Manufacturing digital twin 3D visualization control method and system

Abstract

The present invention relates to the field of intelligent manufacturing technology, specifically to a three-dimensional visualization management and control method and system for digital twins in the manufacturing industry, comprising the following steps: collecting manufacturing equipment operation data and converting it into a time series, analyzing the trend characteristics of each dynamic subsequence in the time series, and generating a dynamic operation state sequence set. In the present invention, by dynamically analyzing the time series characteristics in the manufacturing equipment operation data, the trend changes of the equipment operation state can be accurately captured in multiple time dimensions, so as to deeply explore and predict potential abnormal states. Combined with the comparison of the prediction of future operation state parameters with the normal state range, the abnormal trend area and its corresponding fault level can be clearly marked, effectively improving the risk assessment capability of the equipment health status. The marked risk area is mapped in three-dimensional space, and combined with the gradient change and deviation distribution analysis of the three-dimensional coordinates of the equipment, the abnormal components and deviation areas can be accurately located.

Description

Technical Field

[0001] The present invention relates to the field of intelligent manufacturing technology, and in particular to a three-dimensional visualization management and control method and system for digital twins in the manufacturing industry. Background Art

[0002] The field of intelligent manufacturing technology involves combining information technology with advanced manufacturing technology to achieve efficient collaboration and optimized management of the manufacturing process by building a digital, networked, and intelligent manufacturing system. Its core content is digital modeling, real-time monitoring, precise control, and optimization of the entire manufacturing process, including data collection, manufacturing equipment interconnection, production scheduling, quality monitoring, and other links.

[0003] Among them, the manufacturing industry digital twin 3D visualization management and control method refers to the combination of digital twin technology and 3D visualization technology to dynamically simulate and visualize various types of data in the manufacturing system, so as to achieve real-time monitoring and management of production processes, manufacturing equipment status and environmental parameters. This method specifically covers the real-time collection and modeling of manufacturing equipment operating status data in the factory, digital simulation of the execution progress of production tasks, precise modeling and visualization of the 3D space environment, and optimization of production control strategies through multi-dimensional data analysis.

[0004] It is difficult for existing technologies to capture the operating characteristics and trends of equipment under dynamic operating conditions, and they can only analyze based on static data, resulting in a relatively delayed recognition of the law of operating status changes. For example, during the operation of manufacturing equipment, if key components show a trend of gradual deterioration, existing methods often find it difficult to identify abnormal trends in advance through a single state value, which may cause potential problems to develop into serious failures. The existing technology has weak three-dimensional spatial positioning capabilities for equipment failures, and can only provide planar abnormal area information, lacking accurate spatial distribution analysis. For example, when multiple components inside a complex device have problems at the same time, existing methods may not be able to provide accurate abnormal component locations, resulting in a time-consuming and inefficient troubleshooting process. In addition, the existing technology is limited to static state display in terms of data presentation, lacks the ability to display the linkage between the time dimension and spatial distribution, making it difficult for monitoring personnel to intuitively understand the changes in equipment status. For example, during the operation of the equipment, if the deviations of the dynamic characteristics of multiple components change at the same time, a single display method may not be able to quickly reflect the global trend, increasing the difficulty of identifying and handling complex problems. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings existing in the prior art and propose a three-dimensional visualization management and control method and system for digital twins in the manufacturing industry.

[0006] In order to achieve the above purpose, the present invention adopts the following technical solution: a manufacturing digital twin three-dimensional visualization management and control method, comprising the following steps:

[0007] S1: Collect manufacturing equipment operation data and convert it into time series, analyze the trend characteristics of each dynamic subsequence in the time series, and generate a dynamic operation status sequence set;

[0008] S2: predicting the operating state of each time series in the dynamic operating state sequence set to obtain an operating state prediction result, comparing the prediction parameters at future moments in the operating state prediction result with a preset normal state range, marking the risk area and fault level of the abnormal trend according to the comparison result, and generating a manufacturing equipment health state risk analysis result;

[0009] S3: Mapping the manufacturing equipment component positions corresponding to the risk areas marked in the manufacturing equipment health status risk analysis results to obtain the three-dimensional spatial data of the manufacturing equipment, analyzing the gradient changes and deviation distribution in the three-dimensional spatial data of the manufacturing equipment to obtain the three-dimensional mapping results of the manufacturing equipment deviation;

[0010] S4: Associating the behavioral characteristic data of manufacturing equipment with the three-dimensional spatial data of manufacturing equipment in the three-dimensional mapping result of manufacturing equipment deviation, constructing a digital twin model of manufacturing equipment, and generating a three-dimensional visualization management and control result of the digital twin of the manufacturing equipment by superimposing and displaying the linkage between the dynamic change characteristics of the manufacturing equipment during the process execution and the deviation distribution in the digital twin model of the manufacturing equipment.

[0011] As a further solution of the present invention, the step of acquiring the dynamic running state sequence set is specifically:

[0012] S111: collecting manufacturing equipment operation data and converting it into a time series, dividing the time series of the manufacturing equipment operation data into multiple dynamic subsequences through a sliding window, and obtaining a dynamic subsequence set;

[0013] S112: Based on the dynamic subsequence set, compare the real-time data in each dynamic subsequence with the state data of the previous period, analyze the trend characteristics of each dynamic subsequence, and generate a dynamic operation state sequence set.

[0014] As a further solution of the present invention, the step of obtaining the running state prediction result is specifically:

[0015] S211: Use a piecewise regression model to predict the operating state of each time series in the dynamic operating state series set, using the formula:

[0016] ;

[0017] Calculate dimensionless predictions of equipment operating status ;

[0018] in, It is Eigenvalues The standardized value of It is Eigenvalues The mean of It is Eigenvalues The standard deviation of It is External environment variables The standardized value of It is external environment variables, Is an environment variable The mean of Is an environment variable The standard deviation of It is Eigenvalues The weight coefficient of Is an external environment variable The weight coefficient of is the bias value of the piecewise regression model, Indicates the total number of feature values ​​of the first feature group run by the device. The total number of feature values ​​of the second feature group representing the operation of the device;

[0019] S212: Optimizing the dimensionless prediction value of the equipment operation status by dynamically adjusting the regression order of the segmented regression model to obtain an operation status prediction result.

[0020] As a further solution of the present invention, the steps for obtaining the manufacturing equipment health status risk analysis result are specifically as follows:

[0021] S221: comparing the prediction parameters at the future time in the operation state prediction result with the preset normal state range, judging whether there is an abnormal trend in the predicted state according to the comparison result, and obtaining an abnormal trend analysis result;

[0022] S222: Count the fluctuation range and deviation degree of the characteristic value within the abnormal time period, mark the corresponding risk area and fault level in the abnormal trend analysis result according to the statistical result, and generate the manufacturing equipment health status risk analysis result.

[0023] As a further solution of the present invention, the steps for obtaining the three-dimensional mapping result of the manufacturing equipment deviation are specifically as follows:

[0024] S311: Mapping the manufacturing equipment component positions corresponding to the risk areas marked in the manufacturing equipment health status risk analysis results, matching the component information marked in the risk areas with the corresponding three-dimensional coordinate points, extracting all coordinate data including the risk area range, and obtaining the three-dimensional spatial data of the manufacturing equipment;

[0025] S312: Based on the three-dimensional spatial data of the manufacturing equipment, the formula is used:

[0026] ;

[0027] Calculating state variables The gradient vector ;

[0028] in, is a state variable, is the coordinate value in three-dimensional space, It represents the state variable exist , , The partial derivatives in three directions reflect the rate of change of the state variables along the spatial direction, Corresponding to , , The unit length of the direction represents the directional component of the gradient vector in the three-dimensional coordinate system;

[0029] S313: According to the gradient vector of the state variable, the gradient change in the three-dimensional spatial data of the equipment is matched with the equipment structure position. By setting the threshold range of the deviation distribution, the deviation area and distribution of the equipment components are analyzed, and the deviation area is marked according to multiple categories to generate a three-dimensional mapping result of the manufacturing equipment deviation.

[0030] As a further solution of the present invention, the steps of constructing a digital twin model of manufacturing equipment are specifically:

[0031] S411: extracting behavior feature data of the manufacturing equipment when executing a specified process from the manufacturing equipment operation data, associating the behavior feature data with the three-dimensional spatial data of the manufacturing equipment in the three-dimensional mapping result of the manufacturing equipment deviation, and generating a structured data set of the behavior feature and the three-dimensional deviation data;

[0032] S412: Extract the dynamic change features in the structured data set of the behavior features and three-dimensional deviation data by partition, fit the dynamic curve, and superimpose the fitting results to construct a digital twin model of the manufacturing equipment.

[0033] As a further solution of the present invention, the steps for obtaining the three-dimensional visualization management and control results of the manufacturing digital twin are specifically as follows:

[0034] S421: Based on the digital twin model of the manufacturing equipment, the dynamic behavior data of the manufacturing equipment in the specified process and the equipment deviation distribution data are uniformly mapped into a three-dimensional coordinate system, and the dynamic change characteristics of key components and the deviation distribution linkage are dynamically displayed by time axis superposition to obtain a dynamically adjusted three-dimensional digital twin model;

[0035] S422: Based on the dynamically adjusted three-dimensional digital twin model, the key dynamic characteristics and deviation distribution of the target area of ​​the equipment are dynamically displayed in combination with the time axis and spatial distribution, and the data is displayed in layers through color changes and area magnification to generate three-dimensional visualization management and control results of the digital twin of the manufacturing industry.

[0036] A manufacturing digital twin three-dimensional visualization management and control system, the manufacturing digital twin three-dimensional visualization management and control system is used to execute the above-mentioned manufacturing digital twin three-dimensional visualization management and control method, the system includes:

[0037] The data collection module collects the operating data of the manufacturing equipment and converts it into a time series, analyzes the trend characteristics of each dynamic subsequence in the time series, and generates a dynamic operating state sequence set;

[0038] The state prediction module predicts the operation state of each time series in the dynamic operation state sequence set to obtain the operation state prediction result, compares the prediction parameters at future moments in the operation state prediction result with the preset normal state range, marks the risk area and fault level of abnormal trend according to the comparison result, and generates the manufacturing equipment health state risk analysis result;

[0039] The health analysis module maps the manufacturing equipment component positions corresponding to the risk areas marked in the manufacturing equipment health status risk analysis results to obtain the three-dimensional spatial data of the manufacturing equipment, analyzes the gradient changes and deviation distribution in the three-dimensional spatial data of the manufacturing equipment, and obtains the three-dimensional mapping results of the manufacturing equipment deviation;

[0040] The digital twin module associates the behavioral characteristic data of manufacturing equipment with the three-dimensional spatial data of manufacturing equipment in the three-dimensional mapping results of manufacturing equipment deviations to construct a digital twin model of the manufacturing equipment. By superimposing the linkage between the dynamic change characteristics of the manufacturing equipment during the process execution and the deviation distribution in the digital twin model of the manufacturing equipment, a three-dimensional visualization management and control result of the digital twin of the manufacturing industry is generated.

[0041] Compared with the prior art, the advantages and positive effects of the present invention are:

[0042] In the present invention, by dynamically analyzing the time series characteristics in the manufacturing equipment operation data, the trend changes of the equipment operation status can be accurately captured in multiple time dimensions, so as to deeply explore and predict the potential abnormal state. Combined with the prediction of future operation status parameters and the comparison of the normal state range, the abnormal trend area and its corresponding fault level can be clearly marked, effectively improving the risk assessment ability of the equipment health status. The marked risk area is mapped in three-dimensional space, combined with the gradient change and deviation distribution analysis of the three-dimensional coordinates of the equipment, the abnormal components and deviation areas can be accurately located, the fault diagnosis accuracy can be optimized and the error can be reduced. The equipment behavior characteristics and the three-dimensional deviation distribution are dynamically associated, and the dynamic change characteristics of the equipment operation process are displayed by superimposing dynamic curves, which significantly enhances the real-time perception ability of the equipment status. Combined with the data display method of the time axis and spatial distribution, the equipment operation status changes can be dynamically presented in multiple dimensions, and the visualization effect and data interpretation efficiency can be further improved through hierarchical display methods, such as color changes and area enlargement. It not only strengthens the abnormal early warning ability of the production process, but also optimizes the dynamic control effect of the equipment operation status, and reduces the impact of equipment failure on the production line. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a schematic diagram of the workflow of the present invention;

[0044] Figure 2 A flow chart for obtaining a dynamic running state sequence set for the present invention;

[0045] Figure 3 A flow chart for obtaining the operation status prediction result of the present invention;

[0046] Figure 4 A flow chart for obtaining the health status risk analysis results of manufacturing equipment according to the present invention;

[0047] Figure 5 A flowchart for obtaining three-dimensional mapping results of manufacturing equipment deviations according to the present invention;

[0048] Figure 6 A flow chart for building a digital twin model of a manufacturing device for the present invention;

[0049] Figure 7 This is a flow chart for obtaining the three-dimensional visualization management and control results of the manufacturing industry digital twin in the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0051] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0052] See also Figure 1 The present invention provides a technical solution: a three-dimensional visualization management and control method for digital twins in manufacturing industry, comprising the following steps:

[0053] S1: Collect manufacturing equipment operation data and convert it into time series, analyze the trend characteristics of each dynamic subsequence in the time series, and generate a dynamic operation status sequence set;

[0054] S2: predict the operating status of each time series in the dynamic operating status sequence set to obtain the operating status prediction result, compare the prediction parameters at future moments in the operating status prediction result with the preset normal state range, mark the risk area and fault level of abnormal trend according to the comparison result, and generate the manufacturing equipment health status risk analysis result;

[0055] S3: Map the manufacturing equipment component positions corresponding to the risk areas marked in the manufacturing equipment health status risk analysis results to obtain the three-dimensional spatial data of the manufacturing equipment, analyze the gradient changes and deviation distribution in the three-dimensional spatial data of the manufacturing equipment, and obtain the three-dimensional mapping results of the manufacturing equipment deviation;

[0056] S4: Associating the manufacturing equipment behavior feature data with the manufacturing equipment three-dimensional spatial data in the manufacturing equipment deviation three-dimensional mapping results, building a digital twin model of the manufacturing equipment, and generating a three-dimensional visualization management and control result of the manufacturing industry digital twin by superimposing and displaying the linkage between the dynamic change characteristics of the manufacturing equipment during the process execution and the deviation distribution in the digital twin model of the manufacturing equipment;

[0057] The dynamic operation status sequence set includes dynamic subsequences, trend feature data, and time series status data. The manufacturing equipment health status risk analysis results specifically include operation status prediction parameters, risk area marking information, and fault level information. The three-dimensional mapping results of manufacturing equipment deviations include three-dimensional spatial data of manufacturing equipment, gradient change data, and deviation distribution data. The three-dimensional visualization management and control results of the manufacturing industry digital twin specifically refer to the dynamic change characteristics of manufacturing equipment, deviation distribution linkage, and the digital twin model of manufacturing equipment.

[0058] See also Figure 2 , the specific steps for obtaining the dynamic running state sequence set are:

[0059] S111: collecting manufacturing equipment operation data and converting it into a time series, dividing the time series of the manufacturing equipment operation data into multiple dynamic subsequences through a sliding window, and obtaining a dynamic subsequence set;

[0060] Through the data acquisition equipment, various types of data are aligned according to the unified time axis to form complete time series data. For the missing points in the time series, the linear interpolation method is used to fill the missing data to ensure the continuity of the time series. The time series is segmented using the sliding window method. The window length is set to 5 seconds and the sliding step is set to 1 second. In each time window, the average amplitude of the vibration signal, the temperature change gradient value and the change amplitude of the power consumption are extracted. These characteristic values ​​are combined into dynamic subsequences. Each dynamic subsequence has the time start and end information corresponding to the window. Through continuous segmentation processing of the sliding window, a dynamic subsequence set containing a complete time period is finally generated.

[0061] S112: Based on the dynamic subsequence set, compare the real-time data in each dynamic subsequence with the state data of the previous period, analyze the trend characteristics of each dynamic subsequence, and generate a dynamic operation state sequence set;

[0062] For each dynamic subsequence in the set, extract the real-time data and the status data of the previous period, and compare them item by item according to the extracted core feature values ​​(including the amplitude of the vibration signal, the temperature change gradient and the power consumption amplitude). Calculate the difference rate of each eigenvalue ,in, Indicates the feature value extracted in the current period, such as the amplitude of the vibration signal in the current period, which can be directly obtained by collecting it through the vibration sensor, or by extracting the peak value after bandpass filtering the original signal. For example, the peak value of the vibration signal data collected by a certain device is 10m / s²; It is the characteristic value of the previous period, which means the corresponding characteristic value extracted in the time window before the current period. It is called after storing the historical data. For example, the peak value of the vibration signal in the previous period is 8m / s². It is the absolute difference between the real-time characteristic value and the characteristic value of the previous period, indicating the change between the two periods, which is directly obtained through difference calculation. Indicates the current moment in the time series. It is a point on the time axis and is used to identify the data feature value of the current period. It is usually measured in seconds (s) or milliseconds (ms). For example, in a dynamic subsequence, the real-time data feature value of the vibration signal is 10m / s², the characteristic value of the previous period is 8m / s², substitute it into the formula to calculate the difference rate: , that is, the difference rate is 25%. Combined with the numerical range of the difference rate, classification and judgment are performed. For example, the difference rate between 0%-10% is defined as a low change interval, 10%-30% is a medium change interval, and greater than 30% is a high change interval. According to the above classification criteria, the current difference rate is 25%, which belongs to the medium change interval, indicating that the vibration signal changes significantly in the current period. According to the classification results, the trend features are classified, and the corresponding time range and source data are marked. The trend information of each dynamic subsequence is integrated to generate a dynamic operation status sequence set.

[0063] See also Figure 3 , the specific steps for obtaining the running status prediction results are:

[0064] S211: Use the piecewise regression model to predict the operating status of each time series in the dynamic operating status series set, using the formula:

[0065] ;

[0066] Calculate dimensionless predictions of equipment operating status ;

[0067] in, It is Eigenvalues The standardized value converts the eigenvalues ​​of different units into dimensionless values ​​of a unified scale, eliminating the influence of the dimension difference between the eigenvalues. This is the first time the device is running. characteristic values, such as the spindle vibration signal amplitude (m / s²), power consumption (W), etc., collected by equipment sensors. It is Eigenvalues The mean value is obtained by calculating the historical operation data. It is Eigenvalues The standard deviation is used to characterize the fluctuation range of the characteristic value and is obtained through calculation of historical operation data. Is an environment variable The standardized value is used to convert the external environmental variables into dimensionless values ​​to eliminate unit differences. It is The characteristic values ​​of external environmental variables, such as ambient temperature (℃), humidity (%), etc., are collected through environmental monitoring equipment. Is an environment variable The mean value is calculated through historical environmental data. Is an environment variable The standard deviation of is calculated through historical environmental data. It is Eigenvalues The weight coefficient is used to indicate the contribution ratio of the characteristic value to the comprehensive operating status. Indicates the index of a specific feature value, weight It is set by analyzing the importance of the influence of the characteristic value on the operating status of the equipment. For example, in the operation of a CNC machine tool, by analyzing the historical changes in vibration signals and power consumption, it is found that abnormal fluctuations in vibration signals often directly lead to failures of machine tool parts, while fluctuations in power consumption more reflect the state of load changes. Therefore, through correlation analysis, it is found that the contribution of vibration signals to failures is higher (such as 70%), while the contribution of power consumption is lower (such as 30%). Therefore, it can be set , ,in Indicates vibration signal, Indicates power consumption, Is an external environment variable The weight coefficient is used to reflect the impact of environmental factors on the operating status of the equipment. It is set by analyzing the importance of the impact of environmental variables on the operating status of the equipment. For example, when a CNC machine tool is running, the environmental humidity will affect the performance of electronic components. Excessive humidity may cause a short circuit in the control system, while temperature changes mainly affect the thermal expansion of the machine tool. Therefore, the impact of humidity on the equipment may be set to a higher weight. For example, through regression analysis, it is found that the correlation contribution of humidity is 60% and the contribution of temperature is 40%. Then the value is set , ,in Indicates temperature, Indicates humidity, It is the bias value of the segmented regression model, which is determined by fitting the offset of the historical operating status benchmark value and is used to adjust the range of the predicted value to make it consistent with the actual operating status score range. Indicates the total number of feature values ​​of the first feature group run by the device. Indicates the total number of feature values ​​for the second feature group run on the device.

[0068] Suppose we predict the comprehensive operation status of a CNC machine tool. The parameters obtained are: spindle vibration signal (m / s²), its historical average (m / s²), standard deviation (m / s²); real-time power consumption (W), its historical mean (W), standard deviation (W); ambient temperature (℃), its historical average (℃), standard deviation (℃); ambient humidity (%), its historical average (%), standard deviation (%); Set the weight of the eigenvalue , ; Environment variable weight , ; Bias value .

[0069] Standardized calculation: , , , ;

[0070] Enter the formula to calculate:

[0071] ;

[0072] ;

[0073] Predicted value A comprehensive score indicating the current operating status of the machine tool.

[0074] S212: Optimizing the dimensionless prediction value of the equipment operation status by dynamically adjusting the regression order of the segmented regression model to obtain an operation status prediction result;

[0075] Based on the calculated operating status prediction value , by dynamically adjusting the regression order of the segmented regression model, for the eigenvalue and external environment variables The specific operation includes resetting the number of feature points in the current segmented time series data. For example, when the initial regression order is 2, adding key feature points to expand the regression order to 3, introducing additional time series feature values ​​(such as bearing temperature changes during equipment operation) or external environmental variables (such as air flow velocity), normalizing the new feature values, and recalculating the weights of the newly added feature values ​​based on historical operation data. The least squares method is used to fit the adjusted model parameters to optimize the prediction value. The accuracy of the model is measured, and the optimized prediction value is verified in combination with the historical benchmark range. When the deviation between the prediction value and the benchmark value is within the allowable range (such as the deviation is less than 5%), the regression order and parameters of the current model are determined as optimal, and finally the optimized operation status prediction result is obtained.

[0076] See also Figure 4 ,The specific steps for obtaining the health status risk analysis results of manufacturing equipment are as follows:

[0077] S221: comparing the prediction parameters of the future time in the running state prediction result with the preset normal state range, judging whether there is an abnormal trend in the predicted state according to the comparison result, and obtaining an abnormal trend analysis result;

[0078] First, define the upper and lower limits of the normal state range. For example, based on historical data, the normal range of the spindle vibration signal is 6m / s² to 12m / s², and the normal range of power consumption is 90W to 130W. Compare the predicted value at each time point with this range. For example, at a certain time point, the predicted value of the spindle vibration signal is 13m / s², which exceeds the upper limit of the normal range of 12m / s². At this time, calculate the percentage of deviation from the normal range. If the value deviates from the normal range by more than 5%, it is recorded as moderate abnormality. The power consumption prediction value is 85W, which is lower than the lower limit of the normal range of 90W. At this time, the percentage of deviation from the normal range is calculated. , and is also recorded as a moderate abnormality. These comparison results are stored as an abnormal trend list, which includes the abnormal time period, abnormal characteristic value type and its deviation degree. For example, it is recorded as "the spindle vibration signal exceeds the upper limit by 8.33% at 13:00, which is a moderate abnormality; the power consumption is lower than the lower limit by 5.56% at 13:00, which is a moderate abnormality."

[0079] S222: Counting the fluctuation range and deviation degree of the characteristic value within the abnormal time period, marking the corresponding risk area and fault level in the abnormal trend analysis result according to the statistical result, and generating the manufacturing equipment health status risk analysis result;

[0080] According to the time period, characteristic value type and deviation degree recorded in the abnormal trend list, the abnormal trend is marked as the corresponding risk area. By counting the fluctuation range and deviation degree of the characteristic value in the abnormal time period, it is further refined in combination with the preset risk level standard. For example, a deviation of 5% to 10% from the normal range is marked as a mild risk, 10% to 20% is marked as a moderate risk, and more than 20% is marked as a severe risk. For example, when the deviation percentage of the spindle vibration signal at 13:00 is 8.33%, it corresponds to a mild risk level; when the deviation percentage of the power consumption at 13:30 is 5.56%, it corresponds to a mild risk level. , which also belongs to a mild risk. After matching these risk levels with abnormal time periods, they are further combined with the three-dimensional structure data of the equipment to locate the risk area. For example, the abnormal spindle vibration signal corresponds to the risk area of ​​the equipment spindle, and the abnormal power consumption corresponds to the risk area of ​​the equipment motor unit. The risk area is marked and color-coded. Green indicates a mild risk, yellow indicates a moderate risk, and red indicates a severe risk. For example, the abnormal area of ​​the spindle vibration signal is marked in green, and the abnormal area of ​​power consumption is marked in yellow. Finally, the manufacturing equipment health status risk analysis results including risk areas, risk levels and time periods are generated.

[0081] See also Figure 5 , the specific steps for obtaining the three-dimensional mapping results of manufacturing equipment deviation are:

[0082] S311: Mapping the manufacturing equipment component positions corresponding to the risk areas marked in the manufacturing equipment health status risk analysis results, matching the component information marked in the risk areas with the corresponding three-dimensional coordinate points, extracting all coordinate data including the risk area range, and obtaining the three-dimensional spatial data of the manufacturing equipment;

[0083] First, the physical spatial position of the equipment components is determined using the 3D CAD model in the equipment structure design. For example, the 3D coordinates of the spindle and bearing are: , the three-dimensional coordinates of the shell surface are , match the component information marked in the risk area with the corresponding three-dimensional coordinate points, extract all coordinate data including the risk area range, and further filter abnormal data points in combination with the time period records in the health status analysis results. For example, the main axis coordinate range of the vibration signal abnormal area is And combine these coordinate points with the risk level, for example, the points in the high-risk area are marked in red, and the spatial distribution data of the entire equipment is represented by a point cloud data set, and finally the complete three-dimensional spatial data of the manufacturing equipment is generated.

[0084] S312: Based on the three-dimensional spatial data of manufacturing equipment, the formula is used:

[0085] ;

[0086] Calculating state variables The gradient vector ;

[0087] in, It is a state variable, such as vibration signal amplitude (unit: m / s²), stress value (unit: Pa), and operating status data collected by the equipment sensor at a specific coordinate point. It is the coordinate value of the three-dimensional space, which represents the specific position coordinate of the device in the three-dimensional space. The unit is selected according to the measurement scale of the device, such as millimeter (mm). It is obtained through the three-dimensional structure model of the device and the calibration coordinate system. It represents the state variable exist , , The partial derivatives in three directions reflect the rate of change of the state variable along each direction of space. The partial derivatives are calculated by the built-in function gradient() in MATLAB. For example, the collected state variable data is input in matrix form, and the gradient() function is used to calculate the difference in the spatial coordinate direction to obtain the gradient value. Corresponding to , , The unit length of the direction is used to express the direction component of the gradient vector in the three-dimensional coordinate system.

[0088] For example, if you need to analyze the vibration signal gradient of a CNC machine tool spindle area, the three-dimensional space coordinates Range is mm, mm, mm. Vibration signal data at different positions in the spindle area have been collected , the parameters are: Position 1: ; Position 2: ; Position 3: ; Position 4: ; Coordinate step size: .

[0089] calculate : ;

[0090] calculate : ;

[0091] calculate : ;

[0092] Gradient vector: ;

[0093] Substitute the calculation results: ;

[0094] Gradient Vector Indicates: In direction, the rate of change of the vibration signal is the largest, with an increase of 0.3m / s² for every 1mm; and The rate of change in direction is small, increasing by 0.1m / s² and 0.05m / s² per 1mm respectively.

[0095] S313: According to the gradient vector of the state variable, the gradient change in the three-dimensional spatial data of the equipment is matched with the actual equipment structure position, and the deviation area and distribution of the equipment components are analyzed by setting the threshold range of the deviation distribution, and the deviation area is marked according to multiple categories to generate a three-dimensional mapping result of the manufacturing equipment deviation;

[0096] According to the gradient calculation results, the gradient changes in the three-dimensional spatial data of the equipment are matched with the actual equipment structure position. The threshold range of the deviation distribution is set during the specific analysis. For example, in the spindle area, the high deviation area is set as the gradient value exceeding 0.2m / s² / mm, the medium deviation area is set as the gradient value between 0.1m / s² / mm and 0.2m / s² / mm, and the low deviation area is set as the gradient value less than 0.1m / s² / mm. Combined with the gradient vector calculation results of the spindle area ,analyze The gradient value of the direction is 0.3m / s² / mm, which exceeds the high deviation threshold and is marked as a high deviation area; The directional gradient value is 0.1m / s² / mm, which belongs to the medium deviation area; The directional gradient value is 0.05m / s² / mm, which belongs to the low deviation area. Direction marks are high-skew distributions, The direction is marked as medium deviation distribution, The direction is marked as low deviation distribution, and each deviation area is mapped in combination with the specific three-dimensional spatial position of the main axis (x: 0-50, y: 0-100, z: 0-30). The gradient results of other components are further combined for distribution analysis. For example, in the bearing area (x: 50-70, y: 100-150, z: 10-20), the gradient vector result is ,in The directional gradient value is 0.15m / s² / mm, which belongs to the medium deviation area. and The directional gradient values ​​are 0.05m / s² / mm and 0.02m / s² / mm, both of which belong to the low deviation area. The direction is marked as medium deviation distribution, and The direction is marked as low deviation distribution. After the deviation distribution is marked, it is displayed in the 3D equipment model through color coding, such as marking the high deviation area in red, the medium deviation area in yellow, and the low deviation area in green. Finally, the 3D mapping result of the manufacturing equipment deviation is generated, which includes the specific location of each component, deviation type, gradient value size and distribution range.

[0097] See also Figure 6 , the specific steps to build a digital twin model of manufacturing equipment are:

[0098] S411: extracting behavior feature data of the manufacturing equipment when executing a specified process from the manufacturing equipment operation data, associating the behavior feature data with the three-dimensional spatial data of the manufacturing equipment in the three-dimensional mapping result of the manufacturing equipment deviation, and generating a structured data set of the behavior feature and the three-dimensional deviation data;

[0099] By marking and filtering the key data during the process operation, such as recording the vibration signal, temperature change and power consumption of the spindle, the data is segmented according to the time axis, and characteristic parameters are extracted in each process interval, such as the peak amplitude of the vibration signal, the frequency distribution range and the rate of change of power consumption. At the same time, the classification and structured storage of the behavioral characteristic data are completed by combining the external data of the operating environment such as temperature and humidity. The specific process is as follows: First, extract the corresponding three-dimensional spatial coordinates according to the high deviation area marked in the three-dimensional deviation mapping result of the manufacturing equipment. For example, the high deviation area range of the spindle is , and extract the data range related to the characteristic parameters of the spindle vibration signal behavior, such as the peak vibration signal at time The speed is recorded as 18m / s², and the corresponding coordinate position of the spindle in the three-dimensional space model is located, and the data at this time point is associated with the high deviation area; secondly, according to other process behavior characteristic data, for example, when the change rate of bearing power consumption is 5%, the corresponding coordinate is , the power change rate is associated with the deviation distribution of the bearing, and finally a complete correspondence between the behavior characteristic parameters and the three-dimensional deviation data is formed, the association between the behavior characteristic data and the three-dimensional mapping results of the manufacturing equipment deviation is completed, and a structured result containing time, spatial position and behavior characteristic data is generated.

[0100] S412: extracting dynamic change features in the structured data set of behavioral features and three-dimensional deviation data by partition, fitting dynamic curves, and superimposing fitting results to construct a digital twin model of the manufacturing equipment;

[0101] First, the correspondence between the behavioral feature data and the three-dimensional spatial data is partitioned. For example, the data in the spindle area and the corresponding three-dimensional deviation distribution are used as one partition, and the data in the bearing area and the corresponding three-dimensional deviation distribution are used as another partition. For the data in each partition, the dynamic change characteristics are extracted based on the time series, such as the peak change sequence of the vibration signal in the spindle area and the change sequence of the power consumption in the bearing area. These dynamic change characteristics are normalized to unify the dimensions, and then a parameterized curve (such as a Bezier curve or a spline curve) is selected for dynamic fitting. During the fitting process, the fitting parameters are optimized and the range of the fitting error is controlled by setting the accuracy requirements. The fitted dynamic curve is superimposed on the corresponding position of the three-dimensional equipment model. Combined with the three-dimensional coordinates and time series, a digital twin model of the manufacturing equipment containing dynamic change characteristics is constructed, and finally a visualization model that can reflect the linkage between the dynamic characteristics of the equipment operation and the deviation distribution is formed.

[0102] See also Figure 7 , the specific steps for obtaining the three-dimensional visualization management and control results of the manufacturing digital twin are:

[0103] S421: Based on the digital twin model of the manufacturing equipment, the dynamic behavior data of the manufacturing equipment in the specified process and the equipment deviation distribution data are uniformly mapped into a three-dimensional coordinate system, and the dynamic change characteristics of the key components and the deviation distribution linkage are dynamically displayed through time axis superposition to obtain a dynamically adjusted three-dimensional digital twin model;

[0104] After associating the dynamic behavior data of the manufacturing equipment in the specified process with the equipment deviation distribution data, the two types of data are uniformly mapped to the three-dimensional coordinate system in the digital twin model to determine the relationship between the deviation distribution and dynamic behavior characteristics of the key components of each process. For example, when displaying the dynamic changes of the vibration signal in the spindle area, the time series results of the vibration signal are associated with the three-dimensional spatial position of the spindle, and the time axis is used as the superposition parameter to display the change trend of the spindle vibration amplitude at different time points, and the overlap of the vibration signal peak and the deviation area is displayed through color coding. For the bearing area, the power consumption change rate is superimposed on the three-dimensional model of the bearing area, and the fluctuation of the power consumption rate is represented by the change of line width, and the relationship between the power consumption change rate and the deviation distribution is dynamically displayed. Through the above method, the dynamic change characteristics of the manufacturing equipment in the process execution and the deviation distribution are linked and displayed, and a three-dimensional digital twin model that can change over time is generated.

[0105] S422: Based on the dynamically adjusted three-dimensional digital twin model, the key dynamic characteristics and deviation distribution of the target area of ​​the equipment are dynamically displayed in combination with the time axis and spatial distribution. The layered display of data is achieved through color changes and area magnification, and the three-dimensional visualization management and control results of the manufacturing digital twin are generated;

[0106] Based on the constructed digital twin model, the key numerical changes of the dynamic characteristics of the equipment are presented in the form of two-dimensional projection, such as the vibration signal peak value in the main shaft area and the power consumption fluctuation rate in the bearing area. At the same time, the linkage between the equipment behavior characteristics and the deviation distribution is updated in real time in the three-dimensional scene by using dynamic curve trajectories and color changes. For example, the color gradient from green to red reflects the area with increased deviation, and the specific change process of the vibration signal or power fluctuation is displayed in combination with the time node of the equipment operation. In addition, in the three-dimensional scene, users are supported to choose to zoom in or out of a specific area. For example, after choosing to zoom in on the bearing area, only the dynamic change curve and power change distribution of the area are displayed, thereby realizing separate analysis of key components.

[0107] The manufacturing industry digital twin three-dimensional visualization management and control system is used to execute the above-mentioned manufacturing industry digital twin three-dimensional visualization management and control method, and the system includes:

[0108] The data collection module collects the operating data of the manufacturing equipment and converts it into a time series, analyzes the trend characteristics of each dynamic subsequence in the time series, and generates a dynamic operating state sequence set;

[0109] The state prediction module predicts the operation state of each time series in the dynamic operation state sequence set, obtains the operation state prediction result, compares the prediction parameters of the future time in the operation state prediction result with the preset normal state range, marks the risk area and fault level of abnormal trend according to the comparison result, and generates the health state risk analysis result of manufacturing equipment;

[0110] The health analysis module maps the manufacturing equipment component positions corresponding to the risk areas marked in the manufacturing equipment health status risk analysis results to obtain the three-dimensional spatial data of the manufacturing equipment, analyzes the gradient changes and deviation distribution in the three-dimensional spatial data of the manufacturing equipment, and obtains the three-dimensional mapping results of the manufacturing equipment deviation;

[0111] The digital twin module associates the behavioral characteristic data of manufacturing equipment with the three-dimensional spatial data of manufacturing equipment in the three-dimensional mapping results of manufacturing equipment deviations to construct a digital twin model of manufacturing equipment. By superimposing the linkage between the dynamic change characteristics of manufacturing equipment during process execution and the deviation distribution in the digital twin model of manufacturing equipment, a three-dimensional visualization management and control result of the digital twin of the manufacturing industry is generated.

[0112] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. A three-dimensional visualization management and control method for digital twins in manufacturing industry, characterized in that: The following steps are involved: S1: Collect the operating data of manufacturing equipment and convert it into a time series, analyze the trend characteristics of each dynamic subsequence in the time series, and generate a dynamic operating status sequence set; S2: predicting the operating state of each time series in the dynamic operating state sequence set to obtain an operating state prediction result, comparing the prediction parameters at future moments in the operating state prediction result with a preset normal state range, marking the risk area and fault level of the abnormal trend according to the comparison result, and generating a manufacturing equipment health state risk analysis result; S3: Mapping the manufacturing equipment component positions corresponding to the risk areas marked in the manufacturing equipment health status risk analysis results to obtain the three-dimensional spatial data of the manufacturing equipment, analyzing the gradient changes and deviation distribution in the three-dimensional spatial data of the manufacturing equipment to obtain the three-dimensional mapping results of the manufacturing equipment deviation; S4: Associating the behavioral characteristic data of manufacturing equipment with the three-dimensional spatial data of manufacturing equipment in the three-dimensional mapping result of manufacturing equipment deviation, constructing a digital twin model of manufacturing equipment, and generating a three-dimensional visualization management and control result of the digital twin of the manufacturing equipment by superimposing and displaying the linkage between the dynamic change characteristics of the manufacturing equipment during the process execution and the deviation distribution in the digital twin model of the manufacturing equipment.

2. The manufacturing industry digital twin three-dimensional visualization management and control method according to claim 1 is characterized in that: The steps for obtaining the dynamic running state sequence set are specifically as follows: S111: collecting manufacturing equipment operation data and converting it into a time series, dividing the time series of the manufacturing equipment operation data into multiple dynamic subsequences through a sliding window, and obtaining a dynamic subsequence set; S112: Based on the dynamic subsequence set, compare the real-time data in each dynamic subsequence with the state data of the previous period, analyze the trend characteristics of each dynamic subsequence, and generate a dynamic operation state sequence set.

3. The manufacturing industry digital twin three-dimensional visualization management and control method according to claim 2 is characterized in that: The steps for obtaining the running status prediction result are specifically as follows: S211: Use a piecewise regression model to predict the operating state of each time series in the dynamic operating state series set, using the formula: ； Calculate dimensionless predictions of equipment operating status ; in, It is Eigenvalues The standardized value of It is Eigenvalues The mean of It is Eigenvalues The standard deviation of It is External environment variables The standardized value of It is external environment variables, Is an environment variable The mean of Is an environment variable The standard deviation of It is Eigenvalues The weight coefficient of Is an external environment variable The weight coefficient of is the bias value of the piecewise regression model, Indicates the total number of feature values ​​of the first feature group run by the device. The total number of feature values ​​of the second feature group representing the operation of the device; S212: Optimizing the dimensionless prediction value of the equipment operation status by dynamically adjusting the regression order of the segmented regression model to obtain an operation status prediction result.

4. The manufacturing industry digital twin three-dimensional visualization management and control method according to claim 3 is characterized in that: The steps for obtaining the manufacturing equipment health status risk analysis results are specifically as follows: S221: comparing the prediction parameters at the future time in the operation state prediction result with the preset normal state range, judging whether there is an abnormal trend in the predicted state according to the comparison result, and obtaining an abnormal trend analysis result; S222: Count the fluctuation range and deviation degree of the characteristic value within the abnormal time period, mark the corresponding risk area and fault level in the abnormal trend analysis result according to the statistical result, and generate the manufacturing equipment health status risk analysis result.

5. The manufacturing industry digital twin three-dimensional visualization management and control method according to claim 4 is characterized in that: The steps for obtaining the three-dimensional mapping result of the manufacturing equipment deviation are specifically as follows: S311: Mapping the manufacturing equipment component positions corresponding to the risk areas marked in the manufacturing equipment health status risk analysis results, matching the component information marked in the risk areas with the corresponding three-dimensional coordinate points, extracting all coordinate data including the risk area range, and obtaining the three-dimensional spatial data of the manufacturing equipment; S312: Based on the three-dimensional spatial data of the manufacturing equipment, the formula is used: ； Calculating state variables The gradient vector ; in, is a state variable, is the coordinate value in three-dimensional space, It represents the state variable exist , , The partial derivatives in the three directions reflect the rate of change of the state variables along the spatial direction, Corresponding to , , The unit length of the direction represents the directional component of the gradient vector in the three-dimensional coordinate system; S313: According to the gradient vector of the state variable, the gradient change in the three-dimensional spatial data of the equipment is matched with the equipment structure position. By setting the threshold range of the deviation distribution, the deviation area and distribution of the equipment components are analyzed, and the deviation area is marked according to multiple categories to generate a three-dimensional mapping result of the manufacturing equipment deviation.

6. The manufacturing industry digital twin three-dimensional visualization management and control method according to claim 5 is characterized in that: The steps of constructing the digital twin model of manufacturing equipment are specifically as follows: S411: extracting behavior feature data of the manufacturing equipment when executing a specified process from the manufacturing equipment operation data, associating the behavior feature data with the three-dimensional spatial data of the manufacturing equipment in the three-dimensional mapping result of the manufacturing equipment deviation, and generating a structured data set of the behavior feature and the three-dimensional deviation data; S412: Extract the dynamic change features in the structured data set of the behavior features and three-dimensional deviation data by partition, fit the dynamic curve, and superimpose the fitting results to construct a digital twin model of the manufacturing equipment.

7. The manufacturing industry digital twin three-dimensional visualization management and control method according to claim 6 is characterized in that: The specific steps for obtaining the three-dimensional visualization management and control results of the manufacturing digital twin are as follows: S421: Based on the digital twin model of the manufacturing equipment, the dynamic behavior data of the manufacturing equipment in the specified process and the equipment deviation distribution data are uniformly mapped into a three-dimensional coordinate system, and the dynamic change characteristics of key components and the deviation distribution linkage are dynamically displayed by time axis superposition to obtain a dynamically adjusted three-dimensional digital twin model; S422: Based on the dynamically adjusted three-dimensional digital twin model, the key dynamic characteristics and deviation distribution of the target area of ​​the equipment are dynamically displayed in combination with the time axis and spatial distribution, and the data is displayed in layers through color changes and area magnification to generate three-dimensional visualization management and control results of the digital twin of the manufacturing industry.

8. The manufacturing industry digital twin three-dimensional visualization management and control system is characterized by: According to the manufacturing digital twin three-dimensional visualization management and control method according to any one of claims 1 to 7, the system comprises: The data collection module collects the operating data of the manufacturing equipment and converts it into a time series, analyzes the trend characteristics of each dynamic subsequence in the time series, and generates a dynamic operating state sequence set; The state prediction module predicts the operation state of each time series in the dynamic operation state sequence set to obtain the operation state prediction result, compares the prediction parameters at future moments in the operation state prediction result with the preset normal state range, marks the risk area and fault level of abnormal trend according to the comparison result, and generates the manufacturing equipment health state risk analysis result; The health analysis module maps the manufacturing equipment component positions corresponding to the risk areas marked in the manufacturing equipment health status risk analysis results to obtain the three-dimensional spatial data of the manufacturing equipment, analyzes the gradient changes and deviation distribution in the three-dimensional spatial data of the manufacturing equipment, and obtains the three-dimensional mapping results of the manufacturing equipment deviation; The digital twin module associates the behavioral characteristic data of manufacturing equipment with the three-dimensional spatial data of manufacturing equipment in the three-dimensional mapping results of manufacturing equipment deviations to construct a digital twin model of the manufacturing equipment. By superimposing the linkage between the dynamic change characteristics of the manufacturing equipment during the process execution and the deviation distribution in the digital twin model of the manufacturing equipment, a three-dimensional visualization management and control result of the digital twin of the manufacturing industry is generated.

Images