Numerical control machine tool operation monitoring system and method based on digital spasm

Abstract

The invention relates to the field of numerical control machine tool operation monitoring, and particularly discloses a numerical control machine tool operation monitoring system and method based on digital spasm. Firstly, numerical control machine tool monitoring images, collected by a camera, of a plurality of preset monitoring points, vibration signals, collected by a vibration sensor, of the plurality of preset monitoring points and numerical control machine tool temperature values, collected by a temperature sensor, of the plurality of preset monitoring points are obtained, and then feature extraction and correlation analysis are conducted on the three images through the deep learning technology; and finally, a classification result is obtained through a classifier to judge whether the operation state of the numerical control machine tool is abnormal or not, so that early warning is performed in advance, measures are taken, shutdown or damage of equipment is avoided, and the reliability and production efficiency of the equipment are improved.

Description

Technical Field

[0001] The present application relates to the field of CNC machine tool operation monitoring, and more specifically, to a CNC machine tool operation monitoring system and method based on digital twin. Background Art

[0002] CNC machine tools are short for digitally controlled machine tools, which are automated machine tools equipped with a program control system. CNC machine tools can easily solve complex, precise, small-batch, and multi-variety parts processing problems, and are widely used in various fields, such as defense aviation, automotive industry, mold manufacturing, mechanical processing, parts construction, etc. As long as there are mechanical industries, CNC machine tools will be used.

[0003] However, accurately obtaining the operating status of CNC machine tools is crucial to improving production efficiency. Current monitoring technologies are usually limited to monitoring explicit factors and lack accurate judgment of machine tool wear and specific status. In addition, traditional manual inspections are inefficient and easily limited by the professional quality of inspectors. In addition, a large amount of processing data is often regarded as useless information and wasted, which exacerbates the problem of resource waste.

[0004] Therefore, a CNC machine tool operation monitoring system and method based on digital twin is desired. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a CNC machine tool operation monitoring system and method based on digital twin, which first obtains the CNC machine tool monitoring images of multiple preset monitoring points collected by the camera, the vibration signals of multiple preset monitoring points collected by the vibration sensor, and the CNC machine tool temperature values ​​of multiple preset monitoring points collected by the temperature sensor, and then uses deep learning technology to perform feature extraction and association analysis on the three, and finally obtains the classification result through the classifier to determine whether there is an abnormality in the operation status of the CNC machine tool, so as to give early warning and take measures to avoid equipment shutdown or damage, thereby improving the reliability and production efficiency of the equipment.

[0006] According to one aspect of the present application, a CNC machine tool operation monitoring system based on digital twin is provided, which comprises:

[0007] A CNC machine tool operation monitoring data acquisition module is used to acquire CNC machine tool monitoring images of multiple preset monitoring points acquired by a camera, vibration signals of multiple preset monitoring points acquired by a vibration sensor, and CNC machine tool temperature values ​​of multiple preset monitoring points acquired by a temperature sensor;

[0008] A CNC machine tool operation monitoring data extraction module, used to extract a CNC machine tool monitoring focus feature vector and a CNC machine tool operation status joint feature vector from the CNC machine tool monitoring images of the multiple preset monitoring points collected by the camera, the vibration signals of the multiple preset monitoring points collected by the vibration sensor, and the CNC machine tool temperature values ​​of the multiple preset monitoring points collected by the temperature sensor;

[0009] The CNC machine tool operation state abnormality judgment module is used to judge whether the CNC machine tool operation state is abnormal based on the CNC machine tool monitoring focus feature vector and the CNC machine tool operation state joint feature vector.

[0010] According to another aspect of the present application, a method for monitoring operation of a CNC machine tool based on digital twin is provided, comprising:

[0011] Acquire monitoring images of a CNC machine tool at a plurality of preset monitoring points collected by a camera, vibration signals at a plurality of preset monitoring points collected by a vibration sensor, and temperature values ​​of a CNC machine tool at a plurality of preset monitoring points collected by a temperature sensor;

[0012] Extracting a CNC machine monitoring focus feature vector and a CNC machine operating status joint feature vector from the CNC machine monitoring images of the multiple preset monitoring points collected by the camera, the vibration signals of the multiple preset monitoring points collected by the vibration sensor, and the CNC machine temperature values ​​of the multiple preset monitoring points collected by the temperature sensor;

[0013] Based on the CNC machine tool monitoring focus feature vector and the CNC machine tool operating state joint feature vector, it is determined whether the CNC machine tool operating state is abnormal.

[0014] Compared with the prior art, the present application provides a CNC machine tool operation monitoring system and method based on digital twin, which first obtains the CNC machine tool monitoring images of multiple preset monitoring points collected by a camera, the vibration signals of multiple preset monitoring points collected by a vibration sensor, and the CNC machine tool temperature values ​​of multiple preset monitoring points collected by a temperature sensor, and then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally obtains the classification result through a classifier to determine whether there is any abnormality in the operation status of the CNC machine tool, so as to give early warning and take measures to avoid equipment shutdown or damage, thereby improving the reliability and production efficiency of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1 It is a block diagram of a CNC machine tool operation monitoring system based on digital twin according to an embodiment of the present application.

[0017] Figure 2 It is a block diagram of a CNC machine tool operation monitoring data extraction module in a CNC machine tool operation monitoring system based on digital twin according to an embodiment of the present application.

[0018] Figure 3 It is a block diagram of a CNC machine tool monitoring feature extraction unit in a CNC machine tool operation monitoring system based on digital twin according to an embodiment of the present application.

[0019] Figure 4 It is a block diagram of a CNC machine tool operation status abnormality judgment module in a CNC machine tool operation monitoring system based on digital twin according to an embodiment of the present application.

[0020] Figure 5 It is a flow chart of a method for monitoring the operation of a CNC machine tool based on digital twin according to an embodiment of the present application.

[0021] Figure 6 is a block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0023] Figure 1 FIG. 1 is a block diagram of a CNC machine tool operation monitoring system based on digital twin according to an embodiment of the present application. Figure 1As shown, according to the embodiment of the present application, the CNC machine tool operation monitoring system 100 based on digital twin includes: a CNC machine tool operation monitoring data acquisition module 110, which is used to obtain the CNC machine tool monitoring images of multiple preset monitoring points collected by a camera, the vibration signals of multiple preset monitoring points collected by a vibration sensor, and the CNC machine tool temperature values ​​of multiple preset monitoring points collected by a temperature sensor; a CNC machine tool operation monitoring data extraction module 120, which is used to extract the CNC machine tool monitoring focus feature vector and the CNC machine tool operation status joint feature vector from the CNC machine tool monitoring images of multiple preset monitoring points collected by the camera, the vibration signals of multiple preset monitoring points collected by the vibration sensor, and the CNC machine tool temperature values ​​of multiple preset monitoring points collected by the temperature sensor; a CNC machine tool operation status abnormality judgment module 130, which is used to judge whether there is an abnormality in the CNC machine tool operation status based on the CNC machine tool monitoring focus feature vector and the CNC machine tool operation status joint feature vector.

[0024] In the above-mentioned CNC machine tool operation monitoring system 100 based on digital twin, the CNC machine tool operation monitoring data acquisition module 110 is used to obtain CNC machine tool monitoring images of multiple preset monitoring points collected by cameras, vibration signals of multiple preset monitoring points collected by vibration sensors, and CNC machine tool temperature values ​​of multiple preset monitoring points collected by temperature sensors. It should be understood that CNC machine tools, i.e., digitally controlled machine tools, are automated equipment equipped with program control systems, which solve the difficult problems of complex, precise, small batch and multi-variety parts processing, and are widely used in defense aviation, automobile industry, mold manufacturing, machining and other fields. However, ensuring the precise operation status of CNC machine tools is crucial to improving production efficiency. Current monitoring technologies are usually limited to dominant factors and cannot accurately assess machine tool wear and specific status. Traditional manual inspections are inefficient and easily limited by operator skills. In addition, a large amount of processing data is usually ignored as useless information, resulting in a waste of resources. Therefore, in the technical solution of the present application, by acquiring the CNC machine tool monitoring images of multiple preset monitoring points collected by the camera, the vibration signals of multiple preset monitoring points collected by the vibration sensor, and the CNC machine tool temperature values ​​of multiple preset monitoring points collected by the temperature sensor, and combining with deep learning technology, the stable operation of the CNC machine tool is the key, because it can detect abnormal operating status in advance and issue timely warnings, thereby avoiding equipment shutdown or damage, and thus improving the reliability and production efficiency of the equipment.

[0025] Specifically, the camera provides visual information and can detect the dynamic operation and potential problems of various parts of the machine tool. The vibration sensor captures the vibration characteristics of the machine tool during operation, which can reflect the stability of the machine tool structure and working accuracy. The temperature sensor monitors the temperature changes of various parts of the machine tool to help evaluate the thermal effect and heat dissipation of the machine tool. The comprehensive analysis and processing of these data can help operators and maintenance personnel to promptly discover potential problems and optimize the operating parameters of the machine tool, thereby improving production efficiency, extending the life of the equipment, and ensuring a safe and stable production environment.

[0026] In the above-mentioned CNC machine tool operation monitoring system 100 based on digital twin, the CNC machine tool operation monitoring data extraction module 120 is used to extract the CNC machine tool monitoring focus feature vector and the CNC machine tool operation status joint feature vector from the CNC machine tool monitoring images of multiple preset monitoring points collected by the camera, the vibration signals of multiple preset monitoring points collected by the vibration sensor, and the CNC machine tool temperature values ​​of multiple preset monitoring points collected by the temperature sensor. It should be understood that the CNC machine tool monitoring focus feature vector is a key feature extracted from the camera image, such as workpiece shape, tool position, etc., which is used to monitor and optimize key parameters in the processing process. The CNC machine tool operation status joint feature vector comprehensively considers the spectral characteristics of the vibration signal and the temporal changes of the temperature. Through the processing of the joint encoder, it can more accurately reflect the overall operation status and health of the machine tool, and help to timely discover and predict potential fault risks. The extraction and analysis of these feature vectors not only helps to achieve real-time monitoring and accurate evaluation of the operating status of CNC machine tools, but also provides timely decision-making support for operation and maintenance personnel, optimizes equipment maintenance plans and production scheduling, and improves the reliability, safety and production efficiency of machine tools, thus playing an important role in the field of industrial manufacturing.

[0027] Figure 2 FIG. 4 is a block diagram of a CNC machine tool operation monitoring data extraction module in a CNC machine tool operation monitoring system based on digital twin according to an embodiment of the present application. Figure 2As shown, in a specific embodiment of the present application, the CNC machine operation monitoring data extraction module 120 includes: a CNC machine monitoring feature extraction unit 121, which is used to perform feature extraction on the CNC machine monitoring images of the multiple preset monitoring points collected by the camera to obtain the CNC machine monitoring focus feature vector; a CNC machine vibration signal feature extraction unit 122, which is used to perform feature extraction on the CNC machine vibration signals of the multiple preset monitoring points collected by the vibration sensor to obtain the CNC machine vibration video feature vector; a CNC machine temperature feature extraction unit 123, which is used to perform feature extraction on the CNC machine temperature values ​​of the multiple preset monitoring points collected by the temperature sensor to obtain the CNC machine temperature time series feature vector; a CNC machine operation status feature combination unit 124, which is used to obtain the CNC machine operation status joint feature vector by combining the CNC machine vibration video feature vector and the CNC machine temperature time series feature vector through a joint encoder.

[0028] It should be understood that through image processing and computer vision technology, important features such as workpiece morphology, tool position, cutting status, etc. in the image can be analyzed. For example, the size, shape and surface quality of the workpiece can be detected to verify whether the processing meets the design requirements; the position and status of the tool can also be identified, the tool wear and tool change requirements can be monitored, and whether the tool is abnormally worn or damaged can be detected. In the feature extraction process, image preprocessing is first performed, including steps such as denoising, contrast enhancement and edge detection, to ensure that the features extracted from the image have good quality and accuracy. Then, through specific image processing algorithms and machine learning techniques, such as convolutional neural networks based on deep learning, features that can best reflect the operating status and processing quality of the machine tool are extracted. These feature vectors not only help to monitor the working status and processing effect of the machine tool in real time, but can also be used to optimize the processing process and adjust process parameters to improve processing accuracy and production efficiency. By extracting features from CNC machine tool monitoring images, it is possible to achieve refined control and comprehensive monitoring of the processing process, help manufacturing companies improve product quality and reduce production costs, and provide a scientific basis for equipment maintenance and management, thereby promoting the application and development of intelligent manufacturing technology.

[0029] Furthermore, feature extraction is performed on the vibration signal of the CNC machine tool collected by the vibration sensor, aiming to extract the key feature vectors reflecting the operating status and health of the machine tool by analyzing the vibration data, so as to realize comprehensive monitoring and evaluation of the vibration of the machine tool. During the operation of the CNC machine tool, the vibration signal contains rich information, which can reflect the stability of the machine tool structure, the status of the tool and cutter, the stability of the cutting process, and possible machine tool abnormalities or failures. The feature extraction process of the vibration signal usually involves a variety of signal processing technologies, such as time domain analysis, frequency domain analysis, wavelet transform and other methods. Through these methods, the amplitude, frequency, waveform shape, energy distribution and other characteristics of the vibration signal can be extracted. These feature vectors can help identify and classify different types of vibration modes, such as stable vibration under normal working conditions, abnormal vibration caused by tool wear, and resonant vibration caused by structural looseness. Using the vibration signal feature vector for analysis, the operating status of the machine tool can be monitored in real time, potential mechanical failures or process problems can be discovered in time, and equipment downtime and maintenance costs can be reduced.

[0030] Furthermore, feature extraction is performed on the temperature values ​​of multiple preset monitoring points of the CNC machine tool collected by the temperature sensor, aiming to extract key feature vectors from the temperature time series data. These vectors can reflect the thermal effect, heat dissipation status and possible temperature anomalies of the machine tool during operation. During the processing of the CNC machine tool, the temperature changes in different parts directly affect the thermal stability, processing accuracy and part quality of the machine tool. Feature extraction of the data collected by the temperature sensor of the CNC machine tool to obtain the temperature time series feature vector not only helps to realize real-time monitoring and accurate evaluation of the thermal state of the machine tool, but also provides important support for optimizing the production process, improving equipment utilization and quality. It is one of the important technologies of digital production and intelligent manufacturing in modern manufacturing.

[0031] In particular, the vibration video feature vector and temperature time series feature vector of CNC machine tools are integrated through a joint encoder, aiming to comprehensively utilize the information in vibration and temperature data to obtain a joint feature vector that reflects the operating status of CNC machine tools more comprehensively and accurately. Vibration signals and temperature changes are two important physical quantities in the operation process of machine tools. They are interrelated and jointly affect the working efficiency, precision and health of machine tools. The design of a joint encoder usually involves multiple technologies, such as joint training models or feature fusion methods in deep learning. First, the vibration video feature vector can provide important information about the structural stability of the machine tool, the quality of the cutting process, and possible mechanical problems. The temperature time series feature vector reflects the thermal effect and energy distribution state of the machine tool under different working conditions. Through the joint encoder, these two types of feature vectors can be effectively integrated and fused to construct a comprehensive joint feature vector of the operating status of the CNC machine tool. This joint feature vector can more comprehensively evaluate the working status and performance of the machine tool and improve the sensitivity and accuracy of fault detection.

[0032] Figure 3 FIG. 4 is a block diagram of a CNC machine tool monitoring feature extraction unit in a CNC machine tool operation monitoring system based on digital twin according to an embodiment of the present application. Figure 3 As shown, in a specific embodiment of the present application, the CNC machine monitoring feature extraction unit 121 includes: a CNC machine monitoring data preprocessing subunit 1211, which is used to perform data preprocessing on the CNC machine monitoring images of the multiple preset monitoring points collected by the camera to obtain the CNC machine monitoring input tensor of interest; a CNC machine monitoring attention encoding subunit 1212, which is used to pass the CNC machine monitoring input tensor of interest through a CNC machine monitoring feature of interest encoder based on a three-dimensional convolutional neural network to obtain the CNC machine monitoring attention feature vector.

[0033] It should be understood that the preprocessing process first involves image denoising and enhancement. Denoising can remove random noise in the image through various filtering techniques, such as Gaussian filtering or median filtering, to ensure the stability and accuracy of subsequent processing. Secondly, by enhancing contrast and color vividness, key features in the image can be highlighted, making it easier to be detected and analyzed by the machine learning model. In addition, image standardization and normalization can be performed to ensure consistent brightness and color distribution between different images, thereby improving the generalization ability and robustness of the model. In data preprocessing, image segmentation and object detection techniques can also be applied. Image segmentation can separate different areas or objects in the image, thereby helping the model to focus on the features of specific areas. For example, by segmenting the workpiece and the tool, the model can more accurately analyze the tool wear or the workpiece processing quality. At the same time, object detection can identify specific objects or areas in the image, such as detecting different components on the machine tool, thereby further refining the input data of the model. By performing data preprocessing on the CNC machine tool monitoring image, the performance and effect of the deep learning model can be effectively improved. The optimized input tensor of interest for CNC machine tool monitoring can not only accurately reflect the processing status and working environment of the machine tool, but also provide a more accurate and reliable data basis for applications such as intelligent monitoring, fault diagnosis and process optimization, thus promoting the intelligent development of the manufacturing industry and improving production efficiency.

[0034] Furthermore, the input tensor of interest for CNC machine tool monitoring is processed by a feature encoder based on a three-dimensional convolutional neural network, aiming to effectively extract key machine tool working features from complex three-dimensional monitoring images and convert them into feature vectors of interest for CNC machine tool monitoring. CNC machine tool monitoring images usually contain three-dimensional information of time series, such as workpiece movement, tool position, and dynamic changes during machining, which are crucial for real-time monitoring and control of machine tool status. The feature encoder based on a three-dimensional convolutional neural network can make full use of the spatiotemporal information in the image. Compared with the traditional two-dimensional CNN, it considers the feature changes at different time points and depth directions in the image. First, through the convolution layer and pooling layer, the 3DCNN can effectively extract local and global features in the image and capture the dynamic changes and important details during the machine tool working process. Secondly, by stacking multiple layers of convolution and pooling operations, higher-level features are gradually abstracted and extracted, so that the model can better understand and characterize complex monitoring image data. In the process of the feature encoder, a recurrent neural network (RNN) or a long short-term memory network (LSTM) can also be used in combination to process time series features to capture the temporal dependency and dynamic changes in the image sequence. This combination can further improve the efficiency and accuracy of feature extraction, so that the monitoring feature vector can better reflect the real-time status and change trend of the machine tool processing process. Specifically, each layer of the CNC machine tool monitoring feature encoder based on the three-dimensional convolutional neural network performs convolution processing, mean pooling processing based on the local feature matrix and nonlinear activation processing on the input data in the forward transfer of the layer to output the CNC machine tool monitoring feature vector by the last layer of the CNC machine tool monitoring feature encoder based on the three-dimensional convolutional neural network, wherein the input of the CNC machine tool monitoring feature encoder based on the three-dimensional convolutional neural network is the CNC machine tool monitoring input tensor of interest.

[0035] In a specific embodiment of the present application, the CNC machine tool monitoring data preprocessing subunit 1211 includes: passing the CNC machine tool monitoring images of multiple preset monitoring points captured by the camera through a CNC machine tool monitoring region of interest feature encoder based on a target detection network to obtain multiple CNC machine tool monitoring regions of interest; arranging the multiple CNC machine tool monitoring regions of interest along a sample dimension to obtain the CNC machine tool monitoring input tensor of interest.

[0036] It should be understood that feature encoders based on object detection networks usually include advanced deep learning models such as Faster R-CNN, YOLO (You Only Look Once) or SSD (Single Shot MultiBox Detector). These models can effectively identify and locate objects or regions of interest in images by combining the feature extraction capabilities of convolutional neural networks (CNN) and the accuracy of object detection algorithms. First, through convolutional layers and feature maps, these models can extract high-level semantic features in images and capture information such as the shape, color, and texture of different objects. Secondly, through the regressor and classifier of the object detection network, accurate positioning and recognition of objects in the image can be achieved, thereby generating bounding boxes and category labels for multiple CNC machine monitoring regions of interest. In the process of feature encoders, spatial attention mechanisms or multi-scale processing techniques can also be combined to further improve the detection accuracy and robustness of regions of interest. This method can effectively identify each region of interest and generate accurate feature representations for each region in the presence of complex backgrounds and multiple objects at the same time. Specifically, in a specific example of the present application, the CNC machine tool monitoring region of interest feature encoder based on the target detection network is Fast R-CNN, Faster R-CNN or RetinaNet.

[0037] Furthermore, multiple CNC machine tool monitoring regions of interest are arranged along the sample dimension in order to construct a unified input tensor structure so that they can be effectively input into the deep learning model for comprehensive analysis and processing. In the monitoring and control process of the manufacturing industry, it is usually necessary to process data from multiple monitoring points or regions at the same time. These data may come from different sensors or monitoring devices and involve the status monitoring and feedback of various parts of the machine tool. By arranging the CNC machine tool monitoring regions of interest along the sample dimension, the data of different regions can be integrated into the same tensor in chronological order or at the same time. This structured input method helps to maintain the temporal and spatial correspondence of the data, so that the deep learning model can comprehensively consider the interaction and influence between different regions. For example, if the temperature, vibration or processing status of different positions on the CNC machine tool is monitored, arranging these data along the sample dimension can ensure that the model can take these key features into account at the same time during analysis, so as to more comprehensively understand and predict the overall working status of the machine tool. In addition, arranging the CNC machine tool monitoring regions of interest along the sample dimension can also simplify the data management and processing process. Through a unified input tensor structure, the complexity of data processing and the consumption of computing resources can be reduced, and the data processing efficiency and the speed of model training can be improved. This is of great significance for real-time monitoring and quick decision-making, especially when immediate response and adjustment to the machine tool operating status is required, which can effectively improve the system's response speed and accuracy.

[0038] In a specific embodiment of the present application, the CNC machine tool vibration signal feature extraction unit 122 includes: performing SIFT transformation on the CNC machine tool vibration signals of multiple preset monitoring points collected by the vibration sensor to obtain a CNC machine tool SIFT transformation time-frequency diagram; passing the CNC machine tool SIFT transformation time-frequency diagram through a CNC machine tool image encoder based on the Clip model to obtain the CNC machine tool vibration video feature vector.

[0039] It should be understood that the SIFT transform can effectively map the time domain features of the signal to the frequency domain by converting the vibration signal into a time-frequency diagram, while maintaining its scale invariance and the stability of the feature points. First, the vibration signal is decomposed into the energy distribution of different frequency components using methods such as fast Fourier transform (FFT) or wavelet transform, thereby showing the energy distribution of the vibration signal at different frequencies. Secondly, combined with the concept of scale space, the SIFT transform can capture the local characteristics of the vibration signal at different time scales, so that the model can more accurately analyze and understand the important structures and patterns in the vibration signal. In this way, it is possible to more accurately identify and locate abnormal or fault features in the vibration signal, such as abnormal peaks or energy concentration areas in the spectrum, providing strong data support for early warning and fault diagnosis of equipment status.

[0040] Furthermore, the SIFT transformed time-frequency graph of CNC machine tools is passed through an image encoder based on the Clip model, aiming to use advanced deep learning technology to convert the time-frequency features of vibration signals into semantically rich vibration video feature vectors. The Clip model (Contrastive Language-Image Pre-training) combines contrastive learning between text and image, can effectively capture the mapping relationship between visual and semantic information, and is suitable for encoding tasks that process multimodal data. The Clip model obtains the representation of visual features through pre-training learning, has good generalization ability and multi-semantic understanding ability, and can extract abstract features of vibration signals from complex time-frequency graphs. Secondly, the SIFT transformed time-frequency graph of CNC machine tools is input into the image encoder of the Clip model. Using its deep convolutional neural network structure, the spectral features and time change patterns in the time-frequency graph can be converted into high-level semantic representations, thereby improving the expressiveness and understanding depth of the data. Through the CNC machine tool image encoder based on the Clip model, the semantic representation of the vibration signal of the CNC machine tool can be realized, and the complex time-frequency information can be converted into a feature vector with high interpretability and semantic relevance. Specifically,

[0041] In a specific embodiment of the present application, the CNC machine tool temperature feature extraction unit 123 includes: constructing the CNC machine tool temperature values ​​of multiple preset monitoring points collected by the temperature sensor into a CNC machine tool temperature timing input matrix; passing the CNC machine tool temperature timing input matrix through a CNC machine tool temperature timing encoder based on the Clip model to obtain the CNC machine tool temperature timing feature vector.

[0042] It should be understood that constructing the temperature time series input matrix of CNC machine tools first involves the integration and processing of time series data. The temperature sensors at different preset monitoring points collect a temperature value at intervals. By arranging these data in chronological order, a two-dimensional matrix can be formed, in which each row represents a time point and each column represents the temperature value of a monitoring point. This structured data form helps to maintain the temporal relationship and spatial distribution of the data, so that the model can more effectively understand and analyze the dynamic characteristics of temperature changes. The construction of the temperature time series input matrix of CNC machine tools is not only conducive to real-time monitoring and control, but also supports the following key application aspects: First, it can detect the periodic changes and trends of temperature through time series analysis, and provide data support for the stability evaluation of the equipment operation status; secondly, it can be used for temperature anomaly detection and early warning, and identify abnormal temperature fluctuations that may affect the processing quality and equipment life; thirdly, it supports intelligent equipment management and maintenance strategy formulation, and optimizes equipment operation plans and maintenance cycles through historical data review and trend analysis, reducing maintenance costs and production downtime.

[0043] Furthermore, the temperature time series input matrix of the CNC machine tool is passed through the temperature time series encoder of the CNC machine tool based on the Clip model, aiming to utilize the powerful multimodal feature extraction capability of the Clip model to convert complex temperature time series data into high-level semantic feature vectors. The Clip model combines contrastive learning between text and images, and learns the association between vision and semantics through pre-training, which can be effectively applied to the encoding and feature extraction tasks of time series data. In the temperature time series encoder of the CNC machine tool, the Clip model can capture the complex relationship between different time points and monitoring points in the temperature time series matrix through its deep convolutional neural network structure. The encoder can convert the time dependency and spatial distribution in the time series data into feature representations with semantic meaning, and improve the expression ability and analysis depth of the data. Through the temperature time series encoder of the CNC machine tool based on the Clip model, the temperature time series data of multiple preset monitoring points can be converted into a unified feature vector representation, simplifying the complexity of data processing and model analysis, and improving the interpretability of the data and the generalization ability of the model.

[0044] In the above-mentioned CNC machine tool operation monitoring system 100 based on digital twin, the CNC machine tool operation state abnormality judgment module 130 is used to judge whether there is an abnormality in the operation state of the CNC machine tool based on the CNC machine tool monitoring attention feature vector and the CNC machine tool operation state joint feature vector. It should be understood that by comparing the monitoring attention feature vector and the CNC machine tool operation state joint feature vector with the preset normal mode or historical data, any deviation of the operation state can be detected. Abnormality is usually manifested as a significant deviation of the value in the feature vector from the normal range, or an unusual change in the relationship between the feature vectors. For example, a sudden increase in temperature, abnormal fluctuations in the vibration spectrum, or a sudden increase in current, etc., may indicate a failure of a device component, an abnormal process parameter, or other potential problems. The method based on the CNC machine tool monitoring attention feature vector and the CNC machine tool operation state joint feature vector can realize rapid detection and diagnosis of abnormal equipment operation state through efficient data processing and feature extraction. This intelligent monitoring system not only helps to reduce equipment downtime and maintenance costs, but also improves production efficiency and product quality, and promotes the widespread application and continuous optimization of digital manufacturing technology in industrial production.

[0045] Figure 4 FIG. 1 is a block diagram of a CNC machine tool operation status abnormality judgment module in a CNC machine tool operation monitoring system based on digital twin according to an embodiment of the present application. Figure 4 As shown, in a specific embodiment of the present application, the CNC machine tool operating state abnormality judgment module 130 includes: a CNC machine tool operating state feature fusion unit 131, which is used to fuse the CNC machine tool monitoring focus feature vector and the CNC machine tool operating state joint feature vector to obtain an operating state abnormality judgment classification feature vector; a CNC machine tool operating state feature optimization unit 132, which is used to perform group distribution constraints on the operating state abnormality judgment classification feature vector based on instance disturbance differential adjustment to obtain an optimized operating state abnormality judgment classification feature vector; an operating state abnormality judgment classification unit 133, which is used to pass the optimized operating state abnormality judgment classification feature vector through a classifier to obtain a classification result, and the classification result is used to judge whether there is an abnormality in the operating state of the CNC machine tool.

[0046] It should be understood that the fusion of the CNC machine tool monitoring attention feature vector and the CNC machine tool operating status joint feature vector is intended to use the comprehensive information of multi-source monitoring data to effectively determine whether there is an abnormality in the CNC machine tool operating status. The CNC machine tool monitoring attention feature vector is obtained by specific processing and feature extraction of a single monitoring data (such as temperature, vibration, etc.), which can accurately express the operating status characteristics of the equipment in specific aspects. The CNC machine tool operating status joint feature vector combines the multimodal features of different types of monitoring data, such as integrating temperature, vibration, current and other aspects of data into a comprehensive feature vector, which can more comprehensively describe the overall operating status of the equipment. By fusing these two feature vectors, the information of multiple monitoring indicators can be comprehensively considered to improve the understanding and judgment ability of the overall operating status of the CNC machine tool.

[0047] Furthermore, in the technical solution of the present application, it is considered that in the industrial CNC machine tool monitoring system, the key to classifying the abnormal operation state is to accurately capture and describe the characteristic patterns and abnormal signals in the operation of the CNC machine tool. Through feature mining, key features can be effectively extracted from complex monitoring data to distinguish between normal operation and abnormal state. Among them, the prior information of the neural network model plays a key role in this process, because the prior information of the neural network model can use the known data distribution and pattern to adjust the representation of the classification feature vector of the abnormal operation state judgment. By learning this prior information, the neural network can map the input abnormal operation state judgment classification feature vector to a more discriminative and operable feature space, namely the feature manifold. Using the neural network model to adjust the feature manifold can help the system better understand and handle different operating states, thereby improving the accuracy and robustness of anomaly detection, thereby enhancing the ability to cope with complex industrial environments.

[0048] Among them, the operation state abnormality judgment classification feature vector is subjected to the population distribution constraint based on the instance disturbance differential adjustment to obtain the optimized operation state abnormality judgment classification feature vector, including: calculating the mean and variance of the operation state abnormality judgment classification feature vector; calculating the difference between the eigenvalue of the predetermined position of the operation state abnormality judgment classification feature vector and the mean of the operation state abnormality judgment classification feature vector to obtain a first intermediate calculation value; multiplying the network parameter matrix by the operation state abnormality judgment classification feature vector and then subtracting the network parameter matrix from the operation state abnormality judgment classification feature vector by position to obtain a first intermediate vector, and calculating the Frobenius norm of the first intermediate vector to obtain a first norm value; adding the first norm value to the The variance of the classification feature vector for judging the abnormal operation state is taken and then the square root is taken to obtain a second intermediate calculation value; the first intermediate calculation value is divided by the second intermediate calculation value to obtain a third intermediate calculation value; the squares of each position in the classification feature vector for judging the abnormal operation state are calculated and then accumulated and summed to obtain a first total value; the square root of the first total value is taken and then divided by the scale of the network parameter matrix, and then the natural exponential function value is calculated to obtain a fourth intermediate calculation value, wherein the scale of the network parameter matrix is ​​equal to the product between the width value of the network parameter matrix and the height value thereof; the fourth intermediate calculation value is multiplied by the weighted hyperparameter and then added to the third intermediate calculation value to obtain the eigenvalue of the predetermined position of the optimized classification feature vector for judging the abnormal operation state.

[0049] Among them, the classification feature vector for judging abnormal operation status is subjected to group distribution constraint based on differential adjustment of instance disturbance to obtain the optimized classification feature vector for judging abnormal operation status, including:

[0050]

[0051] Wherein, V is the abnormal operation status judgment classification feature vector, M is the network parameter matrix, β represents the weighted hyperparameter, and v i is the eigenvalue of the i-th feature vector in the abnormal operation status judgment classification feature vector, N is the scale of the network parameter matrix, the scale of the network parameter matrix is ​​equal to the product of the width value and the height value of the network parameter matrix, L is the length of the abnormal operation status judgment classification feature vector, ||·|| F The Frobenius norm of the classification feature vector representing the abnormal operation status judgment, represents matrix multiplication, μ represents the mean of the feature set composed of all eigenvalues ​​of the classification feature vector for judging the abnormal operation status, σ 2 represents the variance of the feature set composed of all eigenvalues ​​of the abnormal operation status judgment classification feature vector, v' iis the ith feature value in the optimization operation state abnormality judgment classification feature vector.

[0052] That is, in the technical solution of the present application, considering that when the neural network model performs feature mining on the input data based on the network parameter matrix, if the inherent implicit association between the network output feature and the network parameter matrix can be further considered, the prior information of the neural network model can be further used to modulate the characteristic manifold of the abnormal operation state judgment classification feature vector to reduce the possibility of special instance disturbance of its characteristic value. Based on this, in a preferred embodiment, the abnormal operation state judgment classification feature vector is subjected to a group distribution constraint based on differential adjustment of instance disturbance, which first maps the abnormal operation state judgment classification feature vector to the high-dimensional feature space where the network parameter matrix is ​​located, and uses the compatibility perception factor between the mapping result and the original input as the inherent implicit association descriptor between the network output feature and the network parameter matrix. Then, the statistical characteristics of the feature group of the abnormal operation state judgment classification feature vector are used as the inherent modal information of the abnormal operation state judgment classification feature vector, and the inherent implicit association descriptor is used as the differential adjustment variable of the instance disturbance, and the attribute values ​​of each position of the abnormal operation state judgment classification feature vector are dynamically standardized. At the same time, the class-standardized modulus data of the abnormal running state judgment classification feature vector is used as a bias item to constrain the acceptable range of the population distribution of the abnormal running state judgment classification feature vector. In this way, the attribute influence of the interference distribution of abnormal samples can be suppressed as much as possible, thereby improving the expression accuracy of the abnormal running state judgment classification feature vector.

[0053] Furthermore, the optimized operation status abnormality judgment classification feature vector is passed through the classifier to obtain the classification result, which is used to judge whether there is an abnormality in the operation status of the CNC machine tool. This is to automatically identify possible problems or abnormal conditions of the equipment through machine learning and data analysis methods. Using the classifier to analyze and predict the feature vector can extract key patterns and features from complex monitoring data, and then make timely decisions and responses. The classification results obtained by the classifier can provide important decision-making basis and early warning information for equipment operators and maintenance personnel. Once the classification results show that there is an abnormality in the operation status of the CNC machine tool, corresponding measures can be taken immediately, such as shutting down for maintenance, adjusting process parameters or replacing equipment parts, to avoid potential equipment failures or production quality problems.

[0054] In summary, the embodiment of the present application first obtains the CNC machine tool monitoring images of multiple preset monitoring points collected by the camera, the vibration signals of multiple preset monitoring points collected by the vibration sensor, and the CNC machine tool temperature values ​​of multiple preset monitoring points collected by the temperature sensor, and then uses deep learning technology to perform feature extraction and association analysis on the three. Finally, a classifier is used to obtain a classification result to determine whether there is any abnormality in the operating status of the CNC machine tool, so as to provide early warning and take measures to avoid equipment shutdown or damage, thereby improving the reliability and production efficiency of the equipment.

[0055] As described above, the CNC machine tool operation monitoring system 100 based on digital twin according to the embodiment of the present application can be implemented in various terminal devices. In one example, the CNC machine tool operation monitoring system 100 based on digital twin can be integrated into the terminal device as a software module and / or hardware module. For example, the CNC machine tool operation monitoring system 100 based on digital twin can be a software module in the operating system of the terminal device, or can be an application developed for the terminal device; of course, the CNC machine tool operation monitoring system 100 based on digital twin can also be one of the many hardware modules of the terminal device.

[0056] Alternatively, in another example, the digital twin-based CNC machine tool operation monitoring system 100 and the terminal device may also be separate devices, and the digital twin-based CNC machine tool operation monitoring system 100 may be connected to the terminal device via a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0057] Figure 5 FIG. 1 is a flow chart of a method for monitoring the operation of a CNC machine tool based on digital twin according to an embodiment of the present application. Figure 5 As shown, according to the CNC machine tool operation monitoring method based on digital twin according to the embodiment of the present application, it includes: S110, acquiring the CNC machine tool monitoring images of multiple preset monitoring points collected by a camera, the vibration signals of multiple preset monitoring points collected by a vibration sensor, and the CNC machine tool temperature values ​​of multiple preset monitoring points collected by a temperature sensor; S120, extracting the CNC machine tool monitoring focus feature vector and the CNC machine tool operation status joint feature vector from the CNC machine tool monitoring images of multiple preset monitoring points collected by the camera, the vibration signals of multiple preset monitoring points collected by the vibration sensor, and the CNC machine tool temperature values ​​of multiple preset monitoring points collected by the temperature sensor; S130, judging whether there is any abnormality in the operation status of the CNC machine tool based on the CNC machine tool monitoring focus feature vector and the CNC machine tool operation status joint feature vector.

[0058] Here, those skilled in the art can understand that the specific operations of each step in the above-mentioned CNC machine tool operation monitoring method based on digital twin have been referred to above. Figures 1 to 4 The description of the CNC machine tool operation monitoring system based on digital twin has been introduced in detail, and therefore, its repeated description will be omitted.

[0059] Below, reference Figure 6 To describe an electronic device according to an embodiment of the present application.

[0060] like Figure 6 As shown, the electronic device 10 includes: a processor 11 and a memory 13. The processor 11 and the memory 13 are connected, such as through a bus 12. Optionally, the electronic device 10 may also include a transceiver 14. It should be noted that in actual applications, the transceiver 14 is not limited to one, and the structure of the electronic device 10 does not constitute a limitation on the embodiments of the present invention.

[0061] The processor 11 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of the present invention. The processor 11 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like.

[0062] The bus 12 may include a path for transmitting information between the above components. The bus 12 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 12 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0063] The memory 13 is used to store a computer program corresponding to the CNC machine tool operation monitoring method based on digital twin of the above embodiment of the present invention, and the computer program is controlled and executed by the processor 11. The processor 11 is used to execute the computer program stored in the memory 13 to implement the contents shown in the above method embodiment.

[0064] Among them, the electronic device 10 includes but is not limited to: mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 6 The electronic device 10 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0065] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0066] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0067] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0068] In the description of the present invention, it is to be understood that the terms “center”, “longitudinal”, “lateral”, “length”, “width”, “thickness”, “up”, “down”, “front”, “back”, “left”, “right”, “vertical”, “horizontal”, “top”, “bottom”, “inside”, “outside”, “clockwise”, “counterclockwise”, “axial”, “radial”, “circumferential”, etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0069] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0070] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements, unless otherwise clearly defined. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0071] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may mean that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, a first feature being "above", "above" or "above" a second feature may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is higher in level than the second feature. A first feature being "below", "below" or "below" a second feature may mean that the first feature is directly below or obliquely below the second feature, or simply means that the first feature is lower in level than the second feature.

[0072] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A CNC machine tool operation monitoring system based on digital twin, characterized in that: include: A CNC machine tool operation monitoring data acquisition module is used to acquire CNC machine tool monitoring images of multiple preset monitoring points acquired by a camera, vibration signals of multiple preset monitoring points acquired by a vibration sensor, and CNC machine tool temperature values ​​of multiple preset monitoring points acquired by a temperature sensor; A CNC machine tool operation monitoring data extraction module, used to extract a CNC machine tool monitoring focus feature vector and a CNC machine tool operation status joint feature vector from the CNC machine tool monitoring images of the multiple preset monitoring points collected by the camera, the vibration signals of the multiple preset monitoring points collected by the vibration sensor, and the CNC machine tool temperature values ​​of the multiple preset monitoring points collected by the temperature sensor; The CNC machine tool operation state abnormality judgment module is used to judge whether the CNC machine tool operation state is abnormal based on the CNC machine tool monitoring focus feature vector and the CNC machine tool operation state joint feature vector.

2. The CNC machine tool operation monitoring system based on digital twin according to claim 1 is characterized in that: The CNC machine tool operation monitoring data extraction module comprises: A CNC machine tool monitoring feature extraction unit, used for extracting features from the CNC machine tool monitoring images of the plurality of preset monitoring points acquired by the camera to obtain the CNC machine tool monitoring focus feature vector; A CNC machine tool vibration signal feature extraction unit, used for extracting features from the CNC machine tool vibration signals of a plurality of preset monitoring points collected by the vibration sensor to obtain a CNC machine tool vibration video feature vector; A CNC machine tool temperature feature extraction unit, used for extracting features from the CNC machine tool temperature values ​​at a plurality of preset monitoring points collected by the temperature sensor to obtain a CNC machine tool temperature time series feature vector; The CNC machine tool operation status feature combining unit is used to obtain the CNC machine tool operation status combined feature vector by combining the CNC machine tool vibration video feature vector and the CNC machine tool temperature time series feature vector through a combined encoder.

3. The CNC machine tool operation monitoring system based on digital twin according to claim 2 is characterized in that: The CNC machine tool monitoring feature extraction unit comprises: A CNC machine tool monitoring data preprocessing subunit, used for performing data preprocessing on the CNC machine tool monitoring images of the plurality of preset monitoring points acquired by the camera to obtain an input tensor of interest for CNC machine tool monitoring; The CNC machine tool monitoring interest encoding subunit is used to obtain the CNC machine tool monitoring interest feature vector by passing the CNC machine tool monitoring interest input tensor through a CNC machine tool monitoring interest feature encoder based on a three-dimensional convolutional neural network.

4. The CNC machine tool operation monitoring system based on digital twin according to claim 3 is characterized in that: The CNC machine tool monitoring data preprocessing subunit comprises: The CNC machine tool monitoring images of the multiple preset monitoring points collected by the camera are passed through a CNC machine tool monitoring region of interest feature encoder based on a target detection network to obtain multiple CNC machine tool monitoring regions of interest; The multiple CNC machine tool monitoring interest regions are arranged along the sample dimension to obtain the CNC machine tool monitoring interest input tensor.

5. The CNC machine tool operation monitoring system based on digital twin according to claim 4 is characterized in that: The CNC machine tool vibration signal feature extraction unit comprises: Performing SIFT transformation on the vibration signals of the CNC machine tool at the plurality of preset monitoring points collected by the vibration sensor to obtain a SIFT transformation time-frequency diagram of the CNC machine tool; The SIFT transformed time-frequency graph of the CNC machine tool is passed through a CNC machine tool image encoder based on the Clip model to obtain a vibration video feature vector of the CNC machine tool.

6. The CNC machine tool operation monitoring system based on digital twin according to claim 5 is characterized in that: The CNC machine tool temperature feature extraction unit comprises: Constructing the CNC machine tool temperature values ​​of the plurality of preset monitoring points collected by the temperature sensor into a CNC machine tool temperature time series input matrix; The temperature timing input matrix of the numerical control machine tool is passed through a temperature timing encoder of the numerical control machine tool based on the Clip model to obtain the temperature timing feature vector of the numerical control machine tool.

7. The CNC machine tool operation monitoring system based on digital twin according to claim 6 is characterized in that: The abnormal operation status judgment module of the numerical control machine tool comprises: A CNC machine tool operation state feature fusion unit, used for fusing the CNC machine tool monitoring focus feature vector and the CNC machine tool operation state joint feature vector to obtain an operation state abnormality judgment classification feature vector; A CNC machine tool operation state feature optimization unit is used to perform group distribution constraints on the operation state abnormality judgment classification feature vector based on instance disturbance differential adjustment to obtain an optimized operation state abnormality judgment classification feature vector; The operation state abnormality judgment classification unit is used to pass the optimized operation state abnormality judgment classification feature vector through a classifier to obtain a classification result, and the classification result is used to judge whether there is an abnormality in the operation state of the CNC machine tool.

8. The CNC machine tool operation monitoring system based on digital twin according to claim 7 is characterized in that: The CNC machine tool operating state characteristic optimization unit comprises: Calculating the mean and variance of the abnormal operation status judgment classification feature vector; Calculating the difference between the feature value of the predetermined position of the abnormal operation state judgment classification feature vector and the mean value of the abnormal operation state judgment classification feature vector to obtain a first intermediate calculation value; Multiplying the network parameter matrix by the running state abnormality judgment classification feature vector and then subtracting the matrix from the running state abnormality judgment classification feature vector by position to obtain a first intermediate vector, and calculating the Frobenius norm of the first intermediate vector to obtain a first norm value; Add the first norm value to the variance of the classification feature vector for abnormal operation status judgment and then take the square root to obtain a second intermediate calculation value; dividing the first intermediate calculated value by the second intermediate calculated value to obtain a third intermediate calculated value; Calculating the square of each position in the abnormal operation status judgment classification feature vector and then accumulating and summing them to obtain a first total value; Calculate a natural exponential function value after taking the square root of the first sum value and dividing it by the scale of the network parameter matrix to obtain a fourth intermediate calculation value, wherein the scale of the network parameter matrix is ​​equal to the product of the width value of the network parameter matrix and the height value thereof; The fourth intermediate calculated value is multiplied by the weighted hyperparameter and then added to the third intermediate calculated value to obtain the feature value of the predetermined position of the optimized operation state abnormality judgment classification feature vector.

9. A method for monitoring the operation of a CNC machine tool based on digital twin, characterized in that: include: Acquire monitoring images of a CNC machine tool at a plurality of preset monitoring points collected by a camera, vibration signals at a plurality of preset monitoring points collected by a vibration sensor, and temperature values ​​of a CNC machine tool at a plurality of preset monitoring points collected by a temperature sensor; Extracting a CNC machine monitoring focus feature vector and a CNC machine operating status joint feature vector from the CNC machine monitoring images of the multiple preset monitoring points collected by the camera, the vibration signals of the multiple preset monitoring points collected by the vibration sensor, and the CNC machine temperature values ​​of the multiple preset monitoring points collected by the temperature sensor; Based on the CNC machine tool monitoring focus feature vector and the CNC machine tool operating state joint feature vector, it is determined whether the CNC machine tool operating state is abnormal.

10. The method for monitoring operation of a CNC machine tool based on digital twin according to claim 9, characterized in that: Extracting a CNC machine monitoring focus feature vector and a CNC machine running state joint feature vector from the CNC machine monitoring images of the multiple preset monitoring points collected by the camera, the vibration signals of the multiple preset monitoring points collected by the vibration sensor, and the CNC machine temperature values ​​of the multiple preset monitoring points collected by the temperature sensor, including: Performing feature extraction on the CNC machine tool monitoring images of the plurality of preset monitoring points collected by the camera to obtain the CNC machine tool monitoring focus feature vector; Extracting features of the vibration signals of the CNC machine tool collected by the vibration sensor at the plurality of preset monitoring points to obtain a CNC machine tool vibration video feature vector; Performing feature extraction on the temperature values ​​of the CNC machine tool at the plurality of preset monitoring points collected by the temperature sensor to obtain a time series feature vector of the temperature of the CNC machine tool; The vibration video feature vector of the CNC machine tool and the temperature time series feature vector of the CNC machine tool are passed through a joint encoder to obtain a joint feature vector of the operation state of the CNC machine tool.

Images