Methods, devices and systems for high-precision assembly of aerospace composite materials

Abstract

This application provides a method, apparatus, and system for high-precision assembly quality control of aerospace composite materials, relating to the fields of aerospace manufacturing assembly and intelligent control technology. The method includes: acquiring the initial shape deviation between the current shape of the section to be assembled and the target shape; controlling each actuator in the actuator array to apply test loads in a preset order, recording the displacement data of each actuator, and identifying the local response matrix between the actuators; inputting the local response matrix into a pre-constructed joint Bayesian prior model to calculate the global response matrix of the section to be assembled, and calculating the optimal actuator control force vector based on the initial shape deviation and the global response matrix. The method, apparatus, and system for high-precision assembly quality control of aerospace composite materials provided by this application can fully utilize the sensing function of the actuators to quickly identify the stiffness characteristics of each section without adding additional external measurements, thereby achieving efficient and high-precision shape control.

Description

Technical Field

[0001] This application relates to the field of aerospace manufacturing assembly and intelligent control technology, and in particular to a method, device and system for high-precision assembly quality control of aerospace composite materials. Background Technology

[0002] In the manufacturing process of modern large passenger and transport aircraft, fuselage section assembly is a critical step in the final assembly stage. Composite material fuselages are characterized by being lightweight, high-strength, and highly flexible. However, due to the combined effects of the thin sheet weight's deflection and manufacturing deviations, certain shape deviations often occur. If the fuselage section's shape is not accurately adjusted to the target shape before assembly, it will lead to excessive interface gaps, misaligned holes, or even overloaded connectors, seriously affecting assembly quality and service performance, and also bringing the risk of rework and scrap.

[0003] In related technologies, the process typically relies on the operator's experience, involving repeated adjustments to the actuator force, measurements of the external shape, and further adjustments until the tolerance requirements are met. This method is inefficient, dependent on the engineer's experience, and has a long debugging cycle in multi-actuator, strongly coupled structures. Summary of the Invention

[0004] The purpose of this application is to provide a method, device and system for high-precision assembly quality control of aerospace composite materials, which can make full use of the sensing function of the actuator to quickly identify the stiffness characteristics of each cylindrical section without adding additional external measurements, thereby achieving efficient and high-precision shape control.

[0005] This application provides a method for high-precision assembly quality control of aerospace composite materials, including: The process involves: acquiring the initial shape deviation between the current shape of the section to be assembled and the target shape; controlling each actuator in the actuator array to apply test loads in a preset order, recording the displacement data of each actuator, and identifying the local response matrix between the actuators based on the displacement data of each actuator; inputting the local response matrix into a pre-constructed joint Bayesian prior model to calculate the global response matrix of the section to be assembled, and calculating the optimal actuator control force vector based on the initial shape deviation and the global response matrix; wherein the joint Bayesian prior model is used to characterize the mapping relationship between the local response matrix between the actuators and the global response matrix of the section; and the optimal actuator control force vector is used to adjust the shape of the section to the target shape.

[0006] Optionally, each actuator in the control actuator array is subjected to a test load in a preset order, and the displacement data of each actuator is recorded, including: selecting one actuator from the actuator array in the preset order, and applying multiple test loads of different amplitudes and within the elastic range to the selected actuator each time; recording the end displacement response measured by the built-in displacement sensors of all actuators each time a test load is applied, so as to obtain the displacement data of all actuators when the test load is applied.

[0007] Optionally, the step of identifying the local response matrix between actuators based on the displacement data of each actuator includes: linearly fitting the test load and displacement data corresponding to each actuator based on the displacement data of each actuator to obtain the linear relationship between the force applied by each actuator and the displacement of all actuators; summarizing the linear relationships corresponding to all actuators to construct the local response matrix between actuators.

[0008] Optionally, the step of inputting the local response matrix into a pre-constructed joint Bayesian prior model to deduce the global response matrix of the cylinder segment to be assembled includes: taking the identified local response matrix as input, substituting it into the joint Bayesian prior model, and obtaining the global response matrix of the cylinder segment to be assembled by calculating the posterior mean of the conditional distribution.

[0009] Optionally, the step of calculating the optimal actuator control force vector based on the initial shape deviation and the global response matrix includes: taking the target shape as the optimization objective and the maximum and minimum forces allowed by each actuator as constraints, solving a constraint optimization problem, and calculating the optimal actuator control force vector that adjusts the cylinder section shape to the target shape.

[0010] Optionally, the method further includes: randomly sampling the manufacturing parameters multiple times to generate multiple finite element simulation samples; each simulation sample corresponds to a cylindrical segment structure model with manufacturing differences; and using the multiple finite element simulation samples, constructing a joint Bayesian prior model between the local response matrix of the actuators and the global response matrix of the cylindrical segment.

[0011] Optionally, the step of constructing a joint Bayesian prior model between the local response matrix of the actuators and the global response matrix of the cylinder segment using the multiple finite element simulation samples includes: for each simulation sample, simulating and calculating the displacement response of all actuators and cylinder segment contact points when each actuator applies a unit test force, generating a local response matrix between the actuators, and extracting the displacement response of multiple key measurement points arranged along the circumference of the cylinder segment to construct a global response matrix; vectorizing and concatenating the local response matrix and global response matrix of all simulation samples to construct an augmented sample vector, and estimating the mean vector and covariance matrix based on the augmented sample vector to obtain a joint Bayesian prior model characterizing the statistical mapping relationship between the local response matrix and the global response matrix.

[0012] This application also provides a high-precision assembly quality control device for aerospace composite materials, comprising: The assembly includes a shape acquisition module for acquiring the initial shape deviation between the current shape of the cylinder segment to be assembled and the target shape; a local identification module for controlling each actuator in the actuator array to apply test loads in a preset order, recording the displacement data of each actuator, and identifying the local response matrix between the actuators based on the displacement data of each actuator; a global calculation module for inputting the local response matrix into a pre-constructed joint Bayesian prior model to calculate the global response matrix of the cylinder segment to be assembled; and a control force calculation module for calculating the optimal actuator control force vector based on the initial shape deviation and the global response matrix. The joint Bayesian prior model characterizes the mapping relationship between the local response matrix between the actuators and the global response matrix of the cylinder segment. The optimal actuator control force vector is used to adjust the shape of the cylinder segment to the target shape.

[0013] Optionally, the local identification module is specifically used to select one actuator from the actuator array in the preset order, and apply multiple test loads of different amplitudes and within the elastic range to the selected actuator each time; the local identification module is also specifically used to record the end displacement response measured by the built-in displacement sensors of all actuators each time a test load is applied, so as to obtain the displacement data of all actuators when the test load is applied.

[0014] Optionally, the local identification module is specifically used to perform linear fitting between the test load and displacement data corresponding to each actuator based on the displacement data of each actuator, so as to obtain the linear relationship between the force applied by each actuator and the displacement of all actuators; the local identification module is also specifically used to summarize the linear relationships corresponding to all actuators to construct a local response matrix between actuators.

[0015] Optionally, the global calculation module is specifically used to take the identified local response matrix as input, substitute it into the joint Bayesian prior model, and obtain the global response matrix of the cylinder segment to be assembled by calculating the posterior mean of the conditional distribution.

[0016] Optionally, the control force calculation module is specifically used to solve a constraint optimization problem with the target shape as the optimization target and the maximum and minimum forces allowed by each actuator as constraints, and to calculate the optimal actuator control force vector that adjusts the cylinder section shape to the target shape.

[0017] Optionally, the device further includes: a sample generation module and a model building module; the sample generation module is used to randomly sample the manufacturing parameters multiple times to generate multiple finite element simulation samples; each simulation sample corresponds to a cylindrical segment structure model with manufacturing differences; the model building module is used to use the multiple finite element simulation samples to construct a joint Bayesian prior model between the local response matrix of the actuators and the global response matrix of the cylindrical segment.

[0018] Optionally, the model building module is specifically used to simulate and calculate the displacement response of all actuators and cylinder contact points when each actuator applies a unit test force for each simulation sample, generate a local response matrix between actuators, and extract the displacement response of multiple key measurement points arranged along the circumference of the cylinder to construct a global response matrix; the model building module is also specifically used to vectorize and concatenate the local response matrix and global response matrix of all simulation samples to construct an augmented sample vector, and estimate the mean vector and covariance matrix based on the augmented sample vector to obtain a joint Bayesian prior model characterizing the statistical mapping relationship between the local response matrix and the global response matrix.

[0019] This application also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the high-precision assembly quality control method for aerospace composites as described above.

[0020] This application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the high-precision assembly quality control method for aerospace composites as described above.

[0021] This application also provides a fuselage section assembly shape control system, including: an actuator array, a shape measurement unit, and a central processing unit; the central processing unit stores a computer program, which, when executed by a processor, implements the steps of the high-precision assembly quality control method for aerospace composites as described above.

[0022] This application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the high-precision assembly quality control method for aerospace composites as described above.

[0023] The high-precision assembly quality control method, apparatus, and system for aerospace composite materials provided in this application first obtain the initial shape deviation between the current shape and the target shape of the cylindrical segment to be assembled. Then, each actuator in the actuator array is controlled to apply test loads in a preset order, and the displacement data of each actuator is recorded. Based on the displacement data of each actuator, the local response matrix between the actuators is identified. Finally, the local response matrix is ​​input into a pre-constructed joint Bayesian prior model to calculate the global response matrix of the cylindrical segment to be assembled. Based on the initial shape deviation and the global response matrix, the optimal actuator control force vector is calculated. The joint Bayesian prior model is used to characterize the mapping relationship between the local response matrix between the actuators and the global response matrix of the cylindrical segment. The optimal actuator control force vector is used to adjust the shape of the cylindrical segment to the target shape. In this way, the sensing function of the actuators can be fully utilized to quickly identify the stiffness characteristics of each cylindrical segment without adding additional external measurements, thereby achieving efficient and high-precision shape control. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is one of the structural schematic diagrams of the fuselage section assembly shape control system provided in this application; Figure 2 This is the second structural schematic diagram of the fuselage section assembly shape control system provided in this application; Figure 3 This is one of the flowcharts illustrating the high-precision assembly quality control method for aerospace composites provided in this application; Figure 4 This is the second flowchart illustrating the high-precision assembly quality control method for aerospace composites provided in this application; Figure 5 This is a schematic diagram of the specific process of actuator sensing provided in this application; Figure 6 This is a flowchart illustrating the process of constructing a joint Bayesian prior model provided in this application; Figure 7This is a structural schematic diagram of the high-precision assembly quality control device for aerospace composite materials provided in this application; Figure 8 This is a schematic diagram of the structure of the electronic device provided in this application. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0027] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, "and / or" in the specification and claims indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship. All actions involving the acquisition of signal information or data in this application are performed in accordance with the relevant data protection laws and policies of the country where the application is located and with authorization from the owner of the relevant device.

[0028] To address the aforementioned technical problems in related technologies, this application provides a high-precision assembly quality control method for aerospace composite materials. This method utilizes actuators with integrated force-displacement sensors to obtain the local mechanical response characteristics between actuators in the current fuselage section through sequential loading tests. Then, it calculates the global mechanical response characteristics based on a pre-constructed statistical mapping relationship and combines this with the initial shape measured by a laser tracker to calculate the optimal actuator force. In actual assembly, the shape adjustment of the fuselage section is usually within a very small deformation range, and the structural response can approximately satisfy the linear elastic law. After the force applied by the actuator is transmitted along the structural path, it will produce a small displacement proportional to the applied force at various positions in the section. This force-displacement relationship can be abstracted as a linear mapping, and its coefficients constitute a response matrix, used to characterize the degree of influence of different actuator forces on the displacement of each measuring point in the section. This is also the core model basis for realizing shape adjustment and force calculation in this invention.

[0029] like Figure 1As shown in the embodiment of this application, a fuselage section assembly shape control system for implementing a high-precision assembly quality control method for aerospace composite materials is provided. The system includes an actuator array, a shape measurement unit, and a central processing unit. The actuator array comprises multiple actuators arranged along the lower half of the fuselage section. Each actuator has a built-in force sensor and displacement sensor, capable of measuring the applied push / pull force and the minute displacement response after loading in real time. The shape measurement unit uses a laser tracker or other three-dimensional measuring equipment to acquire the three-dimensional coordinates of preset key measurement points on the fuselage section, obtaining the initial shape before assembly and the final assembled shape. The central processing unit is the core component, responsible for executing the high-precision assembly quality control method for aerospace composite materials proposed in this application embodiment.

[0030] like Figure 2 The diagram shows the structure of the fuselage section assembly shape control system. The aircraft composite fuselage section (1) to be assembled is placed in a rigid support fixture (3) and supported and loaded by actuators (2) containing force sensors and displacement sensors arranged along the lower half of the section. Measurement nodes (4) are arranged at key positions of the section and used in conjunction with a laser tracker (5) to obtain initial shape data. The actuators and laser trackers are all connected to the control unit (6) for communication. The computer terminal uniformly executes data acquisition, stiffness identification calculation and final actuation control command issuance.

[0031] The high-precision assembly quality control method for aerospace composite materials provided in this application will be described in detail below with reference to the accompanying drawings, through specific embodiments and application scenarios.

[0032] like Figure 3 As shown in the embodiment of this application, a high-precision assembly quality control method for aerospace composites is provided, which may include the following steps 301 to 303: Step 301: Obtain the initial shape deviation between the current shape of the cylindrical section to be assembled and the target shape.

[0033] For example, in the actual assembly process, after the composite fuselage section to be assembled (i.e., the aforementioned section to be assembled) is placed on the assembly bracket or tooling, a laser tracker is used to perform three-dimensional coordinate measurements on the key measurement points pre-arranged on the surface of the section (i.e., such as...). Figure 4 (As shown in the initial shape measurement), the current initial shape deviation vector of this cylinder segment is obtained. It is used to determine the initial deformation of the lower cylinder section at the current position caused by gravity sagging, manufacturing errors, transportation collisions, etc., and to provide target compensation for subsequent actuator control.

[0034] Step 302: Control each actuator in the actuator array to apply the test load in a preset order, record the displacement data of each actuator, and identify the local response matrix between the actuators based on the displacement data of each actuator.

[0035] For example, the local response matrix between actuators can be obtained by applying test loads individually or in groups and calculating the displacement response of each contact point.

[0036] Specifically, step 302 above may also include the following steps 302a1 and 302a2: Step 302a1: Select an actuator from the actuator array in the preset order, and apply multiple test loads with different amplitudes and within the elastic range to the selected actuator each time.

[0037] Step 302a2: When the test load is applied each time, record the end displacement response measured by the built-in displacement sensor of all actuators to obtain the displacement data of all actuators when the test load is applied.

[0038] Exemplarily, in the embodiments of this application, such as Figure 4 As shown, after initial shape measurement and obtaining initial shape deviation, actuator sensing tests can be performed. That is, each actuator is selected in a predetermined order. Apply a set of test loads with small amplitudes that are within the elastic range (such as...) During this process, other actuators are kept unloaded or kept in a constant support state.

[0039] For example, such as Figure 5 As shown, the experiment specifically includes: selecting each actuator in a predetermined order. Apply a sequence of test loads with a small amplitude that is within the elastic range (such as...) During this process, other actuators remain unloaded. Then, for each set of test loads, all actuators (denoted as...) are used. The built-in displacement sensor records the displacement at the contact position between the end and the cylinder section, which can be denoted as: Specifically, step 302 above may also include the following steps 302b1 and 302b2: Step 302b1: Based on the displacement data of each actuator, perform linear fitting between the test load and displacement data corresponding to each actuator to obtain the linear relationship between the force applied by each actuator and the displacement of all actuators.

[0040] Step 302b2: Summarize the linear relationships corresponding to all actuators and construct the local response matrix between actuators.

[0041] For example, after obtaining the displacement data of all actuators, such as Figure 5 As shown, the above experiment also includes: obtaining the results for each actuator Linear fitting was performed based on multiple sets of test loads and measured displacements (the coefficients are the response matrix of the first set of loads). k (Column), to obtain the linear relationship between the force applied by the actuator and the displacements of all actuators. Repeat the above steps until each actuator has been traversed. Then, summarize all the test results and construct the local response matrix, i.e.: in, This represents the current on-site clamping state of the physical cylinder segment, when the first... When the actuator applies a unit test force, the first... The displacement of the actuator contact point.

[0042] Step 303: Input the local response matrix into the pre-constructed joint Bayesian prior model to deduce the global response matrix of the cylinder segment to be assembled, and calculate the optimal actuator control force vector based on the initial shape deviation and the global response matrix.

[0043] The joint Bayesian prior model is used to characterize the mapping relationship between the local response matrix of the actuators and the global response matrix of the cylinder segment; the optimal actuator control force vector is used to adjust the shape of the cylinder segment to the target shape.

[0044] For example, such as Figure 4 As shown, after obtaining the actual measured local response matrix, the global response matrix can be calculated based on the local response matrix using the Bayesian prior model (i.e., the joint Bayesian prior model mentioned above).

[0045] Specifically, step 303 above, which involves calculating the global response matrix of the cylindrical segment to be assembled, may further include the following step 303a: Step 303a: Using the identified local response matrix as input, substitute it into the joint Bayesian prior model, and obtain the global response matrix of the cylinder segment to be assembled by calculating the posterior mean of the conditional distribution.

[0046] For example, by utilizing a pre-built joint Bayesian prior model, the data identified on-site can be... As input, the global response matrix of the current tube segment is calculated. Without considering measurement noise or incorporating measurement noise into the calculation... In this case, according to the conditional distribution formula of the multivariate Gaussian distribution, we can obtain: The posterior mean is: The posterior covariance is: in, Reflecting the residual uncertainty of the model, It is the mean vector. Let be the covariance matrix. , , , ,as well as All of these were obtained when constructing the joint Bayesian prior model.

[0047] For example, such as Figure 4 As shown, in obtaining the global response matrix and initial shape deviation Then, the actuator control force optimization calculation can be performed to obtain the optimal control force.

[0048] Specifically, step 303 above, which involves calculating the optimal actuator control force vector, may further include the following step 303b: Step 303b: Using the target shape as the optimization objective and the maximum and minimum forces allowed by each actuator as constraints, solve the constraint optimization problem to calculate the optimal actuator control force vector that adjusts the cylinder section shape to the target shape.

[0049] For example, such as Figure 4 As shown, in this embodiment of the application, an optimization method is used to solve for the actuator control force vector. To adjust the cylindrical section to the target shape. (For example, designing a baseline shape or a contour that meets the requirements for wiring shape), its basic form is: in, The minimum / maximum force constraints allowed for each actuator can be set according to structural safety and equipment capacity. The above-mentioned constraint quadratic programming can be conveniently solved using a common least squares solver (such as MOSEK, CPLEX, etc.), yielding the optimal actuator force. It is sent to the actuator array for execution, completing the precise adjustment of the cylinder section's shape.

[0050] Optionally, in the embodiments of this application, before application at the assembly site, it is necessary to construct a joint Bayesian prior model of the structural response matrix using finite element simulation in the offline stage.

[0051] For example, prior to step 302 above, the high-precision assembly quality control method for aerospace composites provided in this application embodiment may further include steps 304 and 305: Step 304: Perform multiple random samplings of the manufacturing parameters to generate multiple finite element simulation samples.

[0052] Each simulation sample corresponds to a cylindrical section structure model with manufacturing differences.

[0053] For example, in this application embodiment, manufacturing parameters are sampled and a finite element model is generated. For instance... Figure 5 As shown, due to the randomness of manufacturing parameters (such as ply angle, thickness, etc.), the manufacturing parameters such as composite material ply thickness, ply angle, connection stiffness, and support stiffness can be randomly sampled multiple times to generate N sets of finite element simulation samples. Each sample corresponds to a cylindrical structure model with manufacturing differences.

[0054] Step 305: Using the multiple finite element simulation samples, construct a joint Bayesian prior model between the local response matrix of the actuators and the global response matrix of the cylinder segment.

[0055] For example, after obtaining the above N After creating a set of finite element simulation samples, the local response matrix between actuators can be obtained through simulation.

[0056] Specifically, step 305 above may also include the following steps 305a1 and 305a2: Step 305a1: For each simulation sample, calculate the displacement response of all actuators and cylinder contact points when each actuator applies a unit test force, generate the local response matrix between actuators, and extract the displacement response of multiple key measurement points arranged along the circumference of the cylinder to construct the global response matrix.

[0057] Step 305a2: Vectorize and concatenate the local response matrices and global response matrices of all simulation samples to construct augmented sample vectors. Estimate the mean vector and covariance matrix based on the augmented sample vectors to obtain a joint Bayesian prior model that characterizes the statistical mapping relationship between the local response matrix and the global response matrix.

[0058] For example, such as Figure 6 As shown, on each finite element sample, actuation points are defined according to the location of the actuators in the field. By applying unit forces individually or in groups, the displacement response of each actuation point is calculated, and the local response matrix between the actuators is obtained. in, Number of actuators Indicates the first In the simulated sample, when the first... When the actuator applies a unit test force, the first... The displacement of the actuator contact position.

[0059] For example, after obtaining the local response matrix, the global response matrix can be simulated. (e.g.) Figure 6 As shown, extract the components arranged along the circumference of the cylinder section. The displacement response of each key measurement point (corresponding to the laser measurement point) is used to construct a global response matrix: Finally, a joint statistical mapping is constructed. The local response matrix of each sample is then plotted. and global response matrix Vectorization and concatenation are performed to form augmented sample vectors: Assuming the augmented vector described above follows a multivariate Gaussian statistical model under varying manufacturing differences and boundary conditions, through... N Estimate the mean vector of a set of finite element samples Covariance Matrix Thus, we obtain the joint Bayesian prior: For example, this joint Bayesian prior characterizes the local response. and global response The statistical correlation between them is based on on-site measurements. right This provides the foundation for Bayesian inference.

[0060] The high-precision assembly quality control method for aerospace composite materials provided in this application has the following significant advantages: 1. Fully utilize the actuator's sensing capabilities to achieve personalized stiffness identification: This method upgrades the actuator from a traditional "pure execution unit" to an "integrated execution-sensing unit." By conducting sequential small-amplitude loading tests through the actuator's internal force-displacement sensors, a local response matrix is ​​constructed. This allows for the automatic sensing of the stiffness characteristics of the current cylinder section without adding external measuring equipment, significantly improving the adaptability to manufacturing differences.

[0061] 2. Significantly reduces external testing and measurement costs: This method only requires one complete laser shape measurement to obtain the initial deformation at the assembly site. All other identification work is completed by the actuator sensing, avoiding the repeated testing process of "loading-laser measurement-unloading" in traditional methods, and reducing the time occupied by the measurement station and the investment of personnel.

[0062] 3. Improve the accuracy and stability of shape control: By constructing a local-global response mapping model offline, this method can transform the local response information of the actuator into a global response matrix, making the calculated actuator force more consistent with the actual stiffness characteristics of the current cylinder section and reducing the residual error caused by model mismatch. On this basis, regularization constraints can be further introduced to balance control accuracy and structural safety.

[0063] 4. Clear assembly process, easy integration and promotion: The method has a clear process, including offline simulation modeling, on-site shape measurement, actuator perception and identification, and force optimization solution based on the calculated response matrix, making it easy to integrate into existing fuselage section assembly lines. Since the laser tracker is only used once at the start of operation, subsequent shape adjustments are almost entirely completed autonomously by the actuator system, facilitating its application in multi-station and multi-model assembly environments.

[0064] The high-precision assembly quality control method for aerospace composites provided in this application first obtains the initial shape deviation between the current shape and the target shape of the cylindrical segment to be assembled. Then, it controls each actuator in the actuator array to apply test loads in a preset order, records the displacement data of each actuator, and identifies the local response matrix between the actuators based on the displacement data of each actuator. Finally, it inputs the local response matrix into a pre-constructed joint Bayesian prior model to calculate the global response matrix of the cylindrical segment to be assembled, and calculates the optimal actuator control force vector based on the initial shape deviation and the global response matrix. The joint Bayesian prior model characterizes the mapping relationship between the local response matrix between actuators and the global response matrix of the cylindrical segment; the optimal actuator control force vector is used to adjust the shape of the cylindrical segment to the target shape. In this way, the sensing function of the actuators can be fully utilized to quickly identify the stiffness characteristics of each cylindrical segment without adding additional external measurements, thereby achieving efficient and high-precision shape control.

[0065] It should be noted that the high-precision assembly quality control method for aerospace composite materials provided in this application can be executed by an aerospace composite high-precision assembly quality control device, or a control module within that device for executing the high-precision assembly quality control method. This application uses the execution of the high-precision assembly quality control method by the aerospace composite high-precision assembly quality control device as an example to illustrate the aerospace composite high-precision assembly quality control device provided in this application.

[0066] It should be noted that, in the embodiments of this application, the high-precision assembly quality control methods for aerospace composite materials shown in the accompanying drawings are all illustrated by way of example with reference to one of the accompanying drawings in the embodiments of this application. In specific implementation, the high-precision assembly quality control methods for aerospace composite materials shown in the accompanying drawings of the above methods can also be implemented in conjunction with any other accompanying drawings shown in the above embodiments, which will not be elaborated here.

[0067] The high-precision assembly quality control device for aerospace composites provided in this application is described below. The high-precision assembly quality control method for aerospace composites described below can be referred to in correspondence with the method described above.

[0068] Figure 7 This is a schematic diagram of the structure of the high-precision assembly quality control device for aerospace composites provided in the embodiments of this application, as shown below. Figure 7 As shown, it specifically includes: The shape acquisition module 701 is used to acquire the initial shape deviation between the current shape of the cylinder segment to be assembled and the target shape; the local identification module 702 is used to control each actuator in the actuator array to apply test loads in a preset order, record the displacement data of each actuator, and identify the local response matrix between the actuators based on the displacement data of each actuator; the global calculation module 703 is used to input the local response matrix into a pre-constructed joint Bayesian prior model to calculate the global response matrix of the cylinder segment to be assembled; the control force calculation module 704 is used to calculate the optimal actuator control force vector based on the initial shape deviation and the global response matrix; wherein, the joint Bayesian prior model is used to characterize the mapping relationship between the local response matrix between the actuators and the global response matrix of the cylinder segment; the optimal actuator control force vector is used to adjust the shape of the cylinder segment to the target shape.

[0069] Optionally, the local identification module 702 is specifically used to select one actuator from the actuator array in the preset order, and apply multiple test loads of different amplitudes and within the elastic range to the selected actuator each time; the local identification module 702 is also specifically used to record the end displacement response measured by the built-in displacement sensor of all actuators each time the test load is applied, so as to obtain the displacement data of all actuators when the test load is applied.

[0070] Optionally, the local identification module 702 is specifically used to perform linear fitting between the test load and displacement data corresponding to each actuator based on the displacement data of each actuator, so as to obtain the linear relationship between the force applied by each actuator and the displacement of all actuators; the local identification module 702 is also specifically used to summarize the linear relationships corresponding to all actuators to construct a local response matrix between actuators.

[0071] Optionally, the global calculation module 703 is specifically used to take the identified local response matrix as input, substitute it into the joint Bayesian prior model, and obtain the global response matrix of the cylinder segment to be assembled by calculating the posterior mean of the conditional distribution.

[0072] Optionally, the control force calculation module 704 is specifically used to solve a constraint optimization problem with the target shape as the optimization target and the maximum and minimum forces allowed by each actuator as constraints, and to calculate the optimal actuator control force vector that adjusts the shape of the cylinder section to the target shape.

[0073] Optionally, the device further includes: a sample generation module and a model building module; the sample generation module is used to randomly sample the manufacturing parameters multiple times to generate multiple finite element simulation samples; each simulation sample corresponds to a cylindrical segment structure model with manufacturing differences; the model building module is used to use the multiple finite element simulation samples to construct a joint Bayesian prior model between the local response matrix of the actuators and the global response matrix of the cylindrical segment.

[0074] Optionally, the model building module is specifically used to simulate and calculate the displacement response of all actuators and cylinder contact points when each actuator applies a unit test force for each simulation sample, generate a local response matrix between actuators, and extract the displacement response of multiple key measurement points arranged along the circumference of the cylinder to construct a global response matrix; the model building module is also specifically used to vectorize and concatenate the local response matrix and global response matrix of all simulation samples to construct an augmented sample vector, and estimate the mean vector and covariance matrix based on the augmented sample vector to obtain a joint Bayesian prior model characterizing the statistical mapping relationship between the local response matrix and the global response matrix.

[0075] The high-precision assembly quality control device for aerospace composite materials provided in this application first acquires the initial shape deviation between the current shape and the target shape of the cylindrical segment to be assembled. Then, it controls each actuator in the actuator array to apply test loads in a preset order, records the displacement data of each actuator, and identifies the local response matrix between the actuators based on the displacement data of each actuator. Finally, it inputs the local response matrix into a pre-constructed joint Bayesian prior model to calculate the global response matrix of the cylindrical segment to be assembled, and calculates the optimal actuator control force vector based on the initial shape deviation and the global response matrix. The joint Bayesian prior model is used to characterize the mapping relationship between the local response matrix between actuators and the global response matrix of the cylindrical segment; the optimal actuator control force vector is used to adjust the shape of the cylindrical segment to the target shape. In this way, the sensing function of the actuators can be fully utilized to quickly identify the stiffness characteristics of each cylindrical segment without adding additional external measurements, thereby achieving efficient and high-precision shape control.

[0076] Figure 8 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 8 As shown, the electronic device may include: a processor 810, a communications interface 820, a memory 830, and a communications bus 840, wherein the processor 810, the communications interface 820, and the memory 830 communicate with each other through the communications bus 840. The processor 810 can call logic instructions in the memory 830 to execute a high-precision assembly quality control method for aerospace composites. This method includes: first, acquiring the initial shape deviation between the current shape of the section to be assembled and the target shape; then, controlling each actuator in the actuator array to apply test loads in a preset order, recording the displacement data of each actuator, and identifying the local response matrix between actuators based on the displacement data of each actuator; finally, inputting the local response matrix into a pre-constructed joint Bayesian prior model to calculate the global response matrix of the section to be assembled, and calculating the optimal actuator control force vector based on the initial shape deviation and the global response matrix; wherein, the joint Bayesian prior model is used to characterize the mapping relationship between the local response matrix between actuators and the global response matrix of the section; the optimal actuator control force vector is used to adjust the shape of the section to the target shape. In this way, the sensing function of the actuators can be fully utilized to quickly identify the stiffness characteristics of each section without adding additional external measurements, thereby achieving efficient and high-precision shape control.

[0077] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0078] On the other hand, this application also provides a computer program product, which includes a computer program stored on a computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the high-precision assembly quality control method for aerospace composites provided by the above methods. The method includes: first, obtaining the initial shape deviation between the current shape of the cylinder segment to be assembled and the target shape; then, controlling each actuator in the actuator array to apply test loads in a preset order, recording the displacement data of each actuator, and identifying the local response matrix between the actuators based on the displacement data of each actuator; finally, inputting the local response matrix into a pre-constructed joint Bayesian prior model to calculate the global response matrix of the cylinder segment to be assembled, and calculating the optimal actuator control force vector based on the initial shape deviation and the global response matrix; wherein, the joint Bayesian prior model is used to characterize the mapping relationship between the local response matrix between the actuators and the global response matrix of the cylinder segment; the optimal actuator control force vector is used to adjust the shape of the cylinder segment to the target shape. In this way, the sensing function of the actuator can be fully utilized to quickly identify the stiffness characteristics of each cylinder section without adding additional external measurements, thereby achieving efficient and high-precision shape control.

[0079] In another aspect, this application also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the aforementioned high-precision assembly quality control methods for aerospace composites. The method includes: first, acquiring the initial shape deviation between the current shape of the section to be assembled and the target shape; then, controlling each actuator in the actuator array to apply test loads in a preset order, recording the displacement data of each actuator, and identifying the local response matrix between the actuators based on the displacement data of each actuator; finally, inputting the local response matrix into a pre-constructed joint Bayesian prior model to calculate the global response matrix of the section to be assembled, and calculating the optimal actuator control force vector based on the initial shape deviation and the global response matrix; wherein the joint Bayesian prior model is used to characterize the mapping relationship between the local response matrix between the actuators and the global response matrix of the section; and the optimal actuator control force vector is used to adjust the shape of the section to the target shape. In this way, the sensing function of the actuator can be fully utilized to quickly identify the stiffness characteristics of each cylinder section without adding additional external measurements, thereby achieving efficient and high-precision shape control.

[0080] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0081] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0082] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for high-precision assembly quality control of aerospace composite materials, characterized in that, include: Obtain the initial shape deviation between the current shape of the cylindrical section to be assembled and the target shape; The test load is applied to each actuator in the actuator array in a preset order, the displacement data of each actuator is recorded, and the local response matrix between the actuators is identified based on the displacement data of each actuator. The local response matrix is ​​input into a pre-constructed joint Bayesian prior model to deduce the global response matrix of the cylindrical section to be assembled, and the optimal actuator control force vector is calculated based on the initial shape deviation and the global response matrix. The joint Bayesian prior model is used to characterize the mapping relationship between the local response matrix of the actuators and the global response matrix of the cylinder segment; the optimal actuator control force vector is used to adjust the shape of the cylinder segment to the target shape.

2. The method according to claim 1, characterized in that, Each actuator in the control actuator array applies a test load in a preset sequence, and the displacement data of each actuator is recorded, including: According to the preset order, one actuator is selected from the actuator array in sequence, and multiple test loads with different amplitudes and within the elastic range are applied to the selected actuator each time. Each time a test load is applied, the end displacement response measured by the built-in displacement sensors of all actuators is recorded, thus obtaining the displacement data of all actuators when the test load is applied.

3. The method according to claim 1 or 2, characterized in that, The process of identifying the local response matrix between actuators based on the displacement data of each actuator includes: Based on the displacement data of each actuator, the test load and displacement data corresponding to each actuator are linearly fitted to obtain the linear relationship between the force applied by each actuator and the displacement of all actuators. By summarizing the linear relationships corresponding to all actuators, a local response matrix among the actuators is constructed.

4. The method according to claim 1, characterized in that, The step of inputting the local response matrix into a pre-constructed joint Bayesian prior model to calculate the global response matrix of the section to be assembled includes: The identified local response matrix is ​​used as input and substituted into the joint Bayesian prior model. By calculating the posterior mean of the conditional distribution, the global response matrix of the cylinder segment to be assembled is obtained.

5. The method according to claim 1, characterized in that, The calculation of the optimal actuator control force vector based on the initial shape deviation and the global response matrix includes: Using the target shape as the optimization objective and the maximum and minimum forces allowed by each actuator as constraints, the constraint optimization problem is solved to calculate the optimal actuator control force vector that adjusts the cylinder section shape to the target shape.

6. The method according to claim 1, characterized in that, The method further includes: Multiple random samplings of manufacturing parameters are performed to generate multiple finite element simulation samples; each simulation sample corresponds to a cylindrical segment structure model with manufacturing differences. Using the multiple finite element simulation samples, a joint Bayesian prior model is constructed between the local response matrices of the actuators and the global response matrix of the cylinder segment.

7. The method according to claim 6, characterized in that, The construction of a joint Bayesian prior model between the local response matrices of the actuators and the global response matrix of the cylinder segment using the multiple finite element simulation samples includes: For each simulation sample, the displacement response of all actuators and cylinder contact points is calculated when each actuator applies a unit test force, generating a local response matrix between actuators, and extracting the displacement response of multiple key measurement points arranged along the circumference of the cylinder to construct a global response matrix. The local and global response matrices of all simulation samples are vectorized and concatenated to construct augmented sample vectors. Based on the augmented sample vectors, the mean vector and covariance matrix are estimated to obtain a joint Bayesian prior model that characterizes the statistical mapping relationship between the local and global response matrices.

8. A high-precision assembly quality control device for aerospace composite materials, characterized in that, The device includes: The shape acquisition module is used to acquire the initial shape deviation between the current shape of the cylindrical section to be assembled and the target shape. The local identification module is used to control each actuator in the actuator array to apply the test load in a preset order, record the displacement data of each actuator, and identify the local response matrix between the actuators based on the displacement data of each actuator. The global calculation module is used to input the local response matrix into the pre-constructed joint Bayesian prior model to calculate the global response matrix of the cylinder segment to be assembled. The control force calculation module is used to calculate the optimal actuator control force vector based on the initial shape deviation and the global response matrix. The joint Bayesian prior model is used to characterize the mapping relationship between the local response matrix of the actuators and the global response matrix of the cylinder segment; the optimal actuator control force vector is used to adjust the shape of the cylinder segment to the target shape.

9. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the high-precision assembly quality control method for aerospace composites as described in any one of claims 1 to 7.

10. A fuselage section assembly shape control system, characterized in that, include: An actuator array, a shape measurement unit, and a central processing unit; the central processing unit stores a computer program that, when executed by a processor, implements the steps of the high-precision assembly quality control method for aerospace composites as described in any one of claims 1 to 7.

Images