A Method for Constructing a Fault Propagation Model of CNC Machine Tools Based on Digital Twin

Abstract

The present invention belongs to the field of numerical control machine tools, and specifically discloses a method for constructing a fault propagation model of a numerical control machine tool based on digital twin. Peripheral sensors are arranged, and relevant operation information of the numerical control machine tool is collected through the peripheral sensors; a database for storing the state information of the numerical control machine tool is constructed; the constructed multi-system coupling model and multi-physical-field coupling model are coupled to establish a multi-system-multi-physical-field digital twin model, and the multi-system-multi-physical-field digital twin model is real-time mapped by using the state information of the numerical control machine tool in the database; fault simulation is carried out through the multi-system-multi-physical-field digital twin model to determine the fault propagation path of the numerical control machine tool, and a fault propagation model is established. The present invention constructs a fault propagation model of a numerical control machine tool. When a fault occurs in the numerical control machine tool, the model is used to timely conduct fault traceability analysis through real-time data, effectively improving the fault diagnosis and maintenance efficiency of the numerical control machine tool and reducing the maintenance cost.

Description

Technical Field

[0001] The present invention belongs to the field of numerical control machine tools, and particularly relates to a method for constructing a fault propagation model of a numerical control machine tool based on digital twin. Background Technique

[0002] A numerical control machine tool is a complex system integrating machinery, electricity, heat, and liquid. The strong coupling of its structure leads to complex fault mechanisms and fault path propagation. At the same time, the number of fault samples in the actual working process is usually extremely scarce. Once the machine tool fails, due to the unclear fault propagation path, it is difficult to timely find the cause of the fault, resulting in long maintenance time, high cost, and serious economic losses. Therefore, analyzing the fault propagation path of a numerical control machine tool can timely and effectively maintain the numerical control machine tool, reduce maintenance costs, and improve operation efficiency.

[0003] Currently, the research on the fault propagation path of numerical control machine tools mainly includes qualitative analysis based on expert knowledge and data-driven analysis. Qualitative analysis based on expert knowledge has strong subjectivity and low efficiency, while data-driven analysis ignores the fault mechanism of numerical control machine tools, making it difficult to qualitatively analyze fault propagation and unable to meet the development needs of current numerical control machine tool maintenance. Summary of the Invention

[0004] To solve the above problems, the present invention provides a method for constructing a fault propagation model of a numerical control machine tool based on digital twin. By constructing a fault propagation model of a numerical control machine tool, when the numerical control machine tool fails, the model is used to timely conduct fault traceability analysis through real-time data, effectively improving the efficiency of numerical control machine tool fault diagnosis and maintenance, and reducing maintenance costs.

[0005] The technical solution of the present invention is a method for constructing a fault propagation model of a numerical control machine tool based on digital twin, including the following steps:

[0006] Arrange peripheral sensors to collect relevant operation information of the numerical control machine tool through the peripheral sensors;

[0007] Construct a numerical control machine tool status information database for storing numerical control machine tool status information, where the numerical control machine tool status information includes operation information collected by peripheral sensors;

[0008] Respectively construct a multi-system coupling model and a multi-physical field coupling model, couple the multi-system coupling model and the multi-physical field coupling model to establish a multi-system - multi-physical field digital twin model, and use the numerical control machine tool status information in the database to perform real-time mapping on the multi-system - multi-physical field digital twin model;

[0009] Conduct fault simulation through the multi-system - multi-physical field digital twin model to determine the fault propagation path of the numerical control machine tool and establish a fault propagation model.

[0010] In an alternative embodiment, after arranging the peripheral sensors, the following steps are further included:

[0011] Obtain the operation information collected by the peripheral sensors;

[0012] Obtain the health status indicators of the CNC machine tool in the same period;

[0013] Analyze the correlation between the operation information collected by the peripheral sensors and the health status indicators of the CNC machine tool in the same period;

[0014] Optimize the arrangement position of the peripheral sensors according to the result of the correlation analysis.

[0015] In an alternative embodiment, analyzing the correlation between the operation information collected by the peripheral sensors and the health status indicators of the CNC machine tool in the same period specifically includes calculating the correlation coefficient through the following formula ,

[0016]

[0017] wherein, is the sample standard deviation of the operation information collected by the peripheral sensors, is the sample standard deviation of the health status indicators of the CNC machine tool, is the sample covariance between the sample of the operation information collected by the peripheral sensors and the sample of the health status indicators of the CNC machine tool.

[0018] In an alternative embodiment, constructing a multi-system coupling model specifically includes:

[0019] Obtain the operation information of the CNC machine tool collected by the peripheral sensors, denoted as dataset S;

[0020] Obtain the first target operation information from the numerical control system of the CNC machine tool, denoted as dataset N;

[0021] Obtain the second target operation information from the servo system of the CNC machine tool, denoted as dataset M;

[0022] Obtain the static information data of the CNC machine tool, denoted as dataset I; wherein the static information data includes geometric dimensions, system structure, physical properties, and model;

[0023] Construct a CNC machine tool status information database .

[0024] In an alternative embodiment, constructing a multi-system coupling model specifically includes:

[0025] Select physical parameters for each subsystem, and the subsystems include the mechanical system, electrical system, and hydraulic system of the CNC machine tool;

[0026] Based on the selected physical parameters, digital description models of each subsystem are constructed separately through the Modelica language;

[0027] According to the coupling mechanism between subsystems, the digital description models of each subsystem are connected to form a multi-system coupling model, denoted as , where represents the nth subsystem.

[0028] In an optional implementation, a multi-physical-field coupling model is constructed, specifically including:

[0029] Select the target physical fields, which include vibrations and heat generation affected by loads;

[0030] According to the selected target physical fields, a multi-physical-field coupling model is constructed, denoted as , where represents the mth physical field.

[0031] In an optional implementation, the multi-system coupling model and the multi-physical-field coupling model are coupled to establish a multi-system - multi-physical-field digital twin model, and the state information of the CNC machine tool in the database is used to perform real-time mapping on the multi-system - multi-physical-field digital twin model, specifically including:

[0032] The multi-system coupling model and the multi-physical-field coupling model are encapsulated into an FMU model with the same platform interface by adopting the FMI standard;

[0033] Set corresponding physical parameter interaction interfaces to realize the interaction between the multi-system coupling model and the multi-physical-field coupling model, and obtain the multi-system - multi-physical-field digital twin model, denoted as ;

[0034] The multi-system - multi-physical-field digital twin model is encapsulated, and a data interface is left;

[0035] The database inputs the target CNC machine tool state information into the multi-system - multi-physical-field digital twin model through the data interface, updates the multi-system - multi-physical-field digital twin model, and realizes the real-time mapping of the multi-system - multi-physical-field digital twin model.

[0036] In an optional implementation, fault simulation is carried out through the multi-system - multi-physical-field digital twin model to determine the fault propagation path of the CNC machine tool and establish a fault propagation model, specifically including:

[0037] According to the expert experience and knowledge, a directed graph of CNC machine tool fault propagation is qualitatively established. In the qualitative fault propagation directed graph, nodes and directed edges are used to represent variables and the potential qualitative causal relationships between variables respectively. Nodes represent the components of the CNC machine tool, and the node set is , and the directed edges represent the fault propagation relationships between the components of the CNC machine tool. The directed edge set is . Then the qualitative fault propagation directed graph is represented as ;

[0038] According to the qualitative fault propagation directed graph , fault injection is performed on the multi-system - multi-physical field digital twin model to obtain the state information of each component of the CNC machine tool under different types of fault injection;

[0039] Based on the state information under fault injection, the fault causal relationships between the components of the CNC machine tool are quantitatively analyzed by the KNN method, the fault propagation path of the CNC machine tool is determined, and a fault propagation model is established.

[0040] In an alternative embodiment, based on the state information under fault injection, the fault causal relationships between the components of the CNC machine tool are quantitatively analyzed by the KNN method, the fault propagation path of the CNC machine tool is determined, and a fault propagation model is established, which specifically includes:

[0041] Step 1, assume that and are the i-th sampling point sets of the fault nodes and respectively, and the corresponding data vectors are and respectively. Then The prediction performance of can be expressed by the following distance formula,

[0042]

[0043] where, represents the prediction performance between nodes, is the K value of the KNN method;

[0044] Step 2, The self-prediction performance of

[0045]

[0046] Step 3, after Q iterations, we get

[0047]

[0048] where, represents after Q iterations, For prediction performance, where Q is the number of iterations;

[0049] In step 4, the causal relationship between nodes and is expressed as:

[0050]

[0051] In the formula, represents and the causal relationship value between. If , it indicates that is the dependent variable, is the result variable. Conversely, is the dependent variable, is the result variable;

[0052] In step 5, calculate the causal relationship between any two nodes through steps 1 - 4, obtain the causal relationship between the components of the numerically controlled machine tool for quantitative analysis, determine the fault propagation path of the numerically controlled machine tool, and obtain the fault propagation model.

[0053] A method for constructing a fault propagation model of a numerically controlled machine tool based on digital twin provided by the present invention has the following beneficial effects compared with the prior art: By arranging peripheral sensors to collect operation data, constructing a database for storing the state information of the numerically controlled machine tool, and simultaneously constructing a multi - system - multi - physical - field digital twin model, realizing digital twin based on the state information in the database, which can reflect the change of physical properties during the operation process, improve the model fidelity, conduct fault simulation based on the multi - system - multi - physical - field digital twin model to achieve quantitative analysis of the fault propagation path. When a fault occurs in the numerically controlled machine tool, use this model to timely conduct fault traceability analysis through real - time data, effectively improve the fault diagnosis and maintenance efficiency of the numerically controlled machine tool, and reduce the maintenance cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions of the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0055] Figure 1 is a schematic flow chart of a method for constructing a fault propagation model of a numerically controlled machine tool based on digital twin provided by an embodiment of the present invention.

[0056] Figure 2It is a schematic diagram of the process principle of a specific embodiment of a method for constructing a fault propagation model of a numerically controlled machine tool based on digital twin provided by the present invention. Detailed implementation manners

[0057] In order to enable those skilled in the art of the present technology to better understand the solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific implementation manners. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0058] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments, and are not intended to limit the present invention.

[0059] The following explains the key terms that appear in the present invention.

[0060] FMI: Functional Mock-up Interface, that is, the functional model interface, is an international open software standard that provides a unified model interface for hybrid simulation between simulation software.

[0061] FMU: Functional Mock-up Units, that is, functional model units, is a model standard independent of the development environment

[0062] As a key enabling technology for intelligent manufacturing, digital twin (DT) has been increasingly widely used in the application field of numerically controlled machine tools in recent years. By constructing a numerically controlled machine tool and integrating a large amount of data generated by the digital twin model with the real-time monitoring data of the numerically controlled machine tool, the interaction feedback, data fusion and analysis between the digital space and the physical space can be realized, and a high-fidelity digital twin model can be used to analyze the full life cycle state of the numerically controlled machine tool, providing a solution idea for the analysis of the fault propagation path of the numerically controlled machine tool.

[0063] Aiming at the problem that it is difficult to qualitatively analyze the fault propagation at present and it cannot meet the development needs of the maintenance of numerically controlled machine tools at the present stage, this embodiment provides a method for constructing a fault propagation model of a numerically controlled machine tool based on digital twin. By constructing a fault propagation model of a numerically controlled machine tool, when a numerically controlled machine tool fails, the model is used to perform fault traceability analysis in a timely manner through real-time data, effectively improving the fault diagnosis and maintenance efficiency of the numerically controlled machine tool and reducing the maintenance cost.

[0064] Figure 1It is a schematic flowchart of a method for constructing a fault propagation model of a numerically controlled machine tool based on digital twin provided by an embodiment of the present invention. As Figure 1 shown, the method includes the following steps.

[0065] S1. Arrange peripheral sensors, and collect relevant operation information of the numerically controlled machine tool through the peripheral sensors.

[0066] In this embodiment, according to the mechanical structure characteristics of the numerically controlled machine tool, external sensors are installed, and the correlation between the sensing data and the health state of the numerically controlled machine tool is analyzed. According to the correlation, it is judged whether the measuring point is the optimal measuring point, and a sensing monitoring scheme that can best reflect the health state of the numerically controlled machine tool is formulated without affecting the machining state of the numerically controlled machine tool.

[0067] S2. Construct a numerically controlled machine tool state information database for storing the state information of the numerically controlled machine tool, and the state information of the numerically controlled machine tool includes the operation information collected by the peripheral sensors.

[0068] In this embodiment, according to the formulated sensing monitoring scheme and the numerical control system, relevant information of the numerically controlled machine tool is obtained, and a numerically controlled machine tool state information database is constructed to realize the storage and call of multi-dimensional historical data and real-time data of the numerically controlled machine tool.

[0069] S3. Respectively construct a multi-system coupling model and a multi-physical field coupling model, couple the multi-system coupling model and the multi-physical field coupling model to establish a multi-system - multi-physical field digital twin model, and perform real-time mapping on the multi-system - multi-physical field digital twin model by using the state information of the numerically controlled machine tool in the database.

[0070] In this embodiment, a multi-domain model and a multi-physical field model of the numerically controlled machine tool are respectively constructed and jointly simulated. The two types of models perform parameter interaction through a general interface to construct a multi-domain - multi-physical field digital twin model of the numerically controlled machine tool; the real-time dynamic data of the numerically controlled machine tool state information database is used to realize the real-time mapping between the digital twin model and the physical system of the numerically controlled machine tool.

[0071] S4. Perform fault simulation through the multi-system - multi-physical field digital twin model to determine the fault propagation path of the numerically controlled machine tool and establish a fault propagation model.

[0072] In this embodiment, the traditional fault modes of the numerically controlled machine tool are analyzed, and according to the expert experience knowledge, a directed graph of fault propagation of the numerically controlled machine tool is qualitatively established; through fault simulation by the digital twin model, based on the directed graph, the KNN method is introduced to quantitatively analyze the fault causal relationship between the components of the numerically controlled machine tool, clarify the fault propagation path of the numerically controlled machine tool, and establish a fault propagation model.

[0073] To further understand the present invention, the following provides a specific embodiment to further elaborate on the present invention in detail. Figure 2It is a schematic diagram of the process principle of this specific embodiment. In this specific embodiment, sensor measurement points are first arranged, the correlation between mechanical characteristics such as the machining accuracy of CNC machine tools and the measurement point data is analyzed, the measurement points are optimized, and a sensing monitoring scheme for CNC machine tools is established. At the same time, a state information database for CNC machine tools is constructed to store the dynamic and static information of CNC machine tools. And a digital twin model of multi-domain and multi-physical field coupling for CNC machine tools is constructed, and through the real-time data stored in the database, the interactive mapping between the digital twin model and the physical system is realized. Finally, the fault modes of CNC machine tools are analyzed, and according to the expert experience and knowledge, a directed graph method (Directed Acyclic Graph, DAG) is used to qualitatively establish a directed graph of CNC machine tool fault propagation. The digital twin model is used for fault simulation to obtain the fault data of CNC machine tools, and the K nearest neighbors (KNN) method is introduced to quantitatively analyze and obtain the fault propagation model of CNC machine tools.

[0074] SS1, Sensing monitoring and measurement point optimization of CNC machine tools.

[0075] For most existing CNC machine tools, it is difficult to directly obtain data directly related to faults such as vibration and noise. Therefore, external sensors need to be installed for collection. First, analyze the mechanical structure characteristics of the CNC machine tool, and design the sensor wiring and installation positions without affecting the normal operation of the CNC machine tool. Secondly, use the correlation analysis method to judge the correlation between the collected information and the machine tool health status indicators (machine self-monitoring data), and optimize the sensor layout positions according to the correlation, which specifically includes the following steps.

[0076] Step 1, Obtain the operation information collected by the peripheral sensors.

[0077] Step 2, Obtain the CNC machine tool health status indicators in the same period.

[0078] Step 3, Analyze the correlation between the operation information collected by the peripheral sensors and the CNC machine tool health status indicators in the same period.

[0079] Calculate the correlation coefficient through the following formula ,

[0080]

[0081] In the formula, is the sample standard deviation of the operation information collected by the peripheral sensors, is the sample standard deviation of the CNC machine tool health status indicators, is the sample covariance between the operation information sample collected by the peripheral sensors and the CNC machine tool health status indicator sample.

[0082] Step 4, Optimize the layout positions of the peripheral sensors according to the correlation analysis results.

[0083] In this embodiment, The closer it is to -1, the more negative the correlation between the measurement point and the operating state. The closer it is to 1, the more positive the correlation between the measurement point and the operating state of the CNC machine tool. The closer it is to 0, the poorer the correlation between the measurement point and the operating state of the CNC machine tool. By adjusting the value of to optimize the measurement point and formulating the optimal sensing and monitoring scheme. Exemplarily, when there is a negative correlation, the position of the measurement point is adjusted so that the measurement point has a positive correlation with the operating state of the CNC machine tool.

[0084] SS2, CNC machine tool data acquisition and storage.

[0085] Step 1: Obtain the operating information of the CNC machine tool collected by the peripheral sensor, denoted as dataset S.

[0086] Step 2: Obtain the first target operating information from the numerical control system of the CNC machine tool, denoted as dataset N.

[0087] Step 3: Obtain the second target operating information from the servo system of the CNC machine tool, denoted as dataset M.

[0088] In this embodiment, the CNC machine tool can collect information such as current and voltage using the numerical control system and the servo system. Therefore, the dynamic data of the CNC machine tool is collected through external sensors, the numerical control system, and the servo system, and the obtained datasets are denoted as S, N, and M respectively, including vibration, rotational speed, displacement, feed rate, noise, current, voltage, etc.

[0089] Step 4: Obtain the static information data of the CNC machine tool, denoted as dataset I; where the static information data includes geometric dimensions, system structure, physical properties, and model.

[0090] It should be noted that the static information data is obtained through methods such as machine tool design drawings and on-site investigations.

[0091] Step 5: Construct a CNC machine tool status information database .

[0092] SS3, Construction of a multi-system - multi-physical field digital twin model for CNC machine tools.

[0093] Step 1: Construct a multi-system coupling model.

[0094] Step 1.1: Select physical parameters for each subsystem, where the subsystems include the mechanical system, electrical system, and hydraulic system of the CNC machine tool.

[0095] Step 1.2: Based on the selected physical parameters, construct digital description models for each subsystem using Modelica language.

[0096] Step 1.3: According to the coupling mechanism between subsystems, connect the digital description models of each subsystem to form a multi-system coupling model, denoted as , where represents the nth subsystem.

[0097] A numerically controlled machine tool is a complex electromechanical device composed of multiple subsystems such as a mechanical system, an electrical system, and a hydraulic system. In this embodiment, appropriate physical parameters are selected, and the mathematical description models of each subsystem are constructed separately through Modelica language. Then, according to the coupling mechanism between each subsystem, the subsystems are connected to form a multi-system coupling digital twin model of the numerically controlled machine tool.

[0098] Step 2: Construct a multi-physical field coupling model.

[0099] Step 2.1: Select the target physical field, which includes vibrations generated under the influence of loads.

[0100] Step 2.2: Construct a multi-physical field coupling model according to the selected target physical field, denoted as , where represents the mth physical field.

[0101] The multi-system coupling model expresses the geometric motion, mechanical control, etc. dimensions of the numerically controlled machine tool. For the physical performance dimension, it is expressed by constructing a multi-physical field coupling model. During the operation of the numerically controlled machine tool, vibrations, heat generation, etc. will occur under the influence of loads, impacts, etc. Therefore, a multi-physical field coupling model of the numerically controlled machine tool is constructed.

[0102] Step 3: Couple the multi-system coupling model and the multi-physical field coupling model to establish a multi-system - multi-physical field digital twin model, and use the numerically controlled machine tool status information in the database to perform real-time mapping on the multi-system - multi-physical field digital twin model.

[0103] Step 3.1: Package the multi-system coupling model and the multi-physical field coupling model into an FMU model with the same platform interface using the FMI standard.

[0104] Step 3.2: Set the corresponding physical parameter interaction interface to realize the interaction between the multi-system coupling model and the multi-physical field coupling model, and obtain the multi-system - multi-physical field digital twin model, denoted as .

[0105] Step 3.3: Package the multi-system - multi-physical field digital twin model and leave a data interface.

[0106] This embodiment adopts the FMI (Functional Mockup Interface) standard to encapsulate the above two models into an FMU (Functional Mock-up Units) model with a unified platform interface, and sets the corresponding physical parameter interaction interface (such as machine tool speed, feed speed, etc.) according to needs to realize the interaction between the two models, and finally obtain a multi-system-multi-physical field coupled digital twin model that can characterize the physical performance and motion relationship of CNC machine tools.

[0107] In step 3.4, the database inputs the target CNC machine tool status information into the multi-system-multi-physics field digital twin model through the data interface, updates the multi-system-multi-physics field digital twin model, and realizes real-time mapping of the multi-system-multi-physics field digital twin model.

[0108] Encapsulate the constructed digital twin model of the CNC machine tool, leave relevant data interfaces, and input dynamic data information such as speed, displacement, feed speed, etc. in the database D into the digital twin model , update the model and realize real-time mapping of the model from multiple dimensions such as geometry, physics, behavior, and rules.

[0109] SS4, Construction of fault propagation model for CNC machine tools.

[0110] Step 1: Based on expert experience and knowledge, a directed graph of fault propagation of CNC machine tools is qualitatively established.

[0111] This step implements qualitative analysis of fault propagation of CNC machine tools. In this embodiment, based on expert experience and knowledge, various fault modes of CNC machine tools and fault correlations between various systems of CNC machine tools are analyzed to model the fault propagation process. The qualitative fault propagation directed graph uses nodes and directed edges to represent variables and potential qualitative causal relationships between variables, respectively. Nodes represent parts of CNC machine tools (such as spindle bearings, shafts, etc.), and the node set is , directed edges represent the fault propagation relationship between CNC machine tool parts, and the directed edge set is , then the qualitative fault propagation directed graph is expressed as .

[0112] Step 2: Based on the qualitative fault propagation directed graph , for multi-system-multi-physics digital twin models Perform fault injection to obtain the status information of each CNC machine tool component under different types of fault injection.

[0113] This step implements fault injection into the digital twin model. Due to the complex mechanical structure and various fault modes of CNC machine tools, it is difficult to obtain effective fault data of CNC machines. , for the constructed Fault injection is carried out. As the fault is injected, the digital twin model exhibits the operating states under different fault modes. According to the visualization characteristics of the digital twin model, each system fault is analyzed, the signal characteristics of other system components are judged when a fault is injected into a certain system component, and the relevant variable set is collected.

[0114] Step 3: Based on the state information under fault injection, use the KNN method to quantitatively analyze the fault causality between the components of the CNC machine tool, determine the fault propagation path of the CNC machine tool, and establish a fault propagation model.

[0115] This step realizes the quantitative analysis of the fault propagation of the CNC machine tool. It is a directed graph drawn using expert experience knowledge and has a certain degree of subjectivity. To further clarify the fault causality between the system components of the CNC machine tool, use the digital twin data of fault injection and the monitoring data of database D, introduce the KNN method, further clarify the causal relationship between the components when a fault occurs, and realize the quantitative analysis of the fault propagation path. The specific steps are as follows.

[0116] Suppose and are respectively the i-th sampling point sets of the fault nodes and , and the corresponding data vectors are respectively and , then The prediction performance of can be expressed by the following distance formula,

[0117]

[0118] In the formula, represents the prediction performance between nodes, is the K value of the KNN method.

[0119] The self-prediction performance of

[0120]

[0121] After iterating Q times, we get

[0122]

[0123] In the formula, represents the prediction performance of for after iterating Q times, and Q is the number of iterations.

[0124] Then the nodes and The causal relationship is expressed as:

[0125]

[0126] In the formula, represents and the causal relationship value of , if , it indicates that is the dependent variable, is the independent variable, and vice versa, is the dependent variable,

[0127] By calculating the causal relationship between any two nodes through the above steps, the causal relationship between the components of the numerically controlled machine tool for quantitative analysis is obtained, the fault propagation path of the numerically controlled machine tool is determined, and the fault propagation model is obtained.

[0128] Through the above steps in this embodiment, an accurate directed graph model of the fault propagation of the numerically controlled machine tool is obtained. When a fault occurs in the numerically controlled machine tool, reasoning is performed according to the fault propagation directed graph model to help the staff effectively find the fault source, repair and eliminate the fault in time, and reduce the maintenance cost.

[0129] The above is only the preferred embodiment of the present invention disclosed, but the present invention is not limited thereto. Any non-creative changes that can be thought of by those skilled in the art, as well as several improvements and refinements made without departing from the principle of the present invention, should fall within the protection scope of the present invention.

Claims

1. A method for constructing a fault propagation model of a numerically controlled machine tool based on digital twin, characterized in that, Including the following steps: Arrange peripheral sensors to collect relevant operation information of the CNC machine tool through the peripheral sensors; Construct a CNC machine tool status information database for storing the status information of the CNC machine tool, where the status information of the CNC machine tool includes the operation information collected by the peripheral sensors; Construct a multi-system coupling model and a multi-physical-field coupling model respectively, and couple the multi-system coupling model and the multi-physical-field coupling model to establish a multi-system-multi-physical-field digital twin model, and use the status information of the CNC machine tool in the database to perform real-time mapping on the multi-system-multi-physical-field digital twin model; Perform fault simulation through the multi-system-multi-physical-field digital twin model to determine the fault propagation path of the CNC machine tool and establish a fault propagation model; specifically including: According to the expert's empirical knowledge, a qualitative directed graph of CNC machine tool fault propagation is established. In the qualitative fault propagation directed graph, nodes and directed edges are used to represent variables and the potential qualitative causal relationships between variables respectively. Nodes represent the components of the CNC machine tool, and the node set is Directed edges represent the fault propagation relationships between the components of the CNC machine tool, and the directed edge set is Then the qualitative fault propagation directed graph is represented as According to the qualitative fault propagation directed graph For the multi-system multi-physical field digital twin model M DT =(M s ∪M p ), fault injection is carried out to obtain the state information of each CNC machine tool component under different types of fault injection; Based on the status information under fault injection, quantitatively analyze the fault causal relationship between the components of the CNC machine tool by the KNN method, determine the fault propagation path of the CNC machine tool, and establish a fault propagation model, specifically including: Step 1, assume that and are respectively the sets of the i-th sampling points of the fault nodes and , and the corresponding data vectors are respectively φ 1i (x 11, x 12, x 13, …,x 1d ) and φ 2i (x 21, x 22, x 23, …,x 2d ). Then The prediction performance of can be expressed by the following distance formula: where Z i represents the prediction performance between nodes, and K is the value of K in the KNN method; Step 2, whose self-prediction performance is expressed by the following formula: Step 3, after iterating Q times, obtain In the formula, represents the prediction performance of after Q iterations, where Q is the number of iterations; is the prediction performance after Q iterations, and Q is the number of iterations; Step 4, then the causal relationship between nodes and is represented as: In the formula, represents and 's causal relationship value. If then it indicates that is the dependent variable, is the outcome variable. Conversely, is the dependent variable, is the outcome variable; Step 5, calculate the causal relationship between any two nodes through Steps 1-4 to obtain the quantitatively analyzed fault causal relationship between the components of the CNC machine tool, determine the fault propagation path of the CNC machine tool, and obtain the fault propagation model.

2. The method for constructing a fault propagation model of a numerically controlled machine tool based on digital twin according to claim 1, wherein After arranging the peripheral sensors, the following steps are further included: Obtain the operation information collected by the peripheral sensors; Obtain the CNC machine tool health status indicators in the same period; Analyze the correlation between the operation information collected by the peripheral sensors and the CNC machine tool health status indicators in the same period; Optimize the arrangement position of the peripheral sensors according to the correlation analysis results.

3. The method for constructing a fault propagation model of a numerically controlled machine tool based on digital twins according to claim 2, wherein Analyze the correlation between the operating information collected by the peripheral sensors and the health status indicators of the CNC machine tool during the same period, specifically including calculating the correlation coefficient r through the following formula xy , Where, S x is the sample standard deviation of the operation information collected by the peripheral sensor, and S y is the sample standard deviation of the CNC machine tool health state index, and S xy is the sample covariance between the sample of the operation information collected by the peripheral sensor and the sample of the CNC machine tool health state index.

4. The method for constructing a fault propagation model of a numerically controlled machine tool based on digital twin according to claim 1, wherein Construct a CNC machine tool status information database, specifically including: Obtain the operation information of the CNC machine tool collected by the peripheral sensors, denoted as dataset S; Obtain the first target operation information from the CNC system of the CNC machine tool, denoted as dataset N; Obtain the second target operation information from the servo system of the CNC machine tool, denoted as dataset M; Obtain the static information data of the CNC machine tool, denoted as dataset I; where the static information data includes geometric dimensions, system structure, physical properties, and model; Construct a CNC machine tool status information database D = (S, N, M, I).

5. The method for constructing a fault propagation model of a numerically controlled machine tool based on digital twin according to claim 1, characterized in that, Construct a multi-system coupling model, specifically including: Select physical parameters for each subsystem, where the subsystems include the mechanical system, electrical system, and hydraulic system of the CNC machine tool; Based on the selected physical parameters, respectively construct digital description models of each subsystem through Modelica language; According to the coupling mechanism between subsystems, the digital description models of each subsystem are connected to form a multi-system coupling model, denoted as M s (s1, s2, s3…s n )), where s n represents the nth subsystem.

6. The method for constructing a fault propagation model of a numerically controlled machine tool based on digital twin according to claim 5, characterized in that Construct a multi-physical-field coupling model, specifically including: Select target physical fields, where the target physical fields include vibration and heat generation affected by loads; Construct a multi-physics coupling model according to the selected target physical field, denoted as M p (p1, p2, p3…p m )), where p m represents the m-th physical field.

7. The method for constructing a fault propagation model of a numerically controlled machine tool based on digital twin according to claim 6, characterized in that Couple the multi-system coupling model and the multi-physical-field coupling model to establish a multi-system-multi-physical-field digital twin model, and use the status information of the CNC machine tool in the database to perform real-time mapping on the multi-system-multi-physical-field digital twin model, specifically including: Package the multi-system coupling model and the multi-physical-field coupling model into an FMU model with the same platform interface using the FMI standard; Set the corresponding physical parameter interaction interface to achieve the interaction between the multi-system coupling model and the multi-physics field coupling model, and obtain the multi-system-multi-physics field digital twin model, denoted as M DT =(M s ∪M p ); Package the multi-system-multi-physical-field digital twin model and leave a data interface; The database inputs the target numerical control machine tool status information into the multi-system multi-physical field digital twin model through a data interface, updates the multi-system multi-physical field digital twin model, and realizes the real-time mapping of the multi-system multi-physical field digital twin model.

Images