A synchronous tensioning method and a tensioning structure for an automobile ceiling cloth

Abstract

The present invention relates to a synchronous tensioning method and a tensioning structure for an automobile ceiling cloth, and is characterized in that: it includes a tensioning table, a ceiling cloth clamping module, a tensioning connecting rod and a tensioning transmission plate; high-risk areas prone to local wrinkles and unevenness in the ceiling cloth are identified through data; a topology optimization algorithm is applied to design the tensioning points on the ceiling cloth tensioning structure; a three-dimensional solid model of the ceiling cloth tensioning structure is established in computer-aided design software according to the optimized tensioning point parameters; during the tensioning process, machine vision technology is used to monitor the morphology and stress distribution of the ceiling cloth in real time; feature points on the surface of the ceiling cloth are extracted through an image processing algorithm, and the displacement and strain of these feature points are calculated to determine whether the expected tensioning effect is achieved; if local wrinkles or unevenness appear in the ceiling cloth during monitoring, a convolutional neural network is used to analyze the shape characteristics of the wrinkled area; after the tensioning is completed, a three-dimensional scanner is used to obtain the point cloud data of the ceiling surface.

Description

Technical Field

[0001] The invention relates to the technical field of automobile product processing, and in particular to a synchronous tensioning method for automobile roof cloth and a tensioning structure thereof. Background Art

[0002] When covering the roof cloth at the car sunroof, a tensioning structure is generally used to tension the ceiling cloth for covering the car roof sunroof; a common tensioning method is to use multiple groups of cylinders set on the same tensioning platform, and multiple groups of cylinders synchronously drive the tensioning; this method of tensioning with cylinders has the following problems: first, due to the large number of tensioning directions and positions of the ceiling cloth, a large number of cylinder units are used, resulting in high cost, large installation space, and very inconvenient operation; second, the multi-group structural design of the cylinder cannot meet the tensioning effect and tensioning force control in the control of synchronous tensioning, and it is easy for the ceiling cloth to be torn and wrinkled due to uneven tensioning force during tensioning.

[0003] In addition, in the synchronous tensioning process of the automobile ceiling cloth, the design of the ceiling cloth tensioning structure is crucial. The traditional ceiling cloth tensioning structure design usually adopts tensioning points that are flush with the ceiling surface. Although this design can achieve the tensioning of the ceiling cloth, there are problems with uneven transmission and distribution of tensioning force. When the ceiling cloth is subjected to external forces during the tensioning process, the flush tensioning points cannot provide sufficient support and fixation, causing the ceiling cloth to be prone to local wrinkles, unevenness and other problems. At the same time, the flush tensioning points also lack sufficient space and operational convenience during the installation and fixation of the ceiling cloth, which increases the difficulty and time cost of the operation.

[0004] In response to the above problems, this paper proposes a synchronous tensioning method for automobile roof cloth, the core of which is to design the tensioning structure of the roof cloth. By designing tensioning points on the tensioning structure of the roof cloth, the technical contradictions existing in the traditional flush tensioning points can be effectively solved. The tensioning points can provide greater tensioning force during the tensioning process of the ceiling cloth, ensuring that the ceiling cloth is evenly stressed and avoiding local wrinkles and unevenness. At the same time, the tensioning points also provide a larger operating space for the installation and fixation of the ceiling cloth, which is convenient for the operation of the operators and improves the work efficiency. However, the design of the tensioning structure of the ceiling cloth also brings new technical challenges. How to reasonably arrange the tensioning points in a limited space, ensure the strength and stability of the tensioning points, and the impact on the vehicle styling and aerodynamic performance, etc., all require further research and optimization. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a synchronous tensioning method for an automotive ceiling cloth and its tensioning structure, which can solve the problems that the traditional automotive ceiling cloth uses a cylinder for tensioning, the synchronization cannot be achieved consistently, and the conventional synchronous tensioning structure does not match well with the ceiling cloth in terms of tensile force before tensioning, resulting in poor tensioning effect of the ceiling cloth.

[0006] To solve the above technical problems, the technical solution of the present invention is: A synchronous tensioning method for an automotive ceiling cloth, characterized in that the specific tensioning method is as follows:

[0007] S1: Using the finite element analysis method, according to the three-dimensional model data of the ceiling cloth tensioning structure, simulate the stress distribution of the ceiling cloth under different tensile forces, and obtain the stress nephogram and deformation nephogram of the ceiling cloth;

[0008] S2: Through these data, identify the high-risk areas in the ceiling cloth that are prone to local wrinkles and unevenness; in the identified high-risk areas, apply the topology optimization algorithm to design the tensioning points on the ceiling cloth tensioning structure; optimize the position parameters of these tensioning points to ensure that these points can provide sufficient support and tensile force in the high-risk areas, while meeting the strength and stability requirements of the ceiling cloth tensioning structure; the objective function of the topology optimization algorithm is set to minimize the stress concentration and deformation of the ceiling cloth, and the constraint conditions include the number of tensioning points, the position range, and the structural strength of the ceiling cloth tensioning structure;

[0009] S3: Establish a three-dimensional solid model of the ceiling cloth tensioning structure in computer-aided design software according to the optimized tensioning point parameters; perform interference checking and assembly simulation to ensure the mating relationship between the tensioning points and the ceiling cloth and adjacent components, and ensure the installation space and operation convenience of the ceiling cloth;

[0010] S4: During the tensioning process of the ceiling cloth, use machine vision technology to monitor the morphology and stress distribution of the ceiling cloth in real time; extract the feature points on the surface of the ceiling cloth through image processing algorithms, calculate the displacement and strain of these feature points to determine whether the ceiling cloth reaches the expected tensioning effect; if it is detected that the ceiling cloth is locally wrinkled or uneven, use a convolutional neural network to analyze the shape characteristics of the wrinkled area; this network is trained through pre-collected ceiling cloth wrinkle samples, combined with the layout parameters of the tensioning points of the ceiling cloth tensioning structure, predict the possible causes, and propose suggestions for adjusting the tensile force or optimizing the tensioning points;

[0011] S5: After the ceiling cloth is tensioned, use a three-dimensional scanner to obtain the point cloud data of the ceiling surface; generate a digital model of the ceiling through a surface reconstruction algorithm, compare it with the design model, quantitatively evaluate the overall tensioning quality and dimensional accuracy of the ceiling cloth, and form a quality report.

[0012] An automobile ceiling cloth synchronous tensioning structure, the innovation of which lies in: including a tensioning table, a ceiling cloth clamping module, tensioning connecting rods and a tensioning transmission plate;

[0013] The tensioning table includes a table board and support guide columns; the table board is in a cuboid plate structure, and a transmission hole is provided at the center position of the table board; a driving motor is provided on one side of the transmission hole on the lower surface of the table board, the output end of the driving motor is connected to a gearbox, and a transmission shaft is provided at the output end of the gearbox, and the transmission shaft passes through the transmission hole at the center of the table board and extends to the upper surface of the table board; the tensioning transmission plate is parallel to the table board and is arranged above the table board, the tensioning transmission plate is in a waist-shaped plate structure, and the center position of the tensioning transmission plate is connected to the end of the transmission shaft; several first tensioning points of the tensioning connecting rods are arranged at the edge of the tensioning transmission plate; the support guide columns are vertically arranged at the corner positions of the lower surface of the table board;

[0014] The ceiling cloth clamping module includes slide rails, sliders and clamping plates; the slide rails are installed at the side edges and corner positions of the table board; the sliders are installed on the slide rails; the clamping plates are installed at the top positions of the sliders, and one side edge of the clamping plate extends out of the end of the slider; a longitudinal clamping cylinder is arranged at the end of the slider, and the output end of the longitudinal clamping cylinder cooperates with the lower surface of the clamping plate to form a clamping gap for clamping the automobile ceiling cloth; a second tensioning point of the tensioning connecting rod is arranged on the upper surface of the clamping plate;

[0015] The tensioning connecting rod includes an adjusting screw sleeve, a connecting screw rod and a rotary shaft sleeve; there are a pair of connecting screw rods which are respectively arranged at both ends of the adjusting screw sleeve, and the length of the tensioning connecting rod is adjusted by rotating the adjusting screw sleeve; the rotary shaft sleeve is installed at the ends of the two connecting screw rods; positioning shafts are arranged at both the first tensioning point and the second tensioning point of the tensioning connecting rod, and rotary bearings are arranged on the positioning shafts and cooperate with the rotary shaft sleeves respectively.

[0016] The advantages of the present invention are as follows:

[0017] 1) In the present invention, the stress distribution of the ceiling cloth under different tensile forces is simulated through finite element analysis, high-risk areas are identified, and a topology optimization algorithm is applied to design the tensioning points on the ceiling cloth tensioning structure. The influence of the optimized tensioning points on the aerodynamic performance is analyzed by using the computational fluid dynamics method, and parametric design optimization is carried out. During the tensioning process, machine vision technology is used to monitor the morphology and stress distribution of the ceiling cloth in real time, a convolutional neural network is used to analyze the wrinkle characteristics and put forward adjustment suggestions. Finally, the tensioning quality is evaluated through three-dimensional scanning and surface reconstruction. The present invention realizes the intelligent optimization of the ceiling cloth tensioning process, improves the flatness and tensioning quality of the ceiling cloth, and at the same time takes into account the aerodynamic performance of the vehicle, providing an innovative solution for the design and production of automobile ceiling cloth. Description of the Drawings

[0018] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0019] Figure 1 It is a flow chart of a synchronous tensioning method for an automobile ceiling cloth according to the present invention.

[0020] Figure 2 It is a schematic diagram of a synchronous tensioning method for an automobile ceiling cloth according to the present invention.

[0021] Figure 3 It is another schematic diagram of a synchronous tensioning method for an automobile ceiling cloth according to the present invention.

[0022] Figure 4 It is a schematic structural diagram of a synchronous tensioning of an automobile ceiling cloth according to the present invention. Specific Embodiments

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Usually, the components of the embodiments of the present invention described and illustrated herein can be arranged and designed in various different configurations.

[0024] Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed present invention, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0025] As Figures 1 to 3 shown in a synchronous tensioning method for an automobile ceiling cloth, in step S101, using the finite element analysis method, according to the three-dimensional model data of the ceiling cloth tensioning structure, the stress distribution of the ceiling cloth under different tensile forces is simulated, and the stress nephogram and deformation nephogram of the ceiling cloth are obtained.

[0026] According to the three-dimensional model data of the canopy fabric tension structure, a finite element analysis model is established, and the material properties and boundary conditions are set; for different tensile force conditions, the load types and magnitudes are defined and applied to the finite element model; the finite element analysis method is used to numerically simulate the stress distribution of the canopy fabric, and the stress distribution results under different tensile forces are obtained; the stress distribution results obtained from the simulation calculation are visually processed to generate the stress nephogram of the canopy fabric, intuitively showing the stress distribution; according to the stress distribution results, the deformation of the canopy fabric under different tensile forces is further calculated to obtain the deformation distribution results; the deformation distribution results are visually processed to generate the deformation nephogram of the canopy fabric, intuitively showing the deformation distribution; the stress nephogram and deformation nephogram are comprehensively analyzed to evaluate the mechanical properties of the canopy fabric under different tensile forces, providing a reference basis for the design of the canopy fabric tension structure.

[0027] Exemplarily, according to the three-dimensional model data of the canopy fabric tension structure, a finite element analysis model is established using ANSYS software. The material of the canopy fabric is set as polyester fiber, with an elastic modulus of 5 GPa, a Poisson's ratio of 0.3, and a density of 1400 kg / m³. Fixed constraints are applied at the model boundaries to simulate the connection between the canopy fabric and the support structure. For three different tensile force conditions of 100 N, 200 N, and 300 N, the load type is defined as a uniformly distributed load and applied to the corresponding edges of the canopy fabric in the finite element model respectively. Using the ANSYS finite element analysis solver and based on the linear elastic material constitutive model, the stress distribution of the canopy fabric is numerically simulated. It can be seen from the stress nephogram that under a tensile force of 100 N, the stress distribution of the canopy fabric is relatively uniform, and the maximum von Mises equivalent stress is 2 MPa, concentrated in the central area of the canopy fabric; as the tensile force increases to 200 N and 300 N, the stress gradually concentrates towards the edges of the canopy fabric, and the maximum equivalent stresses reach 18 MPa and 17 MPa respectively. The deformation distribution results of the canopy fabric are extracted in the ANSYS post-processing module to generate a deformation nephogram. The maximum deformation of the canopy fabric under a tensile force of 100 N is 15 mm, and the deformation is mainly along the direction of the tensile force; under the 300 N condition, the deformation further increases, and the maximum deformation reaches 52 mm. Comprehensive analysis of the stress nephogram and deformation nephogram shows that the mechanical properties of the canopy fabric are closely related to the magnitude of the tensile force. The greater the tensile force, the higher the stress level and the greater the deformation of the canopy fabric. The tensile force should be appropriately controlled to obtain better mechanical properties. To further optimize the design of the canopy fabric tension structure, in the ANSYS parametric design module, by changing geometric parameters such as the material parameters of the canopy fabric and the curvature of the canopy fabric tension structure, multi-condition and multi-parameter analysis can be carried out to obtain the mechanical properties of the canopy fabric under different combinations of design parameters, providing a design reference for engineering applications.

[0028] Step S102: Identify the high-risk areas in the ceiling fabric where local wrinkles and unevenness are likely to occur based on this data.

[0029] Obtain the parameter data of each work station and quality inspection data collected during the production process of the ceiling fabric, and establish a production database for the ceiling fabric. Preprocess the production data of the ceiling fabric, including data cleaning, feature extraction, data standardization, etc., to obtain a normalized data set. Use a clustering algorithm to analyze the production data of the ceiling fabric, divide the data points into different categories according to the similarity of each data point, and identify the internal patterns and rules of the data. For the clustering results, analyze the feature distribution of each category of data, and determine the range of production process parameters and quality levels corresponding to different categories. Combine the quality inspection results of the ceiling fabric, compare and analyze the production data characteristics corresponding to the products with defects such as local wrinkles and unevenness, and judge the key influencing factors for the formation of defects. Based on the production data characteristics of the defective products, establish a defect risk assessment model to predict and evaluate the risk levels of local wrinkles and unevenness defects under different production parameter combinations. Map the defect risk levels of the production parameter combinations to the workshop layout diagram to generate a high-risk area distribution map of the ceiling fabric production line, providing visual decision support for quality control and process optimization.

[0030] Exemplarily, first, 12 parameter data such as temperature, pressure, and speed are collected from 20 workstations of the ceiling cloth production line, as well as 8 inspection indicators such as appearance defects and dimensional deviations at the quality inspection workstation. The data is collected every 5 seconds, and about 1 million production records are formed and stored in the MySQL database. Then, the data is cleaned using the Pandas library in Python, fields with missing values exceeding 10% are removed, and the remaining missing values are filled with the mean value. 10 statistical features such as the mean, standard deviation, and kurtosis of each parameter are extracted, and the data is scaled between 0 - 1 using the Min - Max normalization method. Next, the K - Means clustering algorithm is used to group the data, and the optimal number of clusters k = 4 is determined through the elbow method, that is, the data points are divided into 4 categories. Analysis shows that the temperature range corresponding to Category I is 180 - 220 °C, the pressure is 2 - 8 MPa, and the product qualification rate reaches 98%; while Category IV has large temperature fluctuations, generally low pressure, and a high incidence of local wrinkling defects, and the qualification rate is only 85%. Further analysis of the defective product data finds that the wrinkling defect has the greatest correlation with the temperature at Station 3 and the pressure at Station 7. When the temperature is lower than 190 °C or the pressure is lower than 5 MPa, the defect rate increases significantly. Therefore, a logistic regression model is established with these two parameters as independent variables to predict the defect risk, and the regression coefficients are -8 and -2 respectively, that is, for every 10 °C decrease in temperature and every 1 MPa decrease in pressure, the defect risk increases by 8% and 12%. Finally, the defect probability of each parameter combination is mapped to the workshop layout diagram to generate a risk heat map, and high - risk red areas around Stations 3 and 7 are found, which need to be monitored and optimized key points. Through the above big data analysis, the key factors affecting the quality of the ceiling cloth are identified, providing a decision - making basis for targeted improvement of the production process.

[0031] Step S103, in the identified high - risk area, apply the topology optimization algorithm to design the tension points on the ceiling cloth tension structure.

[0032] According to the environmental parameters and load conditions of the high-risk area, establish a three-dimensional geometric model and a finite element model of the ceiling cloth tension structure, and import the topology optimization algorithm for calculation and analysis. Optimize the design of the ceiling cloth tension structure through the topology optimization algorithm to obtain the optimal material distribution and structural layout scheme, and determine the position and quantity of the tension points. Use the simulated annealing algorithm to further optimize the specific layout scheme of the tension points, and search for the optimal coordinate parameters of the tension points by setting the objective function and constraint conditions. According to the optimized tension point layout scheme, perform parametric modeling on the ceiling cloth tension structure to automatically generate a three-dimensional solid model containing the tension points. Import the generated ceiling cloth tension structure model into the finite element analysis software to carry out the static and dynamic performance analysis of the structure, and evaluate the rationality and reliability of the tension point layout. If the analysis results meet the design requirements and specifications, apply the tension point layout scheme to the design of the ceiling cloth tension structure; otherwise, return to the topology optimization and simulated annealing algorithm steps, adjust the optimization parameters until a satisfactory design scheme is obtained.

[0033] Exemplarily, use SolidWorks to establish a three-dimensional geometric model of the automobile ceiling cloth tension structure, and import it into ANSYS to generate a finite element model containing 100,000 elements. Adopt the SIMP topology optimization algorithm, set the volume fraction to 30%, obtain the optimal material distribution, and determine the positions of 7 tension points. Then apply the simulated annealing algorithm to further optimize the tension point layout, with the minimum displacement as the objective function and the tension point spacing greater than 30 cm as the constraint condition, and search for the optimal tension point coordinates. Accordingly, parametrically generate a three-dimensional solid model of the ceiling containing the tension points in SolidWorks, and then import it into ANSYS for static and dynamic analysis. The results show that the maximum displacement of the automobile ceiling cloth is 50 mm and the stress is 100 MPa, meeting the specification requirements.

[0034] Step S104, optimize the position parameters of these tension points to ensure that these points can provide sufficient support and tensile force in the high-risk area, while meeting the strength and stability requirements of the ceiling cloth tension structure.

[0035] Obtain the three-dimensional model data and material property parameters of the ceiling fabric tension structure. Through the finite element analysis method, calculate the stress distribution and deformation of the ceiling fabric tension structure under different tension point position parameters. According to the risk level of the environment where the ceiling fabric tension structure is located, determine the range of high-risk areas and wind load parameters, and add them as constraint conditions to the optimization model. Use the genetic algorithm to optimize and solve the position parameters of the tension points. The objective function is to minimize the maximum stress and deformation of the ceiling fabric tension structure, while meeting the support and tension requirements in the high-risk areas. During the optimization process, generate new tension point position parameters through crossover and mutation operations, and evaluate their advantages and disadvantages according to the fitness function, eliminate the inferior solutions, and retain the superior solutions. After multiple iterative optimizations, obtain the optimal combination of tension point position parameters, so that the ceiling fabric tension structure has sufficient support and tension in the high-risk areas while meeting the strength and stability requirements. Apply the optimized tension point position parameters to the design of the ceiling fabric tension structure to ensure the safety and reliability of the ceiling fabric tension structure. Conduct load tests and health monitoring on the completed ceiling fabric tension structure, evaluate its actual stress performance, and reinforce and maintain it if necessary to ensure its long-term use safety.

[0036] Exemplarily, first, 3D laser scanners are used to obtain the point cloud data of the ceiling cloth tension structure, and the point cloud is converted into a solid model through reverse modeling software such as Geomagic Design X. At the same time, parameters such as the elastic modulus and Poisson's ratio of the ceiling cloth tension structure are measured through material tests. Then, the 3D model is imported into the finite element analysis software ANSYS, meshed, and material properties are set. Through parametric modeling, different tension point position parameters are set, such as the horizontal spacing of the tension points ranging from 2 cm to 10 cm, and the stress and displacement nephograms under different working conditions are calculated batch by using the APDL command stream. In the genetic algorithm optimization module, the objective function is set to minimize the Mises equivalent stress and the maximum deformation of the ceiling cloth tension structure, and the constraint conditions are that the supports and tensile forces in the high-risk areas are not less than 5 kN and 20 kN respectively. The initial population size is taken as 50, the crossover probability is taken as 7, the mutation probability is taken as 1, and after 100 iterations, the optimal tension point position is obtained as a horizontal spacing of 16 cm. At this time, the maximum equivalent stress of the ceiling cloth tension structure is 12 MPa, and the maximum deformation is 20 mm, meeting the strength and stiffness requirements. Finally, the optimization results are used to guide the design, and after the ceiling cloth tension structure is built, strain gauges and displacement sensors are used for load tests. It is measured that under the action of 2 times the design load, the errors between the measured stress and deformation and the finite element results are both within 10%, verifying the reliability of the optimized design. At the same time, the stress, vibration and other data of the ceiling cloth tension structure are collected in real time through the online monitoring system to evaluate its health status. When the structure stress exceeds 14 MPa or the vibration frequency is lower than 2 Hz, reinforcement and maintenance are carried out in time to ensure its use safety.

[0037] In step S105, the objective function of the topology optimization algorithm is set to minimize the stress concentration and deformation of the ceiling cloth, and the constraint conditions include the number and position range of the tension points and the structural strength of the ceiling cloth tension structure.

[0038] According to the preset topology optimization algorithm, the stress distribution data and deformation data of the ceiling cloth are obtained and used as the input of the optimization objective function. Through the finite element analysis of the ceiling cloth, the stress concentration and deformation under different numbers and positions of the tension points are calculated, and the feasible solutions that meet the constraint conditions are selected. For each feasible solution, the stress condition of its ceiling cloth tension structure is evaluated to determine whether it meets the preset structural strength requirements. If not, the solution is eliminated. From the feasible solutions that meet the strength requirements, the solution with the minimum objective function value is selected as the optimal topology layout plan. The simulated annealing algorithm is used to further optimize the optimal layout plan, and the tension point position is finely adjusted to reduce the stress concentration degree. Through iterative optimization, the ceiling cloth topology structure that meets the strength requirements and has the minimum stress distribution and deformation is obtained. The optimized ceiling cloth topology structure data is transmitted to the subsequent design and manufacturing processes.

[0039] Exemplarily, first, according to a preset topology optimization algorithm, stress distribution data and deformation data of the ceiling cloth under different layouts of tension points are obtained. The ceiling cloth is modeled and meshed using the finite element analysis software ANSYS. The material parameters are set as polyester fiber cloth, with an elastic modulus of 5 GPa, a Poisson ratio of 3, and a density of 1400 kg / m³. A pre-tension of 10 kN / m is applied to the edge of the ceiling cloth, and the stress distribution and deformation under 10 different layouts of tension points are calculated. Four feasible layouts with a stress concentration coefficient less than 5 and a maximum deformation less than 150 mm are selected. Then, MATLAB programming is used to perform a force analysis on the ceiling cloth tension structure under the four layouts respectively, and the maximum Mises stress is calculated. If it exceeds the allowable stress of the material, 25 MPa, it is excluded. The remaining three feasible solutions are substituted into the objective function, that is, the weighted sum of the stress concentration coefficient and the maximum deformation, with weights of 6 and 4 respectively, and the layout with the minimum objective function value is taken as the optimal topology solution. Then, the simulated annealing algorithm is used to fine-tune the optimal layout. The initial temperature is set to 100, the termination temperature is set to 01, the cooling coefficient is 95, and it is iterated 500 times to obtain an improved layout with an 8% reduction in the stress concentration coefficient and a 5% reduction in the maximum deformation. Finally, the optimized ceiling cloth topology structure data is converted into a CAD model and transferred to the design and manufacturing processes.

[0040] Step S106, establish a three-dimensional solid model of the ceiling cloth tension structure in the computer-aided design software according to the optimized tension point parameters.

[0041] According to the three-dimensional model template of the ceiling cloth tension structure preset in the CAD software, key parameters and structural feature information of the model are obtained; for the obtained key parameters, a parameter optimization algorithm, such as a genetic algorithm, a particle swarm algorithm, or a simulated annealing algorithm, is used to optimize and solve the position coordinates and force magnitudes of the tension points; the optimized parameter values of the tension points obtained are passed into the corresponding parameter variables in the three-dimensional model template of the ceiling cloth tension structure; according to the three-dimensional model template after the parameters are passed in, solid modeling is performed in the CAD software to automatically generate a three-dimensional solid model of the ceiling cloth tension structure; the generated three-dimensional solid model of the ceiling cloth tension structure is subjected to finite element analysis to simulate its force and deformation conditions under various working conditions; according to the results of the finite element analysis, it is judged whether the strength, stiffness, and stability of the ceiling cloth tension structure meet the design requirements. If not, return to step 2 to re-optimize and solve the tension point parameters; when the mechanical properties of the three-dimensional solid model of the ceiling cloth tension structure meet the design requirements, it is exported as a general three-dimensional model file format for subsequent production, processing, and installation.

[0042] Exemplarily, first, a three-dimensional model template of the ceiling fabric tension structure is retrieved from the model library of the CAD software, and its key parameters are obtained, such as dimensional parameters like the length, width, and thickness of the ceiling fabric tension structure, as well as structural characteristic information such as material properties and load conditions. Then, the obtained key parameters are input into the optimization algorithm. Taking the spatial coordinates (x, y, z) and the force magnitude F of the tension points as the optimization variables, and the maximum deflection and stress of the ceiling fabric tension structure as the optimization objectives, the genetic algorithm is used for solution. The genetic algorithm takes the value range of the optimization variables as the constraint conditions, randomly generates the initial population, and through genetic operations such as selection, crossover, and mutation, continuously iterates and evolves until it converges to the optimal solution. The obtained coordinates of the tension points are (5, 8, 2), and the force magnitude is 15 kN. Next, the optimized tension point parameters are passed into the corresponding variables of the three-dimensional model template of the ceiling fabric tension structure, and using the modeling function of the CAD software, a three-dimensional solid model of the ceiling fabric tension structure is automatically generated. Finite element mesh division is performed on the generated three-dimensional solid model, material properties and boundary conditions are defined, and force and deformation analyses are carried out under various working conditions such as dead load, wind load, and earthquake. After calculation, the maximum deflection of the ceiling fabric tension structure is 3 mm, the maximum stress is 95 MPa, which is less than the allowable stress of the material, 105 MPa, meeting the strength requirements; the horizontal displacement at the end of the ceiling fabric tension structure is 5 mm, which is less than 1 / 400 of the span, meeting the stiffness requirements; under wind load and earthquake working conditions, there is no buckling instability phenomenon in the ceiling fabric tension structure, meeting the stability requirements. Finally, the three-dimensional solid model of the ceiling fabric tension structure is exported in STEP or IGES format and transferred to the production and processing department for numerical control machining and on-site installation.

[0043] Step S107, perform interference checking and assembly simulation to ensure the fitting relationship between the tension points and the ceiling fabric and adjacent components, and guarantee the installation space and operation convenience of the ceiling fabric.

[0044] According to the three-dimensional model of the ceiling cloth, obtain the geometric shape and dimensional information of the ceiling cloth, and convert it into a data format that can be used for interference checking and assembly simulation. Obtain the three-dimensional models of the tension points and adjacent components, extract their geometric shape and dimensional information, and convert it into a data format that can be used for interference checking and assembly simulation. Import the three-dimensional models of the ceiling cloth, tension points, and adjacent components into computer-aided design software, and assemble them in a virtual environment according to the actual assembly relationship. Through the computer-aided design software, simulate the installation process of the ceiling cloth, analyze the interference situation between the ceiling cloth and the tension points and adjacent components, and determine whether there are interference problems. If there are interference problems, then according to the results of the interference check, optimize and adjust the design of the ceiling cloth, tension points, or adjacent components until the interference problems are eliminated. On the basis of the assembly simulation, analyze the installation space and operation convenience of the ceiling cloth, and according to the analysis results, further optimize the design of the ceiling cloth, tension points, or adjacent components to ensure that the installation space and operation convenience of the ceiling cloth meet the requirements. According to the optimized design scheme, generate the final three-dimensional models and engineering drawings of the ceiling cloth, tension points, and adjacent components, and apply them to the actual production and assembly process to ensure the installation quality and efficiency of the ceiling cloth.

[0045] Exemplarily, first, a three-dimensional scanner is used to scan the ceiling cloth to obtain its geometric shape and dimensional information, and the scanned point cloud data is converted into a three-dimensional model in STL format through a triangulation algorithm. At the same time, a three-dimensional modeling software such as CATIA is used to perform three-dimensional modeling on the tensioning points and adjacent components, extract their geometric shape and dimensional information, and export the model in STEP format. Then, the three-dimensional models of the ceiling cloth, tensioning points, and adjacent components are imported into the CATIA software, and according to the actual assembly relationship, assembly constraints such as mating and coincidence are used to complete the assembly in a virtual environment. Next, through the assembly simulation module of CATIA, the installation process of the ceiling cloth is simulated, the assembly tolerance is set to 2 mm, and the interference situation between the ceiling cloth and the tensioning points and adjacent components is analyzed through an interference check algorithm. If there are interference problems, according to the interference amount and position, the parametric modeling method is used to optimize the dimensions and shapes of the ceiling cloth, tensioning points, or adjacent components. For example, the edge of the ceiling cloth is shrunk by 5 mm, and the interference check is performed again until the interference amount is less than 1 mm. On the basis of the assembly simulation, an ergonomics software such as Jack is used to analyze the installation space and operation convenience of the ceiling cloth. By simulating the installation actions of a virtual human, the reachability of the hand and the viewing angle during the installation process are calculated. If the reachability is less than 90% or the viewing angle is less than 60°, the layout of the components is optimized. Finally, according to the optimized design scheme, the final three-dimensional models and engineering drawings of the ceiling cloth, tensioning points, and adjacent components are generated, imported into the PDM system for management, and applied to actual production assembly. The assembly information is projected onto the physical object through a laser projector to guide the operator for assembly, thus ensuring the installation accuracy and efficiency of the ceiling cloth.

[0046] Step S108, during the tensioning process of the ceiling cloth, machine vision technology is used to monitor the morphology and stress distribution of the ceiling cloth in real time.

[0047] Obtain the real-time image data of the ceiling cloth during the tensioning process, input the image data into a pre-constructed convolutional neural network model for processing to obtain the shape characteristics and stress distribution characteristics of the ceiling cloth. According to the shape characteristics of the ceiling cloth, use an edge detection algorithm to extract the contour of the ceiling cloth, and compare the contour with a preset standard contour to determine whether the shape of the ceiling cloth meets the requirements. If it does not meet the requirements, output a shape abnormality warning message. According to the stress distribution characteristics of the ceiling cloth, use a numerical analysis method to calculate the stress values of each area of the ceiling cloth, compare the stress values with a preset safety threshold to determine whether the stress exceeds the safe range. If it exceeds the safe range, output a stress abnormality warning message. Perform a fusion analysis on the shape characteristics and stress distribution characteristics of the ceiling cloth to construct a health status evaluation model for the ceiling cloth. According to the output result of the evaluation model, judge the overall tensioning quality of the ceiling cloth to obtain a quality score. Output the shape abnormality warning message, stress abnormality warning message, and overall quality score of the ceiling cloth to the monitoring interface, and at the same time transmit the relevant data to the control unit of the tensioning device to achieve real-time adjustment of the tensioning process through feedback control. After the ceiling cloth is tensioned, use 3D reconstruction technology to generate a 3D model of the ceiling cloth, compare the 3D model with the design model, calculate the shape error, and obtain the final forming quality report of the ceiling cloth. Establish a database for the ceiling cloth tensioning process, record information such as image data, shape characteristics, stress distribution characteristics, and health status scores at each moment, and provide data support for subsequent process optimization and quality analysis.

[0048] For example, during the tensioning process of the ceiling cloth, a high-speed camera is used to obtain real-time image data of the ceiling cloth at a speed of 100 frames per second, and the image resolution is 1920×1080 pixels. The acquired image data is input into a pre-built convolutional neural network model for processing. The model adopts the ResNet-50 network structure and is pre-trained on a dataset of 10,000 ceiling cloth images by transfer learning. The shape characteristics and stress distribution characteristics of the ceiling cloth can be accurately extracted. The ceiling cloth image is processed using the Canny edge detection algorithm to extract the contour of the ceiling cloth. By calculating the Hausdorff distance between the contour and the preset standard contour, when the distance exceeds 10mm, the shape abnormality warning information is output. According to the stress distribution characteristics output by the convolutional neural network model, the finite element analysis method is used to divide the ceiling cloth into 1,000 units, and the Von Mises stress value of each unit is calculated. When the stress value exceeds the yield strength of the material (such as the yield strength of aluminum alloy is 200MPa), the stress abnormality warning information is output. The fuzzy comprehensive evaluation method is used to integrate and analyze the shape characteristics and stress distribution characteristics of the ceiling cloth, and a ceiling cloth health status assessment model is constructed. The model calculates the health status membership of the ceiling cloth by establishing fuzzy membership functions of indicators such as shape error and stress distribution uniformity, and judges the overall tensioning quality of the ceiling cloth according to the size of the membership. When the membership is greater than 8, the quality is considered qualified, otherwise it is considered unqualified. The evaluation results and related warning information are displayed in real time on the monitoring interface, and the data is transmitted to the PLC control unit of the tensioning equipment. The adaptive adjustment of the tensioning force is achieved through the closed-loop control algorithm to ensure the tensioning quality of the ceiling cloth. After the tensioning is completed, the three-dimensional point cloud data of the ceiling cloth is obtained by a structured light three-dimensional scanner. The three-dimensional model of the ceiling cloth is generated by point cloud splicing and mesh reconstruction algorithm. The three-dimensional model is compared with the design CAD model, and the root mean square value of the shape error is calculated. When the root mean square value is less than 1mm, the forming quality of the ceiling cloth is considered to be qualified. Finally, all the data collected during the tensioning process are uploaded to the cloud database, and big data analysis technology is used to explore the correlation between the data, optimize the tensioning process parameters, and improve the forming quality of the ceiling cloth.

[0049] Step S109, extracting characteristic points on the surface of the ceiling cloth by an image processing algorithm, and calculating the displacement and strain of these characteristic points to determine whether the ceiling cloth has achieved the expected tensioning effect.

[0050] Obtain the surface image of the ceiling cloth, and use image preprocessing methods to denoise and enhance the image to improve the image quality. Extract the feature points on the surface of the ceiling cloth from the preprocessed image through a feature extraction algorithm to obtain a set of feature points. According to the set of feature points, use the optical flow method to calculate the displacement of the feature points between two adjacent frames of images to obtain a displacement matrix. According to the displacement matrix, calculate the strain at each feature point through the strain calculation formula to obtain a strain matrix. Conduct statistical analysis on the displacement matrix and the strain matrix to obtain the distribution characteristics of the displacement and the strain. According to the pre-established judgment model for the tensioning effect of the ceiling cloth, use the distribution characteristics of the displacement and the strain as inputs to judge whether the ceiling cloth has achieved the expected tensioning effect. If the ceiling cloth has not achieved the expected tensioning effect, feedback and adjust the tensioning parameters according to the judgment result, and re-perform the tensioning process until the expected effect is achieved.

[0051] Exemplarily, first preprocess the obtained surface image of the ceiling cloth. Use the median filtering algorithm to denoise the image, and set the size of the filtering window to 5×5. Then use the histogram equalization algorithm to enhance the image to improve the contrast of the image. Next, use the SIFT feature extraction algorithm to extract the feature points on the surface of the ceiling cloth from the preprocessed image, and set the number of extracted feature points to 500. According to the set of extracted feature points, use the Lucas-Kanade optical flow method to calculate the displacement of the feature points between two adjacent frames of images. Set the size of the search window to 21×21 and the number of iterations to 3 times to obtain a matrix representing the displacement of the feature points. According to the displacement matrix, calculate the strain at each feature point through the strain calculation formula ε = Δl / l, where Δl represents the displacement of the feature point and l represents the distance from the initial position of the feature point to the fixed point, to obtain a matrix representing the strain of the feature points. Conduct statistical analysis on the displacement matrix and the strain matrix, and calculate statistical features such as the mean, variance, maximum value, and minimum value of the displacement and the strain to obtain the distribution characteristics of the displacement and the strain. According to the pre-established SVM judgment model for the tensioning effect of the ceiling cloth, use the distribution characteristics of the displacement and the strain as inputs, and obtain the judgment result on whether the ceiling cloth has achieved the expected tensioning effect through model prediction. If the judgment result indicates that the ceiling cloth has not achieved the expected tensioning effect, then use the fuzzy control algorithm to calculate the adjustment amount of the tensioning parameters according to the distribution characteristics of the displacement and the strain. Modify parameters such as the tensile force and the tensioning time of the tensioning equipment according to the adjustment amount, and then re-perform the tensioning process. Obtain the surface image of the ceiling cloth again for analysis and judgment until the judgment result indicates that the ceiling cloth has achieved the expected tensioning effect.

[0052] Step S1010, if it is monitored that wrinkles or unevenness appear locally on the ceiling cloth, use a convolutional neural network to analyze the shape characteristics of the wrinkled area.

[0053] The vehicle roof layout image is acquired through the camera, and the image is preprocessed, including grayscale, denoising and other operations, to obtain image data suitable for analysis. The preprocessed image data is input into the pre-trained convolutional neural network model to extract the shape features of the image and obtain the feature vector. According to the feature vector, the support vector machine algorithm is used to classify the wrinkle area and determine the type of wrinkles, such as horizontal wrinkles, longitudinal wrinkles, etc. For each wrinkle area, the degree of wrinkles is quantified by calculating the depth, width and other parameters of the wrinkles to obtain a numerical representation of the degree of wrinkles. If the degree of wrinkles exceeds the preset threshold, the area is judged to be a serious wrinkle area and needs to be repaired or reworked. The location, type, degree and other information of the detected wrinkle area are summarized to generate a test report to provide data support for the subsequent improvement of the quality of the roof layout. According to the test report, the convolutional neural network model is fine-tuned and optimized to improve the accuracy and efficiency of wrinkle detection and achieve continuous improvement of the quality of the roof layout.

[0054] Exemplarily, an image of the vehicle ceiling layout is obtained through a high-definition camera, and the image resolution is 1920×1080 pixels. The obtained image is preprocessed. First, the image is converted into a grayscale image, and then the median filtering algorithm is used to denoise the image. The filtering window size is 5×5 to eliminate salt-and-pepper noise and Gaussian noise in the image, and image data suitable for subsequent analysis is obtained. The preprocessed image data is input into a pre-trained convolutional neural network model. This model adopts the VGG-16 network structure, which includes 13 convolutional layers and 3 fully connected layers. Through convolutional operations and pooling operations, multi-scale shape features of the image are extracted to obtain a 4096-dimensional feature vector. According to the extracted feature vector, the support vector machine algorithm is used to classify the wrinkled areas in the image. The radial basis kernel function is used, and the penalty factor C is taken as 10. The type of wrinkles, such as horizontal wrinkles, vertical wrinkles, and mixed wrinkles, is judged through the one-vs.-rest strategy. For each detected wrinkled area, parameters such as the average depth and maximum width of the wrinkles are calculated to quantify the degree of wrinkles. Among them, the wrinkle depth is obtained by calculating the difference between the pixel gray value in the wrinkled area and the surrounding average gray value, and the wrinkle width is obtained through connected component analysis. If the average depth of the wrinkled area exceeds 20 gray levels, or the maximum width exceeds 100 pixels, it is judged that this area is a severely wrinkled area and needs to be repaired or reworked. The position coordinates, wrinkle types, wrinkle degrees, and other information of all detected wrinkled areas are summarized to generate a detection report, providing data support for the subsequent improvement of the ceiling layout quality. According to the wrinkle distribution in the detection report, the structure and parameters of the convolutional neural network model are fine-tuned and optimized, such as increasing the number of convolutional layers, adjusting the convolutional kernel size, optimizing the loss function, etc., to continuously improve the accuracy and efficiency of wrinkle detection and achieve continuous improvement of the ceiling layout quality. Through this method, automatic detection and quantitative evaluation of the wrinkle defects in the vehicle ceiling layout can be realized, and the detection accuracy reaches more than 95%, greatly improving the efficiency and reliability of the ceiling layout quality detection and providing strong quality guarantee for the production of automotive interior parts.

[0055] Step S1011, this network is trained through pre-collected ceiling cloth wrinkle samples, combined with the layout parameters of the tensioning points of the ceiling cloth tensioning structure, predicts possible causes, and proposes suggestions for adjusting the tension force or optimizing the tensioning points.

[0056] According to the pre - collected data of the ceiling cloth wrinkle samples, a wrinkle sample feature database is established, and feature parameters such as the texture and shape of the wrinkle samples are extracted. Obtain the layout parameters of the tension points of the ceiling cloth tension structure of the current ceiling cloth, including parameters such as the position coordinates of the tension points and the magnitude of the tensile force, and input them into the pre - trained convolutional neural network model. Use the convolutional neural network model to extract and analyze the feature of the input layout parameters of the tension points of the ceiling cloth tension structure. By comparing with the data in the wrinkle sample feature database, predict the possible causes of the wrinkles on the ceiling cloth under the current layout parameters. If the predicted cause of the wrinkle is due to uneven tensile force, then according to the position and shape of the wrinkle, judge the position of the tension point and the magnitude of the tensile force that need to be adjusted, and generate the tensile force adjustment parameters. If the predicted cause of the wrinkle is due to unreasonable layout of the tension points, then use the genetic algorithm to optimize the layout parameters of the tension points. Through operations such as crossover and mutation, explore the optimal layout scheme of the tension points. Output the adjusted tensile force parameters or the optimized layout parameters of the tension points, and apply them to the actual tensioning process of the ceiling cloth to verify and evaluate the wrinkle situation of the ceiling cloth. According to the feedback of the wrinkle situation of the ceiling cloth, further train and optimize the convolutional neural network model to improve the prediction accuracy and adaptability of the model, and provide optimization suggestions for subsequent similar projects.

[0057] Exemplarily, first, a large number of ceiling cloth wrinkle sample data are collected. Texture features of the wrinkle samples such as gray-level co-occurrence matrix, wavelet transform coefficients, etc., and shape features such as wrinkle depth, wrinkle angle and other parameters are extracted to establish a wrinkle feature database containing 500 samples. Then, the coordinates (x, y, z) of 20 tensioning points of the ceiling cloth tensioning structure in the current project and the corresponding tensile force F are obtained, and these parameters are input into the convolutional neural network VGG-16 model pre-trained with 1000 wrinkle images. The model extracts the spatial features of the tensioning layout parameters through the convolutional layer, then maps the features to the wrinkle cause classification through the fully connected layer, and outputs the probabilities of various causes through the softmax function. If the model predicts that the wrinkle is most likely caused by uneven tensile force with a probability of 85%, then it is judged according to the wrinkle position (such as the northeast corner) and shape (such as radial) that the tensile force in the corresponding area (such as the 2nd and 3rd tensioning points) needs to be increased by 20% to 30N. If the model predicts that the wrinkle is most likely caused by unreasonable layout of the tensioning points with a probability of 90%, then the genetic algorithm is used to optimize the layout parameters. First, 50 layout schemes are randomly generated as the initial population, and each scheme consists of 20 tensioning point coordinates (x, y, z). Then, the fitness function f(x) of each scheme is calculated, such as the ratio of the ceiling area S to the volume V. Next, operations such as tournament selection, uniform crossover, and Gaussian mutation are iterated 500 generations, and finally the layout scheme with the highest fitness is obtained, and its tensioning point coordinates (x', y', z') are output and applied to the actual situation. After the actual tensioning is completed, the three-dimensional point cloud data of the ceiling cloth is collected by a laser scanner, its wrinkle features are extracted, and compared with the ideal ceiling model to calculate parameters such as wrinkle depth and area. If the wrinkle depth is less than 1 cm and the area is less than 5 m², it is considered that the optimization effect is good, and the measured data is added to the sample library to continue training the convolutional neural network model to continuously improve its prediction accuracy. Through the above method, the intelligent optimization of the ceiling cloth tensioning can be realized, the occurrence of wrinkles can be reduced, the ceiling quality can be improved, and a reference can be provided for similar projects.

[0058] Step S1012, after the ceiling cloth tensioning is completed, use a three-dimensional scanner to obtain the point cloud data of the ceiling surface.

[0059] According to the actual situation after the tensioning of the ceiling cloth is completed, determine the scanning range and scanning accuracy parameters of the 3D scanner. Start the 3D scanner and conduct an all-round scan of the surface of the ceiling cloth to obtain the original point cloud data. Preprocess the original point cloud data to remove noise points and outliers, and obtain the cleaned point cloud data. According to the characteristics of the scanned object, use the normal vector estimation algorithm to calculate the normal vector of each point, and obtain the point cloud data with normal vector information. Through the density analysis of the point cloud data, judge the integrity and uniformity of the point cloud data. If the requirements are not met, return to step 1 for rescan. Use the point cloud data for triangulation to reconstruct the 3D mesh model of the ceiling cloth surface. Compare and analyze the reconstructed 3D mesh model with the design model to obtain the shape error and dimensional deviation of the ceiling cloth surface, providing data support for subsequent process optimization.

[0060] Exemplarily, according to the actual situation after the tensioning of the ceiling cloth is completed, determine the scanning range of the 3D scanner to be 2m × 2m, and set the scanning accuracy parameter to 1mm. Start the 3D scanner and conduct an all-round scan of the surface of the ceiling cloth using the multi-view scanning method. The scanning angle range is 0° to 360°, and the scanning time is 5 minutes, obtaining approximately 5 million points of original point cloud data. Preprocess the original point cloud data, and use the statistical filtering algorithm to remove noise points and outliers. The noise point threshold is set to 5mm, and the outlier threshold is set to 2mm, obtaining approximately 4.8 million points of cleaned point cloud data. According to the smooth characteristics of the ceiling cloth surface, use the PCA-based normal vector estimation algorithm to calculate the normal vector of each point. When calculating the normal vector, consider 20 neighboring points of the point, and obtain the point cloud data with normal vector information. Through the density analysis of the point cloud data, with 100 points per square centimeter as the standard, judge the integrity and uniformity of the point cloud data. The analysis result shows that the density of the point cloud data meets the requirements and no rescan is required. Use the point cloud data for Poisson surface reconstruction, and the reconstructed 3D mesh model contains 100,000 triangular patches. Compare and analyze the reconstructed 3D mesh model with the design model, use the ICP algorithm for registration, and control the registration error within 5mm. Through deviation analysis, obtain the shape error and dimensional deviation of the ceiling cloth surface. The maximum shape error is 3mm, and the maximum dimensional deviation is 5mm, providing data support for subsequent process optimization.

[0061] Step S1013: Generate a digital model of the ceiling through the surface reconstruction algorithm, compare it with the design model, quantitatively evaluate the overall tensioning quality and dimensional accuracy of the ceiling cloth, and form a quality report.

[0062] Obtain the three-dimensional point cloud data of the ceiling cloth, and generate a digital model of the ceiling cloth through a surface reconstruction algorithm; compare the generated digital model of the ceiling cloth with a preset design model, and calculate the deviation value between the two models; determine the overall tensioning quality of the ceiling cloth according to the deviation value. If the deviation value exceeds the preset threshold, it is judged that the overall tensioning quality of the ceiling cloth is unqualified; evaluate the dimensional accuracy of the ceiling cloth by calculating the dimensional differences between the digital model and the design model at key positions; use the support vector machine algorithm to comprehensively evaluate the overall tensioning quality and dimensional accuracy of the ceiling cloth to obtain a quality score; integrate information such as the deviation value, dimensional difference, and quality score of the ceiling cloth to generate a ceiling cloth quality assessment report; use an incremental learning method to optimize the surface reconstruction algorithm to improve the generation accuracy of the digital model of the ceiling cloth.

[0063] Exemplarily, first, use a three-dimensional scanner to obtain the point cloud data of the surface of the ceiling cloth, set the point cloud density to 5 mm, and control the scanning accuracy within 1 mm. Then import the point cloud data into three-dimensional reconstruction software, and use the Poisson surface reconstruction algorithm to digitally model the surface of the ceiling cloth. Set the resolution of the reconstruction grid to 1024×1024, and the accuracy of the generated digital model can reach 2 mm. Next, align the digital model with the design model, and use the ICP algorithm to calculate the deviation value between the two models. If the deviation value is less than 1 mm, it is considered that the overall tensioning quality of the ceiling cloth is qualified, otherwise it is judged as unqualified. At the same time, by extracting the cross-sectional contours of the digital model and the design model at key positions, calculate the Hausdorff distance between the two contours to evaluate the dimensional accuracy of the ceiling cloth. If the Hausdorff distance is less than 5 mm, it is considered that the dimensional accuracy meets the requirements. Then use the support vector machine algorithm to comprehensively evaluate the overall tensioning quality and dimensional accuracy of the ceiling cloth. The input features are the deviation value and the Hausdorff distance, and the output result is the quality score, with the scoring range from 0 to 100 points. Finally, summarize the evaluation results to generate a report, including the three-dimensional digital model of the ceiling cloth, the deviation value distribution diagram, the dimensional accuracy analysis diagram, and the comprehensive quality score, etc. To further improve the accuracy of the digital model of the ceiling cloth, use an incremental learning method to optimize the Poisson surface reconstruction algorithm, and adaptively adjust the algorithm parameters according to the reconstruction error. After 10 iterations of optimization, the accuracy of the digital model can be improved by more than 20%.

[0064] As Figure 4 shown, a synchronous tensioning structure for an automotive ceiling cloth includes a tensioning table 1, a ceiling cloth clamping module 2, a tensioning connecting rod 3, and a tensioning drive plate 4.

[0065] The tensioning table 1 includes a table board 11 and support guide columns 12; the table board 11 is in the shape of a cuboid plate structure, and a transmission hole is provided at the center position of the table board 11; a driving motor 13 is provided on one side of the transmission hole on the lower surface of the table board 11, the output end of the driving motor 13 is connected with a gearbox, and a transmission shaft is provided at the output end of the gearbox, and the transmission shaft passes through the transmission hole at the center of the table board 11 and extends to the upper surface of the table board; the tensioning transmission plate 4 is parallel to the table board 11 and is arranged above the table board 11, the tensioning transmission plate 4 is in the shape of a waist-shaped plate structure, and the center position of the tensioning transmission plate 4 is connected to the end of the transmission shaft; several first tensioning points of tensioning connecting rods are arranged at the edge of the tensioning transmission plate 4; the support guide columns 12 are vertically arranged at the corner positions of the lower surface of the table board.

[0066] The ceiling cloth clamping module 2 includes a slide rail 21, a slider 22 and a clamping plate 23; the slide rail 21 is installed at the side edge and corner positions of the table board 11; the slider 22 is installed on the slide rail 21; the clamping plate 23 is installed at the top position of the slider 22, and one side edge of the clamping plate 23 extends out of the end of the slider 22; a longitudinal clamping cylinder 24 is provided at the end of the slider 22, and the output end of the longitudinal clamping cylinder 24 cooperates with the lower surface of the clamping plate 23 to form a clamping gap for clamping the automotive ceiling cloth; a second tensioning point of the tensioning connecting rod is arranged on the upper surface of the clamping plate 23.

[0067] The tensioning connecting rod 3 includes an adjusting screw sleeve 31, a connecting screw rod 32 and a rotary shaft sleeve 33; there are a pair of connecting screw rods 32 which are respectively arranged at both ends of the adjusting screw sleeve 31, and the length of the tensioning connecting rod 3 is adjusted by rotating the adjusting screw sleeve 31; the rotary shaft sleeve 33 is installed at the ends of the two connecting screw rods 32; positioning shafts are arranged at both the first tensioning point and the second tensioning point of the tensioning connecting rod, and rotary bearings are arranged on the positioning shafts and cooperate with the rotary shaft sleeve 33 respectively.

[0068] Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principle of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A synchronous tensioning method for automobile roof cloth, characterized in that: The specific tensioning method is as follows: S1: Using the finite element analysis method, based on the three-dimensional model data of the ceiling cloth tensioning structure, simulate the stress distribution of the ceiling cloth under different tensioning forces, and obtain the stress cloud map and deformation cloud map of the ceiling cloth; S2: Using these data, identify high-risk areas in the ceiling fabric that are prone to local wrinkles and unevenness; In the identified high-risk areas, the topology optimization algorithm is used to design the tensioning points on the ceiling fabric tensioning structure; Optimize the location parameters of these tensioning points to ensure that they can provide sufficient support and tensioning force in high-risk areas while meeting the strength and stability requirements of the ceiling fabric tensioning structure; The objective function of the topology optimization algorithm is set to minimize the stress concentration and deformation of the ceiling fabric. The constraints include the number and position range of tensioning points and the structural strength of the ceiling fabric tensioning structure. S3: Establish a three-dimensional solid model of the ceiling cloth tensioning structure according to the optimized tensioning point parameters in the computer-aided design software; perform interference checking and assembly simulation to ensure the matching relationship between the tensioning points and the ceiling cloth and adjacent components, and to ensure the installation space and operation convenience of the ceiling cloth; S4: During the tensioning process of the ceiling cloth, the morphology and stress distribution of the ceiling cloth are monitored in real time using machine vision technology; feature points on the surface of the ceiling cloth are extracted using image processing algorithms, and the displacement and strain of these feature points are calculated to determine whether the ceiling cloth has achieved the expected tensioning effect; if wrinkles or unevenness are detected in the ceiling layout, a convolutional neural network is used to analyze the shape characteristics of the wrinkled area; the network is trained using pre-collected ceiling cloth wrinkle samples, combined with the layout parameters of the tensioning points of the ceiling cloth tensioning structure, to predict possible causes, and to propose suggestions for adjusting the tensioning force or optimizing the tensioning points; S5: After the ceiling fabric is tensioned, use a 3D scanner to obtain point cloud data of the ceiling surface; generate a digital model of the ceiling through a surface reconstruction algorithm, compare it with the design model, quantitatively evaluate the overall tensioning quality and dimensional accuracy of the ceiling fabric, and form a quality report.

2. A synchronous tensioning method for automobile roof fabric according to claim 1, characterized in that: The specific method in S1 includes: According to the three-dimensional model data of the ceiling cloth tensioning structure, a finite element analysis model is established, and material properties and boundary conditions are set; for different tensioning force conditions, the load type and size are defined and applied to the finite element model; the finite element analysis method is used to perform numerical simulation calculations on the force distribution of the ceiling cloth, and the stress distribution results under different tensioning forces are obtained; the stress distribution results obtained by the simulation calculation are visualized to generate a stress cloud map of the ceiling cloth to intuitively display the stress distribution; based on the stress distribution results, the deformation of the ceiling cloth under different tensioning forces is further calculated to obtain the deformation distribution results; the deformation distribution results are visualized to generate a deformation cloud map of the ceiling cloth to intuitively display the deformation distribution; the stress cloud map and the deformation cloud map are comprehensively analyzed to evaluate the force performance of the ceiling cloth under different tensioning forces, providing a reference for the design of the ceiling cloth tensioning structure.

3. The method for synchronously tensioning a car roof cloth according to claim 1, characterized in that: The specific method in S2 includes: Acquire the parameter data and quality inspection data of each workstation collected during the production process of ceiling cloth, and establish a ceiling cloth production database; pre-process the ceiling cloth production data, including data cleaning, feature extraction and data standardization, to obtain a standardized data set; use clustering algorithm to analyze the ceiling cloth production data, divide it into different categories according to the similarity of each data point, and identify the inherent patterns and laws of the data; analyze the characteristic distribution of each category of data based on the clustering results, and determine the production process parameter range and quality level corresponding to different categories; combine the ceiling cloth quality inspection results, compare and analyze the production data characteristics corresponding to products with local wrinkles and unevenness defects, and judge the key influencing factors of the defects; establish a defect risk assessment model based on the production data characteristics of defective products, and predict and evaluate the risk level of local wrinkles and unevenness defects under different production parameter combinations; map the defect risk level of the production parameter combination with the workshop layout diagram, and generate a high-risk area distribution map of the ceiling cloth production line to provide visual decision support for quality control and process optimization; According to the environmental parameters and load conditions of the high-risk area, a three-dimensional geometric model and a finite element model of the ceiling cloth tensioning structure are established, and a topology optimization algorithm is introduced for calculation and analysis; the ceiling cloth tensioning structure is optimized and designed by the topology optimization algorithm to obtain the optimal material distribution and structural layout scheme, and the location and number of tensioning points are determined; the specific layout scheme of the tensioning points is further optimized by the simulated annealing algorithm, and the optimal tensioning point coordinate parameters are searched by setting the objective function and constraint conditions; according to the optimized tensioning point layout scheme, the ceiling cloth tensioning structure is parametrically modeled, and a three-dimensional solid model containing tensioning points is automatically generated; the generated ceiling cloth tensioning structure model is imported into the finite element analysis software, and the static and dynamic performance analysis of the structure is carried out to evaluate the rationality and reliability of the tensioning point layout; if the analysis results meet the design requirements and specifications, the tensioning point layout scheme is applied to the design of the ceiling cloth tensioning structure; otherwise, return to the topology optimization and simulated annealing algorithm steps, and adjust the optimization parameters until a satisfactory design scheme is obtained; The three-dimensional model data and material property parameters of the ceiling fabric tensioning structure are obtained, and the stress distribution and deformation of the ceiling fabric tensioning structure under different tensioning point position parameters are calculated by the finite element analysis method; according to the risk level of the environment in which the ceiling fabric tensioning structure is located, the scope and wind load parameters of the high-risk area are determined, and they are added to the optimization model as constraints; the genetic algorithm is used to optimize the tensioning point position parameters, and the objective function is to minimize the maximum stress and deformation of the ceiling fabric tensioning structure while meeting the support and tensioning force requirements of the high-risk area; in the optimization process, new tensioning point position parameters are generated through cross-mutation operations, and their advantages and disadvantages are evaluated according to the fitness function, the inferior solutions are eliminated, and the superior solutions are retained; after multiple iterative optimizations, the optimal tensioning point position parameter combination is obtained, so that the ceiling fabric tensioning structure has sufficient support and tensioning force in the high-risk area while meeting the strength and stability requirements; the optimized tensioning point position parameters are applied to the design of the ceiling fabric tensioning structure to ensure the safety and reliability of the ceiling fabric tensioning structure; load tests and health monitoring are carried out on the built ceiling fabric tensioning structure to evaluate its actual stress performance; According to the preset topology optimization algorithm, the stress distribution data and deformation data of the ceiling cloth are obtained and used as the input of the optimization objective function; through the finite element analysis of the ceiling cloth, the stress concentration and deformation at different numbers and positions of tensioning points are calculated, and the feasible solutions that meet the constraints are screened out; for each feasible solution, the force condition of its ceiling cloth tensioning structure is evaluated to determine whether it meets the preset structural strength requirements, and if not, the solution is eliminated; from the feasible solutions that meet the strength requirements, the solution with the smallest objective function value is selected as the optimal topology layout scheme; the simulated annealing algorithm is used to further optimize the optimal layout scheme, and the tensioning point position is fine-tuned to reduce the degree of stress concentration; through iterative optimization, the ceiling cloth topology structure that meets the strength requirements and has the smallest stress distribution and deformation is obtained; the optimized ceiling cloth topology structure data is passed to the subsequent design and processing and manufacturing links.

4. The method for synchronously tensioning a car roof cloth according to claim 1, characterized in that: The specific method in S3 includes: According to the three-dimensional model template of the ceiling cloth tensioning structure preset in the CAD software, the key parameters and structural feature information of the model are obtained; according to the key parameters obtained, the parameter optimization algorithm is used to optimize and solve the position coordinates and force magnitude of the tensioning point; the tensioning point parameter values ​​obtained by the optimization solution are transferred to the corresponding parameter variables in the three-dimensional model template of the ceiling cloth tensioning structure; according to the three-dimensional model template after the parameters are transferred, solid modeling is performed in the CAD software to automatically generate a three-dimensional solid model of the ceiling cloth tensioning structure; finite element analysis is performed on the generated three-dimensional solid model of the ceiling cloth tensioning structure to simulate its force and deformation under various working conditions; according to the results of the finite element analysis, it is judged whether the strength, stiffness and stability of the ceiling cloth tensioning structure meet the design requirements. If not, return to step 2 to re-optimize the tensioning point parameters; when the mechanical properties of the three-dimensional solid model of the ceiling cloth tensioning structure meet the design requirements, it is exported to a general three-dimensional model file format for subsequent production, processing and installation; According to the 3D model of the ceiling cloth, the geometric shape and size information of the ceiling cloth are obtained and converted into a data format that can be used for interference checking and assembly simulation; the 3D models of the tensioning points and adjacent parts are obtained, their geometric shape and size information are extracted, and converted into a data format that can be used for interference checking and assembly simulation; the 3D models of the ceiling cloth, tensioning points and adjacent parts are imported into the computer-aided design software, and they are assembled in a virtual environment according to the actual assembly relationship; the installation process of the ceiling cloth is simulated through the computer-aided design software, and the interference between the ceiling cloth and the tensioning points and adjacent parts is analyzed to determine whether There is an interference problem; if there is an interference problem, the design of the ceiling cloth, tensioning points or adjacent components is optimized and adjusted according to the results of the interference check until the interference problem is eliminated; on the basis of assembly simulation, the installation space and operational convenience of the ceiling cloth are analyzed, and according to the analysis results, the design of the ceiling cloth and tensioning points or adjacent components is further optimized to ensure that the installation space and operational convenience of the ceiling cloth meet the requirements; according to the optimized design plan, the final 3D model and engineering drawing of the ceiling cloth, tensioning points and adjacent components are generated, and applied to the actual production and assembly process to ensure the installation quality and efficiency of the ceiling cloth.

5. The method for synchronously tensioning a car roof cloth according to claim 1, characterized in that: The specific method in S4 includes: The real-time image data of the ceiling cloth during the tensioning process is obtained, and the image data is input into the pre-built convolutional neural network model for processing to obtain the shape characteristics and stress distribution characteristics of the ceiling cloth; according to the shape characteristics of the ceiling cloth, the edge detection algorithm is used to extract the outline of the ceiling cloth, and the outline is compared with the preset standard outline to determine whether the shape of the ceiling cloth meets the requirements. If it does not meet the requirements, the shape abnormality warning information is output; according to the stress distribution characteristics of the ceiling cloth, the numerical analysis method is used to calculate the stress value of each area of ​​the ceiling cloth, and the stress value is compared with the preset safety threshold to determine whether the stress exceeds the safety range. If it exceeds the safety range, the stress abnormality warning information is output; the shape characteristics and stress distribution characteristics of the ceiling cloth are fused and analyzed to construct the ceiling The cloth health status assessment model is used to judge the overall tensioning quality of the ceiling cloth according to the output results of the assessment model and obtain the quality score; the shape abnormality warning information and stress abnormality warning information of the ceiling cloth and the overall quality score are output to the monitoring interface, and the relevant data are transmitted to the control unit of the tensioning equipment at the same time, and the real-time adjustment of the tensioning process is realized through feedback control; after the ceiling cloth is tensioned, the 3D model of the ceiling cloth is generated by 3D reconstruction technology, and the 3D model is compared with the design model to calculate the shape error and obtain the final ceiling cloth forming quality report; a database of the ceiling cloth tensioning process is established to record the image data, shape characteristics, stress distribution characteristics and health status score information at each moment, so as to provide data support for subsequent process optimization and quality analysis; The surface image of the ceiling cloth is obtained, and the image preprocessing method is used to reduce noise and enhance the image to improve the image quality; the feature points of the ceiling cloth surface are extracted from the preprocessed image by a feature extraction algorithm to obtain a feature point set; based on the feature point set, the optical flow method is used to calculate the displacement of the feature points between two adjacent frames of images to obtain a displacement matrix; based on the displacement matrix, the strain at each feature point is calculated by a strain calculation formula to obtain a strain matrix; the displacement matrix and the strain matrix are statistically analyzed to obtain the distribution characteristics of the displacement and the strain; based on the pre-established ceiling cloth tensioning effect judgment model, the displacement and strain distribution characteristics are used as input to judge whether the ceiling cloth achieves the expected tensioning effect; if the ceiling cloth does not achieve the expected tensioning effect, the tensioning parameters are adjusted according to the judgment result feedback, and the tensioning process is performed again until the expected effect is achieved; The vehicle roof layout image is acquired through a camera, and the image is preprocessed, including grayscale and denoising operations, to obtain image data suitable for analysis; the preprocessed image data is input into a pre-trained convolutional neural network model to extract the shape features of the image and obtain a feature vector; based on the feature vector, the support vector machine algorithm is used to classify the wrinkle area and determine the type of wrinkle; for each wrinkle area, the depth and width parameters of the wrinkle are calculated to quantify the degree of wrinkles and obtain a numerical representation of the degree of wrinkles; if the degree of wrinkles exceeds a preset threshold, the area is judged to be a severe wrinkle area and needs to be repaired or reworked; the location, type and degree information of the detected wrinkle area is summarized to generate a test report to provide data support for subsequent improvement of the quality of the roof layout; based on the test report, the convolutional neural network model is fine-tuned and optimized to improve the accuracy and efficiency of wrinkle detection and achieve continuous improvement of the quality of the roof layout; According to the pre-collected ceiling cloth wrinkle sample data, a wrinkle sample feature database is established to extract the texture and shape feature parameters of the wrinkle samples; the tensioning point layout parameters of the ceiling cloth tensioning structure of the current ceiling cloth are obtained, including the tensioning point position coordinates and tensioning force parameters, and are input into the pre-trained convolutional neural network model; the convolutional neural network model is used to extract and analyze the input ceiling cloth tensioning structure tensioning point layout parameters, and the cause of the wrinkles that may appear in the ceiling cloth under the current layout parameters is predicted by comparing them with the data in the wrinkle sample feature database; if the predicted wrinkle cause is caused by uneven tension force, then the tensioning point layout parameters are extracted and analyzed based on the pre-trained convolutional neural network model ... According to the position and shape of the wrinkles, the position of the tensioning point and the size of the tensioning force that need to be adjusted are determined, and the tensioning force adjustment parameters are generated; if the predicted wrinkle cause is caused by unreasonable tensioning point arrangement, the genetic algorithm is used to optimize the tensioning point arrangement parameters, and the optimal tensioning point arrangement scheme is explored through cross-mutation operation; the adjusted tensioning force parameters or optimized tensioning point arrangement parameters are output and applied to the actual ceiling cloth tensioning process to verify and evaluate the wrinkle condition of the ceiling cloth; according to the feedback of the ceiling cloth wrinkle condition, the convolutional neural network model is further trained and optimized to improve the prediction accuracy and adaptability of the model.

6. The method for synchronously tensioning a car roof cloth according to claim 1, characterized in that: The specific method in S5 includes: According to the actual situation after the ceiling cloth is stretched, determine the scanning range and scanning accuracy parameters of the 3D scanner; start the 3D scanner, perform a full-scale scan of the ceiling cloth surface, and obtain the original point cloud data; pre-process the original point cloud data, remove noise points and outliers, and obtain the cleaned point cloud data; according to the characteristics of the scanned object, use the normal vector estimation algorithm to calculate the normal vector of each point, and obtain the point cloud data with normal vector information; through the density analysis of the point cloud data, judge the integrity and uniformity of the point cloud data, if it does not meet the requirements, return to step 1 and rescan; use the point cloud data to perform triangulation processing and reconstruct the 3D mesh model of the ceiling cloth surface; compare and analyze the reconstructed 3D mesh model with the design model to obtain the shape error and size deviation of the ceiling cloth surface, and provide data support for subsequent process optimization; The three-dimensional point cloud data of the ceiling cloth is obtained, and a digital model of the ceiling cloth is generated through a surface reconstruction algorithm; the generated digital model of the ceiling cloth is compared with the preset design model, and the deviation value between the two models is calculated; the overall tensioning quality of the ceiling cloth is determined according to the deviation value, and if the deviation value exceeds the preset threshold, the overall tensioning quality of the ceiling cloth is judged to be unqualified; the dimensional accuracy of the ceiling cloth is evaluated by calculating the dimensional difference between the digital model and the design model at key positions; the support vector machine algorithm is used to comprehensively evaluate the overall tensioning quality and dimensional accuracy of the ceiling cloth to obtain a quality score; the deviation value, dimensional difference and quality score information of the ceiling cloth are integrated to generate a ceiling cloth quality evaluation report; the incremental learning method is used to optimize the surface reconstruction algorithm to improve the generation accuracy of the digital model of the ceiling cloth.

Images