Civil aircraft wing-body junction contact gap virtual measurement method and system based on manifold space alignment

Abstract

The application discloses a civil aircraft wing-body docking contact gap virtual measurement method and system based on manifold space alignment, constructs a "physical manifold guide-micro texture compensation" collaborative architecture, analyzes the nonlinear mapping reference of macro load and average gap, defines the physical manifold boundary conforming to energy conservation and deformation coordination, and designs a deep space-time network to adaptively extract micro texture residual features caused by surface roughness from a dynamic strain sequence, and then introduces a latent space manifold alignment mechanism, uses a physical reference track to forcibly constrain a data feature track in a low-dimensional manifold space, realizes deep fusion of mechanism priori and data features, effectively solves the problems of physical consistency loss and non-physical oscillation of a pure data-driven model under sparse samples, performs high-precision and strong-robustness cross-scale quantitative state reconstruction on a non-uniform contact gap field, and provides a technical path with mechanical interpretability for digital precise assembly of an aviation structural part.

Description

Technical Field

[0001] This invention belongs to the field of civil aircraft structural component assembly, specifically relating to a virtual measurement method and system for civil aircraft wing-body docking contact gap based on manifold space alignment. Background Technology

[0002] Currently, the methods for obtaining the wing-fuselage joint assembly clearance of civil aircraft are mainly divided into direct measurement and indirect measurement:

[0003] Direct measurement: Geometric depth is captured using digital feeler gauges, laser scanning, or industrial vision. Feeler gauges convert displacement into signals through probe movement; laser and vision measurements are based on optical triangulation or structured light projection principles.

[0004] Indirect measurement includes the capacitance method and the data model-driven method. The capacitance method uses the inverse relationship between capacitance and the distance between plates to characterize the gap; the data model-driven method uses statistical learning to construct a nonlinear mapping between input variables and target physical quantities.

[0005] The problems and causes of existing technologies are as follows:

[0006] Spatial interference and line-of-sight obstruction: The wing-body docking structure has a large size span and a narrow contact space, making it difficult for direct measurement tools to enter the enclosed space, and optical means have a line-of-sight blind spot.

[0007] Edge effect interference: Capacitive sensors are prone to edge effect errors due to the influence of surface micro-roughness and macro-non-parallelism.

[0008] Lack of physical consistency: Pure data-driven models are prone to overfitting under sparse sample conditions, and the prediction results lack contact mechanical constitutive constraints, thus limiting their generalization ability. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to provide a virtual measurement method and system for the contact gap of civil aircraft wing-body docking based on manifold space alignment, which solves the problems of unobservable contact interface of large civil aircraft structural components under complex assembly conditions, nonlinear contact behavior coupling, and limited generalization of virtual measurement under small sample conditions in the prior art.

[0010] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0011] A virtual measurement method for the contact gap between the wing and fuselage of civil aircraft based on manifold space alignment is proposed. By constructing a collaborative architecture of "physical manifold guidance and microtexture compensation", the nonlinear mapping benchmark between macroscopic load and average gap is analyzed, the physical manifold boundary that conforms to energy conservation and deformation coordination is defined, and a deep spatiotemporal network is designed to adaptively extract the microtexture residual features caused by surface roughness from the dynamic strain sequence. Then, a latent space manifold alignment mechanism is introduced, and the physical benchmark trajectory is used to force the data feature trajectory in the low-dimensional manifold space to achieve deep fusion of mechanism prior and data features.

[0012] Specifically, the process includes the following steps:

[0013] Physical manifold boundary characterization: By analyzing the statistical characteristics of the elastoplastic deformation of micro-protrusions on fractal rough surfaces, a nonlinear physical mapping benchmark for macroscopic normal load and interface proximity is established, and a physical manifold boundary that conforms to the laws of energy conservation and deformation compatibility is established.

[0014] Spatiotemporal feature extraction: A deep spatiotemporal network is constructed, which consists of a multi-scale stress feature encoder, a convolutional long short-term memory network unit, and a gap field reconstruction decoder. The encoder is used to extract high-dimensional features from the dynamic stress tensor sequence, and the convolutional long short-term memory network unit uses the load evolution process as a dynamic constraint to adaptively extract the micro-texture residual information induced by surface roughness.

[0015] Latent space manifold alignment: A feature decoupling mechanism is introduced into the bottleneck layer of the network to decompose the latent vector into physical principal components and data residual components. The physical principal components are constrained by the MB mechanism model, while the data residual components are adaptively extracted by the neural network. By constructing a latent space alignment loss function, the physical state predicted by the neural network is forced to converge to the physical reference trajectory calculated by the mechanism model in the low-dimensional manifold space.

[0016] Multi-scale collaborative reconstruction: The final contact gap field is generated by nonlinear weighted coupling of the physical reference field and the micro-texture residual field. At the same time, spatiotemporal evolution regularization constraints based on continuum mechanics are introduced to ensure that the rate of change of the gap field with load is compatible with the normal velocity field of the contact interface.

[0017] The process of establishing the nonlinear physical mapping benchmark is as follows:

[0018] Real-time load sequence Input the MB fractal contact physical mechanism model to establish a macroscopic dimensionless gap. With dimensionless load Mapping relationship:

[0019] Inverse mapping function of execution mechanism model Analytical calculation of theoretical average gap generates a homogenized physical reference field. The physical reference field constitutes the zeroth-order approximate estimate of the macroscopic equilibrium position of the contact interface.

[0020] The adaptive extraction process of micro-texture residual information is as follows:

[0021] The preprocessed stress cloud map sequence Input a deep spatiotemporal network, and the multi-scale stress feature encoder of the deep spatiotemporal network extracts low-dimensional abstract feature vectors. Using an aggregated overall deformation mode, the convolutional long short-term memory network receives the current features and the hidden state from the previous time step to perform state transition updates, and separates the micro-geometric texture residual feature vector in the hidden feature space. .

[0022] The latent space alignment loss function is as follows:

[0023]

[0024] Align the physical state vector and the data feature vector within the characteristic manifold, and synthesize the corrected latent variables. .

[0025] The corrected latent variables The input decoder generates a gap field, and constraints are introduced to perform post-processing corrections on the output gap field, specifically including:

[0026] Apply nonnegativity constraints Applying spatiotemporal evolution regularization constraints based on continuum mechanics To ensure that the rate of change of the gap field between adjacent load steps is compatible with the normal velocity field of the contact interface, a contact state monotonicity constraint is introduced. It handles the irreversibility caused by plastic deformation; outputs contact state characterization results, and completes the virtual measurement process of contact gap.

[0027] A virtual measurement system for the wing-body docking contact gap of a civil aircraft based on manifold space alignment includes a data acquisition module, a collaborative robot, and a computing workstation. The data acquisition module is used to acquire the stress characteristic tensor sequence of structural components and the instantaneous external load parameter sequence. The collaborative robot carries the data acquisition module as a motion carrier and cruises along the normal direction of the curved surface. The computing workstation uses the method described above to obtain virtual measurement data of the wing-body docking contact gap of the civil aircraft.

[0028] The data acquisition module includes a 3D optical scanner, a distributed strain sensor array, and a static acquisition device. The 3D optical scanner is used for path planning algorithms to automatically navigate along the surface normal to obtain high-fidelity geometric true values. The distributed strain sensor array is arranged at the key connection interfaces of the structural components to capture weak deformation responses. The static acquisition device is used to connect to the distributed strain sensor array to capture deformation responses.

[0029] A computer-readable storage medium storing computer-readable instructions that, when executed by a processor, invoke the steps of the method.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. Achieving High-Precision Cross-Scale Reconstruction: This invention effectively decouples macroscopic surface trends from microscopic roughness textures through a reconstruction strategy of "macroscopic physical benchmark + microscopic texture residual". Experimental results show that the mean absolute error (MAE) of the virtual gap measurement on the wing-body docking test rig is as low as 0.2 mm, which is about 81.4% higher than the pure mechanistic model and about 65.4% higher than the pure data-driven model.

[0032] 2. Significantly Enhanced Physical Consistency and Interpretability: A latent space manifold alignment mechanism is introduced, forcing the predicted trajectory of the neural network to converge to the mechanical constitutive relation. The Physical Consistency Error (PCE) is reduced by 83.5% compared to the pure data-driven model, effectively filtering out non-physical high-frequency noise caused by the black-box characteristics of the mathematical model, and ensuring that the reconstructed gap field conforms to the laws of energy conservation and deformation compatibility.

[0033] 3. Overcoming the generalization bottleneck under sparse samples: Under small sample conditions, this invention effectively suppresses the overfitting tendency of deep learning models by relying on strong prior constraints of physical mechanisms. For complex topological regions such as the wing leading edge and wing-body junction, the model's coefficient of determination... It remains above 0.92, compared to the performance of purely data-driven models in complex regions. With a value of only 0.62, a relative performance improvement of approximately 48.4% was achieved.

[0034] 4. High robustness and engineering adaptability: Through spatiotemporal evolution regularization constraints and latent space feature alignment, this invention can maintain system stability even under strong background noise interference. When the input signal signal-to-noise ratio drops to the extreme operating condition, this model can still maintain... The goodness of fit was verified, confirming its practical engineering value in harsh industrial electromagnetic interference environments.

[0035] 5. Significantly improved feature fusion efficiency: The intra-cluster compactness (ICC) of features after adopting the manifold alignment strategy reached 0.91. This represents a 75.0% improvement compared to the traditional direct splicing strategy and a 33.8% improvement compared to the canonical correlation analysis (CCA) strategy, effectively eliminating the differences between physical mechanisms and heterogeneous data features. Attached Figure Description

[0036] Figure 1 This is a schematic diagram of the integrated prediction model module of the present invention.

[0037] Figure 2 This is a schematic diagram illustrating the implementation mechanism of manifold alignment and feature decomposition of the present invention – data fusion.

[0038] Figure 3 This is a schematic diagram of a specific embodiment of the present invention. Detailed Implementation

[0039] The structure and working process of the present invention will be further described below with reference to the accompanying drawings.

[0040] like Figure 1 As shown.

[0041] The purpose of this invention is to address the problems of unobservable contact interfaces, coupled nonlinear contact behaviors, and limited generalization of virtual measurements in large civil aircraft structural components under complex assembly conditions. It proposes a virtual measurement method for wing-body docking contact gaps based on manifold space alignment. This method constructs a collaborative architecture of "physical manifold guidance - microtexture compensation," utilizes Majumdar-Bhushan fractal contact theory to analyze the nonlinear mapping benchmark between macroscopic loads and average gaps, defines physical manifold boundaries that conform to energy conservation and deformation compatibility, and designs ST-GapNet. Deep spatiotemporal networks adaptively extract micro-texture residual features caused by surface roughness from dynamic strain sequences, and then introduce a latent space manifold alignment mechanism. In low-dimensional manifold space, physical reference trajectories are used to forcibly constrain data feature trajectories, achieving deep fusion of mechanism priors and data features. This aims to effectively solve the problems of lack of physical consistency and non-physical oscillation in pure data-driven models under sparse samples. Ultimately, it achieves high-precision and robust cross-scale quantitative state reconstruction of non-uniform contact gap fields under background noise interference and complex topological conditions such as variable curvature at wing-body joints, providing a mechanically interpretable technical path for the digital precision assembly of aerospace structural components.

[0042] A virtual measurement method for the contact gap between the wing and fuselage of civil aircraft based on manifold space alignment is proposed. By constructing a collaborative architecture of "physical manifold guidance and microtexture compensation", the nonlinear mapping benchmark between macroscopic load and average gap is analyzed, the physical manifold boundary that conforms to energy conservation and deformation coordination is defined, and a deep spatiotemporal network is designed to adaptively extract the microtexture residual features caused by surface roughness from the dynamic strain sequence. Then, a latent space manifold alignment mechanism is introduced, and the physical benchmark trajectory is used to force the data feature trajectory in the low-dimensional manifold space to achieve deep fusion of mechanism prior and data features.

[0043] Physical manifold boundary characterization: By analyzing the statistical characteristics of the elastoplastic deformation of micro-protrusions on fractal rough surfaces, a nonlinear physical mapping benchmark for macroscopic normal load and interface proximity is established, and a physical manifold boundary that conforms to the laws of energy conservation and deformation compatibility is established.

[0044] Spatiotemporal feature extraction: A deep spatiotemporal network is constructed, which consists of a multi-scale stress feature encoder, a convolutional long short-term memory network unit, and a gap field reconstruction decoder. The encoder is used to extract high-dimensional features from the dynamic stress tensor sequence, and the convolutional long short-term memory network unit uses the load evolution process as a dynamic constraint to adaptively extract the micro-texture residual information induced by surface roughness.

[0045] Latent space manifold alignment: A feature decoupling mechanism is introduced into the bottleneck layer of the network to decompose the latent vector into physical principal components and data residual components. The physical principal components are constrained by the MB mechanism model, while the data residual components are adaptively extracted by the neural network. By constructing a latent space alignment loss function, the physical state predicted by the neural network is forced to converge to the physical reference trajectory calculated by the mechanism model in the low-dimensional manifold space.

[0046] Multi-scale collaborative reconstruction: The final contact gap field is generated by nonlinear weighted coupling of the physical reference field and the micro-texture residual field. At the same time, spatiotemporal evolution regularization constraints based on continuum mechanics are introduced to ensure that the rate of change of the gap field with load is compatible with the normal velocity field of the contact interface.

[0047] Specific embodiments, such as Figure 1 , Figure 2 , Figure 3 As shown:

[0048] This embodiment relies on an automated testing platform. The platform's hardware system consists of a collaborative robot, a 3D optical scanner, a distributed strain sensor array, a static data acquisition instrument, and a computing workstation, as shown in the attached figure. Figure 3 As shown, the structural connections and interactions are as follows: A distributed strain sensor array is positioned at the critical connection interfaces of the structural components to capture subtle deformation responses. The distributed strain sensor array establishes a signal connection with the static data acquisition unit. The static data acquisition unit establishes a data transmission connection with the computing workstation for synchronous control and recording.

[0049] The implementation process of the virtual measurement method follows strict physical and data reasoning logic. Each step is executed in a forward time sequence. Since feature deduction depends on data input and benchmark establishment, the order of each step is not interchangeable. The specific steps are as follows:

[0050] Step 1: Contact System Initialization and Multi-Source Data Acquisition

[0051] By using a distributed strain sensing array and a static acquisition instrument, the stress characteristic tensor sequence evolving with the discrete load step of the pseudo-time variable is obtained. Obtain the instantaneous external load parameter sequence.

[0052] Step 2: Calculation of the macroscopic physical reference field:

[0053] Real-time load sequence Input the Majumdar-Bhushan (MB) fractal contact physics mechanism model from the computing workstation, and use the Majumdar-Bhushan fractal contact mechanism model to establish a macroscopic dimensionless gap. With dimensionless load Mapping relationship:

[0054] , execution mechanism model inverse mapping function Analytical calculation of theoretical average gap to generate homogenized physical reference field The zeroth-order approximate estimate of the macroscopic equilibrium position of the contact interface is formed by the physical reference field.

[0055] Majumdar-Bhushan fractal contact theory provides mechanical constitutive constraints for calculating the contact gap at rough interfaces. Actual engineering interfaces exhibit microscopic roughness and undulations, with contact occurring at discrete groups of micro-protrusions. This theory utilizes the self-affine properties of fractal geometry to construct the statistical distribution characteristics of the micro-morphology. The model defines the critical geometric scale for the elastoplastic deformation of the micro-protrusions and integrates the elastic and plastic responses within the contact surface. The total normal load of the system equals the sum of the elastic and plastic loads. This mechanical mechanism establishes a nonlinear mapping relationship between the macroscopic normal load and the interface normal approximation. In virtual measurement tasks, the model receives real-time load sequences and uses an inverse mapping function to analyze the theoretical average contact gap of the interface. The theoretical average gap constitutes the macroscopic equilibrium position reference for the contact state, defining the physical manifold boundary of the gap field reconstruction.

[0056] Step 3: Deduction of Microscopic Residual Characteristics:

[0057] The preprocessed stress cloud map sequence Inputting the ST-GapNet deep spatiotemporal network, the ST-GapNet deep spatiotemporal network's multi-scale stress feature encoder extracts low-dimensional abstract feature vectors. Using an aggregated overall deformation mode, the convolutional long short-term memory network receives the current features and the hidden state from the previous time step to perform state transition updates, and separates the micro-geometric texture residual feature vector in the hidden feature space. .

[0058] Step 4: Latent Space Manifold Alignment and Feature Fusion:

[0059] In latent feature space Internal heterogeneous feature alignment is performed using a physical consistency loss function:

[0060]

[0061] Forced data features to approximate the baseline physical coordinates mapped by the dimensionality reduction of the mechanistic model Align the physical state vector and the data feature vector within the characteristic manifold to synthesize the corrected latent variables. .

[0062] Step 5: Full-field decoding and reconstruction, and spatiotemporal evolution correction:

[0063] The corrected latent variables The input decoder generates a gap field, and constraints are introduced to perform post-processing corrections on the output gap field, specifically including:

[0064] Apply nonnegativity constraints Applying spatiotemporal evolution regularization constraints based on continuum mechanics To ensure that the rate of change of the gap field between adjacent load steps is compatible with the normal velocity field of the contact interface, a contact state monotonicity constraint is introduced. It handles the irreversibility caused by plastic deformation; outputs contact state characterization results, and completes the virtual measurement process of contact gap.

[0065] It should be understood that this solution is not limited to the specific embodiments described above. Devices and structures not described in detail herein should be understood as being implemented in a manner common to the art. Any person skilled in the art can make many possible variations and modifications to this solution, or modify it into equivalent embodiments, without departing from the scope of this solution, using the methods and techniques disclosed above. This does not affect the substantive content of this solution. Therefore, any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of this solution, without departing from its scope, still fall within the protection scope of this solution.

[0066] A virtual measurement system for the wing-body docking contact gap of a civil aircraft based on manifold space alignment includes a data acquisition module, a collaborative robot, and a computing workstation. The data acquisition module is used to acquire the stress characteristic tensor sequence of structural components and the instantaneous external load parameter sequence. The collaborative robot, as a motion carrier, carries a three-dimensional optical scanner and cruises along the normal direction of the curved surface. The computing workstation uses the method described above to obtain virtual measurement data of the wing-body docking contact gap of the civil aircraft.

[0067] The data acquisition module includes a 3D optical scanner, a distributed strain sensor array, and a static acquisition device. The 3D optical scanner is used for path planning algorithms to automatically navigate along the surface normal to obtain high-fidelity geometric true values. The distributed strain sensor array is arranged at the key connection interfaces of the structural components to capture weak deformation responses. The static acquisition device is used to connect to the distributed strain sensor array to capture deformation responses.

[0068] The experimental platform architecture for the civil aircraft wing-body docking mission is divided into two parallel computational domains: a physical space and a digital twin space. The physical space integrates a scaled-down passenger aircraft model and a collaborative robotic arm. A sensor array on the docking surface performs stress acquisition, capturing structural deformation parameters and outputting stress time-series data. Based on the Majumdar-Bhushan fractal contact theory, it analyzes the nonlinear mapping benchmark between macroscopic loads and average clearance, establishing the physical manifold boundary. In the digital twin space, the gasket modeling module receives physical contact parameters to construct a geometric model, while the Ansys simulation module performs numerical calculations of the structural stress state based on the geometric model and boundary conditions. The system applies a manifold space alignment mechanism, using physical reference trajectories to forcibly constrain time-series data features within a low-dimensional latent space, achieving the fusion calculation of prior mechanisms and multi-source stress data. The Unity twin system receives the simulated stress cloud map and renders it onto the virtual model surface, completing the digital mapping of the physical state during the assembly process. The measurement characteristics of the physical domain and the analytical results of the virtual domain together constitute a data closed loop.

[0069] A computer-readable storage medium storing computer-readable instructions that, when executed by a processor, invoke the steps of the method.

[0070] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A virtual measurement method for wing-body docking contact gap of civil aircraft based on manifold space alignment, characterized in that: By constructing a collaborative architecture of "physical manifold guidance-microtexture compensation", the nonlinear mapping benchmark between macroscopic load and average gap is analyzed, the physical manifold boundary that conforms to energy conservation and deformation coordination is defined, and a deep spatiotemporal network is designed to adaptively extract the microtexture residual features caused by surface roughness from the dynamic strain sequence. Then, a latent space manifold alignment mechanism is introduced, and the physical benchmark trajectory is used to force the data feature trajectory in the low-dimensional manifold space to achieve deep fusion of mechanism prior and data features.

2. The virtual measurement method for wing-body docking contact gap of civil aircraft based on manifold space alignment according to claim 1, characterized in that: Specifically, the process includes the following steps: Physical manifold boundary characterization: By analyzing the statistical characteristics of the elastoplastic deformation of micro-protrusions on fractal rough surfaces, a nonlinear physical mapping benchmark for macroscopic normal load and interface proximity is established, and a physical manifold boundary that conforms to the laws of energy conservation and deformation compatibility is established. Spatiotemporal feature extraction: A deep spatiotemporal network is constructed, which consists of a multi-scale stress feature encoder, a convolutional long short-term memory network unit, and a gap field reconstruction decoder. The encoder is used to extract high-dimensional features from the dynamic stress tensor sequence, and the convolutional long short-term memory network unit uses the load evolution process as a dynamic constraint to adaptively extract the micro-texture residual information induced by surface roughness. Latent space manifold alignment: A feature decoupling mechanism is introduced into the bottleneck layer of the network to decompose the latent vector into physical principal components and data residual components. The physical principal components are constrained by the MB mechanism model, while the data residual components are adaptively extracted by the neural network. By constructing a latent space alignment loss function, the physical state predicted by the neural network is forced to converge to the physical reference trajectory calculated by the mechanism model in the low-dimensional manifold space. Multi-scale collaborative reconstruction: The final contact gap field is generated by nonlinear weighted coupling of the physical reference field and the micro-texture residual field. At the same time, spatiotemporal evolution regularization constraints based on continuum mechanics are introduced to ensure that the rate of change of the gap field with load is compatible with the normal velocity field of the contact interface.

3. The virtual measurement method for wing-body docking contact gap of civil aircraft based on manifold space alignment according to claim 2, characterized in that: The process of establishing the nonlinear physical mapping benchmark is as follows: Real-time load sequence Input the MB fractal contact physical mechanism model to establish a macroscopic dimensionless gap. With dimensionless load Mapping relationship: , execution mechanism model inverse mapping function Analytical calculation of theoretical average gap generates a homogenized physical reference field. The physical reference field constitutes the zeroth-order approximate estimate of the macroscopic equilibrium position of the contact interface.

4. The virtual measurement method for wing-body docking contact gap of civil aircraft based on manifold space alignment according to claim 2, characterized in that: The adaptive extraction process of micro-texture residual information is as follows: The preprocessed stress cloud map sequence Input a deep spatiotemporal network, and the multi-scale stress feature encoder of the deep spatiotemporal network extracts low-dimensional abstract feature vectors. Using an aggregated overall deformation mode, the convolutional long short-term memory network receives the current features and the hidden state from the previous time step to perform state transition updates, and separates the micro-geometric texture residual feature vector in the hidden feature space. .

5. The virtual measurement method for wing-body docking contact gap of civil aircraft based on manifold space alignment according to claim 2, characterized in that: The latent space alignment loss function is as follows: Align the physical state vector and the data feature vector within the characteristic manifold, and synthesize the corrected latent variables. .

6. The virtual measurement method for wing-body docking contact gap of civil aircraft based on manifold space alignment according to claim 5, characterized in that: The corrected latent variables The input decoder generates a gap field, and constraints are introduced to perform post-processing corrections on the output gap field, specifically including: Apply nonnegativity constraints Applying spatiotemporal evolution regularization constraints based on continuum mechanics To ensure that the rate of change of the gap field between adjacent load steps is compatible with the normal velocity field of the contact interface, a contact state monotonicity constraint is introduced. It handles the irreversibility caused by plastic deformation; outputs contact state characterization results, and completes the virtual measurement process of contact gap.

7. A virtual measurement system for wing-body docking contact gap of civil aircraft based on manifold space alignment, characterized in that: The system includes a data acquisition module, a collaborative robot, and a computing workstation. The data acquisition module is used to acquire the stress characteristic tensor sequence of structural components and the instantaneous external load parameter sequence. The collaborative robot carries the data acquisition module as a motion carrier and cruises along the normal direction of the curved surface. The computing workstation uses any one of the methods in claims 1 to 6 to obtain virtual measurement data of the wing-body docking contact gap of a civil aircraft.

8. The virtual measurement system for wing-body docking contact gap of civil aircraft based on manifold space alignment according to claim 7, characterized in that: The data acquisition module includes a 3D optical scanner, a distributed strain sensor array, and a static acquisition device. The 3D optical scanner is used for path planning algorithms to automatically navigate along the surface normal to obtain high-fidelity geometric true values. The distributed strain sensor array is arranged at the key connection interfaces of the structural components to capture weak deformation responses. The static acquisition device is used to connect to the distributed strain sensor array to capture deformation responses.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions that, when executed by a processor, invoke the steps of the method according to any one of claims 1 to 6.

Images