Intelligent plant digital twin modeling method

Abstract

The invention relates to the technical field of digital twinborn modeling, and discloses a smart plant digital twinborn modeling method, which comprises the following steps of: acquiring static point cloud data of machine tool equipment in different operation states and preprocessing the static point cloud data; performing three-dimensional reconstruction on the preprocessed static point cloud data of the machine tool equipment; operating data of the machine tool equipment in different operating states are collected, and a machine tool equipment operating state recognition model is constructed; and real-time operation data of the machine tool equipment are identified, and analog simulation of a three-dimensional modeling result of the machine tool equipment is carried out. According to the method, the plane equipment components and the cylinder equipment components in the machine tool equipment are subjected to modeling combination, outlier cloud data detection and noise information removal in a complex scene are carried out, point cloud data are reconstructed in a polygonal mesh reconstruction mode, three-dimensional modeling results of the machine tool equipment in different operation states are obtained, and the modeling efficiency is improved. And digital analog simulation of the operation process of machine tool equipment in a plant is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of digital twin modeling, and in particular to a digital twin modeling method for a smart factory. Background Art

[0002] Smart factory is a new production model built by using modern information technology, automation technology, intelligent technology, etc. Its core goal is to realize the automation, informatization, intelligence and flexibility of the production process. In the process of realizing smart factory, digital twin technology, as one of its basic technologies, plays an extremely important role. The basic idea of ​​digital twin is to establish a virtual entity model through real-time data exchange and feedback between physical entities and digital models, so as to carry out real-time monitoring, prediction, optimization and decision-making. Specifically in the construction of smart factory, digital twin technology can digitally map and dynamically monitor the production process of physical factory by building a virtual "factory twin", providing data support for the operation, management and optimization of the factory. However, the existing digital twin technology applied to smart factory cannot realize the automatic operation of smart factory, that is, the twin model cannot run dynamically. Summary of the invention

[0003] In view of this, the present invention provides a digital twin modeling method for a smart factory, which realizes the intelligent and automated twin operation of factory equipment through a dynamic and static combined modeling method.

[0004] To achieve the above object, the present invention provides a method for modeling a digital twin of a smart plant, comprising the following steps: S1: Use a laser scanner to perform three-dimensional scanning on the machine tools in the factory, obtain static point cloud data of the machine tools in different operating states, and pre-process the static point cloud data of the machine tools; S2: Perform three-dimensional reconstruction on the pre-processed static point cloud data of the machine tool equipment to obtain three-dimensional modeling results of the machine tool equipment under different operating conditions; S3: Collecting operation data of machine tools in different operation states, and constructing an operation state recognition model for machine tools, wherein the operation state recognition model for machine tools takes the operation data of machine tools as input and takes the operation state of machine tools as output; S4: Based on the three-dimensional modeling results of machine tools and equipment and the machine tool equipment operation status identification model, the digital twin modeling of the factory is carried out, the real-time operation data of the machine tools and equipment is identified, the real-time operation status of the machine tools and equipment is obtained, and the three-dimensional modeling results of the machine tools and equipment are simulated.

[0005] As a further improvement method of the present invention: Optionally, in step S1, a laser scanner is used to perform a three-dimensional scan of the machine tool equipment in the factory, including: Adjust the operating status of the machine tools in the factory, start the laser scanner to perform a three-dimensional scan of the machine tools, the laser scanner emits a laser beam to the machine tools, the laser beam reaches the surface of the machine tools to form a reflected beam, the reflected beam reaches the receiver of the laser scanner, the receiver extracts the beam information of the reflected beam, obtains the flight time of the laser beam and the angle of the laser beam emission, and calculates the point cloud data of the surface of the machine tools. The flight time of the laser beam is the time difference between the laser scanner emitting the laser beam and receiving the reflected beam; The operating status of the machine tool equipment includes a processing execution status, a standby status, a maintenance status, a fault status and a shutdown status.

[0006] Optionally, the preprocessing of static point cloud data of machine tool equipment in different operating states includes: All point cloud data of the machine tool equipment in the running state are used as static point cloud data of the machine tool equipment in the running state; Modeling equipment components in the machine tool equipment using static point cloud data of the machine tool equipment, calculating the distance between the point cloud data in the static point cloud data of the machine tool equipment and the modeling result, marking the point cloud data with a distance less than a preset threshold as an inlier, marking the point cloud data with a distance greater than the preset threshold as an outlier point cloud data, and performing outlier point cloud data detection on the static point cloud data of the machine tool equipment, wherein the outlier point cloud data is point cloud data that is not on the surface of the machine tool equipment due to the influence of environmental noise during the point cloud data collection process, and the sources of the environmental noise include electromagnetic interference, temperature changes, and equipment vibration; The machine tool equipment mainly includes a bed, a worktable, a spindle and a feed transmission system. The bed is a planar equipment component of the machine tool equipment, which is used to support and fix other components. The worktable is a planar equipment component used to fix the processed workpiece. The spindle is a cylindrical equipment component used to install and drive the processed workpiece to rotate. The feed transmission system is a system for controlling the movement of equipment components and processed workpieces in the machine tool equipment, including a planar equipment component and a plurality of shafts, each of which is a cylindrical equipment component. The shafts in the feed transmission system are divided into an X-axis, a Y-axis, a Z-axis, an A-axis and a B-axis. The X-axis, the Y-axis and the Z-axis are used to control the movement of equipment components and processed workpieces in the machine tool equipment in the horizontal direction, in the direction perpendicular to the X-axis and in the vertical direction. The A-axis is used to control the rotation of equipment components and processed workpieces in the machine tool equipment around the X-axis. The B-axis is used to control the rotation of equipment components and processed workpieces in the machine tool equipment around the Y-axis. The detected outlier point cloud data are removed from the static point cloud data of the machine tool equipment to obtain the preprocessed static point cloud data of the machine tool equipment.

[0007] Optionally, in step S2, three-dimensionally reconstructing the preprocessed static point cloud data of the machine tool equipment includes: The three-dimensional reconstruction process is as follows: Select plane point cloud data from the pre-processed static point cloud data of the machine tool equipment, and use the plane point cloud data as the three-dimensional contour surface of the bed and the worktable in the machine tool equipment; Based on the three-dimensional contour surface, the Poisson surface reconstruction method is used to reconstruct the pre-processed static point cloud data of machine tool equipment into the form of polygonal mesh.

[0008] Optionally, collecting operation data of the machine tool equipment in different operation states in step S3 includes: The operation data of the machine tool equipment is divided into spindle data, feed transmission system data, processing parameters, electrical data and time data, wherein the spindle data includes the spindle speed per minute, spindle temperature, spindle torque and spindle vibration amplitude; the feed transmission system data includes the feed speed of the processed workpiece and the moving speed of each axis; the processing parameters include the type of the current processed workpiece, the speed and depth of the tool cutting the processed workpiece; the electrical data includes the current value and voltage value of the machine tool equipment; and the time data includes the processing time of each processed workpiece.

[0009] Optionally, constructing a machine tool equipment operation status identification model based on the collected operation data includes: The machine tool equipment operation status recognition model is a random forest model structure, which includes several decision trees. Each decision tree takes the operation data of the machine tool equipment as input and outputs the probability values ​​of different operation statuses. The collected operation data is constructed as a training data set data for the machine tool equipment operation status recognition model: ; in: represents the nth group of operating data collected, and N represents the total number of operating data collected; Indicates running data The corresponding operating status; The number of decision trees M and the maximum depth of the decision trees deep in the machine tool equipment operation status recognition model are solved by using strategy enhancement method; The training data set data is randomly sampled with replacement to generate M different sub-training sets, each of which is used to train a decision tree; During the training process of the decision tree, for the current child node, multiple features are selected from all features, and a threshold is found so that after splitting according to these features and the threshold, the purity of the child node after splitting is the highest; according to the selected threshold and feature, the training data of the node is divided into two groups of sub-training sets as the training data of the left and right child nodes, and the feature and threshold selection are repeated until the preset termination condition is met.

[0010] Optionally, the method of solving the number M of decision trees and the maximum depth deep of the decision trees in the machine tool equipment operation state recognition model by adopting a strategy enhancement method includes: Divide the training data set data into K subsets, use one of the subsets as the validation set in turn, and use the other subsets as the training set to train decision trees under different model parameters, where the model parameters include the number of decision trees and the maximum depth; Based on the training results of decision trees under different model parameters, the model parameters are optimized by using a heuristic optimization strategy to obtain the number of decision trees M and the maximum depth deep of the decision tree in the machine tool equipment operation status recognition model that makes the training results optimal. The training results of the decision tree are the verification accuracy of all decision trees obtained based on the model parameter training in the verification set. The model parameters The training result of the decision tree is : ; in: Represents model parameters The number of decision trees in Represents the operating status set of machine tool equipment. , j represents the running state set Any running state in Represents model parameters based on The constructed s-th decision tree receives the i-th running data in the validation set, and outputs the probability value that the running state of the i-th running data in the validation set is j; Indicates the true recognition value of the running state of the i-th running data in the validation set as j, Indicates that the i-th running data in the validation set is not in running state j. Indicates that the i-th running data in the validation set is in running state j; Indicates the number of runs in the validation set.

[0011] Optionally, in step S4, the digital twin modeling of the plant is performed based on the three-dimensional modeling result of the machine tool equipment and the machine tool equipment operation status identification model, including: Digital twin modeling of the factory is performed based on the three-dimensional modeling results of machine tool equipment and the machine tool equipment operation status identification model, wherein the digital twin modeling includes monitoring results and simulation results. The monitoring results are used to display the real-time operation data of the machine tool equipment in the form of charts, and the simulation results identify the real-time operation data of the machine tool equipment to obtain the real-time operation status of the machine tool equipment. The identification process of the operation data is as follows: the real-time operation data is input into the machine tool equipment operation status identification model, the output result mean of the last layer of sub-nodes in all decision trees is integrated, and the operation status with the highest probability is selected as the identification result; and according to the three-dimensional modeling results of the machine tool equipment under different operation status in step S1 and step S2, the three-dimensional modeling results of the machine tool equipment corresponding to the real-time operation status of the machine tool equipment are obtained by simulation.

[0012] In order to solve the above problem, the present invention provides an electronic device, the electronic device comprising: A memory storing at least one instruction; Communication interface, enabling electronic equipment to communicate; and The processor executes the instructions stored in the memory to implement the above-mentioned smart factory digital twin modeling method.

[0013] In order to solve the above problems, the present invention also provides a computer-readable storage medium, in which at least one instruction is stored. The at least one instruction is executed by a processor in an electronic device to implement the above-mentioned smart factory digital twin modeling method.

[0014] Compared with the prior art, the present invention proposes a digital twin modeling method for a smart plant, which has the following advantages: First, this scheme proposes a three-dimensional modeling method, which uses a laser scanner to scan the machine tools and equipment in the factory, obtains the point cloud data of the machine tools and equipment, and performs discrete point cloud data detection. By modeling and fitting the planar equipment components and cylindrical equipment components in the machine tools and equipment, the distance between the point cloud data and the modeling results is used to perform more robust outlier point cloud data detection in complex scenes, remove noise information, and then reconstruct the point cloud data using polygonal mesh reconstruction to obtain the three-dimensional modeling results of the machine tools and equipment under different operating states, thereby realizing the operation process modeling of the machine tools and equipment in the factory.

[0015] At the same time, this solution proposes a digital twin modeling method, which adopts a random forest structure to build a machine tool equipment operation status recognition model, and adopts a heuristic algorithm strategy to iteratively guide the model parameters to quickly obtain a feasible solution. In the iterative guidance process, the discoverer and the tracker are divided according to the training results of the model parameters. The discoverer is responsible for exploring the search space and finding higher quality solutions. In order to improve the diversity of the discoverer, adaptive iterative weights and spiral change strategies are further introduced to further improve the search efficiency. The tracker performs local search based on the mean of the model parameters corresponding to the discoverer to improve the accuracy of the solution. The efficient optimization of the complex search space is achieved through joint collaboration, so that the solved model parameters have strong adaptability and robustness, and the training results are achieved. The number of decision trees and the maximum depth of the decision trees in the optimal machine tool equipment operation status identification model are determined. Digital twin modeling of the factory is performed based on the three-dimensional modeling results of the machine tool equipment and the machine tool equipment operation status identification model. The real-time operation data of the machine tool equipment is identified to obtain the real-time operation status of the machine tool equipment, and the three-dimensional modeling results of the machine tool equipment are simulated. The digital twin modeling includes monitoring results and simulation results. The monitoring results are used to display the real-time operation data of the machine tool equipment in the form of charts. The simulation results identify the real-time operation data of the machine tool equipment to obtain the real-time operation status of the machine tool equipment. According to the three-dimensional modeling results of the machine tool equipment under different operation statuses, the three-dimensional modeling results of the machine tool equipment corresponding to the real-time operation status of the machine tool equipment are simulated. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic flow chart of a method for digital twin modeling of a smart factory provided in accordance with an embodiment of the present invention.

[0017] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0018] 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.

[0019] The embodiment of the present application provides a method for digital twin modeling of a smart factory. The execution subject of the method for digital twin modeling of a smart factory includes but is not limited to at least one of the electronic devices such as a server and a terminal that can be configured to execute the method provided in the embodiment of the present application. In other words, the method for digital twin modeling of a smart factory can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0020] A method for modeling a digital twin of a smart plant includes the following steps: S1: Use a laser scanner to perform three-dimensional scanning on the machine tools in the factory, obtain static point cloud data of the machine tools under different operating conditions, and pre-process the static point cloud data of the machine tools.

[0021] In step S1, a laser scanner is used to perform a three-dimensional scan of the machine tool equipment in the factory, including: Adjust the operating status of the machine tools in the factory, start the laser scanner to perform a three-dimensional scan of the machine tools, the laser scanner emits a laser beam to the machine tools, the laser beam reaches the surface of the machine tools to form a reflected light beam, the reflected light beam reaches the receiver of the laser scanner, the receiver extracts the beam information of the reflected light beam, obtains the flight time of the laser beam and the angle of emission of the laser beam, and calculates the point cloud data of the surface of the machine tools, the flight time of the laser beam is the time difference between the laser scanner emitting the laser beam and receiving the reflected light beam; as an embodiment of the present invention, the point cloud data is three-dimensional coordinate data, the flight time t of the laser beam, the horizontal angle of the laser beam emission and the vertical angle at which the laser beam is emitted The corresponding point cloud data is : ; ; ; in: c represents the speed of light; The horizontal angle of the laser beam emission is the rotation angle of the laser scanner in the horizontal plane, and the vertical angle of the laser beam emission is the rotation angle of the laser scanner in the vertical plane; Point cloud data The distance between the corresponding machine tool surface coordinates and the laser scanner; The operation status of the machine tool equipment includes the processing status, standby status, maintenance status, fault status and shutdown status. Specifically, the processing status is divided into the processing status of various processing tasks, and the processing tasks of the machine tool equipment include turning processing tasks, milling processing tasks, drilling processing tasks, grinding processing tasks, electric spark processing tasks, laser processing tasks and CNC processing tasks.

[0022] All point cloud data of the machine tool equipment in the running state are used as static point cloud data of the machine tool equipment in the running state; Modeling equipment components in the machine tool equipment using static point cloud data of the machine tool equipment, calculating the distance between the point cloud data in the static point cloud data of the machine tool equipment and the modeling result, marking the point cloud data with a distance less than a preset threshold as an inlier, marking the point cloud data with a distance greater than the preset threshold as an outlier point cloud data, and performing outlier point cloud data detection on the static point cloud data of the machine tool equipment, wherein the outlier point cloud data is point cloud data that is not on the surface of the machine tool equipment due to the influence of environmental noise during the point cloud data collection process, and the sources of the environmental noise include electromagnetic interference, temperature changes, and equipment vibration; The machine tool equipment mainly includes a bed, a worktable, a spindle and a feed transmission system. The bed is a planar equipment component of the machine tool equipment, which is used to support and fix other components. The worktable is a planar equipment component used to fix the processed workpiece. The spindle is a cylindrical equipment component used to install and drive the processed workpiece to rotate. The feed transmission system is a system for controlling the movement of equipment components and processed workpieces in the machine tool equipment, including a planar equipment component and a plurality of shafts, each of which is a cylindrical equipment component. The shafts in the feed transmission system are divided into an X-axis, a Y-axis, a Z-axis, an A-axis and a B-axis. The X-axis, the Y-axis and the Z-axis are used to control the movement of equipment components and processed workpieces in the machine tool equipment in the horizontal direction, in the direction perpendicular to the X-axis and in the vertical direction. The A-axis is used to control the rotation of equipment components and processed workpieces in the machine tool equipment around the X-axis. The B-axis is used to control the rotation of equipment components and processed workpieces in the machine tool equipment around the Y-axis. As a preferred embodiment of the present invention, by modeling and fitting the planar equipment components and cylindrical equipment components in the machine tool equipment, outlier point cloud data detection is performed using the distance between the point cloud data and the modeling result, and robust discrete point cloud data detection is performed in complex modeling scenarios. The outlier point cloud data detection process of the static point cloud data of the machine tool equipment is as follows: Select any three point cloud data from the static point cloud data of the machine tool equipment : ; The modeling results of the plane equipment components and the cylindrical equipment components in the machine tool equipment are obtained by fitting the selected point cloud data; Calculate the distance from other point cloud data in the static point cloud data of the machine tool equipment to the modeling result, where the point cloud data The distance to the modeling result is : ; ; ; ; ; in: Indicates the radius of the cylindrical equipment component obtained by fitting the selected point cloud data. Representing point cloud data The distance to the modeling result of the cylindrical equipment component, represents the L2 norm; Representing point cloud data The distance to the modeling result of the planar equipment component; Represents the normal vectors of the plane and the axis plane in the planar device component and the cylindrical device component; Representing point cloud data The vertical distance to the centerline of the cylindrical equipment component; The distance between the static point cloud data of the machine tool equipment and the modeling result is less than the preset threshold The point cloud data is marked as discrete point cloud data, and the proportion of the current discrete point cloud data is counted. If the proportion is less than the preset ratio threshold rate, the marking result of the current discrete point cloud data is accepted, otherwise the point cloud data is reselected for modeling and fitting of planar equipment components and cylindrical equipment components; The detected outlier point cloud data are removed from the static point cloud data of the machine tool equipment to obtain the preprocessed static point cloud data of the machine tool equipment.

[0023] S2: Perform three-dimensional reconstruction on the pre-processed static point cloud data of the machine tool equipment to obtain three-dimensional modeling results of the machine tool equipment under different operating conditions.

[0024] The three-dimensional reconstruction process is as follows: Select plane point cloud data from the preprocessed static point cloud data of the machine tool equipment, and use the plane point cloud data as the three-dimensional contour surface of the bed and the worktable in the machine tool equipment; specifically, this scheme calculates the distance between the point cloud data in the preprocessed static point cloud data of the machine tool equipment and the modeling result of the plane equipment component, and uses the point cloud data with a distance lower than a preset threshold as the plane point cloud data; On the basis of the three-dimensional contour surface, the Poisson surface reconstruction method is used to reconstruct the pre-processed static point cloud data of the machine tool equipment into the form of a polygonal mesh; specifically, the normal vector of the point cloud data in the pre-processed static point cloud data of the machine tool equipment is calculated, and the Poisson equation of the point cloud data is constructed according to the point cloud data and the normal vector of the point cloud data. The Poisson equation is solved by the multi-grid method to obtain the isosurface of the point cloud data, and the pre-processed static point cloud data of the machine tool equipment is reconstructed into the form of a polygonal mesh as the three-dimensional modeling result of the machine tool equipment in the operating state corresponding to the pre-processed static point cloud data of the machine tool equipment.

[0025] S3: Collecting operation data of machine tool equipment under different operation states, and constructing a machine tool equipment operation state recognition model, wherein the machine tool equipment operation state recognition model takes the operation data of the machine tool equipment as input and takes the operation state of the machine tool equipment as output.

[0026] The operation data of the machine tool equipment under different operation states are collected in step S3, including: The operation data of the machine tool equipment is divided into spindle data, feed transmission system data, processing parameters, electrical data and time data, wherein the spindle data includes the spindle speed per minute, spindle temperature, spindle torque and spindle vibration amplitude; the feed transmission system data includes the feed speed of the processed workpiece and the moving speed of each axis; the processing parameters include the type of the current processed workpiece, the speed and depth of the tool cutting the processed workpiece; the electrical data includes the current value and voltage value of the machine tool equipment; and the time data includes the processing time of each processed workpiece.

[0027] The machine tool equipment operation status recognition model is a random forest model structure, which includes several decision trees. Each decision tree takes the operation data of the machine tool equipment as input and outputs the probability values ​​of different operation statuses. The collected operation data is constructed as a training data set data for the machine tool equipment operation status recognition model: ; in: represents the nth group of operating data collected, and N represents the total number of operating data collected; Indicates running data The corresponding operating status; The number of decision trees M and the maximum depth of the decision trees deep in the machine tool equipment operation status recognition model are solved by using strategy enhancement method; The training data set data is randomly sampled with replacement to generate M different sub-training sets, each of which is used to train a decision tree; During the training process of the decision tree, for the current child node, multiple features are selected from all features, and a threshold is found so that after splitting according to these features and the threshold, the purity of the child node after splitting is the highest; according to the selected threshold and feature, the training data of the node is divided into two groups of sub-training sets as the training data of the left and right child nodes, and the feature and threshold selection are repeated until the preset termination condition is met.

[0028] As an embodiment of the present invention, the termination conditions include: 1. The depth of the current decision tree reaches the maximum depth deep; 2. The number of child node training data is lower than the preset number threshold num; 3. The training data of the child nodes all belong to the same operating state. Specifically, the Gini coefficient and information gain can be used as the purity of the child nodes after splitting.

[0029] Divide the training data set data into K subsets, use one of the subsets as the validation set in turn, and use the other subsets as the training set to train decision trees under different model parameters, where the model parameters include the number of decision trees and the maximum depth; Based on the training results of decision trees under different model parameters, the model parameters are optimized by using a heuristic optimization strategy to obtain the number of decision trees M and the maximum depth deep of the decision tree in the machine tool equipment operation status recognition model that makes the training results optimal. The training results of the decision tree are the verification accuracy of all decision trees obtained based on the model parameter training in the verification set. The model parameters The training result of the decision tree is : ; in: Represents model parameters The number of decision trees in Represents the operating status set of machine tool equipment. , j represents the running state set Any running state in Represents model parameters based on The constructed s-th decision tree receives the i-th running data in the validation set, and outputs the probability value that the running state of the i-th running data in the validation set is j; Indicates the true recognition value of the running state of the i-th running data in the validation set as j, Indicates that the i-th running data in the validation set is not in running state j. Indicates that the i-th running data in the validation set is in running state j; Indicates the number of runs in the validation set.

[0030] As a preferred embodiment of the present invention, the present invention adopts a heuristic algorithm strategy to iteratively guide the model parameters to quickly obtain a feasible solution, and in the iterative guidance process, the discoverer and the tracker are divided according to the training results of the model parameters, wherein the discoverer is responsible for exploring the search space and finding a higher quality solution. In order to improve the diversity of the discoverer, an adaptive iteration weight and a spiral change strategy are further introduced to further improve the search efficiency. The tracker performs a local search based on the mean of the model parameters corresponding to the discoverer to improve the accuracy of the solution. The efficient optimization of the complex search space is achieved by joint collaboration, so that the solved model parameters have strong adaptability and robustness. Specifically, the iterative process of the model parameters is as follows: S31: Initialize and generate Q groups of model parameters, the initialization formula of the qth group of model parameters is: ; in: represents the qth group of model parameters generated by initialization, , Indicates the preset control parameters; S32: Set the current iteration number of the model parameters to , the maximum number of iterations is Max, then the qth group of model parameters The result of the iteration is ; S33: Calculate the training result of the decision tree under the model parameters, where the model parameters The training result of the decision tree is ; S34: Select the After the iteration, the 30% model parameters with the highest training results are used as discoverers, and the other model parameters are used as trackers. The model parameters are iterated, and the iteration formula of the discoverer is: ; ; in: Indicates The iteration weight of the iteration, Represents a random number between -1 and 1. represents an exponential function with a natural constant as base; Represents a random number between 0 and 1; Represents random numbers that conform to the normal distribution; The iteration formula of the tracker is: ; in: For the The mean of the finders after iterations; Repeat step S34 until the maximum number of iterations is reached, and select the model parameter with the highest training result , as the number of decision trees M and the maximum depth of the decision tree deep, where M= ,deep= .

[0031] S4: Based on the three-dimensional modeling results of machine tools and equipment and the machine tool equipment operation status identification model, the digital twin modeling of the factory is carried out, the real-time operation data of the machine tools and equipment is identified, the real-time operation status of the machine tools and equipment is obtained, and the three-dimensional modeling results of the machine tools and equipment are simulated.

[0032] In the step S4, digital twin modeling of the plant is performed based on the three-dimensional modeling results of the machine tool equipment and the machine tool equipment operation status identification model, including: Based on the three-dimensional modeling results of machine tool equipment and the machine tool equipment operation status identification model, the digital twin modeling of the factory is carried out, wherein the digital twin modeling includes monitoring results and simulation results. The monitoring results are used to display the real-time operation data of the machine tool equipment in the form of charts, and the simulation results identify the real-time operation data of the machine tool equipment to obtain the real-time operation status of the machine tool equipment. The identification process of the operation data is as follows: the real-time operation data is input into the machine tool equipment operation status identification model, the output result mean of the last layer of sub-nodes in all decision trees is integrated, and the operation status with the highest probability is selected as the identification result; and according to the three-dimensional modeling results of the machine tool equipment under different operation status in step S1 and step S2, the three-dimensional modeling results of the machine tool equipment corresponding to the real-time operation status of the machine tool equipment are obtained by simulation.

[0033] As a preferred embodiment of the present invention, the output value of each layer of sub-nodes in the decision tree is optimized by smoothing optimization, wherein the output value optimization result of the g-th layer of the m-th decision tree is : ; in: represents the output value of the g-th layer of the m-th decision tree, , represents the depth of the mth decision tree, ; Represents the control parameter, Set to 20; Indicates the number of training data used to train the g-th layer child nodes of the m-th decision tree; As an embodiment of the present invention, the operating data of the machine tool equipment in the digital twin modeling can be adjusted in advance before the machine tool equipment is operated to obtain the corresponding three-dimensional modeling results of the machine tool equipment, and observe whether the three-dimensional modeling results of the machine tool equipment cause irreversible damage to the machine tool equipment, thereby adjusting the operating data of the machine tool equipment and improving the operating life of the machine tool equipment in the factory.

[0034] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0035] It should be noted that the serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages and disadvantages of the embodiments. And the terms "including", "comprising" or any other variants thereof in this article are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "including a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0036] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0037] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A digital twin modeling method for a smart factory, characterized in that: The method comprises: S1: Use a laser scanner to perform three-dimensional scanning on the machine tools in the factory, obtain static point cloud data of the machine tools in different operating states, and pre-process the static point cloud data of the machine tools; S2: Perform three-dimensional reconstruction on the pre-processed static point cloud data of the machine tool equipment to obtain three-dimensional modeling results of the machine tool equipment under different operating conditions; S3: Collecting operation data of machine tools in different operation states, and constructing an operation state recognition model for machine tools, wherein the operation state recognition model for machine tools takes the operation data of machine tools as input and takes the operation state of machine tools as output; S4: Based on the three-dimensional modeling results of machine tools and equipment and the machine tool equipment operation status identification model, the digital twin modeling of the factory is carried out, the real-time operation data of the machine tools and equipment is identified, the real-time operation status of the machine tools and equipment is obtained, and the three-dimensional modeling results of the machine tools and equipment are simulated.

2. A method for modeling a digital twin of a smart factory building according to claim 1, characterized in that: In step S1, a laser scanner is used to perform a three-dimensional scan of the machine tool equipment in the factory, including: Adjust the operating status of the machine tools in the factory, start the laser scanner to perform a three-dimensional scan of the machine tools, the laser scanner emits a laser beam to the machine tools, the laser beam reaches the surface of the machine tools to form a reflected beam, the reflected beam reaches the receiver of the laser scanner, the receiver extracts the beam information of the reflected beam, obtains the flight time of the laser beam and the angle of the laser beam emission, and calculates the point cloud data of the surface of the machine tools. The flight time of the laser beam is the time difference between the laser scanner emitting the laser beam and receiving the reflected beam; The operating status of the machine tool equipment includes a processing execution status, a standby status, a maintenance status, a fault status and a shutdown status.

3. A method for modeling a digital twin of a smart factory building according to claim 2, characterized in that: The preprocessing of static point cloud data of machine tool equipment in different operating states includes: All point cloud data of the machine tool equipment in the running state are used as static point cloud data of the machine tool equipment in the running state; Modeling equipment components in the machine tool equipment using static point cloud data of the machine tool equipment, calculating the distance between the point cloud data in the static point cloud data of the machine tool equipment and the modeling result, marking the point cloud data with a distance less than a preset threshold as an inlier, marking the point cloud data with a distance greater than the preset threshold as an outlier point cloud data, and performing outlier point cloud data detection on the static point cloud data of the machine tool equipment, wherein the outlier point cloud data is point cloud data that is not on the surface of the machine tool equipment due to the influence of environmental noise during the point cloud data collection process, and the sources of the environmental noise include electromagnetic interference, temperature changes, and equipment vibration; The machine tool equipment mainly includes a bed, a worktable, a spindle and a feed transmission system. The bed is a planar equipment component of the machine tool equipment, which is used to support and fix other components. The worktable is a planar equipment component used to fix the workpiece to be processed. The spindle is a cylindrical equipment component used to install and drive the workpiece to rotate. The feed transmission system is a system that controls the movement of equipment components and workpieces in the machine tool equipment. The detected outlier point cloud data are removed from the static point cloud data of the machine tool equipment to obtain the preprocessed static point cloud data of the machine tool equipment.

4. A method for modeling a digital twin of a smart factory building as claimed in claim 3, characterized in that: The step S2 performs three-dimensional reconstruction on the pre-processed static point cloud data of the machine tool equipment, including: The three-dimensional reconstruction process is as follows: Select plane point cloud data from the pre-processed static point cloud data of the machine tool equipment, and use the plane point cloud data as the three-dimensional contour surface of the bed and the worktable in the machine tool equipment; Based on the three-dimensional contour surface, the Poisson surface reconstruction method is used to reconstruct the pre-processed static point cloud data of machine tool equipment into the form of polygonal mesh.

5. The method for modeling a digital twin of a smart factory building according to claim 1, characterized in that: The operation data of the machine tool equipment under different operation states are collected in step S3, including: The operation data of the machine tool equipment is divided into spindle data, feed transmission system data, processing parameters, electrical data and time data, wherein the spindle data includes the spindle speed per minute, spindle temperature, spindle torque and spindle vibration amplitude; the feed transmission system data includes the feed speed of the processed workpiece and the moving speed of each axis; the processing parameters include the type of the current processed workpiece, the speed and depth of the tool cutting the processed workpiece; the electrical data includes the current value and voltage value of the machine tool equipment; and the time data includes the processing time of each processed workpiece.

6. A method for modeling a digital twin of a smart factory building according to claim 5, characterized in that: The step of constructing a machine tool equipment operation status identification model based on the collected operation data includes: The machine tool equipment operation status recognition model is a random forest model structure, which includes several decision trees. Each decision tree takes the operation data of the machine tool equipment as input and outputs the probability values ​​of different operation statuses. The collected operation data is constructed as a training data set data for the machine tool equipment operation status recognition model: ； in: represents the nth group of operating data collected, and N represents the total number of operating data collected; Indicates running data The corresponding operating status; The number of decision trees M and the maximum depth of the decision trees deep in the machine tool equipment operation status recognition model are solved by using strategy enhancement method; The training data set data is randomly sampled with replacement to generate M different sub-training sets, each of which is used to train a decision tree; During the training process of the decision tree, for the current child node, multiple features are selected from all features, and a threshold is found so that after splitting according to these features and the threshold, the purity of the child node after splitting is the highest; according to the selected threshold and feature, the training data of the node is divided into two groups of sub-training sets as the training data of the left and right child nodes, and the feature and threshold selection are repeated until the preset termination condition is met.

7. A method for modeling a digital twin of a smart factory building according to claim 6, characterized in that: The method of using strategy enhancement to solve the number M of decision trees and the maximum depth deep of the decision trees in the machine tool equipment operation status recognition model includes: Divide the training data set data into K subsets, use one of the subsets as the validation set in turn, and use the other subsets as the training set to train decision trees under different model parameters, where the model parameters include the number of decision trees and the maximum depth; Based on the training results of decision trees under different model parameters, the model parameters are optimized by using a heuristic optimization strategy to obtain the number of decision trees M and the maximum depth deep of the decision tree in the machine tool equipment operation status recognition model that makes the training results optimal. The training results of the decision tree are the verification accuracy of all decision trees obtained based on the model parameter training in the verification set. The model parameters The training result of the decision tree is : ； in: Represents model parameters The number of decision trees in Represents the operating status set of machine tool equipment. , j represents the running state set Any running state in Represents model parameters based on The constructed s-th decision tree receives the i-th running data in the validation set, and outputs the probability value that the running state of the i-th running data in the validation set is j; Indicates the true recognition value of the running state of the i-th running data in the validation set as j, Indicates that the i-th running data in the validation set is not in running state j. Indicates that the i-th running data in the validation set is in running state j; Indicates the number of runs in the validation set.

8. The method for modeling a digital twin of a smart factory building according to claim 1, characterized in that: In the step S4, digital twin modeling of the plant is performed based on the three-dimensional modeling results of the machine tool equipment and the machine tool equipment operation status identification model, including: Digital twin modeling of the factory is performed based on the three-dimensional modeling results of machine tool equipment and the machine tool equipment operation status identification model, wherein the digital twin modeling includes monitoring results and simulation results. The monitoring results are used to display the real-time operation data of the machine tool equipment in the form of charts, and the simulation results identify the real-time operation data of the machine tool equipment to obtain the real-time operation status of the machine tool equipment. The identification process of the operation data is as follows: the real-time operation data is input into the machine tool equipment operation status identification model, the output result mean of the last layer of sub-nodes in all decision trees is integrated, and the operation status with the highest probability is selected as the identification result; and according to the three-dimensional modeling results of the machine tool equipment under different operation status in step S1 and step S2, the three-dimensional modeling results of the machine tool equipment corresponding to the real-time operation status of the machine tool equipment are obtained by simulation.

Images