Digital twin modeling method based on large model

Abstract

The invention relates to the technical field of digital twinning, in particular to a digital twinning modeling method based on a large model. The method comprises the following steps: acquiring three-dimensional point cloud data to form a plurality of initial segmentation areas; correcting the edge of the initial segmentation region to obtain a to-be-modeled region; performing modeling and model matching on each entity target in the to-be-modeled area according to the entity feature set to generate a corresponding three-dimensional model; and carrying out real-time monitoring and early warning on the driving state of the moving trolley based on the large model, verifying the early warning condition according to the simulated driving speed of the moving trolley, and adjusting the size of the initial segmentation region based on the verification result. According to the method, the precision of the model is improved by correcting the edge, it is ensured that the boundary of the to-be-modeled area better conforms to the actual situation, the actual object is converted into the digital model, the model is imported into the virtual environment for deep analysis, the modeling parameters are adjusted according to the analysis result, and the real-time performance and accuracy of modeling are enhanced.

Description

Technical Field

[0001] The present invention relates to the field of digital twin technology, and in particular to a digital twin modeling method based on a large model. Background Art

[0002] Digital twin refers to a virtual model of a physical entity. Through real-time data and analysis technology, it can reflect the state, behavior and performance of the physical entity, and use large-scale machine learning models to achieve modeling and simulation of complex systems. In the existing technology, feature line fitting and regional growing algorithms can be used to achieve rapid modeling point cloud segmentation in the production workshop to solve the problems of insufficient segmentation and over-segmentation. The segmentation integrity is good, but the modeling accuracy and efficiency for complex environments are not ideal. Therefore, there is an urgent need for a rapid modeling point cloud segmentation method that optimizes segmentation parameters through real-time data and virtual data to improve modeling accuracy.

[0003] A Chinese patent document with publication number CN117634210A discloses a digital twin modeling method, including a data collection module, a data storage module, a simulation calculation module, a model building module, a data comparison module and a data modification module. The data collection module is used to collect various data information of the digital twin system, the data storage module is used to store the collected data information, and the simulation calculation module is used to calculate the collected data and cooperate with the model building module to complete the modeling of the virtual project. It can be seen that the existing digital twin modeling method lacks real-time monitoring and early warning of the status of dynamic targets based on a large model, verifies the accuracy of the early warning situation, and adjusts the size of the initial segmentation area according to the verification results, resulting in low model accuracy. Summary of the invention

[0004] To this end, the present invention provides a digital twin modeling method based on a large model to overcome the problem in the prior art of low model accuracy due to the lack of monitoring and verification of the driving state of the mobile vehicle and optimization of the initial segmentation area.

[0005] To achieve the above object, the present invention provides a digital twin modeling method based on a large model, comprising:

[0006] Use the calibrated laser radar to conduct a comprehensive scan of the production workshop, pre-process the scan data to obtain 3D point cloud data, and perform the initial segmentation of the production workshop according to the density distribution of the 3D point cloud data to form several initial segmentation areas;

[0007] Extracting edge contours of any of the initial segmented regions to calculate a real-time contour error, and correcting the edge of the initial segmented region according to the real-time contour error to obtain a region to be modeled;

[0008] Modeling and model matching are performed on each entity target in the area to be modeled according to the entity feature set, and a corresponding three-dimensional model is generated and stored in a model database, wherein the three-dimensional model includes a static three-dimensional model and a dynamic three-dimensional model;

[0009] The three-dimensional model is imported into a simulated production workshop, and the driving status of the mobile vehicle is monitored and warned in real time based on the large model. The warning situation is verified according to the simulated driving speed of the mobile vehicle, and the size of the initial segmentation area is adjusted based on the verification result.

[0010] Furthermore, the initial segmentation of the production workshop is performed according to the density distribution of the three-dimensional point cloud data, including:

[0011] Dividing the three-dimensional point cloud data into a plurality of squares, calculating the density of feature points corresponding to each square, so as to analyze the density distribution of each square;

[0012] Carry out initial segmentation of the production workshop based on the density analysis results;

[0013] The points in each square are counted to calculate the density of feature points corresponding to each square. The density of feature points is the ratio of the number of points in the square to the area of ​​the square.

[0014] Further, modeling and model matching of each entity target in the area to be modeled according to the entity feature set includes:

[0015] Extracting feature information in the area to be modeled as an entity feature set;

[0016] Matching the entity feature set in a feature database, and dividing each entity target into a static target and a dynamic target according to the matching result;

[0017] For static targets, the geometry of the model is defined according to the corresponding entity feature set, and material and texture information are applied to generate a static 3D model;

[0018] For dynamic targets, real-time coordinates and timestamps are obtained to calculate the moving speed of the dynamic targets in the production workshop and to generate dynamic 3D models.

[0019] Further, correcting the edge of the initial segmented area according to the real-time contour error includes:

[0020] Comparing the actual contour error with the standard contour error, and performing smoothing on the initial segmented area based on the comparison result;

[0021] When it is determined that the actual contour error is greater than the standard contour error, the edge is smoothed using a B-spline curve; the real-time contour error is the sum of the squares of the differences between each actual edge point and the corresponding extracted edge point.

[0022] Furthermore, real-time monitoring of the driving status of the mobile car based on the large model includes:

[0023] The simulated driving state of the mobile car is analyzed, and when it is determined that there is an abnormal risk in the driving state, an early warning prompt is issued, and the simulated driving speed of the mobile car is compared with the real-time driving speed to verify the real-time accuracy of the early warning prompt.

[0024] Furthermore, the simulated driving state of the mobile car is analyzed including:

[0025] Obtaining a real-time safety distance between the mobile vehicle and the marked target, and comparing the real-time safety distance with a standard safety distance;

[0026] Based on the comparison result, it is determined that there is no abnormal risk in the driving state, or whether there is an abnormal risk in the driving state is analyzed in combination with the deviation trend of the mobile car.

[0027] Further, judging that there is no abnormal risk in the driving state based on the comparison result includes:

[0028] When the real-time safety distance is less than or equal to the standard safety distance, determine whether the driving state has abnormal risks based on the deviation trend;

[0029] When the real-time safety distance is greater than the standard safety distance, it is determined that there is no abnormal risk in the driving state.

[0030] Further, determining whether the driving state has an abnormal risk according to the deviation trend includes:

[0031] Acquire the simulated offset direction of the mobile vehicle, and determine the offset trend of the mobile vehicle according to the simulated offset direction.

[0032] If the simulated deviation direction does not begin to deviate, it is determined that there is an abnormal risk in the driving state, an early warning prompt is issued, and the simulated driving speed of the mobile car is compared with the real-time driving speed.

[0033] Furthermore, the real-time accuracy of the early warning indication is verified, including:

[0034] Obtain the real-time driving speed of the mobile vehicle in the simulated production workshop, and compare the real-time driving speed with the standard driving speed.

[0035] If the real-time accuracy is lower than the standard accuracy, the verification result is judged to be incorrect;

[0036] If the real-time accuracy is greater than or equal to the standard accuracy, the verification result is determined to be correct.

[0037] Further, adjusting the size of the initial segmented area based on the verification result includes:

[0038] When the verification result is determined to be incorrect, the size of the initial segmented area is adjusted.

[0039] Compared with the prior art, the beneficial effects of the present invention are that data is collected by using a calibrated and adjusted lidar device to ensure the accuracy and reliability of its measurement, preprocessing includes denoising, filtering and point cloud simplification to ensure data quality and improve the efficiency of subsequent processing, the clustering algorithm K-means is used to group the point cloud, identify different areas, and correct the edges to improve the accuracy of the model to ensure that the boundary of the area to be modeled is more in line with the actual situation, by converting the actual object into a digital model and importing the model into a virtual environment for in-depth analysis, by combining real-time data with the model, the three-dimensional model can be dynamically updated to ensure that it always reflects the latest status, enhance the real-time and accuracy of the digital twin, and by real-time monitoring of the driving conditions of the mobile car, the real-time data displayed in the virtual workshop is compared with the dynamic time-varying data in the actual workshop to analyze the error. If the error is large, it means that the model accuracy is low, then the size of the segmented area is adjusted to improve the accuracy of the model, and dynamic monitoring is used to ensure production safety, timely warning and improve the accuracy of model warning, so as to effectively prevent potential safety hazards.

[0040] Furthermore, through the communication between the database and the physical engine, the coordinate information, speed information and environmental information of the dynamic target are mapped into the virtual space, so that the dynamic changes in the workshop are presented in the virtual environment. The virtual space is a simulated production workshop. In a dynamic environment, the three-dimensional model is updated according to real-time data. By storing the three-dimensional model in the model database, dynamic updates are conveniently performed to ensure that the model always reflects the latest status. By combining the real-time data collected by the sensor with the model, the status or position of the model is updated to enhance the real-time and accuracy of the digital twin. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A schematic diagram of a flow chart of a digital twin modeling method based on a large model according to an embodiment of the present invention;

[0042] Figure 2 This is a schematic diagram of the process of initially dividing a production workshop according to an embodiment of the present invention;

[0043] Figure 3 A schematic diagram of a process of correcting the edge of the initial segmented area according to an embodiment of the present invention;

[0044] Figure 4 The figure is a flow chart of modeling and model matching for each entity target in the area to be modeled according to an embodiment of the present invention. DETAILED DESCRIPTION

[0045] In order to make the objects and advantages of the present invention more clearly understood, the present invention is further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0046] The preferred embodiments of the present invention are described below with reference to the accompanying drawings. It should be understood by those skilled in the art that these embodiments are only used to explain the technical principles of the present invention and are not intended to limit the protection scope of the present invention.

[0047] It should be noted that, in the description of the present invention, terms such as "up", "down", "left", "right", "inside" and "outside" indicating directions or positional relationships are based on the directions or positional relationships shown in the drawings. This is merely for the convenience of description and does not indicate or imply that the device or element must have a specific orientation, be constructed and operated in a specific orientation. Therefore, it cannot be understood as a limitation on the present invention.

[0048] In addition, it should be noted that in the description of the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" 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 it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For those skilled in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0049] See also Figure 1 As shown, it is a flow chart of a digital twin modeling method based on a large model according to an embodiment of the present invention. The present invention provides a digital twin modeling method based on a large model, comprising:

[0050] Step S1, using a calibrated laser radar to comprehensively scan the production workshop, pre-processing the scanned data to obtain three-dimensional point cloud data, and performing initial segmentation of the production workshop according to the density distribution of the three-dimensional point cloud data to form a number of initial segmentation areas;

[0051] Step S2, extracting edge contours of any of the initial segmented regions to calculate a real-time contour error, and correcting the edge of the initial segmented region according to the real-time contour error to obtain a region to be modeled;

[0052] Step S3, modeling and model matching are performed on each entity target in the area to be modeled according to the entity feature set, generating a corresponding three-dimensional model, and storing it in a model database, wherein the three-dimensional model includes a static three-dimensional model and a dynamic three-dimensional model;

[0053] Step S4, importing the three-dimensional model into the simulated production workshop, performing real-time monitoring and early warning on the driving status of the mobile vehicle based on the large model, verifying the early warning situation according to the simulated driving speed of the mobile vehicle, and adjusting the size of the initial segmentation area based on the verification result.

[0054] In this embodiment, the calibrated laser radar refers to a laser radar device that has been calibrated and adjusted to ensure the accuracy and reliability of its measurement. The preprocessing includes denoising, filtering and point cloud simplification to ensure data quality and improve the efficiency of subsequent processing. The clustering algorithm K-means is used to group the point cloud and identify different areas. The edge is corrected to improve the accuracy of the model and ensure that the boundary of the area to be modeled is more in line with the actual situation. By converting the actual object into a digital model and importing the model into a virtual environment for in-depth analysis, the three-dimensional model can be dynamically updated by combining real-time data with the model to ensure that it always reflects the latest status and enhance the real-time and accuracy of the digital twin. By real-time monitoring of the driving conditions of the mobile car, the real-time data displayed in the virtual workshop is compared with the dynamic time-varying data in the actual workshop to analyze the error. If the error is large, it means that the model accuracy is low, then the size of the segmented area is adjusted to improve the accuracy of the model. Through dynamic monitoring, production safety is ensured, timely warnings are issued, and the accuracy of model warnings is improved to effectively prevent potential safety hazards.

[0055] See also Figure 2 As shown, it is a schematic diagram of the process of initially dividing a production workshop according to an embodiment of the present invention;

[0056] Specifically, the initial segmentation of the production workshop based on the density distribution of the 3D point cloud data includes:

[0057] Step S101, dividing the three-dimensional point cloud data into a plurality of grids;

[0058] Step S102, calculating the density of feature points corresponding to each square to analyze the density distribution of each square;

[0059] Step S103, performing initial segmentation of the production workshop according to the density analysis result;

[0060] The points in each square are counted to calculate the density of feature points corresponding to each square. The density of feature points is the ratio of the number of points in the square to the area of ​​the square.

[0061] In this embodiment, the K-means clustering algorithm is used to analyze whether the density of feature points of the squares is similar, so that the squares with similar feature point density are classified into the same category to form multiple initial segmentation areas. The initial square size is set to 2×2×2 to achieve preliminary segmentation of the point cloud. According to the distribution of the data, the number of clusters k is selected, and k is 3. First, the feature point density is processed using Z-score standardization, and the K-means clustering algorithm is executed to cluster the standardized feature point density, so that similar squares are classified into the same category to form multiple initial segmentation areas. The initial segmentation areas represent different functional areas or equipment areas in the production workshop; the process of edge contour extraction for any of the initial segmentation areas is to use the OpenCV image processing library, apply the Canny edge detection algorithm to each initial segmentation area, and obtain the edge contour; collect actual edge data through a sensor to obtain real edge information corresponding to the extracted edge contour, use the least squares method to calculate the error between the extracted contour and the actual edge, as the real-time contour error, calculate the sum of the squares of the difference between each actual edge point and the corresponding extracted edge point, and obtain the corresponding real-time contour error.

[0062] See also Figure 3 As shown, it is a schematic diagram of the process of correcting the edge of the initial segmented area according to an embodiment of the present invention;

[0063] Specifically, correcting the edge of the initial segmented area according to the real-time contour error includes:

[0064] Step S201, comparing the actual contour error with the standard contour error;

[0065] Step S202, performing smoothing processing on the initial segmented area based on the comparison result;

[0066] When it is determined that the actual contour error is greater than the standard contour error, the edge is smoothed using a B-spline curve; the real-time contour error is the sum of the squares of the differences between each actual edge point and the corresponding extracted edge point.

[0067] See also Figure 4 As shown, it is a schematic diagram of the process of modeling and model matching of each entity target in the area to be modeled according to an embodiment of the present invention;

[0068] Specifically, modeling and model matching of each entity target in the area to be modeled according to the entity feature set includes:

[0069] Step S301, extracting feature information in the area to be modeled as an entity feature set;

[0070] Step S302, matching the entity feature set in a feature database, and classifying each entity target into a static target and a dynamic target according to the matching result;

[0071] Step S303, for a static target, define the geometric shape of the model according to the corresponding entity feature set, apply material and texture information, and generate a static three-dimensional model;

[0072] For dynamic targets, real-time coordinates and timestamps are obtained to calculate the moving speed of the dynamic targets in the production workshop and to generate dynamic 3D models.

[0073] In this embodiment, static targets include buildings and processing equipment, and dynamic targets include cars and personnel. The identified dynamic targets will be matched with the preset dynamic three-dimensional model library. After successful matching, the model will be instantiated, and the object will be imported into the virtual space to form a virtual model corresponding to the real object. The ultra-wideband (UWB) positioning technology is used to monitor the movement and trajectory of dynamic targets in the production workshop in real time, and the real-time coordinates and timestamps of the dynamic targets are recorded, and their moving speed in the workshop is calculated. The light intensity, temperature and humidity values, and sound decibel values ​​in the workshop are detected by light sensors, temperature and humidity sensors, and sound sensors. The identified dynamic targets The type, perceived movement speed, and environmental monitoring data (light, temperature, humidity, and sound) will be transmitted to the database. Through the communication between the database and the physical engine, the coordinate information, speed information, and environmental information of the dynamic target will be mapped to the virtual space, thereby presenting the dynamic changes in the workshop in the virtual environment. The virtual space is a simulated production workshop. In a dynamic environment, the three-dimensional model is updated according to real-time data. By storing the three-dimensional model in the model database, it is convenient to perform dynamic updates to ensure that the model always reflects the latest status. By combining the real-time data collected by the sensor with the model, the status or position of the model is updated to enhance the real-time and accuracy of the digital twin.

[0074] Specifically, real-time monitoring of the driving status of a mobile vehicle based on a large model includes:

[0075] The simulated driving state of the mobile car is analyzed, and when it is determined that there is an abnormal risk in the driving state, an early warning prompt is issued, and the simulated driving speed of the mobile car is compared with the real-time driving speed to verify the real-time accuracy of the early warning prompt.

[0076] Specifically, the analysis of the simulated driving state of the mobile car includes:

[0077] Obtaining a real-time safety distance between the mobile vehicle and the marked target, and comparing the real-time safety distance with a standard safety distance;

[0078] Based on the comparison result, it is determined that there is no abnormal risk in the driving state, or whether there is an abnormal risk in the driving state is analyzed in combination with the deviation trend of the mobile car.

[0079] The marked target in this embodiment represents equipment with certain danger or special characteristics in the production workshop and cannot collide with or approach the equipment. The standard safety distance represents the safety distance threshold set from the marked target. The set value is larger than the danger distance and is generally set to be greater than 10% of the danger distance. When it is determined that the real-time safety distance is greater than the standard safety distance, the current driving state is normal. When it is determined that the real-time safety distance is less than or equal to the standard safety distance, there may be a trend of reaching the danger range. The turning trend of the mobile vehicle is analyzed, that is, according to the comparison result of the standard offset angle and the real-time offset angle, the simulated offset direction is analyzed to determine whether there is a risk in the driving state.

[0080] Specifically, judging that there is no abnormal risk in the driving state based on the comparison result includes:

[0081] When the real-time safety distance is less than or equal to the standard safety distance, determine whether the driving state has abnormal risks based on the deviation trend;

[0082] When the real-time safety distance is greater than the standard safety distance, it is determined that there is no abnormal risk in the driving state.

[0083] Specifically, determining whether there is an abnormal risk in the driving state based on the deviation trend includes:

[0084] Acquire the simulated offset direction of the mobile vehicle, and determine the offset trend of the mobile vehicle according to the simulated offset direction.

[0085] If the simulated deviation direction begins to deviate, it is determined that there is no abnormal risk in the driving state;

[0086] If the simulated deviation direction does not begin to deviate, it is determined that there is an abnormal risk in the driving state, an early warning prompt is issued, and the simulated driving speed of the mobile car is compared with the real-time driving speed.

[0087] In this embodiment, the standard offset angle of the mobile car is calculated, the standard offset angle is compared with the actual offset angle, and whether the simulated offset direction starts to offset is determined according to the comparison result. Two tangent lines are drawn from the closest point of the mobile car to the marked target to the edge contour of the current marked target, and the straight line where the real-time moving direction of the mobile car is located is used as the axis. The angle between any tangent line and the axis is obtained as the standard offset angle between the mobile target and the current marked target, and the angle between the real-time moving direction of the mobile car and the axis is obtained as the real-time offset angle. The standard offset angle is compared with the real-time offset angle.

[0088] If the real-time offset angle is smaller than the standard offset angle, it is determined that the simulated offset direction has not started to offset;

[0089] If the real-time offset angle is greater than or equal to the standard offset angle, it is determined that the simulated offset direction starts to offset.

[0090] Specifically, the verification of the real-time accuracy of the warning prompt indication includes:

[0091] Obtain the real-time driving speed of the mobile vehicle in the simulated production workshop, and compare the real-time driving speed with the standard driving speed.

[0092] If the real-time accuracy is lower than the standard accuracy, the verification result is judged to be incorrect;

[0093] If the real-time accuracy is greater than or equal to the standard accuracy, the verification result is determined to be correct;

[0094] Among them, real-time accuracy = 1-I (real-time driving speed-standard driving speed) / real-time driving speed I.

[0095] The real-time accuracy in this embodiment indicates the relative closeness between the real-time driving speed and the standard driving speed, and the value range is 0-1. The closer the real-time accuracy is to 1, the closer the real-time speed is to the standard speed.

[0096] Specifically, adjusting the size of the initial segmented area based on the verification result includes:

[0097] When the verification result is determined to be incorrect, adjusting the size of the initial segmented area;

[0098] Among them, the current segmentation area is adjusted to the modified segmentation area to adjust the size of the initial segmentation area; the modified segmentation area = the current segmentation area × [1-j × (standard accuracy - real-time accuracy) / standard accuracy], j is the adjustment coefficient, j = 0.2.

[0099] The accuracy and reliability of the model are improved by dynamically adjusting the size of the initial segmentation area according to the errors monitored in real time.

[0100] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

[0101] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A digital twin modeling method based on a large model, characterized in that: include, Use the calibrated laser radar to conduct a comprehensive scan of the production workshop, pre-process the scan data to obtain 3D point cloud data, and perform the initial segmentation of the production workshop according to the density distribution of the 3D point cloud data to form several initial segmentation areas; Extracting edge contours of any of the initial segmented regions to calculate a real-time contour error, and correcting the edge of the initial segmented region according to the real-time contour error to obtain a region to be modeled; Modeling and model matching are performed on each entity target in the area to be modeled according to the entity feature set, and a corresponding three-dimensional model is generated and stored in a model database, wherein the three-dimensional model includes a static three-dimensional model and a dynamic three-dimensional model; The three-dimensional model is imported into a simulated production workshop, and the driving status of the mobile vehicle is monitored and warned in real time based on the large model. The warning situation is verified according to the simulated driving speed of the mobile vehicle, and the size of the initial segmentation area is adjusted based on the verification result.

2. The digital twin modeling method based on a large model according to claim 1 is characterized in that: The initial segmentation of the production workshop based on the density distribution of the 3D point cloud data includes: Dividing the three-dimensional point cloud data into a plurality of squares, calculating the density of feature points corresponding to each square, so as to analyze the density distribution of each square; Carry out initial segmentation of the production workshop based on the density analysis results; The points in each square are counted to calculate the density of feature points corresponding to each square. The density of feature points is the ratio of the number of points in the square to the area of ​​the square.

3. The digital twin modeling method based on a large model according to claim 1 is characterized in that: Modeling and model matching of each entity target in the area to be modeled according to the entity feature set includes: Extracting feature information in the area to be modeled as an entity feature set; Matching the entity feature set in a feature database, and dividing each entity target into a static target and a dynamic target according to the matching result; For static targets, the geometry of the model is defined according to the corresponding entity feature set, and material and texture information are applied to generate a static 3D model; For dynamic targets, real-time coordinates and timestamps are obtained to calculate the moving speed of the dynamic targets in the production workshop and to generate dynamic 3D models.

4. The digital twin modeling method based on a large model according to claim 1, characterized in that: Correcting the edge of the initial segmented area according to the real-time contour error includes, Comparing the actual contour error with the standard contour error, and performing smoothing on the initial segmented area based on the comparison result; When it is determined that the actual contour error is greater than the standard contour error, the edge is smoothed using a B-spline curve; the real-time contour error is the sum of the squares of the differences between each actual edge point and the corresponding extracted edge point.

5. The digital twin modeling method based on a large model according to claim 1 is characterized in that: Real-time monitoring of the driving status of the mobile car based on the large model includes: The simulated driving state of the mobile car is analyzed, and when it is determined that there is an abnormal risk in the driving state, an early warning prompt is issued, and the simulated driving speed of the mobile car is compared with the real-time driving speed to verify the real-time accuracy of the early warning prompt.

6. The digital twin modeling method based on a large model according to claim 5 is characterized in that: Analysis of the simulated driving state of the mobile car includes: Obtaining a real-time safety distance between the mobile vehicle and the marked target, and comparing the real-time safety distance with a standard safety distance; Based on the comparison result, it is determined that there is no abnormal risk in the driving state, or whether there is an abnormal risk in the driving state is analyzed in combination with the deviation trend of the mobile car.

7. The large-model-based digital twin modeling method according to claim 6, characterized in that: The comparison results show that there is no abnormal risk in the driving state, including: When the real-time safety distance is less than or equal to the standard safety distance, determine whether the driving state has abnormal risks based on the deviation trend; When the real-time safety distance is greater than the standard safety distance, it is determined that there is no abnormal risk in the driving state.

8. The large-model-based digital twin modeling method according to claim 7, characterized in that: Determining whether there is an abnormal risk in the driving state based on the deviation trend includes: Acquire the simulated offset direction of the mobile vehicle, and determine the offset trend of the mobile vehicle according to the simulated offset direction. If the simulated deviation direction does not begin to deviate, it is determined that there is an abnormal risk in the driving state, an early warning prompt is issued, and the simulated driving speed of the mobile car is compared with the real-time driving speed.

9. The large-model-based digital twin modeling method according to claim 5, characterized in that: Verification of the real-time accuracy of the warning prompt indication includes: Obtain the real-time driving speed of the mobile vehicle in the simulated production workshop, and compare the real-time driving speed with the standard driving speed. If the real-time accuracy is lower than the standard accuracy, the verification result is judged to be incorrect; If the real-time accuracy is greater than or equal to the standard accuracy, the verification result is determined to be correct.

10. The digital twin modeling method based on a large model according to claim 9, characterized in that: Adjusting the size of the initial segmented area based on the verification result includes: When the verification result is determined to be incorrect, the size of the initial segmented area is adjusted.

Images