreign enterprises, and most of the ball screw pairs are selected from foreign enterprise products, making the imported products of the ball screw pair in high-end numerical control equipment in a dominant position, which is the key for domestic ball screw by-products to catch up with or surpass foreign high-end product levels. To improve the transmission accuracy of domestic ball screw pairs and enhance product performance, the present invention proposes a method for predicting accuracy loss and its optimization design of ball screw pairs based on digital twin drive, mainly applying the concept of digital twin technology to solve the key scientific problems in the modeling, intelligent perception, data-driven accuracy loss prediction and optimization design of the digital twin body of the ball screw pair, providing theoretical support and research methods for improving the transmission accuracy of the ball screw pair.

[0034] Embodiment 1:

[0035] This embodiment provides an optimization method for a ball screw pair based on digital twin drive, including:

[0036] Obtain the original data and scene perception data of the ball screw pair;

[0037] Based on the original data and the scene perception data, establish a digital twin three-dimensional digital model including a ball screw pair model; based on the digital twin three-dimensional digital model, obtain accuracy prediction indicators;

[0038] Based on the obtained scene perception data and a preset accuracy loss prediction model, obtain an accuracy loss prediction result; wherein, the accuracy loss prediction model is trained according to the scene perception data, the accuracy prediction indicators and a fuzzy neural network.

[0039] Based on the predicted precision loss results and a preset precision optimization model, an operation precision optimization strategy for the ball screw pair is obtained; among them, the precision optimization model is constructed using an artificial neural network;

[0040] Specifically, as Figure 1 shown, first, based on multi-domain modeling, parametric, modular, and mathematical equation-based descriptions of the ball screw pair system are realized, and a digital ball screw pair model is constructed in combination with the original data; a precision loss model of the ball screw pair is designed and embedded in a digital twin three-dimensional digital model (Digital Twin model) including the ball screw pair model. Combining with scenario perception data, a precision loss update mechanism is designed to realize the self-update of the usage period of the ball screw pair model; the precision loss model refers to the calculation method of raceway wear. First, multi-scale characterization and force analysis of the rough raceway surface are carried out, and a wear calculation method with time-varying wear rate of the ball screw pair based on fractal characteristics and the modified Archard theory is proposed; the embedding process can be realized by connecting the time-varying wear amount of the raceway with the three-dimensional Digital Twin model and changing the structural parameter dimensions of the digital model at different time stages;

[0041] Analyze the data acquisition requirements, install sensors such as load and positioning measurement on the ball screw pair test bench to realize data acquisition and transmission; construct a scenario perception module, store the data in the historical scenario database, train the perception model and realize data mining and processing to realize the intelligent perception of the physical machine tool application scenario by Digital Twin; store the mined knowledge in the perception database, and the perception data is used by the Digital Twin model, precision prediction, and intelligent optimization module; the scenario perception module is a real-time measurement system for operating parameters (speed, acceleration, temperature rise, load, preload, etc.), and the main measurement sensors include speed, temperature rise, load, and preload sensors. The measured operating scenario data is uploaded to the industrial control computer for analysis and processing;

[0042] Using an artificial neural network and a fuzzy logic system, a fuzzy neural network precision loss prediction model is constructed, trained with the historical data of scenario perception, and analyzed with the real-time data of scenario perception to realize the prediction of the precision loss of the ball screw pair;

[0043] Finally, construct a multi-parameter precision optimization design module based on Digital Twin scenario perception to realize an intelligent operation precision optimization strategy applicable to different time periods of the ball screw pair based on Digital Twin scenario perception.

[0044] As Figure 2As shown, in this implementation, first, using the object-oriented concept, a parametric, modular, and mathematical equation-based description of the ball screw pair system is achieved through multi-domain modeling. Combining with the original data, a digital system performance model is constructed. Guided by application analysis tasks, self-configuration and self-optimization of the ball screw pair model are realized. A mechanism for updating the accuracy loss model is designed to achieve self-update based on model evaluation during the usage period of the ball screw pair model. An accuracy loss model of the ball screw pair is designed and embedded in the Digital Twin model. A scene perception module for the ball screw pair is constructed, and the perceived data is stored in the historical scene database. The perception model is trained and data mining and processing are realized to achieve intelligent perception of the application scenario of the physical ball screw pair by Digital Twin. The mined knowledge is stored in the perception database, and the perceived data is used by the Digital Twin model.

[0045] As Figure 3 shown, according to the reasons for accuracy loss obtained from the accuracy loss module of the ball screw pair, the establishment rules of the prediction index are derived, and the establishment rules of the prediction index are obtained through simulation based on the Digital Twin model. According to the establishment rules of the prediction index, the parameter status of future data is classified and scored according to importance. Then, based on the historical operation information in the scene perception module, it is inversely mapped to the relevant operation states and parameters. Using the historical data, parameter information, and prediction index, the accuracy loss prediction model is trained to obtain an accurate accuracy loss prediction model. When the ball screw pair is operating, the real-time scene information obtained from the scene perception module is input into the accuracy loss prediction model, and the fuzzy neural network is used to predict the operation parameters. According to the accuracy parameterized score, the loss status of the parameter is determined. If the parameter of a certain operation state belongs to a certain accuracy loss state, combined with the fuzzy inference rules, accuracy loss prediction is realized. The predicted result is provided to the operation optimization module to improve the operation accuracy of the ball screw pair. It can be understood that in the operation scenario, the accuracy loss result of this scenario is obtained through the accuracy loss prediction model, and then the operation parameters and structural parameters of the ball screw pair are coupled and adjusted through the accuracy optimization model. Then, the optimal value of the operation accuracy is obtained based on the accuracy loss prediction model to guide the operation and design strategy of the ball screw pair.

[0046] As Figure 4As shown in the figure, in this embodiment, a multi-parameter precision optimization design model based on Digital Twin time-varying state data is first established; the optimization target model is proposed to be constructed by an artificial neural network. Based on the context decoupled data for multi-parameter optimization, the backpropagation algorithm is used to train the multi-parameter neural network model to obtain a multi-parameter function model; based on the precision loss law of the ball screw pair reflected by the Digital Twin model, the dynamic update mechanism of its constraint conditions is studied, and a multi-parameter constraint function is established; secondly, a multi-parameter optimization solution module is designed to solve the above multi-parameter optimization model; the multi-objective particle swarm algorithm is used to solve the optimal solution set of the model; finally, a decision analysis module is designed to perform multi-parameter optimal decision analysis; subsequently, the multi-parameter precision optimization results are analyzed online on the Digital Twin model for the application scenario, and the simulation results are fed back to the scenario perception module for analysis and storage; and the optimal parameters are loaded onto the physical ball screw pair test bench to guide the actual application of the ball screw pair.

[0047] Embodiment 2:

[0048] This embodiment provides an optimization system for a ball screw pair driven by digital twin, including:

[0049] A data acquisition module, configured to: acquire the original data and scenario perception data of the ball screw pair;

[0050] A digital twin three-dimensional digital model establishment module, configured to: establish a digital twin three-dimensional digital model including a ball screw pair model according to the original data and the scenario perception data; obtain a precision prediction index according to the digital twin three-dimensional digital model;

[0051] A prediction module, configured to: obtain a precision loss prediction result according to the acquired scenario perception data and a preset precision loss prediction model; wherein, the precision loss prediction model is trained according to the scenario perception data, the precision prediction index, and a fuzzy neural network;

[0052] An optimization module, configured to: obtain an operation precision optimization strategy for the ball screw pair according to the precision loss prediction result and a preset precision optimization model; specifically, obtain the precision loss result in this scenario through the precision loss prediction model in the operation scenario, and then couple and adjust the operation parameters and structural parameters of the ball screw pair through the precision optimization model; and then obtain the optimal value of the operation precision through the precision loss prediction model; wherein, the precision optimization model is constructed by an artificial neural network.

[0053] Embodiment 3:

[0054] This embodiment provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the steps of the optimization method of the ball screw pair based on digital twin driving described in Embodiment 1 are implemented.

[0055] Embodiment 4:

[0056] This embodiment provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the steps of the optimization method of the ball screw pair based on digital twin driving described in Embodiment 1 are implemented.

[0057] The above are only the preferred embodiments of this embodiment and are not used to limit this embodiment. For those skilled in the art, this embodiment can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this embodiment shall be included within the protection scope of this embodiment.

Claims

1. An optimization method for ball screw pairs driven by digital twins, characterized in that, Including: Obtain the original data and scene perception data of the ball screw pair; Based on the original data and the scene perception data, establish a digital twin three-dimensional digital model including a ball screw pair model; based on the digital twin three-dimensional digital model, obtain accuracy prediction indicators; Based on the obtained scene perception data and a preset accuracy loss prediction model, obtain an accuracy loss prediction result; wherein, the accuracy loss prediction model is trained according to the scene perception data, the accuracy prediction indicators and a fuzzy neural network; Based on the accuracy loss prediction result and a preset accuracy optimization model, obtain an operation accuracy optimization strategy for the ball screw pair; specifically, in the operation scenario, obtain the accuracy loss result in this scenario through the accuracy loss prediction model, and then couple and adjust the operation parameters and structural parameters of the ball screw pair through the accuracy optimization model; then obtain the optimal value of the operation accuracy through the accuracy loss prediction model; wherein, the accuracy optimization model is constructed using an artificial neural network; Based on the accuracy loss reasons obtained from the accuracy loss model of the ball screw pair, obtain the establishment rules of the prediction indicators, and obtain the establishment rules of the prediction indicators through simulation of the digital twin three-dimensional digital model; divide and score the importance of the parameter states of future data according to the establishment rules of the prediction indicators; then, based on the historical operation information in the scene perception model, inversely map and project it to the relevant operation states and parameters; use the historical data and parameter information and the prediction indicators to train the accuracy loss prediction model to obtain an accurate accuracy loss prediction model; when the ball screw pair is operating, input the real-time scene information obtained by the scene perception model into the accuracy loss prediction model, use the fuzzy neural network to predict the operation parameters, and determine the loss state of the parameters according to the accuracy parameterization score; if the parameters of a certain operation state belong to a certain accuracy loss state, combine the fuzzy inference rules to achieve accuracy loss prediction; the predicted result is provided to the operation optimization model to improve the operation accuracy of the ball screw pair.

2. The optimization method of a ball screw pair driven by digital twin according to claim 1, characterized in that Based on the scene perception data, realize the self-update of the usage cycle of the ball screw pair model.

3. The optimization method of a ball screw pair driven by digital twin according to claim 2, characterized in that, Based on different application working conditions, perform self-configuration and self-optimization on the ball screw pair model.

4. The optimization method of a ball screw pair driven by digital twin according to claim 2, characterized in that, Construct a scene perception model for the ball screw pair, use historical perception data to train the scene perception model for the ball screw pair and realize data mining and processing, and realize the intelligent perception of the application scenario of the physical ball screw pair by the digital twin three-dimensional digital model.

5. The optimization method of a ball screw pair driven by digital twin according to claim 1, characterized in that, When constructing the accuracy optimization model, based on the context decoupled data for multi-parameter optimization, use the backpropagation algorithm to train the multi-parameter neural network model to obtain a multi-parameter function model; based on the accuracy loss law of the ball screw pair reflected by the digital twin three-dimensional digital model, realize the dynamic update of the constraint conditions; adopt the multi-objective particle swarm algorithm to solve the optimal solution set of the model.

6. The optimization method of a ball screw pair driven by digital twin as claimed in claim 4 or 5, characterized in that Analyze the multi-parameter optimal solution set, output the optimal parameters, perform online simulation analysis of the optimal parameters on the digital twin three-dimensional digital model, and feedback the simulation results to the scene perception model.

7. The optimization method of a ball screw pair based on digital twin drive according to claim 6, characterized in that, When analyzing the optimal solution of multiple parameters, it includes normalizing the decision matrix, solving the weighted normalized matrix, determining the interval positive ideal solution and negative ideal solution, calculating the proximity degree of each scheme to the ideal solution, and determining the pros and cons order of the schemes according to the proximity degree to the positive ideal solution.

8. An optimization system for ball screw pairs driven by digital twins, characterized in that, It includes: A data acquisition module, configured to: acquire the original data and scenario perception data of the ball screw pair; A digital twin three-dimensional digital model establishment module, configured to: establish a digital twin three-dimensional digital model including the ball screw pair model based on the original data and the scenario perception data; obtain the accuracy prediction index based on the digital twin three-dimensional digital model; A prediction module, configured to: obtain the accuracy loss prediction result based on the acquired scenario perception data and a preset accuracy loss prediction model; wherein, the accuracy loss prediction model is trained according to the scenario perception data, the accuracy prediction index and a fuzzy neural network; An optimization module, configured to: obtain the operation accuracy optimization strategy of the ball screw pair based on the accuracy loss prediction result and a preset accuracy optimization model; specifically, obtain the accuracy loss result in this scenario through the accuracy loss prediction model in the operation scenario, and then couple and adjust the operation parameters and structural parameters of the ball screw pair through the accuracy optimization model; then obtain the optimal value of the operation accuracy based on the accuracy loss prediction model; wherein, the accuracy optimization model is constructed by an artificial neural network; Based on the accuracy loss reason obtained from the accuracy loss model of the ball screw pair, the establishment rule of the prediction index is obtained, and the establishment rule of the prediction index is obtained through simulation of the digital twin three-dimensional digital model; the parameter status of future data is divided and scored according to the importance based on the establishment rule of the prediction index; then, based on the historical operation information in the scenario perception model, it is inversely mapped to the relevant operation status and parameters; the accuracy loss prediction model is trained using the historical data and parameter information as well as the prediction index to obtain an accurate accuracy loss prediction model; when the ball screw pair is operating, the real-time scenario information obtained by the scenario perception model is input into the accuracy loss prediction model, and the operation parameters are predicted using a fuzzy neural network, and the loss status of the parameter is determined according to the accuracy parameterization score; if the parameter of a certain operation status belongs to a certain accuracy loss status, combined with the fuzzy inference rule, the accuracy loss prediction is realized; the predicted result is provided to the operation optimization model to improve the operation accuracy of the ball screw pair.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it realizes the steps of the digital twin-driven ball screw pair optimization method according to any one of claims 1-7.

10. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it realizes the steps of the digital twin-driven ball screw pair optimization method according to any one of claims 1-7.

Images