An aircraft body structure assembly process design and quality control green degree improvement method

Abstract

The application provides an aircraft body structure assembly process design and quality control green degree improvement method, aiming at the high trial and error cost, high energy consumption and passive quality control problem of the traditional aircraft assembly process, a whole process technical system of "data sensing and preprocessing-assembly process scheme intelligent planning-assembly process parameter optimization-on-site assembly quality intelligent regulation" is constructed, and the "measurable, controllable and visible" of the assembly process is realized. Specifically, it includes: multi-source sensing and standardized preprocessing of assembly process and result quality data; "shape and property collaborative control" process scheme planning based on high-fidelity virtual prototype; parameter optimization design of multi-objective parameter optimization before assembly, process self-adaptive adjustment and knowledge base construction; and on-site quality intelligent regulation of assembly deviation characterization, out-of-tolerance tracing and multi-dimensional regulation. The application can reduce the assembly trial and error and rework cost, reduce material loss and energy consumption, improve the assembly quality and efficiency, and help to realize the green assembly target.

Description

Technical Field

[0001] This invention relates to the field of aviation manufacturing technology, specifically to a method for improving the greenness of aircraft airframe structure assembly process design and quality control. Background Technology

[0002] Green represents health, safety, and environmental protection. Green manufacturing is receiving increasing attention in production and plays a crucial role in achieving "dual carbon" goals. The national standard GB / T 28612-2023 defines green manufacturing as a modern manufacturing model characterized by low consumption, low emissions, high efficiency, and high returns. Its essence lies in comprehensively considering industrial structure, energy resources, ecological environment, and health and safety factors during the development of the manufacturing industry, integrating the concept of green development throughout the entire product lifecycle to promote the green transformation and upgrading of traditional industries. Green manufacturing is committed to solving problems related to green product design, efficient and clean production, resource conservation, and recycling (remanufacturing), ensuring that products meet environmental protection and resource conservation requirements at every stage, from design, manufacturing, packaging, use, maintenance to end-of-life disposal.

[0003] Aircraft airframe assembly is a core component of aerospace manufacturing. Traditional assembly processes often employ a trial-and-error approach and an experience-based, passive repair method, which presents several problems: First, reliance on physical trial-and-error to verify process solutions leads to significant material waste, high energy consumption, and long cycle times. Second, the use of open-loop control overly focuses on geometric precision, neglecting the impact of internal stress and damage on physical properties generated during assembly, resulting in insufficient assembly quality stability. Third, the design of process parameters lacks scientific quantitative basis, and on-site adjustments rely on operator experience, making it difficult to meet the high-precision requirements of assembling complex thin-walled structures and dissimilar materials. Fourth, multi-source heterogeneous quality data is difficult to integrate effectively, failing to provide precise support for process optimization and quality control. These problems result in a low level of green technology in traditional assembly processes, failing to meet the development requirements of "low consumption, high efficiency, and high quality" in green manufacturing.

[0004] Next-generation aircraft products typically exhibit characteristics such as high stealth, lightweight, high reliability, low cost, and long service life. Their airframe structures also face new requirements for lightweight, large-scale / integrated, and structure-function integrated design. For aircraft assembly operations, the primary challenge is to meet the generational improvement requirements for assembly quality control and efficiency. Furthermore, to ensure the service performance of aircraft products and achieve green assembly goals, it is necessary to adhere to the guiding principles of green manufacturing, focusing on the environmental impact of assembly production and reducing carbon emissions and energy consumption during the assembly process. To achieve green delivery of products, the degree of greening must be enhanced in the specific assembly process design and quality control aspects of the airframe structure.

[0005] Green assembly technology should be a process that completely abandons traditional trial-and-error methods. It should be a predictable, integrable, and model- and simulation-based scientific design that ensures assembly quality with the technical characteristics of being "measurable, controllable, and visible." Specifically, to address assembly quality deviations, based on assembly process and result quality data, and from the perspectives of process, precision, and physical characteristics, it employs mathematical modeling and virtual simulation verification during the assembly process. It conducts systematic research on aspects such as establishing a process knowledge base, finite element analysis, constructing an assembly measurement field, data analysis / processing / mining, and process optimization decisions. This establishes a new assembly quality control model oriented towards multidisciplinary and multi-field coupled analysis.

[0006] Therefore, it is necessary to study a method to improve the greenness of aircraft airframe assembly process design and quality control in order to address the shortcomings of existing technologies and solve or mitigate one or more of the above-mentioned problems. Summary of the Invention

[0007] In view of this, the present invention provides a method for improving the greenness of aircraft airframe structure assembly process design and quality control. By constructing a full-process technical system of "assembly data perception and preprocessing - intelligent planning of assembly process scheme - optimization of assembly process parameters - intelligent control of on-site assembly quality", it aims to solve the technical problems of high trial and error costs, high energy consumption, passive quality control and low greenness in traditional aircraft airframe structure assembly processes.

[0008] On the one hand, this invention provides a method for improving the greenness of aircraft airframe structure assembly process design and quality control, including the following steps: S1: Perception, measurement and preprocessing of multi-source quality data of assembly process and results: collect geometric, physical and functional data of the entire assembly process, realize data interoperability through standardized interfaces and communication protocols, classify and store heterogeneous data using a hybrid storage architecture, and build a unified perception data model through time alignment, outlier removal, feature extraction and data fusion preprocessing operations. S2: Intelligent Assembly Process Planning: Construct a high-fidelity virtual assembly prototype, reproduce the multi-physics behavior of assembly, and conduct virtual pre-simulation verification; with "shape-property collaborative control" as the goal, optimize the assembly task arrangement and equipment operating parameters of the assembly process plan to achieve simultaneous assurance of geometric accuracy and physical performance; S3: Assembly process parameter optimization design: Based on the preprocessing data of assembly process and result quality in S1, and the assembly process scheme in S2, a multi-objective optimization model is established before assembly to solve the optimal combination of positioning, hole making, connection and finishing parameters. During the assembly process, a closed-loop control system is built based on real-time sensing data to realize adaptive adjustment of process parameters. The process data is expanded through small sample learning technology to build a self-updating assembly quality data knowledge base. S4: Intelligent process control of on-site assembly quality: The mechanism-data hybrid modeling method is used to characterize assembly deviations. A quantitative correlation model between assembly deviations and process parameters in the assembly quality data knowledge base in S3 is constructed. Global sensitivity analysis is used to trace the source of deviations. A multi-dimensional dynamic control mechanism for process error allocation, process parameter correction and repair pre-control is established by adopting a hierarchical and parallel solution strategy, and the process control scheme is output.

[0009] In addition to the aspects and any possible implementations described above, an implementation is further provided, wherein S1 specifically includes: S11: Multi-source Assembly Quality Data Measurement and Sensing: For the pre-assembly inspection, positioning, hole making, connection, finishing and final inspection stages, three types of data are collected: geometric quantities, physical quantities and functional quantities. The specific measurement equipment and requirements correspond as follows: geometric quantity data is collected using contact and non-contact equipment; physical quantity data is collected through force sensors, strain gauges and temperature sensors; functional quantity data is obtained using universal testing machines and ultrasonic non-destructive testing equipment. S12: Standardization and Storage of Multi-Source Assembly Quality Data: Through standardized interfaces such as RESTful API, OPC UA, and MQTT, and industrial communication protocols such as OPC UA and Modbus TCP, the data transmission formats of JSON and Protobuf are unified, and 32-bit cyclic redundancy check is used to ensure data integrity; a hybrid storage architecture is constructed, using MySQL / PostgreSQL relational databases to store structured data, MongoDB document databases to store semi-structured data, and MinIO / S3 object storage services to store unstructured data, and table partitioning and database partitioning strategies and index structures are designed; S13: Preprocessing and Fusion of Multi-Source Assembly Quality Data: Based on a unified timestamp benchmark, linear interpolation is used to fill in missing time-series data, and data from different acquisition frequencies are synchronized and aligned using a sliding window; outliers are removed using the 3σ principle or box plot method, and missing data are filled with the mean / median or K-nearest neighbor method, and the dimensions are unified by Z-score standardization or Min-Max normalization; correlation is analyzed using Pearson / Spearman coefficients, and data features are extracted using convolutional neural networks and long short-term memory networks; raw data level, feature level, and decision level fusion are achieved based on weighted average method, Kalman filtering, principal component analysis, and DS evidence theory, and a unified perception data model for ontology modeling is constructed, and data visualization is achieved through ECharts, Tableau, and Unity 3D.

[0010] In addition to the aspects and any possible implementations described above, a further implementation is provided, wherein S2 specifically includes: S21: High-fidelity virtual prototype construction: Using digital twin technology, integrating real assembly environment, equipment parameters, and working condition constraints, a high-fidelity real-time mapping between physical entities and virtual models is constructed to reproduce the geometric constraints, force-thermal coupling, and multi-physics behavior of material deformation in the assembly process. The working condition constraints include, but are not limited to, temperature fluctuations and equipment wear. S22: Virtual pre-simulation verification of assembly quality: Conduct dynamic simulation of the entire assembly process in virtual space to identify assembly quality problems such as unreasonable positioning sequence, excessive clamping force causing delamination damage, and warping deformation in advance, and optimize process plans to reduce physical trial and error; S23: Intelligent planning of assembly process scheme: By adopting the method of shape-property synergistic optimization, it breaks through the traditional single dimension of geometric accuracy control, and puts geometric deformation, assembly internal stress, connection strength and physical properties on the same priority. Through model calibration and correction, it optimizes the arrangement of assembly positioning, hole making, connection, and finishing tasks and equipment running time. It adopts closed-loop feedback correction technology to precisely control the assembly process and form an optimized assembly process scheme. The closed-loop feedback correction technology includes, but is not limited to, optimization and correction and flexible positioning control.

[0011] In addition to the aspects and any possible implementations described above, a further implementation is provided, wherein S3 specifically includes: S31: Pre-assembly process multi-objective parameter optimization: Considering workpiece manufacturing error, material properties and on-site working conditions, as well as the pre-processing data of assembly process and result quality in S1, orthogonal experiments are used to determine parameter combinations and establish a multi-objective optimization model composed of geometric accuracy, physical performance and cost. S32: Adaptive adjustment of process parameters in assembly operation: Based on the assembly process scheme in S2 and the real-time data collected by multi-dimensional sensors deployed in tooling equipment in S1, the data is quickly processed by the edge computing unit and analyzed by machine learning and deep reinforcement learning algorithms to solve for the optimal parameters that meet the geometric error constraints and material yield strength threshold. The process parameters of hole drilling speed, riveting force, and feed speed are dynamically adjusted to reduce equipment power consumption and enable the assembly quality indicators to quickly meet the design requirements. S33: Intelligent Expansion of Assembly Test Data and Construction of Process Knowledge Base: Using generative adversarial networks, Bayesian optimization, and transfer learning few-shot learning techniques, high-fidelity virtual samples are generated based on limited process test data to establish a mapping relationship between process parameters and assembly quality; historical data, simulation results, and field feedback are integrated to construct a self-updating process parameter database and knowledge base, forming an intelligent decision chain of "parameter recommendation - effect prediction - deviation correction" to reduce the trial and error cost of process parameters.

[0012] In addition to the aspects and any possible implementations described above, an implementation is further provided, wherein S4 specifically includes: S41: Assembly Deviation Characterization and Correlation Modeling: Based on the time-varying coupling and transmission of geometric-physical assembly state, a mechanism-data hybrid modeling approach is adopted to comprehensively obtain data on dimensional coordination, stress distribution, and damage morphology to characterize the assembly deviation state of the shape target; combined with the Apriori association rule algorithm, a two-way mapping relationship between deviation and core process parameters is constructed. S42: Assembly deviation source tracing: Using the Sobol's index method for global sensitivity analysis, the first-order sensitivity index and total sensitivity index of each process parameter to deviation are calculated, and the contribution is comprehensively formed to quantify the influence weight of each link, analyze the transmission path of key deviation sources in the assembly process, and locate key influencing factors and root causes of deviation. S43: Multi-dimensional dynamic control of assembly quality: By reverse reasoning the parameter adjustment rules of preceding process variables, an intelligent multi-dimensional control mechanism for multi-source assembly quality data is constructed, including error adaptive allowance allocation, secondary correction of process parameters and pre-control design measures for repair and compensation. A complete toolchain of modeling-solving-optimization is developed to output a software-based intelligent process control scheme.

[0013] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the geometric data in S1 includes component manufacturing errors, assembly geometry, and key feature spatial positions, which are collected by a coordinate measuring machine, laser scanning equipment, and laser tracker; the physical data includes assembly load, temperature field, stress and strain, and equipment operating parameters; and the functional data includes stiffness, bearing strength, and fatigue life.

[0014] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the shape-property collaborative control in S2 simultaneously optimizes geometric deformation and assembly internal stress and connection strength physical properties through finite element simulation and multi-objective optimization algorithms.

[0015] In addition to the aspects and any possible implementations described above, a further implementation is provided in which the self-updating function of the process parameter database and knowledge base in S3 is achieved through incremental learning. When new process test data, simulation results, or field feedback data are added, the knowledge base update process is automatically triggered to retrain the process parameter-assembly quality mapping model and update the association rules and decision logic. The knowledge base is stored using a distributed database architecture.

[0016] In addition to the aspects and any possible implementations described above, a further implementation is provided in which, in the mechanism-data hybrid model characterizing assembly deviations in S4, the mechanism model is constructed using theories of materials mechanics, structural mechanics, and thermodynamics to analyze the influence mechanism of process parameters on deviations; the data model adopts neural network and support vector machine machine learning models to learn high-dimensional mapping relationships based on historical data; the hybrid modeling integrates the outputs of the mechanism model and the data model through weighted fusion, with the weight coefficients dynamically adjusted according to the prediction accuracy of the model to ensure the accuracy of the data characterization of geometric fit clearance, stress distribution, and damage morphology deviations.

[0017] In addition to the aspects and any possible implementations described above, a further implementation is provided in which, in step S4, a hierarchical and parallel solution strategy is adopted to establish a multi-dimensional dynamic control mechanism for process error allocation, process parameter correction, and pre-control of adjustment. The output process control scheme specifically includes: The hierarchical solution strategy is divided into global, local, and execution levels. The global level is responsible for optimizing and controlling the overall process plan, the local level is responsible for adjusting the process parameters of individual processes, and the execution level is responsible for controlling the motion of tooling equipment. The parallel solution strategy uses multi-threading technology to achieve parallel processing of control tasks at different levels and in different stages, improving control efficiency. The software output of the toolchain includes process parameter adjustment lists, equipment operation manuals, and repair area and repair quantity descriptions, which can be directly imported into the control system and MES system of tooling equipment to achieve rapid execution of assembly quality control plans.

[0018] Compared with the prior art, the present invention can achieve the following technical effects: 1) Based on the perception and detection of assembly process and result quality data, carry out data standardization and multi-level fusion preprocessing to provide a high-quality data foundation for the formulation of assembly process plans and the optimization of process parameters.

[0019] 2) By using digital virtual modeling and pre-dynamic control of the entire assembly process, and by adopting shape-based collaborative control and closed-loop parameter adjustment, an "intelligent / scientific" assembly process planning method has been formed, which enables the assembly process to be "predictable, optimizable, and traceable".

[0020] 3) By using a combination of "virtual pre-simulation - real-time control - knowledge accumulation", we can carry out the optimization design of assembly process parameters, realize the intelligent recommendation and adaptive adjustment of process parameters, and form an intelligent decision chain of "parameter recommendation - effect prediction - deviation correction", which significantly reduces the number of physical trial and error and rework rate.

[0021] 4) By standardizing data and building a knowledge base throughout the assembly process, we can reduce the ineffective operating time and energy consumption of equipment, lower the overall assembly cost, and meet the core requirements of green manufacturing: "low consumption, high efficiency, and high quality".

[0022] Of course, any product implementing this invention does not necessarily need to achieve all of the technical effects described above at the same time. Attached Figure Description

[0023] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart illustrating the steps of a method for improving the greenness of aircraft airframe structure assembly process design and quality control according to the present invention. Figure 2 This is a flowchart illustrating the planning of an "intelligent / scientific" assembly process scheme according to the present invention. Figure 3 This is a flowchart of the assembly process parameter optimization design according to the present invention; Figure 4 This invention provides a process intelligent control diagram for on-site assembly quality. Detailed Implementation

[0025] To better understand the technical solution of the present invention, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0026] It should be understood that the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0027] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0028] This invention provides a method for improving the greenness of aircraft airframe structure assembly process design and quality control. This method mainly includes four parts, such as... Figure 1As shown, the process involves four steps. First, during and after the assembly of the aircraft fuselage structure, multi-source heterogeneous assembly quality data is collected throughout the entire assembly process, including geometric, physical, and functional data. This data is then stored and pre-processed. Second, a high-fidelity virtual assembly prototype is constructed to reproduce the multi-physics behavior of the assembly process and conduct virtual pre-simulation verification. With "form-property collaborative control" as the goal, intelligent planning of the assembly process scheme is achieved. Third, based on the pre-processed data of the assembly process and results, and the assembly process scheme, a comprehensive approach of "virtual pre-simulation - real-time control - knowledge accumulation" is used to optimize the design of process parameters before and during assembly, and to construct a process knowledge base. Fourth, a mechanism-data hybrid modeling method and a data-driven learning approach are employed to achieve intelligent process control of on-site assembly quality. Ultimately, this improves the green level of assembly process design and assembly quality control, meeting the development needs of green manufacturing characterized by "low consumption, high efficiency, and high quality." Specifically, it includes the following steps: S1: Sensing, measurement and preprocessing of multi-source quality data for assembly process and results: Collect geometric, physical and functional data of the entire assembly process, achieve data interoperability through standardized interfaces and communication protocols, classify and store heterogeneous data using a hybrid storage architecture, and construct a unified sensing data model through time alignment, outlier removal, feature extraction and data fusion preprocessing operations.

[0029] S2: Intelligent Assembly Process Planning: Construct a high-fidelity virtual assembly prototype to reproduce multi-physics behavior during assembly and conduct virtual pre-simulation verification; with "shape-property coordinated control" as the goal, optimize the assembly task arrangement and equipment operating parameters of the assembly process plan to achieve simultaneous assurance of geometric accuracy and physical performance, such as... Figure 2 As shown.

[0030] S3: Assembly Process Parameter Optimization Design: Based on the preprocessed assembly process and result quality data in S1, and the assembly process scheme in S2, a multi-objective optimization model is established before assembly to solve for the optimal combination of positioning, hole making, connection, and finishing parameters. During the assembly process, a closed-loop control system is constructed based on real-time sensing data to achieve adaptive adjustment of process parameters. Process data is expanded through few-sample learning technology to build a self-updating assembly quality data knowledge base, such as... Figure 3 As shown.

[0031] S4: Intelligent Process Control of On-site Assembly Quality: This approach employs a mechanism-data hybrid modeling method to characterize assembly deviations. It constructs a quantitative correlation model between assembly deviations and process parameters in the assembly quality data knowledge base of S3. Global sensitivity analysis is used to trace the source of deviations. Furthermore, a hierarchical and parallel solution strategy is employed to establish a multi-dimensional dynamic control mechanism for process error allocation, process parameter correction, and pre-control of adjustments. The resulting process control scheme is then output, such as... Figure 4 As shown.

[0032] S1 specifically includes: S11: Multi-source assembly quality data measurement and sensing: For the pre-assembly inspection, positioning, hole making, connection, finishing and final inspection stages, three types of data are collected: geometric quantity, physical quantity and functional quantity. The specific measurement equipment and the correspondence with the requirements are as follows: geometric quantity data is collected using contact and non-contact equipment; physical quantity data is collected through force sensors, strain gauges and temperature sensors; functional quantity data is obtained using universal testing machines and ultrasonic non-destructive testing equipment.

[0033] Furthermore, geometric data includes component manufacturing errors, assembly geometry, and key feature spatial positions, collected using coordinate measuring machines, laser scanning equipment, and laser trackers; physical data includes assembly loads, temperature fields, stress and strain, and equipment operating parameters; and functional data includes stiffness, load-bearing capacity, and fatigue life.

[0034] Furthermore, for key stages of the entire aircraft fuselage assembly process, including pre-assembly part inspection, positioning, hole making, connection, finishing, and final assembly inspection, geometric, physical, and functional data were collected. Geometric data included component manufacturing errors, post-assembly geometric state, key feature spatial positions, profile dimensions, edge / normal / contour shape characteristics, hole diameter, hole position, hole perpendicularity, normal vector, burrs, countersink accuracy, hole wall roughness, interference, clearance, rivet head protrusion / concavity, workpiece connection warpage, end face machining dimensions and accuracy, and hole coaxiality. Physical data included assembly site temperature and humidity, tooling load / torque, tooling strain / deformation, product strain / deformation, clamping damage, acceleration, delamination, fiber tearing state, hole wall scratches, cutting force, clamping force, cutting heat, equipment operating parameters, connection load, tightening torque, contact friction, and radial force. Extrusion pressure, preload / torque, nail load and deformation, local stress and deformation of the workpiece, connection damage type and location, processing deformation; functional data includes the stiffness, load-bearing strength, and fatigue life of intermediate assemblies; acquisition equipment is configured according to the data type, geometric data acquisition uses coordinate measuring machines, laser scanning equipment, articulated arm measuring machines, photogrammetry systems, lidar, indoor space measurement systems, laser trackers, grating rulers, motor encoders, digital gauges for aperture, digital gauges for depth, digital gauges for perpendicularity, roughness testers, eddy current sensors, laser rangefinders, dual-camera measurement, (visual) image measurement, digital vernier calipers, tool microscopes, workspace measurement and positioning systems (WMPS), indoor space measurement systems (iGPS), gap guns, convexity and concaveness measurement gauges, and scratch testers; physical data acquisition uses temperature and humidity sensors, motor torque / Torque sensors, multi-dimensional force sensors, strain gauges, gratings, accelerometers, near-field spectrum visualization measurement equipment, barometers, three-dimensional digital imaging (3D-DIC) equipment, ultrasonic testing scanning electron microscopes, washer-type force sensors, torque wrenches, dynamic torque sensors, and infrared thermometers are used for functional quantity data acquisition. Other equipment includes universal testing machines, linear displacement measurement equipment, strain measurement devices, scanning electron microscopes, ultrasonic / infrared non-destructive testing equipment, high-resolution video monitoring crack propagation measurement systems, structural health monitoring systems, and vibration sensors.

[0035] S12: Standardization and Storage of Multi-Source Assembly Quality Data: Define standardized data interfaces such as RESTful API, OPC UA interface, and MQTT interface, adopt industrial communication protocols such as OPC UA, Modbus TCP, and HTTP / HTTPS, unify the data transmission format to JSON or Protobuf, and ensure the integrity of data during transmission through a 32-bit cyclic redundancy check method, so as to realize efficient interoperability and sharing of assembly data between different platforms such as MES system, ERP system, and quality inspection system. In terms of data storage, a hybrid storage architecture is adopted to classify and store the collected heterogeneous data. Structured data, including numerical data such as bolt torque and part dimensions, is stored in relational databases such as MySQL or PostgreSQL. A relational model is constructed through entity-relational graphs in the database, clarifying field types, primary key constraints, and association rules. Semi-structured data, including assembly process documents in XML format and sensor configuration information in JSON format, is stored in a document-oriented database such as MongoDB. The data structure is defined through key-value pairs and document models. Unstructured data, including assembly process images, equipment vibration audio, and defect detection videos, is stored in MinIO or S3 object storage services. It is stored in binary format and associated with metadata such as acquisition time, equipment number, and workstation information. At the same time, a table partitioning and database partitioning strategy and index structure are designed to optimize data read and write performance.

[0036] S13: Multi-source assembly quality data preprocessing and fusion: Through time-series alignment and correlation statistical analysis between data collected at different assembly stages, data preprocessing and mining analysis are achieved, and key data feature parameters are extracted from the raw data. Specifically, in terms of time series alignment, based on a unified timestamp benchmark, linear interpolation and spline interpolation are used to fill in missing time series data. The window size is dynamically adjusted according to the acquisition frequency, and data from different acquisition frequencies are synchronously aligned through a sliding window. In terms of data preprocessing, outliers can be identified and removed using the 3σ principle and box plot method, and missing data can be filled in using mean / median and K-nearest neighbor missing value filling methods. Then, the data dimensions are unified by Z-score standardization and Min-Max normalization methods. In terms of correlation statistical analysis, Pearson / Spearman coefficients are used to analyze the linear / nonlinear correlation between variables, and the potential correlations between data from different assembly stages are extracted using the Apriori-T time series correlation rule mining algorithm. In the feature extraction stage, the mean, variance, peak value, and kurtosis statistical features can be calculated for structured data. For unstructured data, convolutional neural network methods can be used to extract image shape features, and for time series data, a long short-term memory network model can be used to extract dynamic time series features, forming a set of key data feature parameters. Finally, by fusing raw data, feature-level, and decision-level data information, a unified perception data model is constructed to achieve the organic integration of measurement and perception data and the visualization of the data. Specifically, in terms of data fusion and visualization, raw data-level fusion can employ weighted average methods and Kalman filtering algorithms to fuse raw data of the same type from multiple sensors, such as measurements of the same assembly gap from multiple displacement sensors; feature-level fusion integrates heterogeneous features of structured and unstructured data through feature concatenation, principal component analysis dimensionality reduction, and multi-kernel learning methods; decision-level fusion can be based on DS evidence theory and Bayesian inference to fuse quality assessment results from multiple models, such as geometric accuracy assessment models and mechanical stability assessment models. In the construction of a unified perception data model, ontology modeling methods can be used to define the concepts, attributes, relationships, and constraints of assembly quality data, achieving consistency at the semantic level of data; in visualization, tools such as ECharts, Tableau, and Unity 3D can be used to construct real-time trend charts based on changes in time-series data, heatmaps based on the correlation strength of parameters, overlay data of 3D assembly models based on the distribution of geometric deviations, and assembly station data dashboards, thereby supporting multi-dimensional data linkage queries, abnormal data warning pop-ups, and historical data retrospective display.

[0037] S2 specifically includes: S21: High-fidelity virtual prototype construction: Utilizing digital twin technology, this system integrates real-world assembly environment parameters, equipment parameters, and operational constraints to construct a high-fidelity real-time mapping relationship between the physical entity and the virtual model. Real-world assembly environment parameters include on-site temperature, humidity, and spatial layout; equipment parameters include structural, operational, and precision parameters of the tooling equipment; and operational constraints include time-varying parameters related to temperature fluctuations, equipment wear, material property dispersion, hole-making force, riveting force, and the force exerted by the product on the tooling positioning and clamping device. This virtual assembly prototype reproduces the geometric constraints, force-thermal coupling, and multi-physics behavior of material deformation throughout the entire assembly process.

[0038] S22: Virtual Pre-simulation Verification of Assembly Quality: Based on the constructed high-fidelity virtual assembly prototype, a digital test plan is formulated to conduct dynamic simulation and test verification of the entire assembly process in virtual space; for typical assembly scenarios such as composite panel assembly, wing-fuselage docking, and titanium alloy-composite material stacking connection, the assembly process is simulated under different positioning sequences, clamping force magnitudes, and process parameter combinations to identify in advance the risk of delamination damage caused by uneven or excessive clamping force, geometric deviation caused by unreasonable positioning sequence, and insufficient connection strength caused by mismatched process parameters, and optimize the process plan to reduce the number of physical trial and error attempts.

[0039] S23: Intelligent Planning of Assembly Process Scheme: Employing form-property co-optimization, this approach breaks through the traditional single-dimensional control of geometric accuracy, prioritizing assembly physical performance and geometric accuracy. Assembly physical performance includes internal stress, connection strength, fatigue life, and damage state. Utilizing high-fidelity modeling and simulation with model calibration, the system focuses on controlling the geometric deformation and physical performance of the assembly structure. In a virtual assembly environment, it rationally arranges the execution sequence of assembly positioning, drilling, connection, and finishing tasks, optimizing equipment runtime. It pre-acquires and verifies the form-property assembly state, repair area, and repair quantity corresponding to different assembly process schemes and process parameters, employing closed-loop feedback correction techniques such as optimized shape correction and clamping force, gap compensation, real-time flexible positioning control, and pre-control of assembly simulation. For special scenarios involving the lamination of titanium alloys and composite materials, it simultaneously optimizes key parameters such as drilling speed, riveting force, and connection force application time, finding the balance control point of form-property parameters through multi-objective optimization algorithms. For weak-rigidity thin-walled structure assembly scenarios, it focuses on controlling the internal stress of the assembly process to not exceed the specified threshold of the material's yield strength, while ensuring that geometric accuracy meets design requirements.

[0040] Furthermore, shape-property synergistic control simultaneously optimizes geometric deformation, assembly internal stress, and connection strength physical properties through finite element simulation and multi-objective optimization algorithms.

[0041] S3 specifically includes: S31: Pre-assembly process multi-objective parameter optimization: Considering workpiece manufacturing errors, material properties, and actual assembly conditions on site, as well as the pre-processed data of assembly process and result quality in S1, a multi-objective optimization model including geometric accuracy, physical performance, and cost factors is established; during the assembly positioning process, combined with specific positioning and clamping schemes, clamping positions, and load constraints, finite element simulation analysis based on a high-fidelity virtual prototype model and mechanical deformation analysis of the assembly positioning process are used to obtain the deformation of the machine body structure under different positioning schemes, and the optimal positioning combination is solved by combining intelligent optimization algorithms to control the geometric errors of the assembly parts; for the cutting speed, cutting force, spindle speed, feed rate, and feed amount process parameters in the assembly hole drilling process, and the riveting force / Preload, tightening torque, connection speed, connection time, and connection sequence process parameters, as well as injection pressure, feed rate, depth of cut, cutting speed, end trajectory, and stroke process parameters during assembly finishing, are determined through orthogonal experiments. By combining theoretical modeling, simulation analysis, and experimental data processing, a correlation model and optimization solution model are established between different process parameters and connection deformation and the balanced distribution of connection residual stress, so as to obtain an optimized assembly scheme and process parameters before assembly operations.

[0042] S32: Adaptive Adjustment of Assembly Process Parameters: Based on the real-time sensing and measurement data of the entire assembly process in S1 and the assembly process plan in S2, a closed-loop control system of "perception-analysis-decision-execution" is constructed. At the assembly site, multi-dimensional sensors deployed on the tooling equipment collect process parameters such as drilling force, connection temperature, workpiece deformation, and equipment operating parameters in real time. After rapid processing by the edge computing unit, the data is analyzed using machine learning and deep reinforcement learning data mining analysis algorithms. Based on the analysis results, the operating parameters of the tooling equipment are coupled and adjusted to effectively reduce the power consumption of the equipment. Based on the actual working conditions at the assembly site and the collected process quality data, an integrated control model of the form-based assembly state is constructed. Through dynamic adjustment of process parameters, the quality indicators of assembly geometric pose changes and assembly physical state distribution quickly meet the design requirements, while adapting to the assembly requirements of similar types of parts.

[0043] S33: Intelligent Expansion of Assembly Test Data and Construction of Process Knowledge Base: Employing Generative Adversarial Networks (GANs), Bayesian optimization few-shot learning techniques, and transfer learning methods, based on a limited number of process test data, and combined with a quality loss model incorporating physical mechanisms, the system quickly adapts to new model parameters, predicts and generates multiple sets of high-fidelity virtual samples, and accurately establishes the mapping relationship between process parameters and assembly quality under different material combinations and operating conditions. It integrates assembly parameters with historical quality records, simulation results, and on-site feedback data to construct a self-updating process parameter database and knowledge base, forming an intelligent decision-making chain of "parameter recommendation - effect prediction - deviation correction," thereby improving the efficiency of assembly process design and reducing the cost of trial and error in process parameters.

[0044] Furthermore, the self-updating function of the process parameter database and knowledge base is achieved through incremental learning. When new process test data, simulation results, or field feedback data are added, the knowledge base update process is automatically triggered, the process parameter-assembly quality mapping model is retrained, and the association rules and decision logic are updated. The knowledge base storage adopts a distributed database architecture, which supports efficient storage and fast query of massive amounts of data.

[0045] S4 specifically includes: S41: Assembly Deviation Characterization and Correlation Modeling: Based on the time-varying coupling and transmission between real geometric and physical assembly state parameters, this study comprehensively analyzes the geometric fit clearance, stress distribution, and damage morphology data results of the dimensional coordination / positioning clamping / hole making and connection / finishing processes during multi-process assembly to characterize the accuracy and consistency deviation status between assembly form targets. Secondly, a mechanism-data hybrid modeling approach is adopted. The mechanism model is used to analyze the influence mechanism of positioning clamping force and hole making feed rate parameters on deviations (e.g., excessive clamping force on thin-walled parts can lead to warping deformation), as well as the transmission and evolution analysis of the quality state during the assembly process. The data model is based on historical assembly data / actual measurements. Data / virtual simulation data are used to learn high-dimensional mapping relationships, compensating for the insufficient accuracy of mechanistic models under complex assembly conditions. This allows for the construction of a rapid prediction and quantitative correction model for the formation and transmission process of assembly accuracy / mechanics, thereby expanding the amount of small-sample assembly quality data. Subsequently, the high-dimensional mapping function relationships between geometric / physical multi-source heterogeneous data types are explored. Association rule algorithms (such as the Apriori method) are used to mine the bidirectional mapping relationship between core process parameter groups (such as the number of positioning points, drilling speed, and riveting force) and deviations, forming a bidirectional association rule between the accuracy / consistency deviation of form assembly and the core process parameter group, providing a precise basis for rapid on-site adjustment of process parameters.

[0046] Furthermore, the deviation characterization adopts a mechanism-data hybrid modeling approach. The mechanism model is constructed based on the theories of materials mechanics, structural mechanics, and thermodynamics to analyze the influence mechanism of process parameters on deviations. The data model uses neural networks and support vector machines to learn high-dimensional mapping relationships based on historical data. The hybrid modeling integrates the outputs of the mechanism model and the data model through weighted fusion, with the weight coefficients dynamically adjusted according to the prediction accuracy of the models to ensure the accuracy of the data characterization of deviations in geometric fit clearance, stress distribution, and damage morphology.

[0047] S42: Assembly Deviation Source Tracing: Using global sensitivity analysis (such as the Sobol's index method) to calculate the contribution of each process parameter to the deviation, quantifying the influence weight of each link, and based on the process deviation transmission path, studying the source tracing method for the difference in the distribution of deviations between form assembly deviations and interchangeability consistency, obtaining the effect and specific contribution value of key "form / force" coupling factors in the process links on the balance and harmony between form assembly states and the improvement of interchangeability consistency; analyzing the bidirectional effect of holographic process elements on the geometric errors, internal stresses and damage states of different processes, and under the collaborative constraints of form assembly deviation states (such as using the allowable deviation threshold as a constraint), constructing a system-level inversion control model for the coordination and interchangeability of assembly quality parameters for the group of on-site process parameter variables.

[0048] S43: Multidimensional Dynamic Control of Assembly Quality: Employing a hybrid approach of geometric-physical assembly data empowerment, and a hierarchical and parallel solution strategy combining overall and local aspects, this approach uses reverse reasoning to determine the controllability of parameter adjustments for weak process variables in the preceding stages. It constructs a multidimensional dynamic assembly adjustment mechanism encompassing error adaptive allowance allocation, secondary correction of positioning / hole making / connection / finishing process parameters, and pre-control design for repair compensation. Error adaptive allowance allocation reduces quality requirements in manufacturing and assembly processes; secondary correction of process parameters dynamically adjusts on-site assembly process parameters based on real-time measurement data; and pre-control design for repair compensation predicts repair areas and presets repair quantity thresholds, reducing on-site padding / grinding repair operation time. A complete toolchain including modeling, solving, and optimization is developed to achieve software-based output from data input to on-site process control solutions, shortening the on-site processing and calculation time for process control quantities.

[0049] Furthermore, the hierarchical solution strategy of the assembly process integrated control mechanism is divided into global, local, and execution levels. The global level is responsible for optimizing and controlling the overall process plan, the local level is responsible for adjusting the process parameters of individual processes, and the execution level is responsible for controlling the motion of tooling equipment. The parallel solution strategy uses multi-threading technology to achieve parallel processing of control tasks at different levels and in different stages, improving control efficiency. The software output of the toolchain includes process parameter adjustment lists, equipment operation manuals, and repair area and repair quantity descriptions, which can be directly imported into the control system and MES system of tooling equipment to achieve rapid execution of assembly quality control plans.

[0050] In summary, the present invention can achieve the following: 1) Improve the green level of assembly processes and reduce resource consumption and environmental impact. This invention uses virtual pre-simulation verification with a high-fidelity virtual assembly prototype to identify problems such as unreasonable positioning and parameter mismatch in the assembly process in advance, significantly reducing the number of physical trial and error attempts. Combined with small sample learning technology to generate high-fidelity virtual samples, it avoids material waste caused by repeated repairs and experiments in traditional processes. By adaptively adjusting the process parameters of the assembly process and optimizing the coupled parameters of equipment operation, it reduces the ineffective operating time of equipment and lowers energy consumption. At the same time, the repair compensation pre-control design reduces the time spent on on-site padding and grinding repair operations, and reduces the generation of waste during the repair process, fully meeting the core requirements of green manufacturing: "low consumption, low emissions, and high efficiency".

[0051] 2) Achieve precise control and form-property synergy in assembly quality control, thereby improving product reliability. This invention breaks through the traditional single-control mode of geometric precision, placing geometric deformation on an equal footing with assembly internal stress, connection strength, and fatigue life physical properties. Through form-property synergy control and mechanism-data hybrid modeling, it improves the assembly geometric precision pass rate and strictly controls assembly internal stress within the material yield strength threshold. By finding the balance control point of form-property parameters through multi-objective optimization algorithms, the fatigue life of connection parts can be effectively improved. With the help of assembly deviation characterization and out-of-tolerance tracing technology, the root cause of quality problems can be accurately located, avoiding batch quality defects and significantly improving the assembly consistency and operational reliability of aircraft airframe structures.

[0052] 3) Significantly improves assembly efficiency and shortens production cycles and process design costs. This invention achieves efficient interoperability and rapid processing of multi-source heterogeneous data through standardized data interfaces, hybrid storage architecture, and data preprocessing technology, controlling data processing latency and supporting real-time response to process decisions; it constructs a self-updating process knowledge base and an intelligent decision-making chain of "parameter recommendation-effect prediction-deviation correction," which improves process design efficiency, controls knowledge base query response time, and significantly reduces reliance on operator experience; and it constructs a virtual simulation and closed-loop control mechanism, which reduces physical testing and rework time, shortens the overall assembly cycle, reduces process trial-and-error costs, and effectively improves the production efficiency and market responsiveness of aerospace manufacturing.

[0053] 4) Enhance the intelligence and versatility of processes to adapt to complex scenarios and new aircraft requirements. The "perception-analysis-decision-execution" closed-loop control system constructed in this invention, combined with machine learning and deep reinforcement learning algorithms, enables adaptive adjustment of process parameters. It can quickly adapt to various complex assembly scenarios such as composite panel assembly, wing-fuselage docking, and dissimilar material layering. Through transfer learning and small-sample expansion technology, it can quickly adapt to new aircraft parameters without large-scale re-experimentation of processes, reducing the transition costs of new aircraft R&D and mass production. The distributed edge computing architecture and software-based control toolchain support the direct import of control schemes into tooling equipment control systems and MES systems, enabling rapid execution of process control and improving the engineering practicality and promotion value of the technical solution.

[0054] 5) Construct a data-driven, end-to-end traceability system to enhance the transparency of process control. This invention achieves end-to-end traceability of the assembly process from data input to control output through the perception, standardized storage, and visualization of multi-source data throughout the entire process. It supports multi-dimensional data linkage queries, abnormal data early warnings, and historical data backtracking. The process knowledge base integrates historical data, simulation results, and on-site feedback to form reusable and self-updating process data assets, providing precise data support for subsequent process optimization and quality problem tracing, and improving the transparency and replicability of aerospace manufacturing process control.

[0055] Example 1: Green Upgrading of Composite Panel Assembly Process • Data sensing and measurement: Laser scanning equipment is used to collect manufacturing errors and assembly geometry of composite panel components. Strain gauges and multi-dimensional force sensors are used to collect clamping force and stress-strain data. Temperature sensors are used to monitor the temperature and humidity at the assembly site. • Data preprocessing: Data is transmitted via OPC UA interface and JSON format, with 32-bit cyclic redundancy check to ensure integrity, MySQL to store clamping force and dimensional structured data, MongoDB to store sensor configuration information, and MinIO to store assembly process images; abnormal stress data is removed using the 3σ principle, missing temperature data is completed using the K-nearest neighbor method, and Z-score is used for standardization and unified dimensions. • Process planning: Construct a high-fidelity virtual prototype of the composite wall panel, simulate the interlayer stress distribution under different positioning sequences, identify the risk of delamination caused by uneven clamping force, and optimize the positioning scheme to a symmetrical positioning of "edge first, center later"; • Parameter optimization: Before assembly, the combination of clamping force and number of positioning points is determined through orthogonal experiments. Combined with finite element simulation and genetic algorithm, the optimal clamping force is found to be 534N and the number of positioning points is 6. During the assembly process, the clamping force is dynamically adjusted based on real-time strain gauge data through deep reinforcement learning algorithm, and the shape deviation is controlled within ±0.5mm. • Quality control: Mechanism-data hybrid modeling is used to characterize the relationship between assembly clearance and clamping force. The contribution of clamping force to deviation is determined to be 75% by the Sobol' index method. An error tolerance allocation and repair pre-control mechanism is constructed, and the final assembly rework rate is reduced from 18% to 3%.

[0056] Example 2: Assembly of titanium alloy and composite dissimilar materials in laminated connection • Data perception: Laser rangefinder and roughness meter are used to collect data on hole diameter and hole wall roughness. Torque wrench and dynamic torque sensor are used to collect data on riveting force and tightening torque. Ultrasonic detection scanning electron microscope is used to collect data on damage at the connection parts. • Process planning: Construct a high-fidelity virtual prototype for the connection of dissimilar materials, reproduce the force-thermal coupling behavior of the hole making and riveting process, and determine the "form-property synergy" optimization goal as meeting the hole wall quality and the connection strength meeting fatigue requirements; • Parameter optimization: Before assembly, the optimal drilling speed was determined to be 3000 r / min, the riveting force to be 8 kN, and the connection force application time to be 2 s through orthogonal experiments and multi-objective optimization algorithms; during assembly, the riveting force fluctuation range was adaptively adjusted to ±0.3 kN based on real-time data from the force sensor. • Quality control: Apriori-T time series association rules are used to explore the mapping relationship between drilling speed, riveting force and connection damage. Global sensitivity analysis is used to locate drilling speed as the main influencing factor of deviation. A secondary parameter correction mechanism is constructed, which ultimately improves the fatigue life of the connection part by 22%, reduces drilling tool wear by 28%, and reduces energy consumption by 21%.

[0057] The above embodiments demonstrate that this invention, by constructing a closed-loop technical system encompassing "data perception, intelligent planning, parameter optimization, and quality control," addresses the core pain points of traditional aircraft airframe assembly processes. It achieves a synergistic improvement in the greening, precision, and efficiency of the assembly process. This not only solves the technical problems of high trial-and-error costs, passive quality control, and low greening levels in traditional assembly processes, but also realizes an intelligent upgrade of the assembly process to be "measurable, controllable, visible, and traceable." It provides a scientific, efficient, and green technical solution for aircraft airframe assembly, possessing significant engineering application value and market prospects. It can be effectively implemented in the green assembly practices of aviation manufacturing enterprises and exhibits good practicality and operability.

[0058] The foregoing has provided a detailed description of a method for improving the greenness of aircraft airframe structure assembly process design and quality control, as provided in the embodiments of the present invention. The descriptions of the above embodiments are merely for the purpose of helping to understand the method and core ideas of the present invention; furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

[0059] Certain terms are used in the specification and claims to refer to specific components. Those skilled in the art will understand that hardware manufacturers may use different names to refer to the same component. This specification and claims do not distinguish components based on differences in name, but rather on differences in function. The terms "comprising" and "including" used throughout the specification and claims are open-ended and should be interpreted as "comprising / including but not limited to". "Substantially" means within an acceptable margin of error, indicating that those skilled in the art can solve the technical problem and substantially achieve the technical effect within a certain margin of error. The following descriptions are preferred embodiments for carrying out the invention; however, these descriptions are intended to illustrate the general principles of the invention and are not intended to limit the scope of the invention. The scope of protection of this invention is determined by the appended claims.

[0060] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a product or system comprising a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a product or system. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the product or system that includes said element.

[0061] It should be understood that the term "and / or" used in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0062] The foregoing description illustrates and describes several preferred embodiments of the present invention. However, as previously stated, it should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the inventive concept described herein through the foregoing teachings or techniques or knowledge in related fields. Any modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A method for improving the greenness of aircraft fuselage structure assembly process design and quality control, characterized in that, Includes the following steps: S1: Perception, measurement and preprocessing of multi-source quality data of assembly process and results: collect geometric, physical and functional data of the entire assembly process, realize data interoperability through standardized interfaces and communication protocols, classify and store heterogeneous data using a hybrid storage architecture, and build a unified perception data model through time alignment, outlier removal, feature extraction and data fusion preprocessing operations. S2: Intelligent Assembly Process Planning: Construct a high-fidelity virtual assembly prototype, reproduce the multi-physics behavior of assembly, and conduct virtual pre-simulation verification; with "shape-property collaborative control" as the goal, optimize the assembly task arrangement and equipment operating parameters of the assembly process plan to achieve simultaneous assurance of geometric accuracy and physical performance; S3: Assembly process parameter optimization design: Based on the preprocessing data of assembly process and result quality in S1, and the assembly process scheme in S2, a multi-objective optimization model is established before assembly to solve the optimal combination of positioning, hole making, connection and finishing parameters. During the assembly process, a closed-loop control system is built based on real-time sensing data to realize adaptive adjustment of process parameters. Expand process data through few-shot learning techniques and build a self-updating assembly quality data knowledge base; S4: Intelligent process control of on-site assembly quality: The mechanism-data hybrid modeling method is used to characterize assembly deviations. A quantitative correlation model between assembly deviations and process parameters in the assembly quality data knowledge base in S3 is constructed. Global sensitivity analysis is used to trace the source of deviations. A multi-dimensional dynamic control mechanism for process error allocation, process parameter correction and repair pre-control is established by adopting a hierarchical and parallel solution strategy, and the process control scheme is output.

2. The method for improving the greenness of aircraft fuselage structure assembly process design and quality control according to claim 1, characterized in that, S1 specifically includes: S11: Multi-source Assembly Quality Data Measurement and Sensing: For the pre-assembly inspection, positioning, hole making, connection, finishing and final inspection stages, three types of data are collected: geometric quantities, physical quantities and functional quantities. The specific measurement equipment and requirements correspond as follows: geometric quantity data is collected using contact and non-contact equipment; physical quantity data is collected through force sensors, strain gauges and temperature sensors; functional quantity data is obtained using universal testing machines and ultrasonic non-destructive testing equipment. S12: Standardization and Storage of Multi-Source Assembly Quality Data: Through standardized interfaces such as RESTful API, OPC UA, and MQTT, and industrial communication protocols such as OPC UA and Modbus TCP, the data transmission formats of JSON and Protobuf are unified, and 32-bit cyclic redundancy check is used to ensure data integrity; a hybrid storage architecture is constructed, using MySQL / PostgreSQL relational databases to store structured data, MongoDB document databases to store semi-structured data, and MinIO / S3 object storage services to store unstructured data, and table partitioning and database partitioning strategies and index structures are designed; S13: Preprocessing and Fusion of Multi-Source Assembly Quality Data: Based on a unified timestamp benchmark, linear interpolation is used to fill in missing time-series data, and data from different acquisition frequencies are synchronized and aligned using a sliding window; outliers are removed using the 3σ principle or box plot method, and missing data are filled with the mean / median or K-nearest neighbor method, and the dimensions are unified by Z-score standardization or Min-Max normalization; correlation is analyzed using Pearson / Spearman coefficients, and data features are extracted using convolutional neural networks and long short-term memory networks; raw data level, feature level, and decision level fusion are achieved based on weighted average method, Kalman filtering, principal component analysis, and DS evidence theory, and a unified perception data model for ontology modeling is constructed, and data visualization is achieved through ECharts, Tableau, and Unity 3D.

3. The method for improving the greenness of aircraft fuselage structure assembly process design and quality control according to claim 1, characterized in that, S2 specifically includes: S21: High-fidelity virtual prototype construction: Using digital twin technology, integrating real assembly environment, equipment parameters, and working condition constraints, a high-fidelity real-time mapping between physical entities and virtual models is constructed to reproduce the geometric constraints, force-thermal coupling, and multi-physics behavior of material deformation in the assembly process. The working condition constraints include, but are not limited to, temperature fluctuations and equipment wear. S22: Virtual pre-simulation verification of assembly quality: Conduct dynamic simulation of the entire assembly process in virtual space to identify assembly quality problems such as unreasonable positioning sequence, excessive clamping force causing delamination damage, and warping deformation in advance, and optimize process plans to reduce physical trial and error; S23: Intelligent planning of assembly process scheme: By adopting the method of shape-property synergistic optimization, it breaks through the traditional single dimension of geometric accuracy control, and puts geometric deformation, assembly internal stress, connection strength and physical properties on the same priority. Through model calibration and correction, it optimizes the arrangement of assembly positioning, hole making, connection, and finishing tasks and equipment running time. It adopts closed-loop feedback correction technology to precisely control the assembly process and form an optimized assembly process scheme. The closed-loop feedback correction technology includes, but is not limited to, optimization and correction and flexible positioning control.

4. The method for improving the greenness of aircraft fuselage structure assembly process design and quality control according to claim 1, characterized in that, S3 specifically includes: S31: Pre-assembly process multi-objective parameter optimization: Considering workpiece manufacturing error, material properties and on-site working conditions, as well as the pre-processing data of assembly process and result quality in S1, orthogonal experiments are used to determine parameter combinations and establish a multi-objective optimization model composed of geometric accuracy, physical performance and cost. S32: Adaptive adjustment of process parameters in assembly operation: Based on the assembly process scheme in S2 and the real-time data collected by multi-dimensional sensors deployed in tooling equipment in S1, the data is quickly processed by the edge computing unit and analyzed by machine learning and deep reinforcement learning algorithms to solve for the optimal parameters that meet the geometric error constraints and material yield strength threshold. The process parameters of hole drilling speed, riveting force, and feed speed are dynamically adjusted to reduce equipment power consumption and enable the assembly quality indicators to quickly meet the design requirements. S33: Intelligent Expansion of Assembly Test Data and Construction of Process Knowledge Base: Using generative adversarial networks, Bayesian optimization, and transfer learning few-shot learning techniques, high-fidelity virtual samples are generated based on limited process test data to establish a mapping relationship between process parameters and assembly quality; historical data, simulation results, and field feedback are integrated to construct a self-updating process parameter database and knowledge base, forming an intelligent decision chain of "parameter recommendation - effect prediction - deviation correction" to reduce the trial and error cost of process parameters.

5. The method for improving the greenness of aircraft fuselage structure assembly process design and quality control according to claim 1, characterized in that, S4 specifically includes: S41: Assembly Deviation Characterization and Correlation Modeling: Based on the time-varying coupling and transmission of geometric-physical assembly state, a mechanism-data hybrid modeling approach is adopted to comprehensively obtain data on dimensional coordination, stress distribution, and damage morphology to characterize the assembly deviation state of the shape target; combined with the Apriori association rule algorithm, a two-way mapping relationship between deviation and core process parameters is constructed. S42: Assembly deviation source tracing: Using the Sobol's index method for global sensitivity analysis, the first-order sensitivity index and total sensitivity index of each process parameter to deviation are calculated, and the contribution is comprehensively formed to quantify the influence weight of each link, analyze the transmission path of key deviation sources in the assembly process, and locate key influencing factors and root causes of deviation. S43: Multi-dimensional dynamic control of assembly quality: By reverse reasoning the parameter adjustment rules of preceding process variables, an intelligent multi-dimensional control mechanism for multi-source assembly quality data is constructed, including error adaptive allowance allocation, secondary correction of process parameters and pre-control design measures for repair and compensation. A complete toolchain of modeling-solving-optimization is developed to output a software-based intelligent process control scheme.

6. The method for improving the greenness of aircraft fuselage structure assembly process design and quality control according to claim 1, characterized in that, The geometric data in S1 includes component manufacturing errors, assembly geometry, and key feature spatial positions, which are collected by a coordinate measuring machine, laser scanning equipment, and laser tracker. The physical data includes assembly load, temperature field, stress and strain, and equipment operating parameters. The functional data includes stiffness, load-bearing capacity, and fatigue life.

7. The method for improving the greenness of aircraft fuselage structure assembly process design and quality control according to claim 1, characterized in that, The S2 form-property collaborative control uses finite element simulation and multi-objective optimization algorithms to simultaneously optimize geometric deformation, assembly internal stress, and connection strength physical properties.

8. The method for improving the greenness of aircraft fuselage structure assembly process design and quality control according to claim 4, characterized in that, The self-updating function of the process parameter database and knowledge base in S3 is achieved through incremental learning. When new process test data, simulation results or field feedback data are added, the knowledge base update process is automatically triggered, the process parameter-assembly quality mapping model is retrained, and the association rules and decision logic are updated. The knowledge base is stored using a distributed database architecture.

9. The method for improving the greenness of aircraft fuselage structure assembly process design and quality control according to claim 1, characterized in that, In the mechanism-data hybrid model representing assembly deviations in S4, the mechanism model is constructed using the theories of materials mechanics, structural mechanics, and thermodynamics to analyze the influence mechanism of process parameters on deviations; the data model adopts neural network and support vector machine machine learning models to learn high-dimensional mapping relationships based on historical data; the hybrid modeling integrates the output results of the mechanism model and the data model through weighted fusion, and the weight coefficients are dynamically adjusted according to the prediction accuracy of the model to ensure the accuracy of the data representation of geometric fit clearance, stress distribution, and damage morphology deviations.

10. The method for improving the greenness of aircraft fuselage structure assembly process design and quality control according to claim 5, characterized in that, In S4, a hierarchical and parallel solution strategy is adopted to establish a multi-dimensional dynamic control mechanism for process error allocation, process parameter correction, and pre-control of adjustment. The output process control scheme specifically includes: The hierarchical solution strategy is divided into global, local, and execution levels. The global level is responsible for optimizing and controlling the overall process plan, the local level is responsible for adjusting the process parameters of individual processes, and the execution level is responsible for controlling the motion of tooling equipment. The parallel solution strategy uses multi-threading technology to achieve parallel processing of control tasks at different levels and in different stages, improving control efficiency. The software output of the toolchain includes process parameter adjustment lists, equipment operation manuals, and repair area and repair quantity descriptions, which can be directly imported into the control system and MES system of tooling equipment to achieve rapid execution of assembly quality control plans.

Images