Continuous generating grinding tooth surface waviness monitoring system based on digital twinning

Abstract

The invention relates to the technical field of machining surface quality control, and discloses a continuous generating grinding tooth surface waviness monitoring system based on digital twinning. Comprising a data acquisition module which acquires vibration signals of a grinding wheel spindle and a workbench, three-dimensional model data of a numerical control gear grinding machine and numerical control system data in real time, and a simulation module comprises a motion and grinding rendering unit and a UI interface; the waviness online prediction and process parameter optimization module comprises a neural network prediction model, a process parameter optimization model and a process parameter feedback interface; according to the method, the digital twin model of the gear grinding machine is established, the tooth surface waviness condition can be monitored in real time, and when the ghost amplitude of the tooth surface waviness exceeds a threshold value, grinding process parameters can be optimized online, so that the gear surface quality is improved, and the production efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of machining surface quality control, and in particular to a continuous generating grinding tooth surface waviness monitoring system based on digital twins. Background Art

[0002] As a key component of the transmission system, gears are widely used in high-end equipment such as aviation, aerospace, ships, rail transportation, and new energy. In the motor drive system of new energy vehicles, due to the low noise level of the motor, the noise and vibration problems generated by the gear transmission system are becoming more prominent, becoming the main factor affecting the NVH performance of the entire vehicle. The waviness of the tooth surface is a mesoscopic morphological feature between the microscopic and macroscopic aspects of the tooth surface. It reflects the unevenness of the mesoscopic geometric morphology of the tooth surface and manifests as periodic or quasi-periodic fluctuations. When the tooth surface has a large waviness, periodic changes in meshing stiffness will occur during the meshing process, causing meshing shock and vibration, which in turn leads to a significant increase in transmission noise.

[0003] Existing research has shown that tooth surface waviness can be effectively controlled by adjusting machining parameters. However, tooth surface waviness is affected by many factors, including worm wheel wear, grinding process parameters, worm wheel dressing defects, machine tool vibration, and gear installation errors. Precise control of tooth surface waviness is difficult in actual machining.

[0004] As a key enabling technology for intelligent manufacturing, digital twin technology builds digital models in the virtual world that are highly matched with physical entities, enabling real-time interaction and synchronization of data between physical entities and virtual models, thereby reducing production costs, improving production efficiency, and improving product quality. Summary of the Invention

[0005] The present invention aims to provide a continuous generation grinding tooth surface waviness monitoring system based on digital twins, establish a digital twin model of the gear grinding machine, and monitor the tooth surface waviness condition in real time. When the ghost step amplitude of the tooth surface waviness exceeds the threshold, the grinding process parameters can be optimized online, thereby improving the gear surface quality and increasing production efficiency.

[0006] In order to achieve the above object, the present invention provides the following method:

[0007] The present invention provides a continuous generating grinding tooth surface waviness monitoring system based on digital twins:

[0008] Data acquisition module: real-time acquisition of vibration signals of the grinding wheel spindle and worktable, 3D model data of the CNC gear grinding machine, and CNC system data;

[0009] The vibration signals of the grinding wheel spindle and the workbench are collected by acceleration sensors installed on the grinding wheel spindle and the workbench;

[0010] The three-dimensional model data of the CNC gear grinding machine is read from the constructed three-dimensional model of the gear grinding machine through Unity3D;

[0011] The CNC system data includes the motion parameters of the six motion axes of the CNC gear grinding machine, the grinding wheel speed, and the radial feed and axial feed rates of the grinding wheel. The data is obtained from the machine tool via the OPC UA protocol and uploaded to a MySQL database. Unity3D obtains the data by accessing the MySQL database.

[0012] Simulation module: including motion and grinding rendering units and UI interface;

[0013] The motion and grinding rendering unit realizes real-time simulation display of the digital twin system based on Unity3D motion rendering and Unity3D cutting rendering;

[0014] The UI interface realizes data chart drawing and process parameter information display based on the UGUI system and supports interactive operation with the machine tool;

[0015] Waviness online prediction and process parameter optimization module: including neural network prediction model, process parameter optimization model and process parameter feedback interface;

[0016] The neural network model uses the vibration signal and grinding process parameters as input to predict the maximum ghost order amplitude of the tooth surface waviness;

[0017] The process parameter optimization model takes the grinding wheel speed, feed amount, feed rate and the ghost step amplitude as input and outputs the optimized speed, feed amount and feed rate optimization process parameters;

[0018] The process parameter feedback interface sends the obtained optimized process parameters to the numerical control system to adjust the processing parameters of the numerical control system.

[0019] Preferably, the CNC system data includes the motion parameters of the six motion axes in the CNC gear grinding machine, the grinding wheel speed, the grinding wheel radial feed and the axial feed rate; the CNC acquisition module establishes a connection with the CNC gear grinding machine through the OPC UA protocol and can read the CNC system data in real time and upload the data to the MySQL database. By writing a corresponding script and mounting it on the model on Unity3D, the data in the database can be accessed, and the data of the motion axis can be transmitted to the transform of the motion axis in the model on Unity3D, thereby driving the model motion simulation in real time.

[0020] Preferably, the motion simulation of the CNC gear grinding machine takes the geometric dimensions and motion parameters of the CNC gear grinding machine as input, and controls each motion axis individually through a script so that each motion axis can only perform linear or rotational motion in one direction, thereby constructing a geometric digital model of the CNC gear grinding machine and realizing the motion simulation of the CNC gear grinding machine.

[0021] Preferably, the geometric dimensions of the CNC gear grinding machine include the geometric parameters of the gear grinding machine; the motion parameters of the CNC gear grinding machine include the motion parameters of the six motion axes of the gear grinding machine, the grinding wheel speed, the grinding wheel radial feed amount and the axial feed rate.

[0022] Preferably, the control principle of the script is: separate the components of each axis of the gear grinder and divide them into parent objects and child objects, wherein the child objects follow the parent objects to move, and assign a linear or rotational transform to each motion axis through the script according to the received motion parameters of the gear grinder, thereby realizing the six-axis linkage of the gear grinder.

[0023] Preferably, the motion rendering based on Unity3D realizes the motion simulation of the CNC gear grinding machine by controlling the transform of Unity3D in real time through six-axis motion parameters; the cutting rendering based on Unity3D realizes the grinding simulation of the gear grinding machine by calculating the geometric model of the tool and the workpiece in real time through Boolean operations.

[0024] Preferably, the grinding simulation of the gear grinder takes the motion parameters of the gear grinder, the geometric parameters of the workpiece to be processed and the physical dimension parameters as input, and generates a voxel aggregate of the workpiece to be processed through the three-dimensional graphics rendering and voxel modeling mechanism in the Unity3D engine; by comparing the motion trajectory of the gear grinder with the spatial position of the workpiece voxels in real time, the corresponding voxel units are eliminated when spatial overlap occurs, thereby simulating the material removal process during the grinding process and realizing a visual simulation of the dynamic cutting effect; the ghost order amplitude of the tooth surface waviness predicted by the neural network and sent through the physical dimension is divided into levels according to the size of the ghost order amplitude, and each level corresponds to a different color, and the workpiece surface is rendered through Unity3D; the physical dimension parameters include vibration and ghost order amplitude.

[0025] Preferably, the neural network prediction model consists of an input layer, a hidden layer and an output layer; the input layer includes an average pooling and a convolution block, and inputs a vibration signal and process parameters with a length of 1024; the hidden layer consists of 4 feature extraction layers, each of which includes a multi-scale channel attention block and a multi-head attention block; the output layer consists of an average pooling and a linear transformation, and finally predicts the ghost order amplitude with the highest tooth surface waviness.

[0026] Preferably, the process parameter optimization model uses a multi-objective particle swarm to optimize the grinding process parameters under the constraints of the three optimization objectives of grinding processing time, surface waviness and vibration root mean square, uses the shortest spatial distance to optimize the process parameter combination, and outputs the optimized grinding process parameters.

[0027] Preferably, the process parameter feedback interface sends the grinding wheel speed, grinding wheel radial feed and axial feed rate to the CNC system through the OPC UA protocol, and adjusts the processing parameter information in real time according to the grinding situation.

[0028] The beneficial effects of the present invention are embodied in: the present invention is based on a digital twin continuous expansion grinding tooth surface waviness monitoring system, which uses a data acquisition module to collect the vibration data of the spindle, the three-dimensional model data of the CNC gear grinding machine and the CNC system data in real time, and inputs the collected vibration signal and the machine tool process parameters into the neural network prediction model to predict the ghost step amplitude of the workpiece surface waviness; finally, the predicted ghost step amplitude is input into the parameter optimization model to obtain the recommended grinding wheel speed, feed amount and feed rate, and the optimized process parameters are fed back to the CNC system, thereby realizing online monitoring and regulation of the tooth surface waviness during the processing process, improving the surface quality of the gear, completing and establishing a digital twin model of the CNC gear grinding machine during the processing process, which can accurately control the tooth surface waviness, reduce production costs, improve production efficiency and improve product quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly describes the drawings required for the specific embodiments or the description of the prior art. Similar elements or parts are generally identified by similar reference numerals throughout the drawings. Elements or parts in the drawings are not necessarily drawn to scale.

[0030] Figure 1 Schematic diagram of the principle framework of the digital twin-based continuous generating grinding tooth surface waviness monitoring system provided in an embodiment of the present invention;

[0031] Figure 2 A tooth surface waviness monitoring UI interface built in Unity3D provided in an embodiment of the present invention;

[0032] Figure 3 A flowchart of the tooth surface waviness prediction model training based on CNN-Attention provided in an embodiment of the present invention;

[0033] Figure 4 A flow chart of using a particle swarm optimization algorithm to control machining parameters and thereby obtain stable tooth surface waviness is provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0034] In order to help those skilled in the art better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0035] The terms "first," "second," and so on, in the description and claims of the present invention and the accompanying drawings are used to distinguish between different items, not to describe a specific order. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product, or end comprising a series of steps or elements is not limited to the listed steps or elements but may optionally include steps or elements not listed therein, or may optionally include other steps or elements inherent to such process, method, product, or end.

[0036] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present invention. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute a separate or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0037] Existing research has shown that tooth surface waviness can be effectively controlled by adjusting machining parameters. However, tooth surface waviness is affected by many factors, including worm wheel wear, grinding process parameters, worm wheel dressing defects, machine tool vibration, and gear installation errors. Precise control of tooth surface waviness is difficult in actual machining.

[0038] As a key enabling technology for intelligent manufacturing, digital twin technology builds digital models in the virtual world that are highly matched with physical entities, enabling real-time interaction and synchronization of data between physical entities and virtual models, thereby reducing production costs, improving production efficiency, and improving product quality.

[0039] The present invention aims to provide a continuous generation grinding tooth surface waviness monitoring system based on digital twins, establish a digital twin model of the gear grinding machine, and monitor the tooth surface waviness condition in real time. When the ghost step amplitude of the tooth surface waviness exceeds the threshold, the grinding process parameters can be optimized online, thereby improving the gear surface quality and increasing production efficiency.

[0040] like Figure 1-4 As shown, a specific embodiment of the present invention provides a continuous generating grinding tooth surface waviness monitoring system based on digital twin, comprising:

[0041] Data acquisition module: Real-time collection of vibration signals from the grinding wheel spindle and worktable, 3D model data of the CNC gear grinder, and CNC system data. The vibration signals of the grinding wheel spindle and worktable are collected via acceleration sensors installed on the grinding wheel spindle and worktable. The 3D model data of the CNC gear grinder is read from the constructed 3D model of the gear grinder via Unity3D. The CNC system data, including the motion parameters of the six motion axes in the CNC gear grinder, the grinding wheel speed, and the radial feed and axial feed rates of the grinding wheel, are obtained from the machine tool via the OPC UA protocol and uploaded to the MySQL database. Unity3D obtains the data by accessing the MySQL database.

[0042] In an embodiment of the present invention, the CNC system data includes the motion parameters of the six motion axes in the CNC gear grinding machine, the grinding wheel speed, the grinding wheel radial feed and the axial feed rate; the CNC acquisition module establishes a connection with the CNC gear grinding machine through the OPC UA protocol, and can read the CNC system data in real time and upload the data to the MySQL database. By writing a corresponding script and mounting it on the model on Unity3D, the data in the database can be accessed, and the data of the motion axis can be transmitted to the transform of the motion axis in the model on Unity3D, driving the model motion simulation in real time; the motion simulation of the CNC gear grinding machine uses the geometric dimensions and motion parameters of the CNC gear grinding machine as input, and controls each motion axis individually through the script so that each motion axis can only perform linear or rotational motion in one direction The script is used to build a geometric digital model of the CNC gear grinding machine and realize the motion simulation of the CNC gear grinding machine. The geometric dimensions of the CNC gear grinding machine include the geometric parameters of the gear grinding machine. The motion parameters of the CNC gear grinding machine include the motion parameters of the six motion axes of the gear grinding machine, the grinding wheel speed, the radial feed amount and the axial feed rate of the grinding wheel. The control principle of the script is to separate the components of each axis of the gear grinding machine and divide them into parent objects and child objects, where the child objects move with the parent objects, and the received motion parameters of the gear grinding machine are assigned a linear or rotational transform to each motion axis through the script to realize the six-axis linkage of the gear grinding machine.

[0043] Simulation module: includes motion and grinding rendering units and UI interface; the motion and grinding rendering units implement real-time simulation display of the digital twin system based on Unity3D motion rendering and Unity3D cutting rendering; the UI interface implements data charting and process parameter information display based on the UGUI system and supports interactive operations with the machine tool;

[0044] In an embodiment of the present invention, motion rendering based on Unity3D realizes motion simulation of CNC gear grinding machine by controlling Unity3D transform in real time through six-axis motion parameters; cutting rendering based on Unity3D realizes grinding simulation of gear grinding machine by real-time calculation of geometric models of tool and workpiece through Boolean operation; the grinding simulation of gear grinding machine takes gear grinding machine motion parameters, geometric parameters of workpiece and physical dimension parameters as input, and generates voxel aggregate of workpiece through three-dimensional graphics rendering and voxel modeling mechanism in Unity3D engine; by comparing the motion trajectory of gear grinding machine with the spatial position of workpiece voxel in real time, the corresponding voxel units are eliminated when spatial overlap occurs, thereby simulating the material removal process during grinding and realizing visual simulation of dynamic cutting effect; the ghost order amplitude of tooth surface waviness predicted by neural network sent through physical dimension is divided into levels according to the size of ghost order amplitude, and each level corresponds to a different color, and the workpiece surface is rendered through Unity3D; physical dimension parameters include vibration and ghost order amplitude.

[0045] Waviness online prediction and process parameter optimization module: includes a neural network prediction model, a process parameter optimization model and a process parameter feedback interface; the neural network model uses the vibration signal and grinding process parameters as input to predict the maximum ghost step amplitude of the tooth surface waviness; the process parameter optimization model uses the grinding wheel speed, feed amount, feed rate and ghost step amplitude as input and outputs the optimized speed, feed amount and feed rate to optimize the process parameters; the process parameter feedback interface sends the obtained optimized process parameters to the CNC system to adjust the processing parameters of the CNC system.

[0046] In an embodiment of the present invention, a neural network prediction model consists of an input layer, a hidden layer and an output layer; the input layer includes an average pooling and a convolution block, and inputs a vibration signal and process parameters with a length of 1024; the hidden layer consists of 4 feature extraction layers, each of which includes a multi-scale channel attention block and a multi-head attention block; the output layer consists of an average pooling and a linear transformation, and finally predicts the ghost order amplitude with the highest tooth surface waviness; the process parameter optimization model uses a multi-objective particle swarm to optimize the grinding process parameters under the constraints of the three optimization objectives of grinding processing time, surface waviness and vibration root mean square, adopts the shortest spatial distance to optimize the process parameter combination, and outputs the optimized grinding process parameters; the process parameter feedback interface sends the grinding wheel speed, grinding wheel radial feed and axial feed rate to the CNC system through the OPC UA protocol, and adjusts the processing process parameter information in real time according to the grinding situation.

[0047] like Figure 1The digital twin system shown here consists of four dimensions: geometric, kinematic, data-driven, and physical. The geometric dimension is derived from the three-dimensional model of the CNC gear grinder. The data-driven dimension captures data from each axis in the CNC system, as well as process parameters such as grinding wheel speed, axis feed rate, and radial feed rate, and physical signals collected by sensors during machining. The kinematic dimension drives the gear grinder's motion simulation by acquiring the kinematic parameters of the six axes and grinding process parameters from the data-driven dimension. The physical dimension is the core of the CNC gear grinder's digital twin system. By extracting features from the signals collected in real time during machining and combining them with the grinding process parameters, these features are input into a pre-trained neural network model to predict the tooth surface waviness amplitude. If the tooth surface waviness is acceptable, no process parameter adjustment is required. If the tooth surface waviness is unacceptable, the current grinding process parameters must be adjusted.

[0048] Specifically, the digital twin-based continuous generating grinding tooth surface waviness monitoring system of this embodiment includes a data acquisition module, an online simulation module, and a waviness online prediction and process parameter optimization module.

[0049] The data acquisition module of this embodiment collects the vibration signals of the grinding wheel spindle and worktable, the three-dimensional model data of the CNC gear grinder and the CNC system data in real time; among them, the vibration signals are collected by acceleration sensors installed on the grinding wheel spindle and worktable respectively; the three-dimensional model data of the CNC gear grinder is read from the constructed three-dimensional model of the gear grinder through Unity3D; the CNC system data includes the motion parameters of the six motion axes in the CNC gear grinder, the grinding wheel speed, the radial feed and the axial feed rate of the grinding wheel, which are obtained from the machine tool through the OPC UA protocol and uploaded to the MySQL database, and Unity3D obtains it by accessing the database.

[0050] The online simulation module of this embodiment includes a motion and grinding rendering unit and a UI interface; the motion and grinding rendering unit is based on Unity3D's motion rendering and cutting rendering to realize real-time simulation display of the digital twin system; wherein, the motion rendering based on Unity3D controls the transform of Unity3D in real time through six-axis motion parameters to realize the motion simulation of the CNC gear grinding machine.

[0051] Specifically, the kinematic simulation of a CNC gear grinding machine uses geometric and kinematic parameters as input. Scripts are installed in the models of each axis to independently control the motion of each axis. A Unity3D-based cutting rendering system uses Boolean operations to perform grinding simulation on the gear grinding machine by calculating the geometric models of the tool and workpiece in real time.

[0052] The script control principle is as follows: based on the actual structure of the CNC gear grinder in the real world, the components of each axis in the twin model of the CNC gear grinder are separated and divided into parent objects and child objects, where the child objects follow the movement of the parent objects. The received gear grinder motion parameters are assigned a linear or rotational transform to each axis through the script, driving the model motion simulation to realize the six-axis linkage of the gear grinder.

[0053] The grinding rendering unit takes the motion dimension parameters, geometric dimension parameters and physical dimension parameters in the digital twin system as input, and generates a voxel aggregate of the processed workpiece through the three-dimensional graphics rendering and voxel modeling mechanism in the Unity3D engine; by comparing the motion trajectory of the gear grinder with the spatial position of the workpiece voxel in real time, the corresponding voxel units are eliminated when spatial overlap occurs, thereby simulating the material removal process during the grinding process and realizing a visual simulation of the dynamic cutting effect; the ghost order and ghost order amplitude of the tooth surface waviness predicted by the neural network sent through the physical dimension are divided into levels according to the amplitude size, and each level corresponds to a different color, and the workpiece surface is rendered through Unity3D; the physical dimension parameters include vibration and ghost order amplitude.

[0054] like Figure 2 As shown in the figure, the UI interface realizes data chart drawing and process parameter and other information display based on the UGUI system and supports interactive operation with the machine tool; the collected vibration signal is feature extracted and the extracted features are displayed in the form of a curve graph, the basic information of the gear grinding machine and the real-time prediction value of the tooth surface waviness during the processing process are displayed in the form of text, and whether the current process parameters need to be optimized is determined based on whether the waviness value exceeds the specified threshold. If the threshold is exceeded, the optimized process parameters are obtained and displayed in the UI interface. The process parameters will be transmitted to the CNC system through the OPC UA protocol to control the grinding process parameters.

[0055] In this embodiment, the motion simulation interface and the parameter optimization interface can be jumped through the buttons at the bottom of the interface, and the historical processing data can be viewed by clicking the historical data button.

[0056] The waviness online prediction and process parameter optimization module of this embodiment includes a CNN-Attention neural network prediction model, a particle swarm process parameter optimization model and a process parameter feedback interface; the CNN-Attention neural network model uses the vibration signal and the grinding process parameters as input to predict the maximum ghost step amplitude of the tooth surface waviness; the particle swarm process parameter optimization model uses the grinding wheel speed, feed rate, feed rate and ghost step amplitude as input and outputs the optimized speed, feed rate and feed rate; the process parameter feedback interface sends the obtained optimized process parameters to the CNC system to adjust the processing parameters of the CNC system.

[0057] The CNN-Attention neural network prediction model consists of an input layer, a hidden layer, and an output layer. The input layer includes an average pooling and a convolution block, and the input length is 1024. The vibration signal and process parameters are input. The hidden layer consists of four feature extraction layers, each of which includes a multi-scale channel attention block and a multi-head attention block. The multi-scale channel attention block consists of a multi-branch depthwise separable convolution and a channel attention layer. Through the dynamic multi-branch depthwise separable convolution and lightweight channel attention mechanism, efficient time series feature extraction is achieved. The multi-head attention mechanism is applied to capture signal features in the global range of the signal. The output layer consists of an average pooling and a linear transformation, and finally the predicted ghost order amplitude is obtained.

[0058] Figure 3 This is a flowchart for training a CNN-Attention neural network prediction model. First, the vibration data from the grinding experiment is normalized and divided into training, validation, and test sets. The model is trained and evaluated to obtain the final model, which is then embedded in the digital twin system.

[0059] Figure 4 To achieve a stable tooth surface waviness flow chart by using a particle swarm optimization algorithm to control machining parameters, a particle swarm process parameter optimization model uses a multi-objective particle swarm to optimize grinding process parameters under the constraints of three optimization objectives: grinding time, surface waviness, and root mean square vibration. The model then uses the shortest spatial distance to find the optimal process parameter combination and outputs the optimized grinding process parameters. After obtaining the optimized process parameters, the process parameter feedback interface transmits the grinding wheel speed, feed rate, and feed rate to the CNC system via the OPC UA protocol, allowing the process parameters to be adjusted in real time based on the grinding performance.

[0060] The beneficial effects of the present invention are embodied in: the present invention is based on a digital twin continuous expansion grinding tooth surface waviness monitoring system, which uses a data acquisition module to collect the vibration data of the spindle, the three-dimensional model data of the CNC gear grinding machine and the CNC system data in real time, and inputs the collected vibration signal and the machine tool process parameters into the neural network prediction model to predict the ghost step amplitude of the workpiece surface waviness; finally, the predicted ghost step amplitude is input into the parameter optimization model to obtain the recommended grinding wheel speed, feed amount and feed rate, and the optimized process parameters are fed back to the CNC system, thereby realizing online monitoring and regulation of the tooth surface waviness during the processing process, improving the surface quality of the gear, completing and establishing a digital twin model of the CNC gear grinding machine during the processing process, which can accurately control the tooth surface waviness, reduce production costs, improve production efficiency and improve product quality.

[0061] The above description is merely an embodiment of the present invention. Common knowledge such as the specific technical solutions or features of the solutions is not described in detail here. It should be noted that those skilled in the art may make several modifications and improvements without departing from the solution of the present invention, and these modifications and improvements should also be considered as the scope of protection of the present invention. These modifications and improvements will not affect the effects of the present invention and the practicality of the patent. The scope of protection claimed in this application shall be based on the content of the claims, and the specific embodiments and other descriptions in the specification may be used to interpret the content of the claims.

Claims

1. A continuous generating grinding tooth surface waviness monitoring system based on digital twin, characterized by: The system comprises: Data acquisition module: real-time acquisition of vibration signals of the grinding wheel spindle and worktable, 3D model data of the CNC gear grinding machine, and CNC system data; The vibration signals of the grinding wheel spindle and the workbench are collected by acceleration sensors installed on the grinding wheel spindle and the workbench; The three-dimensional model data of the CNC gear grinding machine is read from the constructed three-dimensional model of the gear grinding machine through Unity3D; The CNC system data includes the motion parameters of the six motion axes of the CNC gear grinding machine, the grinding wheel speed, and the radial feed and axial feed rates of the grinding wheel. The data is obtained from the machine tool via the OPC UA protocol and uploaded to a MySQL database. Unity3D obtains the data by accessing the MySQL database. Simulation module: including motion and grinding rendering units and UI interface; The motion and grinding rendering unit realizes real-time simulation display of the digital twin system based on Unity3D motion rendering and Unity3D cutting rendering; The UI interface realizes data chart drawing and process parameter information display based on the UGUI system and supports interactive operation with the machine tool; Waviness online prediction and process parameter optimization module: including neural network prediction model, process parameter optimization model and process parameter feedback interface; The neural network model uses the vibration signal and grinding process parameters as input to predict the maximum ghost order amplitude of the tooth surface waviness; The process parameter optimization model takes the grinding wheel speed, feed amount, feed rate and the ghost step amplitude as input and outputs the optimized speed, feed amount and feed rate optimization process parameters; The process parameter feedback interface sends the obtained optimized process parameters to the numerical control system to adjust the processing parameters of the numerical control system.

2. The digital twin-based continuous generating grinding tooth surface waviness monitoring system according to claim 1, characterized in that: The numerical control system data includes the motion parameters of the six motion axes in the numerical control gear grinding machine, the grinding wheel speed, the grinding wheel radial feed and the axial feed rate; The CNC acquisition module establishes a connection with the CNC gear grinding machine through the OPC UA protocol, can read the CNC system data in real time and upload the data to the MySQL database. By writing the corresponding script and mounting it on the model on Unity3D, the data in the database can be accessed, and the data of the motion axis can be transmitted to the transform of the motion axis in the model on Unity3D, driving the model motion simulation in real time.

3. The digital twin-based continuous generating grinding tooth surface waviness monitoring system according to claim 1, characterized in that: The motion simulation of the CNC gear grinding machine uses the geometric dimensions and motion parameters of the CNC gear grinding machine as input, and controls each motion axis individually through a script so that each motion axis can only perform linear or rotational motion in one direction. A geometric digital model of the CNC gear grinding machine is constructed to realize the motion simulation of the CNC gear grinding machine.

4. The digital twin-based continuous generating grinding tooth surface waviness monitoring system according to claim 3, characterized in that: The geometric dimensions of the CNC gear grinding machine include geometric parameters of the gear grinding machine; The motion parameters of the CNC gear grinding machine include motion parameters of the six motion axes of the gear grinding machine, a grinding wheel rotation speed, a grinding wheel radial feed amount, and an axial feed rate.

5. The digital twin-based continuous generating grinding tooth surface waviness monitoring system according to claim 4, characterized in that: The control principle of the script is: separate the components of each axis of the gear grinder and divide them into parent objects and child objects, where the child objects follow the movement of the parent objects, and assign a linear or rotational transform to each motion axis through the script based on the received motion parameters of the gear grinder, thereby realizing the six-axis linkage of the gear grinder.

6. The digital twin-based continuous generating grinding tooth surface waviness monitoring system according to claim 1, characterized in that: The Unity3D-based motion rendering realizes motion simulation of the CNC gear grinding machine by controlling the transform of Unity3D in real time through six-axis motion parameters; The cutting rendering based on Unity3D realizes the grinding simulation of the gear grinding machine by calculating the geometric models of the tool and the workpiece in real time through Boolean operations.

7. The digital twin-based continuous generating grinding tooth surface waviness monitoring system according to claim 6, characterized in that: The grinding simulation of the gear grinding machine uses the gear grinding machine motion parameters, the geometric parameters and physical dimension parameters of the workpiece as input, and generates a voxel aggregate of the workpiece through the 3D graphics rendering and voxel modeling mechanism in the Unity3D engine; By comparing the motion trajectory of the gear grinding machine with the spatial position of the workpiece voxels in real time, the corresponding voxel units are eliminated when spatial overlap occurs, thereby simulating the material removal process during the grinding process and achieving a visual simulation of the dynamic cutting effect; The ghost order amplitude of the tooth surface waviness predicted by the neural network and sent through the physical dimension is divided into levels according to the size of the ghost order amplitude. Each level corresponds to a different color, and the workpiece surface is rendered through Unity3D; the physical dimension parameters include vibration and ghost order amplitude.

8. The digital twin-based continuous generating grinding tooth surface waviness monitoring system according to claim 1, characterized in that: The neural network prediction model consists of an input layer, a hidden layer and an output layer; The input layer includes an average pooling block and a convolution block, and inputs a vibration signal and process parameters with a length of 1024; The hidden layer consists of four feature extraction layers, each of which includes a multi-scale channel attention block and a multi-head attention block; The output layer is composed of an average pooling and a linear transformation, and finally predicts the ghost order amplitude with the highest tooth surface waviness.

9. The digital twin-based continuous generating grinding tooth surface waviness monitoring system according to claim 1, characterized in that: The process parameter optimization model uses a multi-objective particle swarm to optimize the grinding process parameters under the constraints of the three optimization objectives of grinding processing time, surface waviness and vibration root mean square, uses the shortest spatial distance to find the optimal process parameter combination, and outputs the optimized grinding process parameters.

10. The digital twin-based continuous generating grinding tooth surface waviness monitoring system according to claim 1, characterized in that: The process parameter feedback interface sends the grinding wheel speed, grinding wheel radial feed and axial feed rate to the CNC system through the OPC UA protocol, and adjusts the processing parameter information in real time according to the grinding situation.

Images