Real-time data-driven intelligent digital twin system for aerospace component docking

Abstract

This invention belongs to the field of intelligent manufacturing technology and discloses a real-time data-driven intelligent digital twin system for the docking of aerospace components. First, a standardized communication interface is established between multi-source sensors and various modules of the system to collect and transmit sensor data in real time. Second, based on the real-time multi-axis displacement data of the positioner, the real-time motion trajectory of the component is derived through kinematic modeling, completing the spatial pose parameter calculation of the component. Finally, the component pose parameters and multi-dimensional force sensing data are fused to construct a dynamic and interactive three-dimensional digital twin interface, realizing twin monitoring of the component's motion state and internal forces during the docking process. This system, based on the Unity3D engine, achieves real-time synchronous mapping of multi-dimensional force information and motion attitude parameters. Through a closed-loop processing flow of data acquisition, analysis, and visualization, it realizes quality assessment and quantitative analysis of the aerospace component assembly process, significantly improving the controllability of the docking process.

Description

Technical Field

[0001] This invention belongs to the field of intelligent manufacturing technology and relates to a real-time data-driven intelligent digital twin system for docking aerospace components. Background Technology

[0002] Aerospace component assembly is a complex engineering process in aerospace equipment manufacturing, accounting for over 40% of production costs and over 50% of total workload. Due to the large size and complex structure of aerospace components, coupled with the low correlation between measurement data and process flow, traditional manual trial-and-error and repeated debugging methods are insufficient for real-time monitoring of the docking process and accurate assessment of docking quality. Addressing these technical bottlenecks, the Unity 3D platform can be utilized. Leveraging its efficient real-time 3D rendering capabilities, multi-source data fusion interfaces, C# extensible scripting functionality, and rich plugin ecosystem, an intelligent digital twin system covering the entire aerospace component docking process can be constructed by applying multi-parameter collaborative technology driven by 3D digital models and real-time sensor data. This significantly improves the efficiency of assembly process monitoring and the final assembly quality.

[0003] The real-time digital twin system constructed by Guan Yuxiang et al. in their patent "A Real-Time Data-Driven FMS Digital Twin System" (CN116360354A) comprises a three-layer architecture: the physical entity layer collects equipment status data in real time through a PLC and sensor network; the digital twin layer parses the data and drives the synchronous movement of the 3D model; and the functional service layer integrates a visual interface and a virtual debugging module, enabling debugging personnel to locate control program defects without on-site inspection. However, this system only achieves a one-way mapping from state to model, failing to construct motion constraint relationships between components and lacking dynamic simulation capabilities for assembly paths. Consequently, the system cannot fully simulate the multi-structure collaborative motion and interaction during the assembly process.

[0004] Tang Dunbing et al. proposed a three-database fusion architecture based on digital twins in their patent "A Visual Real-Time Monitoring Method Based on Digital Twins" (CN 115439267A). This architecture constructs a digital resource pool based on a 3D geometric model library, a manufacturing attribute library, and a behavioral rule library. Combined with the physical layout of the workshop, a Unity 3D virtual scene is generated. A standardized communication interface is used to dynamically map operational data to the digital twin model, ultimately building a digital twin of the discrete manufacturing workshop and a real-time monitoring platform. While this system achieves digital control of equipment status and production tasks, the visualization interaction is limited to parameter panel display. It lacks real-time data-driven 3D dynamic simulation capabilities, resulting in the inability to visualize and reproduce the trajectory planning and motion process of assembly actions through the digital twin model.

[0005] Although the above two methods have achieved the application of digital twin technology, they have failed to effectively present the motion correlation characteristics between components and lack a model dynamic visualization mechanism driven by real-time sensor data. Therefore, they are difficult to meet the real-time twin monitoring requirements of the docking process of aerospace components. Summary of the Invention

[0006] This invention addresses the challenges of real-time monitoring of component motion status and the lag and lack of intuitive presentation of component attitude information during the docking and assembly of aerospace components. It establishes an intelligent digital twin system for aerospace component docking driven by real-time sensor data. The system achieves its functions through a three-layer architecture: First, a standardized communication interface is established between multi-source sensors and various modules of the system to collect and transmit sensor data in real time. Second, based on the multi-axis real-time displacement data of the positioner, the real-time motion trajectory of the component is derived through kinematic modeling, completing the spatial pose parameter calculation of the component. Finally, the component pose parameters and multi-dimensional force sensing data are fused to construct a dynamic and interactive three-dimensional digital twin interface, realizing twin monitoring of the component motion status and attitude adjustment internal forces during the docking process.

[0007] The technical solution of the present invention:

[0008] A real-time data-driven intelligent digital twin system for aircraft component docking is proposed. Based on the process requirements of large-scale aircraft component docking, it acquires physical and geometric parameter data in real time through multiple types of sensor interfaces at the docking site. Combined with attitude adjustment algorithms, it calculates the aircraft component's pose parameters in real time, driving the synchronous movement of the 3D model of the component in digital space, thus achieving real-time digital twin of the docked component's state information. The system consists of three main functional modules: a data acquisition module analyzes sensor communication protocols, designs standardized communication interfaces between sensors and system modules, and establishes an efficient data acquisition channel; a data analysis module uses the displacement sensor data stream and applies the SVD singular value decomposition algorithm to solve the pose transformation matrix, constructing an accurate component pose data analysis model; and a digital twin module updates the pose transformation matrix in real time, drives the movement of the virtual space component, and integrates force / displacement sensor data to generate a dynamic interface, achieving visualized monitoring of the docking process. The specific steps are as follows:

[0009] Step 1: Building the data acquisition module;

[0010] First, the force sensor is installed and fixed to the support part of the positioner to ensure that it accurately measures the force on the aerospace components during the positioner's movement. Then, the force sensor is connected to a signal amplifier, and the mechanical signal converted by the signal amplifier is input to the programmable logic controller (PLC). At the same time, the position signal of the positioner's servo motor is input to the PLC, and the PLC is connected to the attitude adjustment system's industrial control computer. Communication is then established between the attitude adjustment system's industrial control computer and the data acquisition industrial control computer. Finally, the force and position information is acquired and processed in real time by the data acquisition industrial control computer.

[0011] The attitude adjustment and docking process of aerospace components requires the use of multiple types of sensors for control and measurement. Therefore, a timestamp synchronization mechanism for these multiple sensor types is designed to achieve data time alignment. The timestamp analysis of the multiple sensor types is as follows:

[0012] τ1[n]=nT1+δ1,τ2[m]=mT2+δ2,… (1)

[0013] Where τ1[n] and τ2[m] represent the timestamps of the nth sample of the first sensor and the mth sample of the second sensor, respectively; T1 and T2 represent the sampling periods of the first and second sensors, respectively; δ1 and δ2 are their respective system delay compensation values, and so on.

[0014] To ensure data consistency, a data synchronization strategy is used for every two sensors. A time interpolation algorithm is used to calculate the value of the second sensor corresponding to the sampling time of the first sensor, thus achieving data matching in the time domain. The expression is as follows:

[0015]

[0016] Among them, F2 * (τ1[n]) is the estimated value of the second sensor at the nth sampling time of the first sensor after time interpolation; F2[m+1] and F2[m] are the (m+1)th and mth data of the second sensor, respectively; τ2[n] is the timestamp of the second sensor closest to τ1[m].

[0017] After time alignment, the data from multiple types of sensors are kept consistent in the time dimension, achieving synchronization. Subsequently, the multi-source sensing data collected by the aligned multiple types of sensors are transmitted to the pose data parsing module and the digital twin module through the OPC UA protocol.

[0018] Step 2: Construction of the pose data parsing module;

[0019] During attitude adjustment, the ball head and the positioner are rigidly connected by bolts and move as a whole rigid body. The attitude of the aerospace component is adjusted synchronously with the change of the ball head position. In order to obtain the attitude information of the aerospace component in real time and accurately, a attitude data analysis module is constructed. This attitude data analysis module performs real-time attitude calculation by processing the multi-source sensor data collected in step 1.

[0020] First, the pose data parsing module defines the pose of the aerospace components:

[0021]

[0022] In this context, it is assumed that an aircraft component has N locators, where X, Y, Z are the position data of the aircraft component, and α, β, γ are the attitude data of the aircraft component. i ,y i ,z i Data from displacement sensors in each direction for each positioner For data parsing algorithms;

[0023] The Unity coordinate system is a left-handed coordinate system. The X-axis represents the horizontal direction and points to the right, the Y-axis represents the vertical direction and points upward, and the Z-axis represents the depth direction and points forward. The global coordinate system is a right-handed coordinate system. The X-axis represents the horizontal direction and points to the left or right, the Y-axis represents the depth direction and points forward or backward, and the Z-axis represents the vertical direction and points upward. Because the Unity coordinate system and the global coordinate system are different, a coordinate transformation is required to convert the position data in the Unity coordinate system to the global coordinate system. A transformation matrix conv is defined from the global coordinate system to the Unity coordinate system. This transformation matrix conv is initially a 3x3 identity matrix. When the coordinate axes of the Unity coordinate system and the global coordinate system are swapped, the corresponding two columns of the transformation matrix are swapped. When the coordinate axes of the Unity coordinate system and the global coordinate system are opposite, the corresponding columns of the transformation matrix are inverted.

[0024] Therefore, in the global coordinate system, the position coordinates of the aerospace component are:

[0025] [X set Y set Z set ] = conv -1 ·[X U Y U Z U ] T (4)

[0026] Among them, [X U Y U Z U [X] represents the position data of the aerospace component in the Unity coordinate system. set Yset Z set [This refers to the position data of the aircraft components in the global coordinate system;]

[0027] Therefore, the coordinates of the center of the locator's ball head in the global coordinate system are:

[0028] [X target Y target Z target ] = [X orginal Y orginal Z orginal ]+[ΔX, ΔY, ΔZ] (5)

[0029] [X target Y target Z target [X] represents the coordinates of the ball's head and center during the motion. original Y original Z original [ΔX, ΔY, ΔZ] represents the coordinates of the ball head and center in the initial state after homing, and [ΔX, ΔY, ΔZ] represents the displacement sensor data.

[0030] Therefore, the coordinates of the center of the ball on the locator are:

[0031]

[0032] Where A is the initial coordinate of the ball's center, B is the coordinate of the ball's center during the motion, [X original1 Y original1 Z original1 ] T Let [X] be the coordinates of the center of the ball on the first locator in the initial state after returning to zero. target1 Y target1 Z target1 ] T The coordinates of the center of the ball on the first locator during the movement; and so on;

[0033] Next, we will solve the attitude problem of the aerospace components. First, we will calculate the centroid and decentralize it.

[0034] Centroid calculation is performed for each point cloud, which consists of matrix A and matrix B, i.e.:

[0035]

[0036] Among them, A cen [i] is the centroid of matrix A in the i-th direction, B cen [i] is the centroid of matrix B in the i-th direction, A[i,j] is the data in the i-th row and j-th column of matrix A, and B[i,j] is the data in the i-th row and j-th column of matrix B, where i = x, y, z. Construct decentralized data:

[0037]

[0038] in, and Let A and B be the decentralized matrices, and e be a vector of all 1s with a length equal to the number of columns in matrix A.

[0039] Furthermore, in order to calculate the rotation matrix of the aerospace component, the covariance matrix is ​​constructed:

[0040]

[0041] Perform singular value decomposition on H, i.e.:

[0042] H=U∑V T (11)

[0043] Where U is the left singular vector matrix; Σ is a diagonal matrix with diagonal elements consisting of non-negative, descending singular values; and V is the right singular vector matrix.

[0044] To prevent reflection, the correction matrix D for removing reflection is constructed as follows:

[0045] in

[0046] The final rotation matrix R of the aerospace component is:

[0047] R = VDU T (13)

[0048] Therefore, the rotation matrix R2 of the aerospace component in the Unity coordinate system is:

[0049] R2 = conv · R· conv T (14)

[0050] The position information of an aerospace component in the Unity coordinate system consists of vector rotation and offset. Let P be the initial centroid coordinates of the centers of mass on all N locators for an aerospace component. A This calculation is performed in the Unity coordinate system, and A needs to be... cen Perform coordinate system transformation:

[0051] P A =conv·A cen (15)

[0052] The vector used for correction is:

[0053] V pos =P original - P A (16)

[0054] Among them, P original Given the initial position coordinates of the aerospace component, the corresponding rotation quaternion q is calculated using the rotation to update the displacement R2, and then the vector V is converted using Unity's built-in functions. pos After rotating, we get:

[0055]

[0056] The standard mathematical expression for the rotation quaternion q is:

[0057] q=w+xi+yj+zk (18)

[0058] When used as a rotation quaternion, it must be a unit quaternion, i.e., a unit norm.

[0059]

[0060] The Unity development platform provides a rotated quaternion structure for vector V. pos When applying quaternion rotations, you can call:

[0061] V rotation =q*V pos (20)

[0062] Equivalent to

[0063] V rotation =q*(0,V) pos )*q -1 (twenty one)

[0064] Thus, the rotated vector V is obtained. rotation ;

[0065] The final position p of the aerospace component in the Unity coordinate system is:

[0066]

[0067] Step 3: Digital Twin Module Construction;

[0068] Based on steps 1 and 2, force-position data twins and aerospace component motion twins are generated on the Unity development platform. The force-position data twin is based on the multi-source sensor data obtained in step 1, including displacement sensor data and force sensor data, and uses the Xchart plugin to monitor the force-position information during the docking process of the aerospace components. The aerospace component motion twin is based on the pose information of the aerospace components obtained in step 2, including the rotation matrix R2 and the position vector p, to drive the digital 3D model, and finally realize the intuitive display of motion information during the docking of aerospace components.

[0069] The beneficial effects of this invention are as follows: The constructed real-time sensor data-driven digital twin system for aerospace component docking acquires displacement and force sensing signals during the assembly process in real time, and calculates the three-dimensional pose transformation matrix of the aerospace components in real time based on the singular value decomposition (SVD) algorithm, ensuring high-precision dynamic mapping between physical space data and digital models. By mapping the pose transformation matrix to the digital twin environment in real time, a digital twin model of the assembly's motion state is established, enabling full-dimensional dynamic monitoring of the assembly process and effectively replacing traditional manual inspection methods. Based on the Unity3D engine, this system achieves real-time synchronous mapping of multi-dimensional force information and motion attitude parameters. Through a closed-loop processing flow of data acquisition, analysis, and visualization, it realizes quality assessment and quantitative analysis of the aerospace component assembly process, significantly improving the controllability of the docking process. Attached Figure Description

[0070] Figure 1 The diagram shows the layout of the docking hardware system.

[0071] Figure 2 The diagram shown is a system flowchart of the present invention.

[0072] Figure 3 The image shown is a schematic diagram of a digital twin interface.

[0073] In the diagram: 1-positioner, 2-force sensor, 3-signal amplifier, 4-programmable logic controller, 5-attitude adjustment system industrial computer, 6-data acquisition industrial computer. Detailed Implementation

[0074] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings and technical solutions.

[0075] Taking the fuselage and wing docking assembly of a large civil airliner as an example, the force sensor and displacement sensor of the positioner are used in the attitude adjustment docking process of the sections.

[0076] (1) Construction of data acquisition module

[0077] The aircraft wing-body attitude adjustment and docking process requires the use of force sensors and displacement sensors for control and measurement. Due to the different sampling frequencies of the two sensors, timestamp misalignment can occur. Therefore, a sensor timestamp synchronization mechanism needs to be designed to achieve data time alignment, thereby avoiding information deviation and data mismatch. The sensor timestamp analysis is as follows:

[0078] τ f [n] = nT f +δ f ,τ d [m]=nT d +δ d (twenty three)

[0079] Where, τf [n] and τ d [m] represents the timestamp of the nth sample from the force sensor and the timestamp of the mth sample from the displacement sensor, respectively. f and T d δ represents the sampling period of the force sensor and the displacement sensor, respectively. f and δ d These are the delay compensation values ​​for their respective systems.

[0080] To ensure data consistency, a data synchronization strategy is required. This involves using a time interpolation algorithm to calculate the displacement value corresponding to the force sensor's sampling time, thus achieving data matching in the time domain. The expression for this is:

[0081]

[0082] D * (τ f [n]) represents the time interpolation result at the nth sampling time τ of the force sensor. f [n] corresponds to the displacement sensor estimate, D[m+1] and D[m] are the (m+1)th and mth data points of the displacement sensor, respectively, τ d [m] represents the distance τ f [n] The most recent displacement sensor timestamp enables time-domain matching of force and potential data. The timestamped sensor data is then transmitted to the data parsing and visualization modules via the OPC UA protocol.

[0083] In the experiment, the force sensor took two consecutive samples at 0.05 seconds and 0.06 seconds. The displacement sensor's sample value was 9.2 at 0.03 seconds, 10 at 0.045 seconds, and 10.5 at 0.06 seconds. Therefore, the estimated value of the displacement sensor at 0.5 seconds is:

[0084]

[0085] (2) Data parsing module construction

[0086] During the aircraft wing-body attitude adjustment and docking process, the ball joint and the component are rigidly connected by bolts, and they move together as a rigid body during the adjustment process. The attitude of the wing-body component adjusts synchronously with the change in the position of the ball joint. To obtain the real-time attitude of the aircraft wing-body component during the attitude adjustment process, a pose analysis module is constructed to calculate the displacement sensor data extracted above. The data analysis module is as follows:

[0087] Because the Unity coordinate system differs from the global coordinate system, a coordinate transformation is required to convert the data in the Unity coordinate system to the global coordinate system. This is done using the transformation matrix `conv`, where the Unity coordinate system X-axis is the inverse of the global coordinate system X-axis, the Unity coordinate system Y-axis is the inverse of the global coordinate system Z-axis, and the Unity coordinate system Z-axis is the inverse of the global coordinate system Y-axis. Therefore, the transformation matrix is:

[0088]

[0089] In the global coordinate system, the position coordinates are:

[0090] [X global ,Y global Z global ] = conv -1 ·[X unity ,Y unity Z unity ] T =[-X unity ,-Z unity ,Y unity (27)

[0091] Among them, [X unity ,Y unity Z unity [X] represents the position data in the Unity coordinate system. global ,Y global Z global [] represents the position data in the global coordinate system.

[0092] Therefore, the coordinates of the center of the locator ball head in the global coordinate system are obtained as follows:

[0093] [X target ,Y target Z target ] = [X orginal ,Y orginal Z orginal ]+[ΔX,ΔY,ΔZ] (28)

[0094] [X target ,Y target Z target [X] represents the coordinates of the ball's head and center during the motion. original ,Y original Z original [ΔX, ΔY, ΔZ] represents the coordinates of the ball's center in the initial state after homing, and [ΔX, ΔY, ΔZ] represents the displacement sensor data, i.e., the offset of the ball's center coordinates in the global coordinate system.

[0095] Therefore, the coordinates of the center of the ball head on the fuselage component locator in the experiment are:

[0096]

[0097] Among them, A fuselage Let B be the coordinate matrix of the center of the sphere on the fuselage in the initial state. fuselage Let [X] be the coordinate matrix of the center of gravity of the ball on the fuselage during the motion. original1 ,Y original1 Z original1 ] T For the initial zeroing state, the coordinates of the ball center on the fuselage positioner 1 are [X] target1 ,Y target1 Z target1 ] T The coordinates of the ball head center on the fuselage positioner 1 during the movement are given, and so on.

[0098] Next, we will solve the fuselage attitude problem, starting with the calculation and decentralization of the center of mass.

[0099] Calculate the centroid for each point cloud (matrices A and B), i.e.:

[0100]

[0101] Where A fuselagecen [i] represents the centroid of the aircraft fuselage component in the i-th direction, B fuselagecen [i] represents matrix B fuselage Given the centroid along the direction i, i = (x, y, z), we get:

[0102] A fuselagecen =[0,0,1000],B fuselagecen =[-100,-20,1100] (32)

[0103] Construct a decentralized matrix:

[0104]

[0105] in and Let A be the decentralized matrix. fuselage Sum matrix B fuselage , where e is a vector of length 4 consisting entirely of 1s.

[0106] To calculate the rotation matrix of the computer body components, construct the covariance matrix:

[0107]

[0108] Perform singular value decomposition (SVD) on H, that is:

[0109]

[0110] Where U is the left singular vector matrix, Σ is the diagonal matrix with diagonal elements consisting of non-negative, descending singular values, and V is the right singular vector matrix.

[0111] To prevent reflection, the correction matrix D for removing reflection is constructed as follows:

[0112] in

[0113] The final obtained fuselage rotation matrix is:

[0114]

[0115] Therefore, the fuselage rotation matrix in the Unity coordinate system is:

[0116] R fuselage2 =conv ·R fuselage conv T (38)

[0117] The position information of the fuselage components in the Unity coordinate system consists of vector rotation plus offset. The initial centroid position coordinates of the spheres on all locators under the fuselage are defined as P. fuselageA This calculation is performed in the Unity coordinate system, with A... fuselagecen Perform coordinate system transformation

[0118] P fuselageA =conv·A fuselagecen (39)

[0119] The vector used for correction is

[0120] V fuselage =P fuselage -P fuselageA =[0,0,100] T (40)

[0121] Where P fuselage Let R be the initial position coordinates of the fuselage component. Then, update the displacement using rotation, starting from R. fuselage2 Calculate the corresponding rotation quaternion q fuselage And use Unity's built-in functions to transform vector V fuselage Rotate to obtain

[0122]

[0123] The final position P of the fuselage in the Unity coordinate system fuselage for:

[0124]

[0125] The measured fuselage position is [97.13015, 1095.93024, 119.88032] and the rotation angle is [-1.71887, 1.71887, 2.29183], both of which are consistent with the calculated results.

[0126] For the wing component, the only difference from the fuselage component is that the number of locators under the wing component is 3. Only the size of the matrix and vector needs to be changed. The rest of the process is exactly the same and will not be repeated.

[0127] The data parsing module is built based on displacement sensor data of the locator assembled during the aircraft wing-body docking, and its output provides data support for the subsequent digital twin interface.

[0128] 1. Construction of Digital Twin Module

[0129] Based on the force-position data and the pose data of the aircraft wings and fuselage mentioned above, force-position data twins are completed using the Xchart plugin in the Unity development platform. This drives the three-dimensional digital models of the fuselage and wings to complete the motion twins of the components, intuitively displaying the docking status information of the aircraft wing and fuselage components.

[0130] The specific embodiments of the present invention have been described in detail above. The present invention proposes a real-time data-driven intelligent digital twin system for the docking of aerospace components. By real-time acquisition of the output signals from displacement and force sensors during assembly, and based on the singular value decomposition (SVD) algorithm, the system calculates the three-dimensional pose transformation matrix of the aerospace components, establishing a high-precision dynamic mapping relationship between physical space data and the digital model, thereby achieving real-time tracking of the component assembly motion state. The system uses the Unity 3D platform to construct a digital mapping model of force and position information and motion attitude. Through a closed-loop processing flow of data acquisition, analysis, and twinning, it completes real-time monitoring of the entire docking process of large aerospace components and provides visualization evaluation and analysis functions for assembly quality, effectively improving the safety and quality control accuracy of the docking process.

Claims

1. A real-time data-driven intelligent digital twin system for docking aerospace components, characterized in that, The steps are as follows: Step 1: Building the data acquisition module; The force sensor is fixed to the support part of the positioner and connected to the signal amplifier. The force signal converted by the signal amplifier and the position signal of the positioner servo motor are connected to the programmable logic controller (PLC). The PLC is connected to the industrial control computer of the attitude adjustment system to establish communication between the industrial control computer of the attitude adjustment system and the industrial control computer of the data acquisition system. The force and position information is collected and processed in real time through the industrial control computer of the data acquisition system. Step 2: Construction of the pose data parsing module; During attitude adjustment, the ball head is connected to the positioner, and the attitude of the aerospace component is adjusted synchronously with the change in the position of the ball head; a pose data analysis module is constructed, which performs real-time pose calculation by processing the multi-source sensor data collected in step 1; First, the pose data parsing module defines the pose of the aerospace components; The Unity coordinate system is a left-handed coordinate system, and the global coordinate system is a right-handed coordinate system. Therefore, the position coordinates of the aircraft component in the global coordinate system can be obtained, and then the coordinates of the center of the locator ball head in the global coordinate system can be obtained. Next, the attitude of the aerospace components is solved, the centroid is calculated and decentralized, and the covariance matrix is ​​constructed. Perform singular value decomposition on 𝐻; Construct the correction matrix for removing reflections D for: (12) Find the rotation matrix of the aerospace component R for: (13) Rotation matrix of aerospace components in Unity coordinate system R 2 for: (14) The position information of an aerospace component in the Unity coordinate system consists of vector rotation and offset. This defines all positions of an aerospace component. N The initial centroid coordinates of the spheres on each locator are: P A This calculation is performed in the Unity coordinate system and requires... A cen Perform coordinate system transformation: (15) The vector used for correction is: (16) in, P original Given the initial position coordinates of the aerospace component, the displacement is updated using rotation. R 2 Calculate the corresponding rotation quaternion q And use Unity's built-in functions to convert vectors V pos After rotating, we get: (17) Among them, rotation quaternions q The standard mathematical expression is: (18) When used as a rotation quaternion, it must be a unit quaternion, i.e., a unit norm. (19) The Unity development platform provides a rotated quaternion structure for vectors. V pos When applying quaternion rotations, you can call: (20) Equivalent to (21) Thus, the rotated vector is obtained. V rotation ; Final position of aircraft components in Unity coordinate system p for: (22) ； Step 3: Digital Twin Module Construction; Force-position data twins are generated based on the multi-source sensor data acquired in step 1, including displacement sensor data and force sensor data, and use the Xchart plugin to monitor force-position information during the docking process of aerospace components; motion twins of aerospace components are generated based on the pose information of aerospace components acquired in step 2, including rotation matrices. R 2 and position vector p It drives digital 3D models to intuitively display motion information during the docking of aerospace components.

2. The real-time data-driven intelligent digital twin system for docking aviation components according to claim 1, characterized in that, In step 1: The attitude adjustment and docking process of aerospace components requires the use of multiple types of sensors for control and measurement. Therefore, a timestamp synchronization mechanism for these multiple sensor types is designed to achieve data time alignment. The timestamp analysis of the multiple sensor types is as follows: (1) in, [ n ]and [ m ] respectively represent the first sensor and the second sensor. n The timestamp of the first sample and the second sensor's... m Each sampling timestamp T 1 and T 2 These represent the sampling periods of the first and second sensors, respectively. and Each of these represents a system latency compensation value, and so on. To ensure data consistency, a data synchronization strategy is used for every two sensors. A time interpolation algorithm is used to calculate the value of the second sensor corresponding to the sampling time of the first sensor, thus achieving data matching in the time domain. The expression is as follows: (2) in, After time interpolation, in the first sensor... n Each sampling time The corresponding second sensor estimate; F 2[ m +1] and F 2[ m [The second sensor's] number is [the number of]... m +1 and the m One data point; Distance The most recent second sensor timestamp; After time alignment, the data from multiple types of sensors remain consistent in the time dimension, achieving synchronization. Subsequently, the multi-source sensing data collected by the aligned multiple types of sensors is transmitted to the pose data parsing module and the digital twin module via the OPC UA protocol.

3. The real-time data-driven intelligent digital twin system for docking aviation components according to claim 1, characterized in that, The pose data parsing module defines the pose of aerospace components as follows: (3) In this case, assuming an aircraft component has N One locator, X , Y , Z For aircraft component location data, For attitude data of aerospace components, x i , y i , z i Data from displacement sensors in each direction for each positioner This is a data parsing algorithm; in the global coordinate system, the position coordinates of the aerospace component are: (4) in,[ X U , Y U , Z U [This refers to the position data of the aircraft component in the Unity coordinate system.] X set , Y set , Z set [This refers to the position data of the aircraft component in the global coordinate system; therefore, the coordinates of the center of the locator ball head in the global coordinate system are:] (5) [ X target , Y target , Z target [ represents the coordinates of the ball's head and center during the motion.] X original , Y original , Z original [This represents the coordinates of the ball's center in the initial state after returning to zero.] For displacement sensor data; The coordinates of the center of the ball on the locator are: (6) (7) in, A Let these be the coordinates of the ball's center in the initial state. B Let [the coordinates of the ball's center during the motion] be [the coordinates of the ball's head]. X original1 , Y original1 , Z original1 ] T The coordinates of the center of the ball on the first locator in the initial state of returning to zero are: X target1 , Y target1 , Z target1 ] T The coordinates of the center of the ball on the first locator during the movement; and so on; Constructing the covariance matrix for: (10) The singular value decomposition is: (11) in, U It is a left singular vector matrix; Σ It is a diagonal matrix, and the diagonal elements are non-negative, descending singular values; V It is a right singular vector matrix; First, centroid calculation and decentralization are performed; Centroid calculation is performed for each point cloud element, which includes a matrix. A sum matrix B， Right now: (8) in, A cen [ i ] is a matrix A of i Directional center of mass, B cen [ i ] is a matrix B of i Directional center of mass, A [ i,j ] is a matrix A The first in i Line number j Column data, B [ i,j ] is a matrix B The first in i Line number j Column data, i = x , y , z Constructing decentralized data: (9) in, and For the decentralized matrix A sum matrix B , e It is a vector consisting entirely of 1s, and its length is a matrix. A The number of columns.

Images