A digital twin tool wear life prediction method and system based on a physical information neural network

Abstract

The present application relates to a kind of digital twin tool wear life prediction method and system based on physical information neural network, comprising:s1: tool data processing;S2: resultant force calculation;S3: cutting speed calculation;S4: sliding distance calculation;S5: load calculation;S6: neural network establishment;Physical information neural network is established based on main cutting force model and Archard wear model;S7: loss function establishment;S8: feedback and iteration;S9: calculate tool wear;Tool wear is calculated by physical information neural network.The present application integrates the physical formula related to the tool wear stage of numerical control machine tool into physical information neural network as part of loss function, can guarantee that data is iteratively trained under the premise of relevant physical law, and the process is enhanced Explainability.

Description

Technical Field

[0001] This invention relates to a digital twin tool wear life prediction method and system based on physical information neural networks, belonging to the field of intelligent manufacturing tool fault diagnosis and health management technology. Background Technology

[0002] Tool wear is a key factor affecting many aspects of metal cutting, including surface quality, machining efficiency, and tool life. During machining, tool wear reduces part dimensional accuracy and surface integrity; severe wear can even lead to tool breakage, resulting in workpiece scrap and machine tool damage. Predicting tool wear life allows for more accurate production planning, avoiding production interruptions due to excessive tool wear. This is crucial for improving production efficiency and reducing downtime. Furthermore, tool wear life prediction is an important component of intelligent manufacturing. By combining tool wear life prediction with other intelligent manufacturing technologies (such as machine learning and big data analytics), automation and intelligence of the production process can be achieved, improving production efficiency, reducing production costs, and enhancing product quality. However, existing methods mostly employ data-driven prediction methods or combine finite element simulation with data prediction to predict tool wear life, which have limitations in interpretability and computational speed. Summary of the Invention

[0003] To address the shortcomings of existing technologies, this invention provides a digital twin tool wear life prediction method based on physical information neural networks;

[0004] This invention integrates physical formulas related to tool wear into a neural network, enabling the prediction of tool wear data while adhering to physical laws. This ensures physical consistency and achieves a dual-driven approach of data mechanism model in digital twin model construction. First, the physical laws of tool wear are clearly defined and embedded into a physical information neural network as the loss function. Since the calculation of tool wear life prediction is crucial, this invention combines the feed force model and the Archard wear model, using the physical information neural network to predict tool wear. The feed force model provides force information during the cutting process, while the Archard wear model converts these forces into wear amounts. Therefore, during training, the loss function includes both data error and physical error, allowing the model to fit data while adhering to physical laws. This increases interpretability and reduces computation time compared to methods combining finite element simulation and data prediction.

[0005] This invention provides a digital twin tool wear life prediction system based on physical information neural networks;

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A digital twin tool wear life prediction method based on physical information neural networks includes:

[0008] s1: Tool data processing; tool data refers to the tool wear dataset, including milling force signals from the X, Y, and Z axes, in order to determine the force (F) of each axis. x ,F y ,F z The role of (F) in the feed force model and wear model; x ,F y ,F z These refer to the milling force signals of the X, Y, and Z axes, respectively.

[0009] s2: Calculation of resultant force;

[0010] s3: Cutting speed calculation;

[0011] s4: Sliding distance calculation;

[0012] s5: Load calculation;

[0013] s6: Neural Network Construction; A physical information neural network was built based on the main cutting force model and the Archard wear model;

[0014] s7: Loss function construction;

[0015] The tool wear is calculated by combining the main cutting force model and the Archard wear model, thereby determining the physical loss function in the physical information neural network. The calculation formula is as follows:

[0016]

[0017] Among them, F f The main cutting force, K f The main cutting force constant, A c Where n is the cutting area, v is the cutting speed, and n is the cutting velocity. e Where is the speed index, W is the wear amount, k is the wear coefficient, F is the load, s is the sliding distance, and H is the material hardness.

[0018] s8: Feedback and iteration; that is, model evaluation and validation, which evaluates and validates the updated physical information neural network.

[0019] s9: Calculate tool wear; calculate tool wear using a physical information neural network.

[0020] According to a preferred embodiment of the present invention, the resultant force calculation includes:

[0021] The resultant force F is obtained by generalizing the forces along each axis. cThe calculation formula is as follows:

[0022]

[0023] According to a preferred embodiment of the present invention, the cutting speed calculation includes:

[0024] For rotating workpieces, the formula for calculating the cutting speed v is as follows:

[0025] v=π·D·n(3);

[0026] Where v is the cutting speed in meters per minute, D is the diameter of the workpiece in meters, and n is the spindle speed in revolutions per minute.

[0027] According to a preferred embodiment of the present invention, the sliding distance calculation includes:

[0028] The formula for calculating the sliding distance is as follows:

[0029]

[0030] Where s is the sliding distance in mm, f is the feed per pass in mm / rev, N is the total number of revolutions, n is the spindle speed, and t is the machining time in minutes.

[0031] According to a preferred embodiment of the present invention, the neural network construction refers to: constructing a physical information neural network based on the main cutting force model and the Archard wear model, incorporating physical constraints into the loss function of the physical information neural network, and realizing the prediction of wear amount; including:

[0032] 1) Data preparation; the data includes sensor data and wear data;

[0033] Sensor data: including force / vibration signals along the X, Y, and Z axes, i.e. (F... x ,F y ,F z );

[0034] Wear data: the actual wear value of the tool;

[0035] 2) Definition of physical constraints;

[0036] The physical constraint definition part is completed by combining the main cutting force model and the Archard wear model; the main cutting force model is Equation (2), and the Archard wear model is Equation (1);

[0037] 3) Define a custom loss function;

[0038] The custom loss function is the differential equation derived from the Archard wear formula over time, as shown below:

[0039]

[0040] 4) Construction of physical information neural network;

[0041] A physical information neural network was built based on the main cutting force model and the Arcard wear model;

[0042] 5) Train the model;

[0043] The force / vibration signals along the X, Y, and Z axes, along with their corresponding wear data, are input into a physical information neural network for training. The physical information neural network continuously adjusts its parameters by optimizing a custom loss function until it converges; this includes:

[0044] 6) Wear and tear prediction process;

[0045] After the physical information neural network is trained, it enters the prediction stage; input force / vibration signals of the X, Y, and Z axes: acquire new force or vibration signals from the sensors, and use the trained physical information neural network to predict the amount of wear;

[0046] 7) Comparison with actual wear values

[0047] The model's predicted wear values ​​are compared with the actual wear values ​​by plotting images.

[0048] Further preferred methods include comparing with actual wear values; including:

[0049] A. Data Acquisition and Classification:

[0050] Take a portion of the prepared dataset as test or validation data;

[0051] B. Obtaining model prediction results:

[0052] The input features of the test data or verification data are fed into the trained physical information neural network, and the corresponding predicted wear value is obtained through forward propagation.

[0053] C. Comparison Curve Chart: Using the index of the data point or time as the horizontal axis, plot the actual wear value and the predicted wear value as two curves for comparison.

[0054] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of a digital twin tool wear life prediction method based on a physical information neural network.

[0055] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a digital twin tool wear life prediction method based on a physical information neural network.

[0056] A digital twin tool wear life prediction system based on physical information neural networks includes:

[0057] The tool data processing module is configured to process tool data, specifically tool wear datasets, including milling force signals from the X, Y, and Z axes, and to determine the force (F) for each axis. x ,F y ,F z The role of (F) in the feed force model and wear model; x ,F y ,F z These refer to the milling force signals of the X, Y, and Z axes, respectively.

[0058] The calculation module is configured to calculate the resultant force, cutting speed, sliding distance, and load.

[0059] The neural network construction module is configured to build a physical information neural network based on the main cutting force model and the Archard wear model.

[0060] The loss function construction module is configured to combine the main cutting force model and the Archard wear model to calculate the tool wear, thereby determining the physical loss function in the physical information neural network.

[0061] The tool wear calculation module is configured to calculate tool wear through a physical information neural network.

[0062] The beneficial effects of this invention are as follows:

[0063] 1. Interpretability: By incorporating the physical formulas related to the wear stage of CNC machine tool tools into the physical information neural network as part of the loss function, it is possible to ensure that the data is iteratively trained under the premise of relevant physical laws, thereby enhancing the interpretability of the process.

[0064] 2. Improve prediction accuracy: By combining data-driven and mechanism-driven approaches, mechanism-driven approaches can be used to compensate for errors in data-driven approaches, thereby further improving prediction accuracy.

[0065] 3. Improved computation speed: Compared with traditional finite element simulation for mechanism prediction, this invention can improve the computation speed of the mechanism part and perform data mechanism fusion in the overall large network architecture. Attached Figure Description

[0066] Figure 1This is a schematic diagram of the process for predicting the wear life of a digital twin tool based on a physical information neural network according to the present invention.

[0067] Figure 2 This is an architecture diagram of the physical information neural network of the present invention;

[0068] Figure 3 This is a schematic diagram of the physical information neural network built based on the main cutting force model and the Arcard wear model of the present invention. Detailed Implementation

[0069] The present invention will be further defined below with reference to the accompanying drawings and embodiments, but is not limited thereto.

[0070] Example 1

[0071] A digital twin tool wear life prediction method based on physical information neural networks includes:

[0072] s1: Tool data processing; tool data refers to the tool wear dataset, including milling force signals from the X, Y, and Z axes, in order to determine the force (F) of each axis. x ,F y ,F z The role of feed force in feed force and wear models; generally, feed force is a resultant force, which can be obtained by combining forces from various directions. (F) x ,F y ,F z These refer to the milling force signals of the X, Y, and Z axes, respectively.

[0073] s2: Calculation of resultant force;

[0074] s3: Cutting speed calculation;

[0075] s4: Sliding distance calculation;

[0076] s5: Load calculation; such as Figure 1 As shown, the load calculation process is as follows: the resultant force Fc calculated from the vibration force signals of the X, Y, and Z axes and the main cutting force Ff are squared, added together, and then the square root is taken.

[0077] s6: Neural Network Construction; A physical information neural network was built based on the main cutting force model and the Archard wear model;

[0078] s7: Loss function construction;

[0079] The tool wear is calculated by combining the main cutting force model and the Archard wear model, thereby determining the physical loss function in the physical information neural network. The calculation formula is as follows:

[0080]

[0081] Among them, F f The main cutting force, K f The main cutting force constant, A c The cutting area is (depth of cut × width of cut), v is the cutting speed, and n is the cutting speed. e Where is the speed index (empirical value), W is the wear amount, k is the wear coefficient, F is the load, s is the sliding distance, and H is the material hardness.

[0082] S8: Feedback and iteration; i.e., model evaluation and validation, evaluating and validating the updated physical information neural network; the evaluation method is: during training, the performance of the physical information neural network is evaluated using the PINN loss metric, the smaller the PINN loss value, the better the performance; the validation method is: on the validation dataset, the difference between the predicted value and the true value is validated, when the difference is small enough, the physical information neural network meets the requirements (please explain and supplement); to ensure that the performance and accuracy of the model meet the requirements.

[0083] s9: Calculate tool wear; calculate tool wear using a physical information neural network.

[0084] Following S9, a digital twin system for tool wear was built based on physical information neural networks. The CNC machine tool was modeled using Unreal Engine to complete the visualization part of the digital twin system. The PINN neural network was embedded into Unreal Engine through the Unreal Engine Blueprint function, and the data input to Unreal Engine was completed using socket communication. This integrated the neural network into Unreal Engine to complete the data interaction and visualization display.

[0085] Example 2

[0086] The difference between the digital twin tool wear life prediction method based on physical information neural network described in Example 1 and the method described in Example 1 is as follows:

[0087] Resultant force calculation; including:

[0088] The resultant force F is obtained by generalizing the forces along each axis. c The calculation formula is as follows:

[0089]

[0090] Cutting speed calculation; including:

[0091] Cutting speed is the speed at which the cutting tool moves relative to the workpiece, usually expressed in meters per minute (m / min). During machining, the calculation of cutting speed depends on the workpiece's rotational speed (spindle speed) and the workpiece's diameter. For rotating workpieces (such as in turning), the formula for calculating the cutting speed v is as follows:

[0092] v=π·D·n(3);

[0093] Where v is the cutting speed in meters per minute (m / min), D is the diameter of the workpiece in meters (m), and n is the spindle speed in revolutions per minute (RPM).

[0094] Sliding distance calculation; including:

[0095] The sliding distance *s* is directly related to the feed rate during machining. It is calculated using the feed rate, representing the total distance the tool travels relative to the workpiece throughout the machining process. The formula for calculating the sliding distance is as follows:

[0096]

[0097] Where s is the sliding distance in mm, f is the feed per pass in mm / rev, N is the total number of revolutions, n is the spindle speed, and t is the machining time in minutes.

[0098] Neural network construction refers to: building a physical information neural network based on the main cutting force model and the Archard wear model, incorporating physical constraints into the loss function of the physical information neural network, and achieving prediction of wear amount; including:

[0099] In this invention, since the tool wear life is predicted, the initial condition is 0, and the boundary condition is the maximum wear value when machining is impossible.

[0100] The first part is the data training section, where the neural network is trained using the input data. The second part is unique to PINN, and it includes the PINN loss, which is composed of the main cutting force model and the Archard model. Its function is to complete the data mechanism fusion by continuously fitting the results with the PINN loss based on the data training in the first part.

[0101] 1) Data preparation; the data includes sensor data and wear data;

[0102] Sensor data: Receives sensor data such as force signals and vibration signals from machine operation, including force / vibration signals along the X, Y, and Z axes (F...). x ,F y ,F z );

[0103] Wear data: the actual wear value of the tool; obtained through measurement or experimentation, serving as a monitoring signal.

[0104] 2) Definition of physical constraints;

[0105] The physical constraint definition part is completed by combining the main cutting force model and the Archard wear model; the main cutting force model is Equation (2), and the Archard wear model is Equation (1);

[0106] 3) Define a custom loss function;

[0107] PINN combines loss functions for both data error and physical error to optimize the model during training: Data loss: For example, using mean squared error (MSE) or mean absolute error (MAE) to measure the difference between the model's predicted wear values ​​and the actual measured values. Physical loss: Calculating the error based on physical formulas and incorporating it into the optimization objective through a custom loss function.

[0108] The custom loss function is the differential equation derived from the Archard wear formula over time, as shown below:

[0109]

[0110] 4) Construction of physical information neural network;

[0111] A physical information neural network is built based on the main cutting force model and the Archard wear model; a neural network model, such as an LSTM network, is constructed to process time series data and combine historical force signals to predict wear.

[0112] like Figure 2 and Figure 3 As shown, an RNN network composed of LSTM is used to predict wear values. The mechanism module guides the data to be trained under physical constraints according to physical formulas in the network. Physical knowledge is integrated in this process. The data loss is the loss obtained by training the data in the data-driven model. The final loss is composed of both.

[0113] 5) Train the model;

[0114] The force / vibration signals along the X, Y, and Z axes, along with their corresponding wear data, are input into a physical information neural network for training. The physical information neural network continuously adjusts its parameters by optimizing a custom loss function (which incorporates both data error and physical constraints) until it converges. This includes:

[0115] Parameter initialization: Randomly initialize the neural network weights and biases (e.g., Xavier initialization).

[0116] Forward Pass: The pre-processed X, Y, and Z vibration / force signals are input into PINN to obtain the predicted wear amount.

[0117] Calculate the loss function:

[0118] The data error LData is calculated based on the predicted and actual wear amounts. The physical residual Lphysics is calculated based on the network output and the built-in physical equations. The total loss Loss is obtained by weighted summation of the two.

[0119] Backpropagation: Calculates the gradient of the total loss with respect to the network parameters using automatic differentiation techniques.

[0120] Parameter Update: Use optimization algorithms (such as Adam or LBFGS) to update network parameters based on gradient information, so that the total loss is continuously reduced.

[0121] Epochs: Repeat the above steps for several rounds until the network parameters converge, that is, the total loss or validation set error no longer decreases significantly, or the preset training stopping condition is reached.

[0122] Convergence and Verification:

[0123] In the early stages of training, the network parameters are relatively disordered, failing to fit the data well or strictly satisfy the physical equations. As training iterates, under the joint constraints of Ldata and Lphysics, the network parameters are gradually adjusted, enabling the prediction results to simultaneously match the measurement data and satisfy the physical laws.

[0124] The network can be considered converged when the loss function stabilizes, the gradient changes become stationary, or the performance on the validation set reaches a satisfactory level. At this point, the PINN model not only fits the training data well but can also extrapolate to unseen data to a certain extent, while maintaining an inherent consistency with physical laws.

[0125] 6) Wear and tear prediction process;

[0126] After the physical information neural network is trained, it enters the prediction stage; input force / vibration signals of the X, Y, and Z axes: acquire new force or vibration signals from the sensors, and use the trained physical information neural network to predict the amount of wear;

[0127] 7) Comparison with actual wear values

[0128] The model's predicted wear values ​​are compared with the actual wear values ​​by plotting images.

[0129] Comparison with actual wear values; including:

[0130] A. Data Acquisition and Classification:

[0131] Take a portion of the prepared dataset as test or validation data; this data was not used to update the model parameters during training, so it can be used to objectively evaluate the model's predictive performance.

[0132] B. Obtaining model prediction results:

[0133] The input features of the test or validation data are fed into a pre-trained Physical Information Neural Network (PINN), and the predicted wear values ​​are obtained through forward propagation; this generates an array or list of "predicted wear values". At the same time, the corresponding "actual wear values" (true measurements) from the test dataset are also available.

[0134] C. Line / Comparison Plot: Using the index of the data point or time as the horizontal axis, the actual wear value and the predicted wear value are plotted as two curves for comparison. If the two curves are relatively close, it indicates that the prediction is relatively accurate.

[0135] The PINN-based data mechanism-driven dual-drive digital twin tool wear life prediction method has broad prospects in practical industrial applications.

[0136] Here are some specific application examples:

[0137] 1. Aviation manufacturing industry:

[0138] Turbine blade machining: Machining turbine blades requires high precision and stability. Using PINN to predict tool wear allows for real-time adjustment of machining parameters, ensuring machining quality and extending tool life.

[0139] Composite material machining: The machining of composite materials places high demands on tool wear. Digital twin technology allows for real-time monitoring and prediction of tool wear, optimizing machining processes and reducing downtime.

[0140] 2. Automobile manufacturing industry:

[0141] Engine component machining: For high-precision parts such as crankshafts and camshafts, controlling tool wear is crucial. The PINN model allows for real-time prediction of tool wear during production, enabling timely tool replacement and ensuring product quality.

[0142] Body manufacturing: In the body manufacturing process, using PINN to predict tool wear can optimize cutting and stamping processes, improving production efficiency and product consistency.

[0143] 3. Mold manufacturing:

[0144] Injection mold processing: The processing of injection molds places high demands on the cutting tools. Through PINN technology, the status of the cutting tools can be monitored in real time, their wear life can be predicted, thereby optimizing processing parameters and extending the tool life.

[0145] Stamping Die Machining: During the stamping die machining process, tool wear directly affects product quality. Using digital twin technology and the PINN model, real-time monitoring and prediction of tool status can be achieved, ensuring efficient and stable die machining.

[0146] 4. Energy sector:

[0147] Wind turbine blade manufacturing: The manufacturing of wind turbine blades involves a large amount of composite materials, and tool wear is a critical issue. PINN technology can predict tool wear in real time, optimize the machining process, and improve production efficiency.

[0148] Oil drilling equipment manufacturing: In the manufacture of oil drilling equipment, tool wear directly affects the equipment's lifespan. Using digital twins and PINN models, tool wear can be predicted, manufacturing processes optimized, and high equipment quality ensured.

[0149] 5. Electronics Manufacturing Industry:

[0150] Semiconductor Manufacturing: In semiconductor manufacturing, the precision and lifespan of cutting tools are crucial. The PINN model allows for real-time prediction of tool wear, enabling timely maintenance and replacement, ensuring production continuity and product consistency.

[0151] Circuit board processing: Circuit board processing requires high-precision cutting tools. Digital twin technology can be used to monitor the tool status in real time and predict its wear life, thereby optimizing the processing technology and improving product quality.

[0152] These examples demonstrate the application potential of PINN-based data mechanism-driven dual-drive digital twin tool wear and life prediction methods across various industries. By combining physical information with data-driven models, accurate prediction and optimization of tool conditions can be achieved, improving production efficiency and product quality.

[0153] Example 3

[0154] A computer device includes a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the digital twin tool wear life prediction method based on physical information neural network described in Embodiment 1 or 2.

[0155] Example 4

[0156] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the digital twin tool wear life prediction method based on a physical information neural network as described in Embodiment 1 or 2.

[0157] Example 5

[0158] A digital twin tool wear life prediction system based on physical information neural networks includes:

[0159] The tool data processing module is configured to process tool data, specifically tool wear datasets, including milling force signals from the X, Y, and Z axes, and to determine the force (F) for each axis. x ,F y ,F z The role of (F) in the feed force model and wear model; x ,F y ,F z These refer to the milling force signals of the X, Y, and Z axes, respectively.

[0160] The calculation module is configured to calculate the resultant force, cutting speed, sliding distance, and load.

[0161] The neural network construction module is configured to build a physical information neural network based on the main cutting force model and the Archard wear model.

[0162] The loss function construction module is configured to combine the main cutting force model and the Archard wear model to calculate the tool wear, thereby determining the physical loss function in the physical information neural network.

[0163] The tool wear calculation module is configured to calculate tool wear through a physical information neural network.

Claims

1. A digital twin tool wear life prediction method based on physical information neural networks, characterized in that, include: s1: Tool data processing; tool data refers to the tool wear dataset, including force and vibration signals along the X, Y, and Z axes (F... x ,F y ,F z ), in determining (F x ,F y ,F z Its role in the feed force model and wear model; s2: Calculation of resultant force; s3: Cutting speed calculation; s4: Sliding distance calculation; s5: Load calculation; S6: Neural Network Construction; refers to: building a physical information neural network based on the main cutting force model and the Archard wear model, incorporating physical constraints into the loss function of the physical information neural network to predict wear amount; including: 1) Data preparation; the data includes sensor data and wear data; Sensor data: including force and vibration signals along the X, Y, and Z axes, i.e. (F... x ,F y ,F z ); Wear data: the actual wear value of the tool; 2) Definition of physical constraints; The physical constraint definition part is completed by combining the main cutting force model and the Archard wear model; the main cutting force model is Equation (2), and the Archard wear model is Equation (1); 3) Define a custom loss function; The custom loss function is the differential equation derived from the Archard wear formula over time, as shown below: 4) Construction of physical information neural network; A physical information neural network was built based on the main cutting force model and the Arcard wear model; 5) Train the model; The force / vibration signals along the X, Y, and Z axes, along with their corresponding wear data, are input into a physical information neural network for training. The physical information neural network continuously adjusts its parameters by optimizing a custom loss function until it converges; this includes: 6) Wear and tear prediction process; After the physical information neural network is trained, it enters the prediction stage; input force / vibration signals of the X, Y, and Z axes: acquire new force or vibration signals from the sensors, and use the trained physical information neural network to predict the amount of wear; 7) Comparison with actual wear values The model's predicted wear values ​​are compared with the actual wear values ​​by plotting images. S7: Building the loss function; The tool wear is calculated by combining the main cutting force model and the Archard wear model, thereby determining the physical loss function in the physical information neural network. The calculation formula is as follows: Among them, F f The main cutting force, K f The main cutting force constant, A c Where n is the cutting area, v is the cutting speed, and n is the cutting velocity. e Where is the speed index, W is the wear amount, k is the wear coefficient, F is the load, s is the sliding distance, and H is the material hardness; Resultant force calculation; including: The resultant force F is obtained by generalizing the forces along each axis. c The calculation formula is as follows: S8: Feedback and iteration; namely, model evaluation and validation, evaluating and validating the updated physical information neural network; S9: Calculate tool wear; calculate tool wear using a physical information neural network.

2. The method for predicting the wear life of a digital twin tool based on a physical information neural network according to claim 1, characterized in that, Cutting speed calculation; including: For rotating workpieces, the formula for calculating the cutting speed v is as follows: v=π·D·n(3); Where v is the cutting speed in meters per minute, D is the diameter of the workpiece in meters, and n is the spindle speed in revolutions per minute.

3. The method for predicting the wear life of a digital twin tool based on a physical information neural network according to claim 1, characterized in that, Sliding distance calculation; including: The formula for calculating the sliding distance is as follows: Where s is the sliding distance in mm, f is the feed per pass in mm / rev, N is the total number of revolutions, n is the spindle speed, and t is the machining time in minutes.

4. The method for predicting the wear life of a digital twin tool based on a physical information neural network according to claim 1, characterized in that, Comparison with actual wear values; including: A. Data Acquisition and Classification: Take a portion of the prepared dataset as test data; B. Obtaining model prediction results: The input features of the test data are fed into the trained physical information neural network, and the corresponding predicted wear value is obtained through forward propagation. C. Comparison Curve Chart: Using the index of the data point or time as the horizontal axis, plot the actual wear value and the predicted wear value as two curves for comparison.

5. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the digital twin tool wear life prediction method based on physical information neural network as described in any one of claims 1-4.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the digital twin tool wear life prediction method based on physical information neural network as described in any one of claims 1-4.

7. A digital twin tool wear life prediction system based on a physical information neural network, used to implement the digital twin tool wear life prediction method based on a physical information neural network as described in any one of claims 1-4, characterized in that, include: The tool data processing module is configured to process tool data, i.e., tool wear datasets, including force and vibration signals along the X, Y, and Z axes (F...). x ,F y ,F z ), in determining (F x ,F y ,F z Its role in the feed force model and wear model; The calculation module is configured to calculate the resultant force, cutting speed, sliding distance, and load. The neural network construction module is configured to build a physical information neural network based on the main cutting force model and the Archard wear model. The loss function construction module is configured to combine the main cutting force model and the Archard wear model to calculate the tool wear, thereby determining the physical loss function in the physical information neural network. The tool wear calculation module is configured to calculate tool wear through a physical information neural network.

Images