A method for optimizing machine tool cutting process parameters based on digital twin

Abstract

The present invention claims protection for a method for optimizing machine tool cutting process parameters based on digital twins, which includes the following steps: realizing multi-dimensional expression including characteristics, functions, and behaviors of the actual cutting process of the machine tool through three-dimensional modeling software and a virtual platform; transmitting text data based on JSON format to reduce the complexity of the underlying code, collecting the process data of the machine tool during the cutting process in real time, and analyzing the influence weight of the process data on the cutting force through grey correlation theory; calling the encapsulated digital twin cutting force prediction model based on the unity3D visualization mapping method, and performing real-time calling to analyze the dynamic changes of the cutting force; generating an optimization plan for the cutting force fluctuation in real time according to the influencing factor weight and the digital twin prediction model, and realizing the relative stability of the cutting force through process parameters to ensure the quality of the mechanical products during the processing.

Description

Technical Field

[0001] The present invention relates to the field of advanced manufacturing and automation technology, and in particular, provides a method for optimizing machine tool cutting process parameters based on digital twin. Background Art

[0002] With the introduction of strategies such as Germany's "Industry 4.0", the United States' "Industrial Internet", and "Made in China 2025", intelligent manufacturing technology has become the focus of world attention. The development of the manufacturing industry has become a strategic goal at present. Large-scale mechanical equipment in the industrial, military, aerospace, and transportation sectors is an important economic pillar. As the mother machine of industry, machine tools are of vital importance for their safety, stability, and process quality. At present, the level of intelligence and digitization of machine tool cutting process optimization is low. The quality error caused by the instability of cutting force during the cutting process usually requires human judgment, which has defects such as strong subjectivity and experience judgment. In the traditional process parameter optimization, the control variable method is often used. After the processing is completed, the factors affecting the fluctuation of cutting force are collected and analyzed to form a processing optimization plan, which restricts the optimization adjustment time. Some scholars also use intelligent algorithms to predict characteristic parameters or analyze the real-time reading of CNC system parameters, ignoring the dynamic simulation process during the cutting process. The key to the optimization of machine tool cutting process lies not only in considering the interference of various factors during the cutting process, but also in the real-time process data monitoring, prediction, and analysis and processing during cutting. Therefore, on the basis of traditional virtual simulation monitoring, the virtual-reality mapping technology of digital twins is used to realize the dynamic changes of the machine tool cutting process. At the same time, the process parameters are analyzed and predicted through the digital twin prediction model. The real-time formation of optimization plans is of great significance to the quality and efficiency of processing.

[0003] After searching, the application publication number CN111695734A, a multi-process planning comprehensive evaluation system and method based on digital twin and deep learning, belongs to the field of process planning virtual simulation. The system includes a physical equipment layer, a data perception layer, an information processing layer, a virtual space layer and a digital twin layer; the data perception layer collects relevant real-time data and historical data of the physical equipment layer during the processing process, and sends the data to the information processing layer for data fusion analysis and processing; at the same time, the collected data is transmitted to the virtual space layer, and a dynamic virtual model corresponding to the physical entity is constructed under the guidance of the requirements of the relevant model in the digital twin layer; the digital twin layer leads the remaining layers, performs virtual-real interactive feedback, fusion analysis and iterative optimization, and carries out comprehensive evaluation and analysis of different process schemes for parts to be processed, so as to achieve process parameter optimization and process route improvement. The present invention improves the flexibility and dynamic adaptability of process planning, can improve resource utilization, shorten processing cycle and reduce production costs.

[0004] (CN111695734A) A multi-process planning comprehensive evaluation system and method based on digital twins and deep learning is an advanced technical concept. It is true that the concept of digital twins has solved many research bottlenecks, but digital twins are not just a process of data collection and data processing. It is more about realizing other functions on the basis of virtual dynamic analysis. The implementation approach based on digital twins must be based on specific entity analysis and research. This patent integrates many data sources, but ignores how such a large data set needs to be collected or what protocols need to be connected, how different data structures interact, and other issues. At the same time, based on the way data is processed, which specific link uses which algorithm, and how should intelligent algorithms interact with virtual scenes. Therefore, this patent is more of a concept integration. Without an object, it is difficult to achieve practical application.

[0005] This patent aims to analyze the factors affecting product quality based on experimental verification, form a prediction model and verify its reliability and the weight of the impact, and realize real-time monitoring of the influencing factors in the realization of virtual scenes. Before the processing is realized, the numerical changes of the cutting force during the processing can be predicted, and the processing plan can be adjusted and optimized in advance. The processing process can realize three-dimensional and two-dimensional dynamic display, and the dynamic fluctuations of the factors affecting product quality (cutting force) can be monitored in real time, providing guidance for further optimization in the later stage. Summary of the invention

[0006] The present invention aims to solve the above problems of the prior art. A method for optimizing machine tool cutting process parameters based on digital twin is proposed. The technical solution of the present invention is as follows:

[0007] A method for optimizing machine tool cutting process parameters based on digital twins, comprising the following steps:

[0008] Through 3D modeling software and virtual platform, the multi-dimensional expression of the characteristics, functions and behaviors of the actual cutting process of the machine tool can be realized;

[0009] Based on the text data transmission in JSON format, the complexity of the underlying code is reduced, the process data of the machine tool during the cutting process is collected in real time, and the influence weight of the process data on the cutting force is analyzed through the grey correlation theory;

[0010] Based on the visual mapping method of unity3D, the encapsulated digital twin cutting force prediction model is called to analyze the dynamic changes of cutting force in real time;

[0011] Based on the weights of influencing factors and the digital twin prediction model, an optimization plan is generated in real time for the cutting force fluctuation, and the relative stability of the cutting force is achieved through process parameters to ensure the quality of mechanical products during the processing.

[0012] Furthermore, the multi-dimensional expression of the characteristics, functions, and behaviors of the actual cutting process of the machine tool is realized through the three-dimensional modeling software and the virtual platform, specifically including:

[0013] Establish a digital twin virtual model of the machine tool through 3D modeling software;

[0014] Lightweight the digital twin virtual model of the machine tool, that is, reduce the number of meshes for the monitored parts, i.e. the workpiece and the tool cutting process, while ensuring the accuracy of the model to avoid distortion during the rendering process; reduce the model accuracy requirements and the number of meshes for non-monitored parts;

[0015] When the lightweight model is imported into the unity3D simulation platform, the virtual model is given the same physical properties and mapping effects as the entity, ensuring that the structure and properties of the virtual model are consistent with the entity and ensuring the visualization effect during cutting processing.

[0016] Furthermore, the text data transmission based on JSON format reduces the complexity of the underlying code, collects the process data of the machine tool in the cutting process in real time, and analyzes the influence weight of the process data on the cutting force through the grey correlation theory, specifically including:

[0017] The OPC UA data interaction architecture is used for data collection. At the same time, the integrated OPC UA client is encapsulated into the digital twin virtual model, and the server of the machine tool equipment registered on the server is accessed through the OPC UA client to obtain the terminal server; at the same time, a connection mechanism between the client and the server is created, and data is read by browsing the address space of the server and the obtained data is stored in the database;

[0018] The collected signals are preprocessed, cleaned, denoised and filtered before being integrated with the dynamic information of the machine tool. The JSON format is used as the transmission format from the machine tool entity to the virtual model in the digital twin.

[0019] The client obtains data by subscription. When the server receives the client's subscription request, it will query the database and feed back the searched machine tool cutting data, namely feed speed, rotation speed, and feed amount to the server to capture the data and ensure its real-time nature. Finally, the actual operation data collected will be converted into a script to initialize and update the virtual model.

[0020] Furthermore, the grey correlation theory is used to analyze the influence weight of process data on cutting force, specifically including:

[0021] Using the C# calling algorithm in the digital twin virtual platform and the experimental data in the database, the gray correlation degree of the cutting force of the digital twin prediction model is calculated, and the influence weights of the process parameters obtained in the three experimental processes are sorted. The specific steps are as follows:

[0022] ① Establish a reference sequence that reflects process data and cutting force system:

[0023] X0(k)=(x0(1),x0(2),…x0(n))

[0024] Where X0(k) represents the cutting force data sequence; x0 is the cutting force data, and n is the number of times;

[0025] Comparison sequence based on the impact of process data factors on the system:

[0026] X i (k)=(x i (1),x i (2),…x i (n)),i=1,2,…m

[0027] Where X i (k) represents a sequence; i=1 is the feed rate; i=2 is the spindle speed; and i=3 is the cutting depth.

[0028] Based on the reference sequence and the comparison sequence, determine the correlation coefficient:

[0029]

[0030] Among them, ξ is the resolution coefficient of the maximum difference between the two levels;

[0031] By the correlation coefficient X i and X o The available grey relational degree is:

[0032]

[0033] Among them, γ oi is the resolution of the system, with ξ∈[0,1], It is called the minimum difference between the two levels. It is called the maximum difference between the two levels, x0(k)-x i (k) is the difference sequence of the system; the grey relational degree γ(X0,X i ) Arrange the sequence from large to small to get the associated sequence set, and use it to judge the sequence X i With X o The degree of correlation; using the displacement difference |X0(k)-X i (k)|Reflects the similarity of the development process or magnitude between two sequences;

[0034] ② The covariance matrix formed by the process data processed by the averaging method reflects the differences in the degree of variation of each indicator in the original process data, and also contains information on the differences in the degree of mutual influence of each process data.

[0035] The transformed data sequence is:

[0036] XD1=(x(1)d1,x(2)d1,…,x(n)d1)

[0037] in,

[0038] is the mean of all data in each sequence; XD1 represents the first mean sequence; in x(n)d1,

[0039] n represents the number of values, d1 is the mean of the first mean sequence

[0040] ③ Select appropriate ξ to reduce the impact of various process data factors in the system on the entire system association:

[0041] Note Δ v is the mean of the absolute values ​​of all difference sequences in the system,

[0042] Δmax is the maximum difference between the two levels, and

[0043] According to the resolution coefficient ξ, the correlation ranking of the process data cutting depth, feed speed and spindle speed is obtained, and finally the influence weight of the influencing factors on the process data is determined.

[0044] Furthermore, the value of ξ is: Δ ≤ξ≤2ε Δ , should meet the following requirements:

[0045] a. When Δmax>3Δ v When Δ ≤ξ≤1.5ε Δ ;

[0046] b. When Δmax≤3Δ v When 1.5ε Δ ≤ξ≤2ε Δ ;

[0047] Calculate Δmax and Δ according to experimental data v The value of , thus obtaining the resolution coefficient ξ.

[0048] Furthermore, the unity3D-based visualization mapping method calls the packaged digital twin cutting force prediction model to perform real-time call analysis on the dynamic changes of the cutting force;

[0049] According to the influencing factors of process parameters, the calculation equation of the physical cutting force of the digital twin machine tool is established:

[0050]

[0051] In the formula: a represents the constant term of the fitting; n represents the spindle speed of the machine tool, in r / min; a p represents the radial depth of cutting, in mm; f is the feed rate of the machine tool, in mm / min; F is the milling force to be obtained, in N; a, b, c, d are the parameters to be fitted;

[0052] First, convert the power function into a multivariate function polynomial, and perform linear fitting on the polynomial; take the natural logarithm of both sides:

[0053] ln F=ln a+bln n+cln a p +dln f

[0054] Substituting the experimental data into the empirical formula, we can finally obtain the corresponding specific values ​​of the four parameters a, b, c, and d, and finally obtain the prediction model F of the physical cutting force of the digital twin machine tool.

[0055] Furthermore, the optimization scheme is generated in real time for the cutting force fluctuation according to the influencing factor weight and the digital twin prediction model, and the relative stability of the cutting force is achieved through the process parameters to ensure the quality of the mechanical products during the processing, specifically including:

[0056] Use programming language to encapsulate and save the final digital twin virtual model and package it into a dLL file, and use C# to call the prediction model in the dLL file in the digital twin virtual simulation platform;

[0057] When the client in unity3D collects data, the digital twin prediction model first predicts the size of the cutting force, and then transfers the data volume to the UI chart component in unity3D. At the same time, the real-time time function is called to create a dynamic chart of the cutting force and time changes for real-time display;

[0058] For machine tools of different specifications, different process parameters have different influence weights on cutting force. The real-time predicted cutting force size and the ranking of weight influencing factors are important references for the control quantity of subsequent processing optimization, i.e., the selection of process parameters, so as to achieve relative stability of cutting force and ensure the quality and efficiency of the processing process.

[0059] The advantages and beneficial effects of the present invention are as follows:

[0060] In traditional cutting process optimization, analytical data models or offline simulation models are often used to analyze the fluctuation of cutting force. The data has hysteresis, the data information is not closely connected, the fluidity is poor, and it is easy to form information islands; therefore, this patent is based on the virtual simulation technology of digital twins, integrating models, data, intelligent prediction, and optimization, and proposes a method for optimizing machine tool cutting process parameters based on digital twins. Taking the machine tool cutting process as the research object, the cutting force in the machine tool entity is described and analyzed in real time through the digital twin virtual model, and the monitoring, prediction and process optimization of the machine tool cutting process are realized in the context of networking and intelligent manufacturing.

[0061] As a functional module of the digital twin virtual simulation platform, the cutting force prediction model is encapsulated in the virtual platform. The prediction model established based on experimental data is predictable and can predict the size of the cutting force before cutting. At the same time, through the analysis of grey correlation, the weight of the influence of process parameters on the cutting force is determined, and guidance is provided for the formation and adjustment of the processing plan, such as the adjustment range of process parameters and the correct use time of cutting fluid. At the same time, the encapsulated prediction model can collect machine tool processing data information in real time for prediction, predict the cutting force fluctuation according to the time function, and combine the dynamic analysis of the virtual model in the simulation platform to cite the reasons for the fluctuation. Ensure the stability of the cutting force during the processing and further improve the processing quality and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 It is a framework diagram of a method for optimizing machine tool cutting process parameters based on digital twins in a preferred embodiment provided by the present invention. DETAILED DESCRIPTION

[0063] The following will describe the technical solutions in the embodiments of the present invention in detail in conjunction with the accompanying drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention.

[0064] The technical solution of the present invention to solve the above technical problems is:

[0065] It should be pointed out first that the quality of mechanical products is closely related to the processing process. One of its important evaluation indicators is cutting force, which is determined by process parameters such as spindle speed, radial depth, and processing feed rate during the cutting process of the machine tool. Certain processing errors will inevitably occur during the cutting process, and its optimization scheme and the speed of generating the optimization scheme have always been a major problem in actual production and processing. At the same time, frequent fluctuations in cutting force will reduce the life of the shaft and increase tool wear, restricting part quality and processing efficiency. Based on the summary of the influencing factors of the machine tool cutting process and the current status of research on traditional optimization schemes, this patent conducts in-depth research on digital modeling of the cutting process, efficient transmission of data information, and cutting process influence weight issues, and introduces a cutting force prediction model. A machine tool cutting process parameter optimization method based on digital twins is proposed, which is of great significance for guiding the quality and efficiency of machine tool cutting processing and improving the quality of mechanical products.

[0066] In the traditional cutting process optimization, analytical data models or offline simulation models are often used to analyze the fluctuation of cutting force. The data has hysteresis, the data information is not closely connected, the fluidity is poor, and it is easy to form information islands; therefore, this patent is based on the virtual simulation technology of digital twins, integrating models, data, intelligent prediction, and optimization, and proposes a method for optimizing machine tool cutting process parameters based on digital twins. Taking the machine tool cutting process as the research object, the cutting force in the machine tool entity is described and analyzed in real time through the digital twin virtual model, and the monitoring, prediction and process optimization of the machine tool cutting process are studied under the background of networking and intelligent manufacturing. First, the multi-dimensional expression of the characteristics, functions, and behaviors of the actual cutting process of the machine tool is realized through three-dimensional modeling software and virtual platform. Secondly, based on the text data transmission in JSON format, the complexity of the underlying code is reduced, the process data of the machine tool in the cutting process is collected in real time, and the influence weight of the process data on the cutting force is analyzed through the grey correlation theory. Based on the visual mapping method of unity3D, the encapsulated digital twin cutting force prediction model is called to analyze the dynamic changes of the cutting force in real time. Finally, an optimization plan is generated in real time for the cutting force fluctuation based on the influencing factor weights and the digital twin prediction model, and the relative stability of the cutting force is achieved through process parameters to ensure the quality of mechanical products during the processing.

[0067] like Figure 1 As shown, a real-time optimization system for machine tool cutting process parameters based on digital twins includes a physical layer, a virtual layer, a data layer, and a functional layer;

[0068] Physical layer: As the basic component of the digital twin and the source of static and dynamic data, the physical layer mainly consists of two parts: the physical entity of the machine tool processing equipment and the physical sensor. The physical sensor is a sensor installed on various parts of the machine tool, which is used to collect various states, motion signals and process data information of the machine tool during the cutting process.

[0069] Virtual layer: As a mirror reflection of the physical entity in the digital twin, the virtual layer needs to be consistent with the physical entity in terms of function, geometry, physical properties, etc. to ensure the accuracy of cutting process data fitting and mapping.

[0070] Step 1: Create a digital twin virtual model of the machine tool through 3D modeling software.

[0071] Step 2: To ensure the efficiency of the digital twin virtual model in data collection and fitting, the digital twin virtual model in the previous step is lightweighted to reduce the amount of rendering data in the simulation process, with two focuses: reducing the number of meshes for monitored parts (workpieces, cutting tools, etc.) while ensuring the accuracy of the model to avoid distortion during rendering; and reducing the model accuracy requirements and the number of meshes for non-monitored parts.

[0072] Step 3: Unity3D is a virtual platform with good encapsulation and strong 3D interactivity in the open source market. When the lightweight model is imported into the unity3D simulation platform, the virtual model is given the same physical properties and mapping effects as the entity, ensuring that the structure and properties of the virtual model are consistent with the entity, and ensuring the visualization effect during cutting processing.

[0073] Data Layer:

[0074] Step 4:

[0075] The OPC UA data interaction architecture is used to ensure efficient data collection. At the same time, the integrated OPC UA client is encapsulated into the digital twin virtual model, and the server of the machine tool equipment registered on the server is accessed through the OPC UA client to obtain the terminal server. At the same time, a connection mechanism between the client and the server is created, and data is read by browsing the address space of the server and the obtained data is stored in the database.

[0076] Step 5: Due to the presence of various interference factors such as noise and distortion in the process of signal acquisition and transmission, the collected signals need to be preprocessed, cleaned, denoised and filtered before integrating the dynamic information of the machine tool. At the same time, JSON has the characteristics of simple format and convenience, which can reduce the code of the data transmission process. Using JSON format as the transmission format from the machine tool entity to the virtual model in the digital twin can improve the data transmission efficiency.

[0077] Step 6: The client obtains data by subscription. When the server receives the client's subscription request, it will query the database and feed back the machine tool cutting data (feed speed, rotation speed, feed amount) to the server to capture the data and ensure its real-time performance. Finally, the actual operation data collected will be converted into a script to initialize and update the virtual model.

[0078] Functional layer:

[0079] Step 7: While keeping other factors the same, set a certain cutting depth and spindle speed for the machine tool, develop the UGUI interface based on the digital twin virtual simulation platform Unity3D, and display the fluctuation of cutting force at different feed speeds in real time through the digital twin virtual model and UGUI interface;

[0080] Step 8: While keeping other factors the same, set the machine tool spindle speed and feed speed unchanged, develop the UGUI interface based on the digital twin virtual simulation platform Unity3D, and display the fluctuation of cutting force at different cutting depths in real time through the digital twin virtual model and UGUI interface;

[0081] Step 9: While keeping other factors the same, set the machine tool feed depth and feed speed unchanged, develop the UGUI interface based on the digital twin virtual simulation platform Unity3D, and display the fluctuation of cutting force at different spindle speeds in real time through the digital twin virtual model and UGUI interface;

[0082] Step 10: According to the above analysis, the process parameters of a single factor have a certain influence on the change of cutting force. Therefore, analyze the influence weight of each factor on the cutting force. The gray correlation theory algorithm process is encapsulated by programming language. After each experiment in steps 8 to 10 is completed, the C# algorithm in the digital twin virtual platform and the experimental data in the database are used to calculate the gray correlation degree of the cutting force of the digital twin prediction model. Sort by the influence weight of the process parameters obtained in the three experimental processes. The specific steps are as follows:

[0083] ① Establish a reference sequence that reflects process data and cutting force system:

[0084] X0(k)=(x0(1),x0(2),…x0(n))

[0085] Comparison sequence based on the impact of process data factors on the system:

[0086] X i (k)=(x i (1),x i (2),…xi (n)),i=1,2,…m

[0087] Based on the reference sequence and the comparison sequence, the correlation coefficient was determined:

[0088]

[0089] Among them, ξ is the resolution coefficient of the maximum difference between the two levels;

[0090] By the correlation coefficient X i and X o The available grey relational degree is:

[0091]

[0092] Among them, γ oi is the system resolution, and It is called the minimum difference between the two levels. It is called the maximum difference between the two levels, x0(k)-x i (k) is the difference sequence of the system. The grey correlation degree γ(X0,X i ) Arrange the sequence from large to small to get the associated sequence set, and use it to judge the sequence X i With X o The degree of correlation is determined by using the displacement difference |X0(k)-X i (k)|Reflects the similarity of the development process or magnitude between the two sequences.

[0093] ④ The covariance evidence formed by the process data processed by the averaging method can reflect the differences in the degree of variation of each indicator in the original process data, and also contains information on the differences in the degree of mutual influence of each process data.

[0094] The transformed data sequence is:

[0095] XD1=(x(1)d1,x(2)d1,…,x(n)d1)

[0096] in,

[0097] ⑤ Select the appropriate ξ to reduce the impact of various process data factors within the system on the entire system association:

[0098] Note Δ v is the mean of the absolute values ​​of all difference sequences in the system,

[0099] Δmax is the maximum difference between the two levels, and Then the value of ξ is: Δ ≤ξ≤2ε Δ , should meet the following requirements:

[0100] a. When Δmax>3Δ v When Δ ≤ξ≤1.5ε Δ ;

[0101] b. When Δmax≤3Δ v When 1.5ε Δ ≤ξ≤2ε Δ .

[0102] Calculate Δmax and Δ according to experimental data v The value of , thus obtaining the resolution coefficient ξ.

[0103] According to the resolution coefficient ξ, the correlation ranking of the process data (cutting depth, feed speed, spindle speed) is obtained, and finally the influence weight of the influencing factor on the process data is determined.

[0104] ⑥According to the influencing factors of process parameters, the calculation equation of the physical cutting force of the digital twin machine tool is established:

[0105]

[0106] In the formula: n represents the spindle speed of the machine tool, the unit is r / min; a p represents the radial depth of cutting, in mm; f is the feed rate of the machine tool, in mm / min; F is the milling force to be obtained, in N; a, b, c, d are the parameters to be fitted

[0107] First, convert the power function into a multivariate function polynomial and perform linear fitting on the polynomial. Take the natural logarithm of both sides:

[0108] ln F=ln a+bln n+cln a p +dln f

[0109] Secondly, substitute the experimental data into the empirical formula, and finally get the corresponding specific values ​​of the four parameters a, b, c, and d, and finally obtain the prediction model F of the physical cutting force of the digital twin machine tool.

[0110] Step 11: Use programming language to encapsulate and save the final digital twin virtual model and package it into a dLL file, and use C# to call the prediction model in the dLL file in the digital twin virtual simulation platform.

[0111] Step 12: When the client in unity3D collects data, the digital twin prediction model first predicts the size of the cutting force, and then passes the data volume to the UI chart component in unity3D. At the same time, the real-time time function is called to create a dynamic chart of the cutting force and time changes for real-time display.

[0112] Step 13: For machine tools of different specifications, different process parameters have different influence weights on cutting force. The ranking of cutting force size and weight influence factors based on real-time prediction is an important reference for the selection of control quantity (process parameters) for subsequent processing optimization, thereby achieving relative stability of cutting force and ensuring the quality and efficiency of the processing process.

[0113] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0114] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0115] The above embodiments should be understood to be only used to illustrate the present invention and not to limit the protection scope of the present invention. After reading the contents of the present invention, technicians can make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims

1. A method for optimizing machine tool cutting process parameters based on digital twin, characterized in that: The following steps are involved: Through 3D modeling software and virtual platform, the multi-dimensional expression of the characteristics, functions and behaviors of the actual cutting process of the machine tool can be realized; Based on the text data transmission in JSON format, the complexity of the underlying code is reduced, the process data of the machine tool during the cutting process is collected in real time, and the influence weight of the process data on the cutting force is analyzed through the grey correlation theory; Based on the visual mapping method of unity3D, the encapsulated digital twin cutting force prediction model is called to analyze the dynamic changes of cutting force in real time; Generate an optimization plan for cutting force fluctuations in real time based on the weight of influencing factors and the digital twin prediction model, and use process parameters to achieve relative stability of cutting force to ensure the quality of mechanical products during processing; The text data transmission based on JSON format reduces the complexity of the underlying code, collects the process data of the machine tool in the cutting process in real time, and analyzes the influence weight of the process data on the cutting force through the grey correlation theory, specifically including: The OPC UA data interaction architecture is used for data collection. The integrated OPC UA client is encapsulated into the digital twin virtual model. The server of the machine tool equipment registered on the server is accessed through the OPC UA client to obtain the terminal server. At the same time, a connection mechanism between the client and the server is created to read data by browsing the address space of the server and store the obtained data in the database. The collected signals are preprocessed, cleaned, denoised and filtered before being integrated with the dynamic information of the machine tool. The JSON format is used as the transmission format from the machine tool entity to the virtual model in the digital twin. The client obtains data by subscription. When the server receives the subscription request from the client, it will query the database and feed back the machine tool cutting data, i.e. feed speed, rotation speed and feed amount, to the server to capture the data and ensure its real-time performance. Finally, the actual operation data collected will be converted into a script to initialize and update the virtual model. The grey correlation theory is used to analyze the influence weight of process data on cutting force, specifically including: Using the C# calling algorithm in the digital twin virtual platform and the experimental data in the database, the gray correlation degree of the cutting force of the digital twin prediction model is calculated, and the influence weights of the process parameters obtained in the three experimental processes are sorted. The specific steps are as follows: ① Establish a reference sequence that reflects process data and cutting force system: X0(k)=(x0(1),x0(2),…x0(n)) Where X0(k) represents the cutting force data sequence; x0 is the cutting force data, and n is the number of times; Comparison sequence based on the impact of process data factors on the system: X i (k)=(x i (1),x i (2),…x i (n)),i=1,2,3 Where X i (k) represents a sequence; i=1 is the feed rate; i=2 is the spindle speed; i=3 is the cutting depth; Based on the reference sequence and the comparison sequence, determine the correlation coefficient: Among them, ξ is the resolution coefficient of the maximum difference between the two levels; By the correlation coefficient X i and X o The available grey relational degree is: Among them, γ oi is the resolution of the system, with ξ∈[0,1], It is called the minimum difference between the two levels. It is called the maximum difference between the two levels, x0(k)-x i (k) is the difference sequence of the system; the grey relational degree γ(X0,X i ) Arrange the sequence from large to small to get the associated sequence set, and use it to judge the sequence X i With X o The degree of correlation; using the displacement difference |X0(k)-X i (k)|Reflects the similarity of the development process or magnitude between two sequences; ② The covariance matrix of each process data processed by the mean method reflects the difference in the degree of variation of each indicator in the original process data, and also contains the information on the difference in the degree of mutual influence of each process data. The transformed data sequence is: XD1=(x(1)d1,x(2)d1,…,x(n)d1) in, is the mean of all data in each sequence; XD1 represents the first mean sequence; in x(n)d1, n represents the number of values, and d1 represents the mean of the first mean sequence ③ Select appropriate ξ to reduce the impact of various process data factors in the system on the entire system association: Note Δ v is the mean of the absolute values ​​of all difference sequences in the system, Δmax is the maximum difference between the two levels, and According to the resolution coefficient ξ, the correlation ranking of the process data cutting depth, feed speed and spindle speed is obtained, and finally the influence weight of the influencing factors on the process data is determined.

2. The method for optimizing machine tool cutting process parameters based on digital twinning according to claim 1, characterized in that: The multi-dimensional expression of the characteristics, functions, and behaviors of the actual cutting process of the machine tool is achieved through the three-dimensional modeling software and the virtual platform, specifically including: Establish a digital twin virtual model of the machine tool through 3D modeling software; Lightweight the digital twin virtual model of the machine tool, that is, reduce the number of meshes for the monitored parts, i.e. the workpiece and the tool cutting process, while ensuring the accuracy of the model to avoid distortion during the rendering process; reduce the model accuracy requirements and the number of meshes for non-monitored parts; When the lightweight model is imported into the unity3D simulation platform, the virtual model is given the same physical properties and mapping effects as the entity, ensuring that the structure and properties of the virtual model are consistent with the entity and ensuring the visualization effect during cutting processing.

3. The method for optimizing machine tool cutting process parameters based on digital twinning according to claim 1, characterized in that: Then the value of ξ is: Δ ≤ξ≤2ε Δ , should meet the following requirements: a. When Δmax > 3Δ v At this time, ε Δ ≤ ξ ≤ 1.5ε Δ ; b. When Δmax≤3Δ v When 1.5ε Δ ≤ξ≤2ε Δ ; Calculate Δmax and Δ according to experimental data v The value of , thus obtaining the resolution coefficient ξ.

4. The method for optimizing machine tool cutting process parameters based on digital twinning according to claim 3 is characterized in that: The unity3D-based visualization mapping method calls the packaged digital twin cutting force prediction model to perform real-time call analysis on the dynamic changes of the cutting force; The calculation equation of the physical cutting force of the digital twin machine tool is established based on the influencing factors of the process parameters: In the formula: a represents the constant term of the fitting; n represents the spindle speed of the machine tool, in r / min; a p represents the radial depth of cutting, in mm; f is the feed rate of the machine tool, in mm / min; F is the milling force to be obtained, in N; a, b, c, d are the parameters to be fitted; First, convert the power function into a multivariate function polynomial, and perform linear fitting on the polynomial; take the natural logarithm of both sides: lnF=lna+blnn+clna p +dlnf Substituting the experimental data into the empirical formula, we can finally obtain the corresponding specific values ​​of the four parameters a, b, c, and d, and finally obtain the prediction model F of the physical cutting force of the digital twin machine tool.

5. The method for optimizing machine tool cutting process parameters based on digital twinning according to claim 4 is characterized in that: The optimization scheme is generated in real time based on the weight of the influencing factors and the digital twin prediction model for the cutting force fluctuation, and the relative stability of the cutting force is achieved through the process parameters to ensure the quality of the mechanical products during the processing, specifically including: Use programming language to encapsulate and save the final digital twin virtual model and package it into a dLL file, and use C# to call the prediction model in the dLL file in the digital twin virtual simulation platform; When the client in unity3D collects data, the digital twin prediction model first predicts the size of the cutting force, and then transfers the data volume to the UI chart component in unity3D. At the same time, the real-time time function is called to create a dynamic chart of the cutting force and time changes for real-time display; For machine tools of different specifications, different process parameters have different influence weights on cutting force. The real-time predicted cutting force size and the ranking of weight influencing factors are important references for the control quantity of subsequent processing optimization, i.e., the selection of process parameters, so as to achieve relative stability of cutting force and ensure the quality and efficiency of the processing process.

Images