CN-120469365-B - Real-time data-driven aviation component docking intelligent digital twin system

CN120469365BCN 120469365 BCN120469365 BCN 120469365BCN-120469365-B

Abstract

The invention belongs to the technical field of intelligent manufacturing, and discloses a real-time data-driven intelligent digital twin system for docking aviation components. Firstly, establishing a standardized communication interface between a multisource sensor and each module of a system, acquiring and synchronously transmitting sensing data in real time, secondly, deducing a real-time motion track of a component through kinematic modeling based on multi-axis real-time displacement data of a positioner to complete the calculation processing of spatial pose parameters of the component, and finally, fusing the pose parameters of the component and multidimensional force sensing data to construct a dynamic interactable three-dimensional digital twin interface so as to realize twin monitoring of the motion state of the component and the force in the pose adjustment of the component in the butt joint process. The system realizes real-time synchronous mapping of multidimensional force information and motion attitude parameters based on the Unity3D engine, realizes quality evaluation and quantitative analysis of the assembly process of the aviation components through closed-loop processing flows of data acquisition, analysis and visualization, and remarkably improves the controllability of the butt joint process.

Inventors

ZHANG YANG
WU HAOYU
LIU RUNZE
XU XIAO
LU YONGKANG
LIU WEI
LU GU

Assignees

大连理工大学

Dates

Publication Date: 20260505
Application Date: 20250509

Claims (3)

1. The real-time data driven aviation component docking intelligent digital twin system is characterized by comprising the following steps: step 1, constructing a data acquisition module; the force sensor is fixed at the supporting part of the positioner, the force sensor is connected to the signal amplifier, and the mechanical signal converted by the signal amplifier and the position signal of the servo motor of the positioner are connected to the programmable logic controller which is connected with the industrial personal computer of the gesture adjusting system to establish communication between the industrial personal computer of the gesture adjusting system and the industrial personal computer of the data acquisition; Step2, constructing a pose data analysis module; In the process of adjusting the pose, the ball head is connected with a positioner, and the pose of the aviation component is synchronously adjusted along with the position change of the ball head; firstly, a pose data analysis module defines the pose of an aviation component; the Unity coordinate system is a left-hand coordinate system, and the global coordinate system is a right-hand coordinate system, so that the position coordinates of the aviation component in the global coordinate system are obtained, and the spherical center coordinates of the positioner ball head in the global coordinate system are obtained; Then carrying out attitude solving on the aviation component, carrying out centroid calculation and decentration, and constructing a covariance matrix Singular value decomposition is carried out on H; Construction to remove reflections the correction matrix D is: (12) The rotation matrix R of the aircraft component is determined as: (13) The rotation matrix R 2 of the aerospace component under the Unity coordinate system is: (14) the position information of the aviation component in the Unity coordinate system is formed by vector rotation and offset, the initial centroid position coordinates of the centers of spheres on all N positioners of the aviation component are defined as P A , the calculation is performed in the Unity coordinate system, and the A cen is required to be subjected to coordinate system conversion: (15) The vectors for correction are: (16) Wherein, P original is the initial position coordinate of the aviation component, the corresponding rotation quaternion q is calculated by using the rotation update displacement R 2 , and the vector V pos is rotated by using the Unity built-in function, so as to obtain: (17) The rotation quaternion q standard mathematical expression is: (18) When a rotation quaternion is used, it is required to be a unit quaternion, that is, a unit norm: (19) Providing a rotation quaternion structure in the Unity development platform, when quaternion rotation is applied to vector V pos , it can be invoked: (20) Equivalent to (21) Thereby obtaining a rotated vector V rotation ; The final position p of the aviation component in the Unity coordinate system is: (22) ; step 3, constructing a digital twin module; the force position data twinning is based on the multisource sensing data acquired in the step 1 and comprises displacement sensor data and force sensor data, xchart plug-in units are used for monitoring force position information in the docking process of the aviation components, and the aviation component motion twinning is based on the aviation component pose information acquired in the step 2 and comprises a rotation matrix R 2 and a position vector p, so that a digital three-dimensional model is driven, and visual display of motion information in the docking process of the aviation components is realized.
2. The real time data driven aerospace component docking intelligent digital twinning system of claim 1, wherein in step 1: The attitude-adjusting docking process of the aviation component needs to apply a multi-type sensor to control and measure, and therefore, a multi-type sensor timestamp synchronization mechanism is designed to realize data time alignment, and the timestamp analysis of the multi-type sensor is firstly as follows: (1) Wherein, the [ N ] and the method [ M ] represents the time stamp of the nth sample of the 1 st sensor and the time stamp of the mth sample of the 2 nd sensor, respectively, T 1 and T 2 represent the sampling periods of the 1 st sensor and the 2 nd sensor, respectively, And Compensating values for respective system delays, and so on; In order to ensure data consistency, a data synchronization strategy is used for every two sensors, and the numerical value of the 2 nd sensor corresponding to the 1 st sensor sampling time is calculated through a time interpolation algorithm, so that the matching of the data in the time domain is achieved, and the expression is as follows: (2) Wherein, the After time interpolation, at the 1 st sensor nth sampling time F 2 [ m+1] and F 2 [ m ] are the m+1 and m data of the 2 nd sensor respectively; Is at a distance of The timestamp of the nearest 2 nd sensor; and then, transmitting the multisource sensing data acquired by the aligned multisensor to the pose data analysis module and the digital twin module through an OPC UA protocol.
3. The real-time data driven aviation component docking intelligent digital twin system of claim 1, wherein the pose data parsing module defines the pose of the aviation component as: (3) Wherein, N positioners are arranged under one aviation component, X, Y and Z are aviation component position data, For the attitude data of the aircraft component, x i , y i , z i is the displacement sensor data of each direction in each positioner, In the global coordinate system, the position coordinates of the aviation component are as follows: (4) wherein [ X U , Y U , Z U ] is the position data of the aviation component in the Unity coordinate system, and [ X set , Y set , Z set ] is the position data of the aviation component in the global coordinate system, so the spherical center coordinates of the locator ball head in the global coordinate system are as follows: (5) [ X target , Y target , Z target ] is the spherical center coordinate of the ball head in the motion process, [ X original , Y original , Z original ] is the spherical center coordinate of the ball head in the zero return initial state, Is displacement sensor data; The coordinates of the sphere center of the bulb on the locator are as follows: (6) (7) Wherein A is the center of sphere coordinate of the ball head in the initial state, B is the center of sphere coordinate of the ball head in the motion process, [ X original1 , Y original1 , Z original1 ] T is the center of sphere coordinate of the ball head on the first positioner in the return-to-zero initial state, [ X target1 , Y target1 , Z target1 ] T is the center of sphere coordinate of the ball head on the first positioner in the motion process, and so on; constructing covariance matrix The method comprises the following steps: (10) Singular value decomposition is: (11) wherein U is a left singular vector matrix, sigma is a diagonal matrix, diagonal elements are singular values arranged in a non-negative descending order, V is a right singular vector matrix; Firstly, calculating a mass center and decentering; centroid calculation is performed on each point cloud, wherein the point cloud comprises a matrix A and a matrix B, namely: (8) Wherein A cen [ i ] is the i-direction centroid of matrix A, B cen [ i ] is the i-direction centroid of matrix B, A [ i, j ] is the ith row and jth column data in matrix A, B [ i, j ] is the ith row and jth column data in matrix B, i=x, y, z, and the de-centering data are constructed: (9) Wherein, the And For the decentered matrix a and matrix B, e is a full 1 vector, and the vector length is the number of columns of matrix a.

Description

Real-time data-driven aviation component docking intelligent digital twin system Technical Field The invention belongs to the technical field of intelligent manufacturing, and relates to a real-time data-driven intelligent digital twin system for docking aviation components. Background Aeronautical component assembly is a complex project in the manufacture of aeronautical equipment, and the cost and workload of the complex project respectively account for 40% of the production cost and more than 50% of the total workload. Because of the huge size and complex structure of the aviation components, and the low relevance of the measurement data and the technological process, the traditional manual trial and error and repeated debugging method is difficult to realize the real-time monitoring and the accurate assessment of the butt joint quality in the butt joint process. Based on the technical bottleneck, the Unity 3D platform can be applied, and by means of the efficient real-time 3D rendering capability, the multi-source data fusion interface, the C# extensible script function and the rich plug-in ecological system, an intelligent digital twin system covering the whole butt joint process of aviation components is constructed by applying a multi-parameter cooperation technology driven by a three-dimensional digital model and real-time sensing data, so that the monitoring efficiency and the final assembly quality of the assembly process are remarkably improved. Guan Yuxiang et al in patent, "FMS digital twin system based on real-time data driving" (CN 116360354A) constructed a real-time digital twin system comprising a three-layer architecture, wherein a physical layer collects device status data in real time through a PLC and a sensor network, the digital twin layer parses the data and drives a three-dimensional model to move synchronously, and a functional service layer integrates a visual interface and a virtual debugging module, so that a debugger can locate a control program defect without checking on site. However, the system only realizes unidirectional mapping from the state to the model, not only does not construct a motion constraint relation between the components, but also lacks dynamic simulation capability of an assembly path, so that the system cannot completely simulate the cooperative motion and interaction of multiple structures in the assembly process. Tang Duibing et al in patent "a visual real-time monitoring method based on digital twin" (CN 115439267A) propose a three-library fusion architecture based on digital twin, wherein a digital resource pool is constructed based on a three-dimensional geometric model library, a manufacturing attribute library and a behavior rule library, a Unity 3D three-dimensional virtual scene is generated by combining workshop physical layout, dynamic mapping of operation data and a digital twin model is realized through a standardized communication interface, and finally a digital twin body and a real-time monitoring platform of a discrete manufacturing workshop are built. Although the system realizes the digital control of the equipment state and the production task, the visual interaction only stays at the display level of the parameter panel, and the three-dimensional dynamic simulation capability of real-time data driving is not established, so that the track planning and the movement process of the assembly action cannot be visually reproduced through a digital twin model. The two methods have realized digital twin technology application, but cannot effectively present the motion correlation characteristics among the components, and lack a model dynamic visualization mechanism driven by real-time sensing data, so that the real-time twin monitoring requirement of the docking process of the aviation components is difficult to meet. Disclosure of Invention Aiming at the problems that the real-time monitoring of the motion state of the part is difficult, the acquisition of the gesture information of the part is lagged and visual presentation is lacking in the process of the butt-joint assembly of the aviation part, the invention builds an intelligent digital twin system for the butt-joint of the aviation part based on real-time sensing data driving. The system achieves the functions through a three-layer architecture, firstly, a standardized communication interface between a multi-source sensor and each module of the system is established, real-time acquisition and synchronous transmission are carried out on sensing data, secondly, real-time motion tracks of the components are deduced through kinematic modeling based on multi-axis real-time displacement data of a positioner, space pose parameter resolving processing of the components is completed, finally, the pose parameters of the components and multi-dimensional force sensing data are fused, a dynamic interactive three-dimensional digital twin interface is constructed, and twin monitoring of the motion sta