A digital-twin-based numerical control machine tool fault diagnosis and operation maintenance platform

Abstract

A numerical control machine tool fault diagnosis and operation maintenance platform based on digital twinning, characterized in that the diagnosis and maintenance platform comprises: a machine tool operation and maintenance data template management module for managing information of different types, types and models of service machine tools; a machine tool operation data management module for synchronously mapping a numerical control machine tool physical entity and a digital model to improve the degree of state visualization; a fault diagnosis module for fusing multi-source data and digital twinning simulation to improve the accuracy and positioning efficiency of numerical control machine tool fault diagnosis; a life prediction module for collecting edge test data and simulation data to realize residual life prediction of key machine tool components; an operation and maintenance management module for providing intelligent operation and maintenance records, realizing integrated maintenance and management functions, and reducing operation and maintenance costs.

Description

Technical Field

[0001] This invention relates to the field of intelligent operation and maintenance technology for CNC machine tools, specifically to a CNC machine tool fault diagnosis and operation and maintenance platform based on digital twins. Background Technology

[0002] CNC machine tools are core equipment in high-end manufacturing and key carriers for realizing intelligent manufacturing. Their operational stability directly affects production efficiency and product quality. As the manufacturing industry transforms towards intelligence and flexibility, traditional CNC machine tool maintenance mostly adopts reactive repair or periodic maintenance models to ensure the normal and stable operation of CNC machine tools.

[0003] Typically, machine tool monitoring systems are set up to monitor the real-time operating status of CNC machine tools. However, existing machine tool monitoring systems can only collect single-dimensional operating data and lack real-time linkage with the machine tool's virtual model. They cannot achieve a visual mapping of the physical machine tool's operating status, making it difficult for maintenance personnel to intuitively grasp the operating status of the machine tool's internal components and to detect potential faults in advance. At the same time, traditional CNC machine tool fault diagnosis relies heavily on on-site inspections by maintenance personnel. Due to the limitations of personnel's experience level, it is impossible to quickly locate the root cause of the fault, resulting in excessive machine tool downtime, which seriously affects production progress. Manual diagnosis is easily affected by subjective factors, with a high rate of misjudgment, which can easily lead to over-maintenance or incomplete maintenance.

[0004] Because the operating data, fault records, maintenance files, and other information of CNC machine tools are stored in a scattered manner and lack a unified management platform, the operation and maintenance data cannot be effectively reused and remote operation and maintenance collaboration cannot be achieved. When complex faults occur in machine tools, it is difficult to quickly connect with technical personnel for remote diagnosis and guidance, thus failing to meet the high-precision, high-reliability, and full-process operation and maintenance needs of CNC machine tools in intelligent manufacturing scenarios. Summary of the Invention

[0005] This invention provides a digital twin-based fault diagnosis and operation and maintenance platform for CNC machine tools. With a rational structural design, it utilizes the digital twin platform to recreate the machine tool's structure and operating status, achieving real-time synchronous mapping between the physical CNC machine tool and the virtual model. This provides a clear view of the fault location, integrates multi-source data for online fault diagnosis and reasoning, improves fault identification accuracy, and shortens location time. Furthermore, based on degraded data, degradation models, and trend analysis, it provides early warnings of potential faults, reducing unplanned downtime. Maintenance suggestions are generated based on the fault diagnosis results, and the maintenance progress is recorded and managed. This completely eliminates the shortcomings of traditional operation and maintenance models, improves the efficiency of CNC machine tool operation and maintenance, reduces operation and maintenance costs, ensures production continuity, and solves the problems existing in the prior art.

[0006] The technical solution adopted by the present invention to solve the above-mentioned technical problems is as follows: A digital twin-based fault diagnosis and operation maintenance platform for CNC machine tools, the platform comprising: A machine tool operation and maintenance data template management module is used to manage information on machine tools in service of different types, models, and types. The machine tool operation data management module is used to synchronously map the physical entity of the CNC machine tool with the digital model, thereby improving the visibility of its status. The fault diagnosis module is used to integrate multi-source data and digital twin simulation to improve the accuracy of fault diagnosis and positioning efficiency of CNC machine tools. A life prediction module is used to aggregate edge test data and simulation data to predict the remaining life of key machine tool components. The operation and maintenance management module is used to provide intelligent operation and maintenance records, realize the integration of maintenance and management functions, and reduce operation and maintenance costs.

[0007] The machine tool operation and maintenance data template management module is set on the platform server and works in conjunction with the data storage layer to realize the full life cycle management of machine tool templates. It is compatible with all types of CNC machine tools and supports template configuration for national standard or enterprise-customized machine tool models. The machine tool operation and maintenance data template management module includes a template basic configuration unit, an operation and maintenance parameter management unit, a machine tool model association unit, a template version management unit, and an access control unit. The template basic configuration unit provides batch import and export functions for templates, supporting XML for import and Excel / PDF for export, facilitating batch configuration and archiving for enterprises. The operation and maintenance parameter management unit enables fine-grained addition, deletion, modification, and querying of operation and maintenance parameter names for different machine tool configurations, with parameter classification matching the hierarchical structure of the machine tool operation data management module. The machine tool model association unit enables one-to-one and one-to-many binding between templates and specific machine tool models, meaning one basic template can be associated with multiple machine tools of the same configuration but different brands and models, requiring only minor adjustments to differentiated parameters. The template version management unit records version changes to machine tool templates. The permission control unit sets three levels of operation permissions to adapt to different operation and maintenance roles within the enterprise.

[0008] The fault diagnosis module includes a known fault data unit and a field acquisition data unit. The known fault data unit is configured with fault labels for easy training, while the field acquisition data unit is configured with unlabeled inference data and requires manual description of the acquisition scenario.

[0009] The lifespan prediction module includes a degradation data upload unit, a model training unit, a model import and inference unit, and a diagnostic output unit. The degradation data includes spindle life, guide rail life, bearing life, hydraulic system life, and overall machine life; the model training unit has functions such as training data overview, adjusting training parameters, training logs, and training record summaries; The model import inference unit performs inference on the data based on the trained model or the imported model, including known lifetime data and field-collected data, and executes lifetime inference function to display lifetime inference results. The diagnostic output unit visualizes the fault location based on a digital twin model, outputs diagnostic results to the operation and maintenance management module, and stores the diagnostic data in the database.

[0010] The operation and maintenance management module includes a fault category selection unit, a fault sub-category selection unit, a problem recording unit, and an operation and maintenance data feedback unit; it selects relevant fault sub-categories based on different fault categories, customizes and improves operation and maintenance records, and generates historical operation and maintenance records through the platform.

[0011] This invention employs the aforementioned structure, managing information on in-service machine tools of different types, models, and specifications through a machine tool operation and maintenance data template management module; synchronously mapping the physical entities of CNC machine tools with digital models through a machine tool operation data management module, enhancing the visualization of their status; integrating multi-source data and digital twin simulation through a fault diagnosis module, improving the accuracy and efficiency of CNC machine tool fault diagnosis; summarizing edge-end test data and simulation data through a life prediction module, enabling the prediction of the remaining life of key machine tool components; and providing intelligent operation and maintenance records through an operation and maintenance management module, integrating maintenance and management functions, reducing operation and maintenance costs, completely resolving the shortcomings of traditional operation and maintenance models, improving the efficiency of CNC machine tool operation and maintenance, reducing operation and maintenance costs, and ensuring production continuity. It possesses the advantages of being economical, practical, precise, and efficient. Attached Figure Description

[0012] Figure 1 This is a schematic diagram of the structure of the present invention.

[0013] Figure 2 This is a schematic diagram of the workflow of the present invention.

[0014] Figure 3 This is a functional block diagram of the machine tool operation and maintenance data template management module of the present invention.

[0015] Figure 4 This is a functional block diagram of the fault diagnosis module of the present invention.

[0016] Figure 5 This is a functional block diagram of the lifetime prediction module of the present invention.

[0017] Figure 6 This is a functional block diagram of the operation and maintenance management module of the present invention. Detailed Implementation

[0018] To clearly illustrate the technical features of this solution, the invention will be described in detail below through specific implementation methods and in conjunction with the accompanying drawings.

[0019] like Figures 1-6 As shown, a digital twin-based CNC machine tool fault diagnosis and operation maintenance platform includes: A machine tool operation and maintenance data template management module is used to manage information on machine tools in service of different types, models, and types. The machine tool operation data management module is used to synchronously map the physical entity of the CNC machine tool with the digital model, thereby improving the visibility of its status. The fault diagnosis module is used to integrate multi-source data and digital twin simulation to improve the accuracy of fault diagnosis and positioning efficiency of CNC machine tools. A life prediction module is used to aggregate edge test data and simulation data to predict the remaining life of key machine tool components. The operation and maintenance management module is used to provide intelligent operation and maintenance records, realize the integration of maintenance and management functions, and reduce operation and maintenance costs.

[0020] The machine tool operation and maintenance data template management module is set on the platform server and works in conjunction with the data storage layer to realize the full life cycle management of machine tool templates. It is compatible with all types of CNC machine tools and supports template configuration for national standard or enterprise-customized machine tool models. The machine tool operation and maintenance data template management module includes a template basic configuration unit, an operation and maintenance parameter management unit, a machine tool model association unit, a template version management unit, and an access control unit. The template basic configuration unit provides batch import and export functions for templates, supporting XML for import and Excel / PDF for export, facilitating batch configuration and archiving for enterprises. The operation and maintenance parameter management unit enables fine-grained addition, deletion, modification, and querying of operation and maintenance parameter names for different machine tool configurations, with parameter classification matching the hierarchical structure of the machine tool operation data management module. The machine tool model association unit enables one-to-one and one-to-many binding between templates and specific machine tool models, meaning one basic template can be associated with multiple machine tools of the same configuration but different brands and models, requiring only minor adjustments to differentiated parameters. The template version management unit records version changes to machine tool templates. The permission control unit sets three levels of operation permissions to adapt to different operation and maintenance roles within the enterprise.

[0021] The fault diagnosis module includes a known fault data unit and a field acquisition data unit. The known fault data unit is configured with fault labels for easy training, while the field acquisition data unit is configured with unlabeled inference data and requires manual description of the acquisition scenario.

[0022] The lifespan prediction module includes a degradation data upload unit, a model training unit, a model import and inference unit, and a diagnostic output unit. The degradation data includes spindle life, guide rail life, bearing life, hydraulic system life, and overall machine life; the model training unit has functions such as training data overview, adjusting training parameters, training logs, and training record summaries; The model import inference unit performs inference on the data based on the trained model or the imported model, including known lifetime data and field-collected data, and executes lifetime inference function to display lifetime inference results. The diagnostic output unit visualizes the fault location based on a digital twin model, outputs diagnostic results to the operation and maintenance management module, and stores the diagnostic data in the database.

[0023] The operation and maintenance management module includes a fault category selection unit, a fault sub-category selection unit, a problem recording unit, and an operation and maintenance data feedback unit; it selects relevant fault sub-categories based on different fault categories, customizes and improves operation and maintenance records, and generates historical operation and maintenance records through the platform.

[0024] The working principle of a digital twin-based CNC machine tool fault diagnosis and operation and maintenance platform in this invention embodiment is as follows: The digital twin platform is used to reconstruct the machine tool structure and operating status, achieving real-time synchronous mapping between the physical CNC machine tool and the virtual model. This intuitively displays the fault location, integrates multi-source data for online fault diagnosis and reasoning, improves fault identification accuracy, and shortens location time. Simultaneously, based on degraded data, degraded models, and trend analysis, potential faults are warned in advance, reducing unplanned downtime. Maintenance suggestions are generated based on the fault diagnosis results, and the maintenance progress is recorded and managed, completely eliminating the defects of traditional operation and maintenance models, improving the efficiency of CNC machine tool operation and maintenance, reducing operation and maintenance costs, and ensuring production continuity.

[0025] In existing technologies, traditional CNC machine tool maintenance mostly adopts reactive repair or periodic maintenance methods, which have the following shortcomings: The physical machine tool is disconnected from the virtual model, resulting in incomplete status perception. In existing technologies, some machine tool monitoring systems can only collect single-dimensional operational data, lacking real-time linkage with the machine tool's virtual model. This makes it impossible to achieve a visual mapping of the physical machine tool's operating status, making it difficult for maintenance personnel to intuitively grasp the operating status of the machine tool's internal components and to detect potential faults in advance. Fault diagnosis relies on human experience, which is inefficient and has a high misjudgment rate. Traditional CNC machine tool fault diagnosis mostly depends on on-site inspection by maintenance personnel. Due to the limitations of personnel experience, it is impossible to quickly locate the root cause of the fault, resulting in excessive machine tool downtime and seriously affecting production progress. At the same time, manual diagnosis is easily affected by subjective factors, with a high misjudgment rate, which can easily lead to over-repair or incomplete repair.

[0026] Data management is fragmented and operation and maintenance collaboration is weak: The operating data, fault records, maintenance files and other information of CNC machine tools are stored in a scattered manner and lack a unified management platform. This makes it impossible to effectively reuse operation and maintenance data and achieve remote operation and maintenance collaboration. When complex faults occur in machine tools, it is difficult to quickly connect with technical personnel for remote diagnosis and guidance.

[0027] In the overall solution, the diagnostic and maintenance platform includes: The machine tool operation and maintenance data template management module is used to manage information on in-service machine tools of different types, models, and specifications; the machine tool operation data management module is used to synchronously map the physical entity of CNC machine tools with digital models, improving the visibility of their status; the fault diagnosis module is used to integrate multi-source data and digital twin simulation to improve the accuracy and efficiency of fault diagnosis and location for CNC machine tools; the life prediction module is used to aggregate edge test data and simulation data to predict the remaining life of key machine tool components; and the operation and maintenance management module is used to provide intelligent operation and maintenance records, integrate maintenance and management functions, and reduce operation and maintenance costs.

[0028] The machine tool operation and maintenance data template management module is set on the platform server and works in conjunction with the data storage layer to realize the full life cycle management of machine tool templates. It is compatible with all types of CNC machine tools and supports template configuration for national standard or enterprise-customized machine tool models. The machine tool operation and maintenance data template management module includes a template basic configuration unit, an operation and maintenance parameter management unit, a machine tool model association unit, a template version management unit, and an access control unit. The template basic configuration unit provides batch import and export functions for templates, supporting XML for import and Excel / PDF for export, facilitating batch configuration and archiving for enterprises. The operation and maintenance parameter management unit enables fine-grained addition, deletion, modification, and querying of operation and maintenance parameter names for different machine tool configurations, with parameter classification matching the hierarchical structure of the machine tool operation data management module. The machine tool model association unit enables one-to-one and one-to-many binding between templates and specific machine tool models, meaning one basic template can be associated with multiple machine tools of the same configuration but different brands and models, requiring only minor adjustments to differentiated parameters. The template version management unit records version changes to machine tool templates. The permission control unit sets three levels of operation permissions to adapt to different operation and maintenance roles within the enterprise.

[0029] With the coordinated efforts of various functional units, it can realize the creation, deletion, modification and query of templates, parameter customization, model binding, version iteration and operation permission management, and supports the editing, deletion and disabling of templates. Editing can modify the basic information of the template, and deletion is only applicable to blank templates that are not associated with machine tool models.

[0030] For different operation and maintenance roles, the administrator has full operation permissions for templates, including creating, deleting, disabling, version management, and permission allocation; the configuration operator has template editing, parameter addition, deletion, modification and query, machine tool model association, and permission removal, but cannot delete templates or allocate permissions; the viewer only has the permission to query and export templates and parameters, and has no editing operation permissions.

[0031] The machine tool operation data management module includes functions for uploading machine tool XML data, browsing machine tool information, and displaying the machine tool hierarchy structure. It parses and imports machine tool type XML files to confirm machine tool model and other information, thereby generating machine tool IDs and summarizing machine tools. Depending on the selected machine tool, it displays details of the selected machine tool, including a parameter list and machine tool hierarchy structure information.

[0032] The XML file covers all dimensions of CNC machine tool operation data, including sensor modules, CNC system interface modules, and edge gateway modules. The sensor modules are arranged in key parts of the CNC machine tool such as the spindle, guide rails, tool magazine, servo motor, cooling system, and hydraulic system to collect status signals such as vibration, temperature, speed, current, voltage, oil pressure, and noise. The collected raw data is preprocessed and then uploaded to the network transmission layer.

[0033] The fault diagnosis module includes a known fault data unit and a field acquisition data unit. The known fault data unit is configured with fault labels for easy training, while the field acquisition data unit is configured with unlabeled inference data and requires manual description of the acquisition scenario.

[0034] The known fault data is labeled with fault tags for easy training purposes. Uploaded fault data should be .NPY files with a size of less than 200MB. Fault types mainly include normal state, spindle fault, guide rail fault, bearing fault, hydraulic fault, pneumatic fault, and electrical fault.

[0035] The platform displays uploaded fault data, including fault type, file name, size, and modification time, and also provides a corresponding data deletion function to reduce platform cache pressure.

[0036] The data collected on-site is used for unlabeled inference and requires manual description of the collection scenario, such as collecting data after the spindle has run for 1000 hours. The uploaded data should be a .NPY file with a size of less than 200MB. The data management module displays the collection scenario, file name, size, and collection time information, and also provides a data deletion operation.

[0037] Furthermore, the model training function includes training data overview, training parameter debugging, training log display, and training record summary functions; the training data overview includes the number of existing fault types, fault type list, number of data files, and preview data shape; training parameters include selecting the diagnostic model, number of training epochs, batch size, learning rate, validation set ratio, and training device; the training log is updated according to the training epoch and includes training number, training machine, fault type, model save directory, number of epochs, training loss, and validation accuracy information.

[0038] The training record summary includes information such as model number, task name, number of training epochs, batch size, learning rate, validation set ratio, training device, training time, and save path. The model import and fault reasoning module mainly includes two functions: importing trained models and fault reasoning.

[0039] Import external trained models for model description, such as a CNN-based main axis fault diagnosis model. The imported model format should be a .pth, .PT, .H5, or .PKL file with a file size of less than 200MB. Fault inference is based on the trained model or the imported model to infer the data, including known fault / normal data and on-site collected data. The fault inference function is executed to display the fault diagnosis results.

[0040] The lifespan prediction module includes a degradation data upload unit, a model training unit, a model import and inference unit, and a diagnostic output unit. The degradation data includes spindle life, guide rail life, bearing life, hydraulic system life, and overall machine life; the model training unit has functions such as training data overview, adjusting training parameters, training logs, and training record summaries; The model import inference unit performs inference on the data based on the trained model or the imported model, including known lifetime data and field-collected data, and executes lifetime inference function to display lifetime inference results. The diagnostic output unit visualizes the fault location based on a digital twin model, outputs diagnostic results to the operation and maintenance management module, and stores the diagnostic data in the database.

[0041] Specifically, the known lifespan data needs to be labeled with lifespan dimensions including spindle lifespan, guide rail lifespan, bearing lifespan, hydraulic system lifespan, and overall machine lifespan. The uploaded data must be a .NPY file under 200MB.

[0042] The uploaded known data display interface shows the lifetime dimension, file name, size, data shape, lifetime value range, and modification time information, and also allows for data deletion.

[0043] Data collected on-site is usually unlabeled for lifetime inference. The collection scenario needs to be filled in, such as data collected after the tool has been working for 1 hour. The uploaded data should be a .NPY file with a size of less than 200MB.

[0044] The lifespan prediction model training includes functions such as training data overview, training parameters, training logs, and training record summary.

[0045] The training data overview displays the number of lifetime dimensions, a list of lifetime dimensions, the number of data files, and preview data information. Training parameters include task name, number of training epochs, batch size, learning rate, validation set ratio, training device, and prediction model. The training log updates training content according to Epoch, mainly including model number, training device, lifetime dimension, model type, model save directory, Epoch, training MAE, validation MAE, and training R... 2 Information. The training record summary displays the model number, task name, model type, number of training epochs, batch size, learning rate, validation set ratio, training device, training time, and save path.

[0046] Import external trained models for model description, such as a random forest-based axis lifetime prediction model. The imported model format should be a .pth, .PT, .H5, or .PKL file with a file size of less than 200MB.

[0047] Lifespan inference is based on a trained or imported model to infer data, including known lifespan data and field-collected data. The lifespan inference function is executed to display the lifespan inference results.

[0048] The operation and maintenance management module includes a fault category selection unit, a fault sub-category selection unit, a problem recording unit, and an operation and maintenance data feedback unit. Based on the selected fault category, relevant fault sub-categories are selected, operation and maintenance records are customized and improved, and historical operation and maintenance records are generated through the platform.

[0049] Specifically, after determining the major and minor categories of faults, the maintenance records are customized and improved. These records mainly include the fault manifestation, such as the spindle abnormal noise frequency of approximately 20Hz and vibration amplitude of 0.5mm; the cause of the fault, such as bearing wear exceeding 0.2mm or mismatched lubricant grade; the repair details, such as replacing the SKF 6205 bearing, adding No. 32 hydraulic oil, and testing the vibration amplitude to <0.1mm; the repair personnel, such as entering the repair personnel's name and employee number; and the repair time.

[0050] By saving the above maintenance records, the platform generates historical maintenance records, including record ID, machine tool ID, machine tool name, fault category, fault subcategory, manifestation, fault cause, maintenance personnel, maintenance time, and creation time information, thereby completing accurate fault diagnosis of CNC machine tools.

[0051] In summary, the digital twin-based CNC machine tool fault diagnosis and operation maintenance platform in this embodiment of the invention utilizes a digital twin platform to reconstruct the machine tool structure and operating status, achieving real-time synchronous mapping between the physical CNC machine tool and the virtual model. This provides an intuitive display of fault locations, integrates multi-source data for online fault diagnosis and reasoning, improves fault identification accuracy, and shortens location time. Furthermore, based on degraded data, degradation models, and trend analysis, it provides early warnings of potential faults, reducing unplanned downtime. Maintenance suggestions are generated based on fault diagnosis results, and maintenance progress is recorded and managed. This completely eliminates the shortcomings of traditional operation and maintenance models, improves CNC machine tool operation and maintenance efficiency, reduces operation and maintenance costs, and ensures production continuity.

[0052] The above specific embodiments should not be construed as limiting the scope of protection of the present invention. For those skilled in the art, any alternative improvements or modifications made to the embodiments of the present invention shall fall within the scope of protection of the present invention.

[0053] Any aspects of this invention not described in detail are well-known to those skilled in the art.

Claims

1. A digital twin-based fault diagnosis and operation maintenance platform for CNC machine tools, characterized in that, The diagnostic and maintenance platform includes: A machine tool operation and maintenance data template management module is used to manage information on machine tools in service of different types, models, and types. The machine tool operation data management module is used to synchronously map the physical entity of the CNC machine tool with the digital model, thereby improving the visibility of its status. The fault diagnosis module is used to integrate multi-source data and digital twin simulation to improve the accuracy of fault diagnosis and positioning efficiency of CNC machine tools. A life prediction module is used to aggregate edge test data and simulation data to predict the remaining life of key machine tool components. The operation and maintenance management module is used to provide intelligent operation and maintenance records, realize the integration of maintenance and management functions, and reduce operation and maintenance costs.

2. The CNC machine tool fault diagnosis and operation maintenance platform based on digital twin as described in claim 1, characterized in that: The machine tool operation and maintenance data template management module is set on the platform server and works in conjunction with the data storage layer to realize the full life cycle management of machine tool templates. It is compatible with all types of CNC machine tools and supports template configuration for national standard or enterprise-customized machine tool models. The machine tool operation and maintenance data template management module includes a template basic configuration unit, an operation and maintenance parameter management unit, a machine tool model association unit, a template version management unit, and an access control unit. The template basic configuration unit is used to provide batch import and export functions for templates. The import format supports XML, and the export format supports Excel / PDF, which facilitates batch configuration and file retention for enterprises. The operation and maintenance parameter management unit is used to enable fine-grained addition, deletion, modification, and querying of operation and maintenance parameter names for different machine tool configurations, and the parameter classification matches the hierarchical structure of the machine tool operation data management module; the machine tool model association unit is used to enable one-to-one and one-to-many binding between templates and specific machine tool models, that is, one basic template can be associated with multiple machine tools of the same configuration but different brands and models, requiring only minor adjustments to differentiated parameters; the template version management unit is used to record version changes to machine tool templates; the permission control unit is used to set three levels of operation permissions to adapt to different operation and maintenance roles within the enterprise.

3. The CNC machine tool fault diagnosis and operation maintenance platform based on digital twin as described in claim 1, characterized in that: The fault diagnosis module includes a known fault data unit and a field acquisition data unit. The known fault data unit is configured with fault labels for easy training, while the field acquisition data unit is configured with unlabeled inference data and requires manual description of the acquisition scenario.

4. The CNC machine tool fault diagnosis and operation maintenance platform based on digital twin as described in claim 1, characterized in that: The lifespan prediction module includes a degradation data upload unit, a model training unit, a model import and inference unit, and a diagnostic output unit. The degradation data includes spindle life, guide rail life, bearing life, hydraulic system life, and overall machine life; the model training unit has functions such as training data overview, adjusting training parameters, training logs, and training record summaries; The model import inference unit performs inference on the data based on the trained model or the imported model, including known lifetime data and field-collected data, and executes lifetime inference function to display lifetime inference results. The diagnostic output unit visualizes the fault location based on a digital twin model, outputs diagnostic results to the operation and maintenance management module, and stores the diagnostic data in the database.

5. The CNC machine tool fault diagnosis and operation maintenance platform based on digital twin as described in claim 1, characterized in that: The operation and maintenance management module includes a fault category selection unit, a fault sub-category selection unit, a problem recording unit, and an operation and maintenance data feedback unit; it selects relevant fault sub-categories based on different fault categories, customizes and improves operation and maintenance records, and generates historical operation and maintenance records through the platform.

Images