A composite material automatic fiber placement path dynamic optimization method, device and medium

Abstract

The application discloses a composite automatic fiber placement path dynamic optimization method, device and medium, and belongs to the technical field of composite automatic fiber placement. The method comprises the following steps: obtaining a three-dimensional model of a component to be laid, and obtaining surface curvature distribution characteristics and fiber laying angle constraints based on the three-dimensional model; processing the surface curvature distribution characteristics based on a path planning algorithm to generate an initial fiber placement path; driving a fiber placement head to perform reverse pre-tension control based on a tension mapping table and a temperature compensation coefficient; obtaining actual fiber orientation data according to a preset sampling period based on a polarized laser polariscope, and extracting fiber angle deviations of the actual fiber orientation data; processing the fiber angle deviations and resin viscoelasticity data based on a Bayesian optimization algorithm to dynamically correct the laying path and generate an update instruction; and adjusting the spatial pose and tension of the fiber placement head according to the update instruction. The application realizes dynamic multi-target fiber placement path optimization and real-time feedback control based on the above method.

Description

Technical Field

[0001] This application relates to the technical field of automated fiber placement of composite materials, and in particular to a method, equipment and medium for dynamic optimization of automated fiber placement path of composite materials. Background Technology

[0002] Automated fiber placement technology for composite materials is a key technology in advanced manufacturing and has wide applications in aerospace and automotive engineering. Current technologies primarily rely on CAD models of components for fiber placement path planning, generating initial paths based on geometric algorithms and employing open-loop control strategies for tension control. This approach struggles to achieve high-precision placement in dynamic environments. Furthermore, traditional path correction methods often utilize rule-based local adjustments, lacking the ability to comprehensively process multi-source heterogeneous data. This results in limited control over fiber angle deviations during placement of complex curved surfaces and difficulty in adapting to the impact of resin viscoelasticity variations and fiber wetting state fluctuations on placement quality.

[0003] Traditional fiber placement methods do not fully consider the real-time coupling effect of environmental data (such as temperature and humidity) and material data (such as resin viscoelasticity and fiber wetting state), resulting in the inability to adjust placement parameters in a timely manner under dynamic conditions, affecting layup quality and forming efficiency. Existing path planning algorithms mostly focus on a single objective (such as shortest path), making it difficult to achieve multi-objective synergistic optimization among fiber continuity, layupability, wetting quality, and layup efficiency, leading to frequent manual intervention to balance various indicators in practical applications. Traditional tension control strategies mostly use fixed parameters or simple feedback mechanisms, failing to establish a multi-dimensional mapping relationship between tension, velocity, temperature, and viscoelastic parameters, making it difficult to adjust control parameters in real time according to the layup state, resulting in tension fluctuations affecting layup uniformity.

[0004] Therefore, how to achieve dynamic multi-objective fiber placement path optimization and real-time feedback control has become a technical problem that urgently needs to be solved. Summary of the Invention

[0005] This application provides a method, device, and medium for dynamic optimization of automatic fiber placement path in composite materials, in order to solve the following technical problem: how to achieve dynamic multi-objective fiber placement path optimization and real-time feedback control.

[0006] In a first aspect, embodiments of this application provide a dynamic optimization method for an automated fiber placement path of composite materials, applied to a fiber placement robot. The fiber placement robot includes a fiber placement head and a polarized laser interferometer. The method comprises: acquiring a three-dimensional model of the component to be placed, and acquiring surface curvature distribution characteristics and fiber placement angle constraints based on the three-dimensional model; wherein the fiber placement angle constraints are dynamically corrected based on environmental and material data of the component to be placed, the environmental data including temperature and humidity data, and the material data including resin viscoelasticity data and fiber wetting state data; and processing the surface curvature distribution characteristics based on a preset path planning algorithm to generate... An initial fiber placement path is established; based on a preset tension mapping table and temperature compensation coefficient, the fiber placement head is driven to perform reverse pretension control; wherein, the reverse pretension value is positively correlated with the square of the real-time placement speed of the fiber placement head; based on the polarization laser polarimeter, actual fiber orientation data is acquired according to a preset sampling period, and the fiber angle deviation of the actual fiber orientation data is extracted; based on a preset Bayesian optimization algorithm, the fiber angle deviation and the resin viscoelastic data are processed to dynamically correct the placement path and generate an update instruction; wherein, the update instruction includes a three-dimensional position compensation vector and a tension compensation coefficient; the spatial pose and tension of the fiber placement head are adjusted according to the update instruction.

[0007] In one implementation of this application, a three-dimensional model of the component to be laid is obtained, and surface curvature distribution characteristics and fiber laying angle constraints are obtained based on the three-dimensional model. Specifically, this includes: scanning the component to be laid using a preset CT scanner to obtain CT scan data; processing the CT scan data and the CAD model of the component to be laid using a multi-source heterogeneous data fusion algorithm to construct a three-dimensional model; extracting Gaussian curvature, average curvature, and principal curvature directions from the three-dimensional model using a preset local surface fitting algorithm to determine surface curvature distribution characteristics; and correcting the allowable range of fiber laying angles based on temperature and humidity data in the environmental data to determine fiber laying angle constraints; wherein the allowable range is non-linearly correlated with the extreme value region in the curvature feature map.

[0008] In one implementation of this application, the surface curvature distribution features are processed based on a preset path planning algorithm to generate an initial fiber placement path. Specifically, this includes: exploring paths within the surface curvature distribution features using a preset random sampling algorithm to generate candidate paths that satisfy fiber continuity and layability constraints; constructing a cost function based on the resin viscoelastic parameters in the material data; wherein the cost function includes a path length optimization objective, a curvature change rate optimization objective, and a wetting quality optimization objective; adjusting the weight coefficients of multiple optimization objectives in the cost function based on feedback from the wetting state data; and determining the initial fiber placement path based on the magnitude of the weight coefficients.

[0009] In one implementation of this application, the filament placement head is driven to perform reverse pretension control based on a preset tension mapping table and a temperature compensation coefficient. Specifically, this includes: constructing a four-dimensional mapping table; wherein the four-dimensional mapping table includes tension, speed, temperature, and viscoelasticity; importing the real-time placement speed of the filament placement head and the temperature data from the environmental data into a preset neural network model to predict the temperature compensation coefficient; determining the reverse pretension value based on the four-dimensional mapping table and the temperature compensation coefficient; and driving the filament placement head based on the reverse pretension value.

[0010] In one implementation of this application, the actual fiber orientation data is acquired based on the polarization laser polarimeter according to a preset sampling period, and the fiber angle deviation of the actual fiber orientation data is extracted. Specifically, this includes: calculating the azimuth and elevation angles of the fiber orientation based on a preset polarization interference fringe analysis algorithm; and performing spatiotemporal joint noise reduction processing on the azimuth and elevation angle data based on a preset extended Kalman filter algorithm to determine the fiber angle deviation. The extended Kalman filter algorithm includes state prediction, covariance update, and measurement update.

[0011] In one implementation of this application, the fiber angle deviation and the resin viscoelastic data are processed based on a preset Bayesian optimization algorithm to dynamically correct the layup path and generate update instructions. Specifically, this includes: constructing a three-objective optimization function; wherein the three-objective optimization function includes minimizing fiber angle deviation, optimizing resin viscoelastic data, and maximizing layup efficiency; modeling the three-objective optimization function based on a preset Gaussian process model and introducing Pareto front analysis to generate a Bayesian optimization algorithm; wherein the Pareto front analysis is used to find the optimal solution set in a multi-objective optimization problem; generating the three-dimensional position compensation vector and tension compensation coefficient based on the Bayesian optimization algorithm; wherein the compensation vector includes position correction, attitude adjustment, and velocity compensation; and determining update instructions based on the three-dimensional position compensation vector and tension compensation coefficient.

[0012] In one implementation of this application, adjusting the spatial pose and tension of the filament-laying head according to the update instruction specifically includes: converting the three-dimensional position compensation vector into a joint spatial trajectory based on a preset robot inverse kinematics model, and optimizing the joint motion trajectory based on a preset weighted least squares method; wherein, the weighted least squares method is used to solve the optimal joint motion trajectory under multiple constraints; and adjusting the tension output based on the tension compensation coefficient.

[0013] In one implementation of this application, the method further includes: constructing a process database; wherein the process knowledge base includes historical process parameters, material performance data, and environmental data; the historical process parameters include tension values, laying speeds, and temperature compensation coefficients under different material systems; based on the material system and geometric features of the component to be laid, comparing with the process knowledge base to determine the search results, and sorting the search results based on a preset similarity algorithm; and generating a pre-layout strategy based on the sorting results.

[0014] Secondly, embodiments of this application also provide a dynamic optimization device for automatic fiber placement paths of composite materials. The device includes: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, which, when executed, enable the at least one processor to: acquire a three-dimensional model of the component to be laid, and acquire surface curvature distribution characteristics and fiber placement angle constraints based on the three-dimensional model. The fiber placement angle constraints are dynamically corrected based on environmental and material data of the component to be laid, the environmental data including temperature and humidity data, and the material data including resin viscoelasticity data and fiber wetting state data. The system processes the surface curvature distribution characteristics based on a preset path planning algorithm to generate an initial fiber placement path. Based on a preset tension mapping table and temperature compensation coefficient, it drives the fiber placement head to perform reverse pretension control; wherein the reverse pretension value is positively correlated with the square of the real-time placement speed of the fiber placement head. Based on a preset sampling period, the system acquires actual fiber orientation data using a polarizing laser polarimeter and extracts the fiber angle deviation from the actual fiber orientation data. Based on a preset Bayesian optimization algorithm, it processes the fiber angle deviation and the resin viscoelasticity data to dynamically correct the placement path and generate an update instruction; wherein the update instruction includes a three-dimensional position compensation vector and a tension compensation coefficient. The system adjusts the spatial position and tension of the fiber placement head according to the update instruction.

[0015] Thirdly, embodiments of this application also provide a non-volatile computer storage medium for dynamic optimization of automatic fiber placement paths in composite materials, storing computer-executable instructions. The computer-executable instructions are characterized by: acquiring a three-dimensional model of the component to be laid, and acquiring surface curvature distribution characteristics and fiber placement angle constraints based on the three-dimensional model; wherein the fiber placement angle constraints are dynamically corrected based on environmental and material data of the component to be laid, the environmental data including temperature and humidity data, and the material data including resin viscoelasticity data and fiber wetting state data; processing the surface curvature distribution characteristics based on a preset path planning algorithm to generate... An initial fiber placement path is established; based on a preset tension mapping table and temperature compensation coefficient, the fiber placement head is driven to perform reverse pretension control; wherein, the reverse pretension value is positively correlated with the square of the real-time placement speed of the fiber placement head; based on the polarization laser polarimeter, actual fiber orientation data is acquired according to a preset sampling period, and the fiber angle deviation of the actual fiber orientation data is extracted; based on a preset Bayesian optimization algorithm, the fiber angle deviation and the resin viscoelastic data are processed to dynamically correct the placement path and generate an update instruction; wherein, the update instruction includes a three-dimensional position compensation vector and a tension compensation coefficient; the spatial pose and tension of the fiber placement head are adjusted according to the update instruction.

[0016] This application provides a method, device, and medium for dynamic optimization of automatic fiber placement paths in composite materials. By integrating multi-source heterogeneous data, it achieves dynamic correction of fiber placement angles, significantly improving the placement accuracy and efficiency of complex curved surface components. The method uses a path planning algorithm to generate an initial fiber placement path and combines it with a reverse pre-tension control strategy to effectively compensate for the impact of temperature fluctuations on tension. A polarized laser interferometer is used to collect fiber orientation data in real time, and a Bayesian optimization algorithm is used to dynamically correct the placement path, generating update instructions containing a three-dimensional position compensation vector and tension compensation coefficients, achieving precise pose and tension adjustment of the fiber placement head. Furthermore, by constructing a process database and combining historical process parameters and material performance data, customized pre-placement strategies are provided for components with different material systems and geometric characteristics, further enhancing the system's adaptability and intelligence. In summary, this application solves the problems of poor dynamic environment adaptability and insufficient multi-objective optimization capabilities of traditional fiber placement methods, improving placement quality and molding efficiency. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0018] Figure 1 A flowchart of an automatic fiber placement path dynamic optimization method for composite materials provided in this application embodiment;

[0019] Figure 2 This is a schematic diagram of the internal structure of a composite material automatic fiber placement path dynamic optimization device provided in an embodiment of this application. Detailed Implementation

[0020] 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 in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0021] This application provides a method, device, and medium for dynamic optimization of automatic fiber placement path in composite materials, in order to solve the following technical problem: how to achieve dynamic multi-objective fiber placement path optimization and real-time feedback control.

[0022] The technical solutions proposed in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0023] Figure 1 This document provides a flowchart for dynamic optimization of an automated fiber placement path for composite materials, as illustrated in an embodiment of this application. Figure 1 As shown in the embodiment of this application, a method for dynamic optimization of automatic fiber placement path of composite materials specifically includes the following steps:

[0024] Step 1: Obtain a 3D model of the component to be laid, and obtain the surface curvature distribution characteristics and fiber laying angle constraints based on the 3D model; wherein, the fiber laying angle constraints are dynamically corrected according to the environmental data and material data of the component to be laid, the environmental data includes temperature data and humidity data, and the material data includes resin viscoelasticity data and fiber impregnation state data.

[0025] A precise three-dimensional model of the component to be laid is obtained through digital means, and its surface curvature distribution characteristics are analyzed based on this model. At the same time, the angle constraints for fiber laying are dynamically determined according to environmental conditions and material properties.

[0026] First, obtain a computer-aided design (CAD) model of the component to be laid using 3D scanning technology or directly from the design department. This model should contain complete geometric information of the component and is the basis for subsequent analysis.

[0027] Based on the three-dimensional model, mathematical methods are used to calculate the curvature distribution of the surface, including Gaussian curvature and mean curvature. These curvature characteristics reflect the degree and directionality of the bending of the component surface.

[0028] The allowable range of fiber placement angles is dynamically adjusted based on environmental data (such as temperature and humidity) and material data (such as resin viscoelasticity and fiber impregnation state) of the component to be laid. Changes in environmental and material data will affect the bonding performance between the fiber and the matrix, and thus the overall performance of the component.

[0029] Step 11: Scan the component to be laid using a preset CT scanner to obtain CT scan data.

[0030] An industrial CT scanner is used to perform a full-range scan of the component to be laid, obtaining detailed image data of its internal structure. CT scanning technology can provide non-destructive information about the internal structure, which is especially important for components with complex structures.

[0031] Step 12: Process the CT scan data and the CAD model of the component to be laid based on the multi-source heterogeneous data fusion algorithm to construct a three-dimensional model.

[0032] By fusing CT scan data with existing CAD model data and using algorithms to automatically match and calibrate the differences between the two, a more accurate 3D model containing internal structural information is constructed.

[0033] Step 13: Based on the preset local surface fitting algorithm, extract the Gaussian curvature, average curvature, and principal curvature direction on the three-dimensional model to determine the surface curvature distribution characteristics.

[0034] Select key points or regions on the 3D model, and apply a local surface fitting algorithm to calculate the curvature characteristics of these points or regions, including Gaussian curvature, mean curvature, and principal curvature directions.

[0035] Step 14: Based on the temperature and humidity data in the environmental data, adjust the allowable range of the fiber laying angle to determine the fiber laying angle constraint; wherein, the allowable range is non-linearly related to the extreme value region in the curvature feature map.

[0036] The temperature and humidity data of the environment to be laid are monitored and recorded in real time, and the allowable range of fiber laying angle is adjusted based on this data. Changes in temperature and humidity affect the resin curing speed and fiber wettability, thus affecting the laying quality.

[0037] In a specific example, firstly, CAD models of the composite material components to be laid are obtained from the design department, and the actual components are scanned and verified using 3D scanning technology to ensure the accuracy of the model. An industrial CT scanner is used to perform a 360-degree scan of the component to obtain internal structural data. Then, a multi-source heterogeneous data fusion algorithm is applied to fuse the CT scan data with the CAD model data, constructing a precise 3D model containing internal structural information. On the fused 3D model, key points or regions are selected, and a local surface fitting algorithm is applied to calculate Gaussian curvature, mean curvature, and principal curvature directions, forming a surface curvature distribution feature map. Temperature and humidity data of the laying environment are monitored and recorded in real time, and material data such as resin viscoelasticity and fiber wetting state are analyzed to provide a basis for correcting the fiber laying angle. Based on environmental and material data, the allowable range of fiber laying angles is dynamically corrected. In particular, extreme regions in the curvature feature map are considered, and the laying angle is adjusted to adapt to the complex shape of the component surface.

[0038] Step 2: Process the surface curvature distribution characteristics based on a preset path planning algorithm to generate an initial filament laying path.

[0039] By using a path planning algorithm and combining the previously extracted surface curvature distribution features, an initial path for automatic fiber placement of composite materials is generated.

[0040] An intelligent path planning algorithm is employed, which can take into account changes in surface curvature and automatically calculate the optimal path from the starting point to the ending point. This type of algorithm is typically based on graph search, heuristic search, or genetic algorithm principles, and can find a satisfactory path in complex curved surface environments.

[0041] Using the previously extracted surface curvature distribution features as input, the path planning algorithm adjusts the path calculation method according to these features to ensure that the path can adapt to the changes in the surface and avoid fiber breakage or wrinkles during the laying process.

[0042] After processing by the path planning algorithm, one or more initial fiber placement paths are generated. These paths form the basis for subsequent optimizations and need to meet basic requirements for fiber continuity and layup capability.

[0043] Step 21: Based on the preset random sampling algorithm, explore paths in the surface curvature distribution characteristics to generate candidate paths that meet the fiber continuity and layability constraints.

[0044] A random sampling technique, such as Rapid Random Exploration Tree (RRT) or Probabilistic Path Map (PRM), is employed to explore paths within the curvature distribution characteristics of a surface. This algorithm can quickly find feasible path solutions in high-dimensional space.

[0045] During path exploration, the algorithm considers fiber continuity and layability constraints to ensure that the generated paths meet the requirements of the actual fiber placement process. For example, the paths cannot have sharp turns or breaks; they must remain smooth and continuous. Through exploration using a random sampling algorithm, a set of candidate paths that satisfy fiber continuity and layability constraints is generated. These paths will serve as the basis for subsequent optimizations.

[0046] Step 22: Construct a cost function based on the resin viscoelastic parameters in the material data; wherein the cost function includes path length optimization objective, curvature change rate optimization objective, and wetting quality optimization objective. Obtain the resin's viscoelastic parameters, such as viscosity and relaxation time, from the material data. These parameters reflect the resin's flowability and curing characteristics during the fiber placement process.

[0047] Based on the resin viscoelastic parameters, a cost function is constructed to evaluate the merits of candidate paths. The cost function typically includes multiple optimization objectives, such as path length, rate of curvature change, and wetting quality.

[0048] The goal of path length optimization is to reduce fiber laying length, thereby reducing material consumption and manufacturing costs.

[0049] Curvature change rate optimization objective: to make the curvature change of the path smoother and reduce the stress and deformation of the fibers during the laying process.

[0050] The goal of wetting quality optimization is to improve the wetting effect of fibers in resin, thereby ensuring the performance and quality of composite materials.

[0051] Step 23: Adjust the weight coefficients of multiple optimization objectives in the cost function based on the feedback from the immersion state data.

[0052] Real-time monitoring of the wetting status during the fiber laying process, such as the fiber-resin contact angle and wetting rate, reflects the fiber's wetting effect and curing progress in the resin.

[0053] Based on feedback from the wetting status data, the weight coefficients of multiple optimization objectives in the cost function are dynamically adjusted. For example, if the wetting effect is poor, the weight of the wetting quality optimization objective is increased, making the algorithm focus more on the fiber wetting effect. By adjusting the weight coefficients, a balance is achieved among multiple optimization objectives. This helps to improve the quality and efficiency of fiber placement while satisfying fiber continuity and layup constraints.

[0054] Step 24: Determine the initial fiber placement path based on the magnitude of the weighting coefficient.

[0055] The adjusted weighting coefficients are applied to the cost function to evaluate and rank the candidate paths. The magnitude of the weighting coefficients determines the importance of each optimization objective in path selection.

[0056] Based on the evaluation results, the optimal candidate path was selected as the initial fiber placement path. This path, while satisfying the constraints of fiber continuity and layability, also considers multiple factors such as resin viscoelasticity, rate of curvature change, and wetting quality, and forms the basis for subsequent fiber placement processes.

[0057] In a specific example, firstly, the curvature distribution features of the surface of the component to be laid are extracted using 3D scanning technology and surface fitting algorithms, including Gaussian curvature and mean curvature. Based on a pre-defined path planning algorithm and a random sampling algorithm, path exploration is performed on the surface curvature distribution features to generate a set of candidate paths that satisfy fiber continuity and layability constraints. Based on the resin viscoelastic parameters in the material data, a cost function is constructed that includes multiple optimization objectives such as path length, rate of curvature change, and wetting quality. Wetting state data during the fiber laying process is monitored in real time, and the weight coefficients of multiple optimization objectives in the cost function are dynamically adjusted based on feedback to achieve a balance between the optimization objectives. The adjusted weight coefficients are applied to the cost function to evaluate and rank the candidate paths, and the optimal candidate path is selected as the initial fiber laying path. Based on the initial fiber laying path, further path optimization is performed, considering more constraints and optimization objectives. Finally, the optimized fiber laying path is input into the automated fiber laying equipment to execute the fiber laying operation.

[0058] Step 3: Based on the preset tension mapping table and temperature compensation coefficient, drive the filament placement head to perform reverse pretension control; wherein, the reverse pretension value is positively correlated with the square of the real-time placement speed of the filament placement head.

[0059] Precise control of the tension at the fiber placement head ensures the stability and quality of the composite material during the layup process. Reverse pretension control is an advanced tension control strategy that considers the real-time layup speed of the fiber placement head and the influence of ambient temperature on material properties.

[0060] The tension mapping table is a pre-defined table that contains pretension values ​​corresponding to different layup speeds, temperatures, and material viscoelasticity conditions. This table is based on extensive experimental data and experience, providing a foundation for reverse pretension control.

[0061] Since changes in ambient temperature affect the viscoelasticity and layup properties of materials, a temperature compensation coefficient is needed to adjust the pretension value. This coefficient is calculated based on real-time temperature data and material performance parameters.

[0062] The reverse pretension value is positively correlated with the square of the laying speed: this is a rule derived from experimental observation, namely, the faster the laying speed, the greater the pretension required to maintain the stability of the material. This positive correlation ensures that appropriate pretension can be applied at different laying speeds.

[0063] Step 31: Construct a four-dimensional mapping table; wherein the four-dimensional mapping table includes tension, velocity, temperature and viscoelasticity.

[0064] The four-dimensional mapping table is a table containing four dimensions: tension, velocity, temperature, and viscoelasticity. Each dimension corresponds to a parameter, and these parameters together determine the magnitude of the pretension.

[0065] First, the relationships between tension, velocity, temperature, and viscoelasticity under different conditions were determined through experiments and data analysis. Then, these relationships were organized into a tabular form, forming a four-dimensional mapping table.

[0066] Step 32: Import the real-time placement speed of the filament placement head and the temperature data in the environmental data into a preset neural network model to predict the temperature compensation coefficient.

[0067] The neural network model is a pre-trained model that can predict the corresponding temperature compensation coefficient based on the input laying speed and temperature data.

[0068] The real-time placement speed of the filament placement head and the temperature data from the environmental data are imported into the neural network model. The model will predict the temperature compensation coefficient based on this data, providing a basis for subsequent pretension control.

[0069] Step 33: Determine the reverse pretension value based on the four-dimensional mapping table and the temperature invariance coefficient.

[0070] Based on the tension and velocity relationship in the four-dimensional mapping table and the predicted temperature compensation coefficient, the reverse pretension value under the current conditions is determined.

[0071] Step 34: Drive the filament placement head based on the reverse pretension value.

[0072] The predetermined reverse pretension value is input into the control system of the filament placement head, which drives the filament placement head to place the filaments according to the preset tension.

[0073] In a specific case study, taking the dynamic optimization of automated fiber placement paths in composite materials as an example, the specific implementation steps are as follows:

[0074] Construct a four-dimensional mapping table, including four dimensions: tension, velocity, temperature, and viscoelasticity.

[0075] Train a neural network model to predict the temperature compensation coefficient.

[0076] The real-time placement speed of the fiber placement head is collected.

[0077] Collect temperature data from environmental data.

[0078] Real-time deployment speed and temperature data are imported into a neural network model to predict the temperature compensation coefficient.

[0079] Based on the four-dimensional mapping table and the predicted temperature compensation coefficient, the reverse pretension value under the current conditions is determined.

[0080] The reverse pretension value is input into the control system of the filament placement head, which drives the filament placement head to place the filaments according to the preset tension.

[0081] During the laying process, the tension and speed changes of the filament laying head, as well as the changes in ambient temperature, are monitored in real time.

[0082] Based on the changes, adjust the parameters of the four-dimensional mapping table and neural network model in a timely manner to optimize the reverse pretension control effect.

[0083] Step 4: Based on the polarization laser polarimeter, acquire the actual fiber orientation data according to the preset sampling period, and extract the fiber angle deviation of the actual fiber orientation data.

[0084] By using a polarizing laser polarimeter to obtain real-time and accurate fiber orientation data during the composite material fiber laying process, and further extracting the fiber angular deviation, key information is provided for subsequent path dynamic optimization.

[0085] A polarized laser polarimeter is a precision instrument that uses the properties of polarized light to measure the orientation of fibers on the surface of an object. It emits polarized laser light and receives the reflected light signals, inferring the fiber orientation by analyzing changes in these signals.

[0086] To ensure that the acquired fiber orientation data has sufficient time-domain resolution, a reasonable sampling period needs to be set. This period should be determined based on the fiber placement speed, fiber material characteristics, and the required control precision.

[0087] The data collected by the polarization laser polarimeter according to the preset sampling period includes information on the fiber's orientation at various times. This data forms the basis for subsequent analysis of fiber angular deviations.

[0088] By processing actual fiber orientation data, the angular deviation of the fibers relative to the ideal orientation is extracted. This deviation is an important indicator for evaluating the quality of fiber placement and also serves as the basis for dynamic path optimization.

[0089] Step 41: Calculate the azimuth and elevation angles of the fiber orientation based on the preset polarized light interference fringe analysis algorithm.

[0090] The polarized light interference fringe analysis algorithm is used to analyze the optical signal collected by a polarizing laser polarimeter to obtain the azimuth and elevation angles of a fiber. This algorithm is based on the interference principle of polarized light and infers the fiber's orientation by analyzing the shape and positional characteristics of the interference fringes.

[0091] The angle between the projection of a fiber onto a horizontal plane and a reference direction. It is an important parameter describing the horizontal orientation of the fiber. The angle between the fiber and the horizontal plane. It is an important parameter describing the degree of inclination of the fiber in the vertical direction.

[0092] The optical signal collected by the polarization laser polarimeter is input into the polarization interference fringe analysis algorithm. The algorithm calculates the azimuth and elevation angles of the fiber according to the preset model and solution method.

[0093] Step 42: Perform spatiotemporal joint noise reduction processing on the azimuth and elevation angle data based on a preset extended Kalman filter algorithm to determine the fiber angle deviation; wherein, the extended Kalman filter algorithm includes state prediction, covariance update and measurement update.

[0094] The Extended Kalman Filter (EKF) algorithm is used for state estimation of nonlinear systems. It achieves noise suppression and signal enhancement by modeling the system in state space and using observational data to estimate and update the system state.

[0095] In the extended Kalman filter algorithm, the state at the next time step is first predicted based on the system's dynamic model. This prediction is calculated based on the current state and the system input.

[0096] To reflect the uncertainty of the predicted state, the covariance matrix of the state needs to be updated. This matrix describes the error range of the state estimate and is an important basis for subsequent measurement updates.

[0097] When new observation data arrives, the extended Kalman filter algorithm uses this data to revise the predicted state, obtaining a more accurate state estimate. Simultaneously, it updates the state's covariance matrix to reflect the revised uncertainty.

[0098] When denoising azimuth and elevation data, the extended Kalman filter algorithm considers information from both the time and spatial domains. By comprehensively utilizing historical and current observation data, the algorithm can more effectively suppress noise and improve the accuracy and reliability of the data.

[0099] After noise reduction using the extended Kalman filter algorithm, more accurate azimuth and elevation angle data are obtained. By comparing these data with the ideal orientation, the angular deviation of the fiber is determined.

[0100] In a specific case summary

[0101] Configure a polarizing laser polarimeter and set a reasonable sampling period. Prepare a polarized light interference fringe analysis algorithm and an extended Kalman filter algorithm. Start the polarizing laser polarimeter and collect actual fiber orientation data according to the preset sampling period. Input the collected optical signal into the polarized light interference fringe analysis algorithm to calculate the fiber azimuth and elevation angles. Input the calculated azimuth and elevation angle data into the extended Kalman filter algorithm for spatiotemporal joint noise reduction processing. Determine the fiber angle deviation based on the noise-reduced azimuth and elevation angle data. Adjust the movement trajectory and speed of the fiber placement head according to the fiber angle deviation to optimize the fiber placement path. Monitor the fiber orientation data in real time and continuously adjust the optimization strategy to ensure fiber placement quality.

[0102] Step 5: Process the fiber angle deviation and the resin viscoelasticity data based on a preset Bayesian optimization algorithm to dynamically correct the layup path and generate update instructions; wherein, the update instructions include a three-dimensional position compensation vector and a tension compensation coefficient.

[0103] By using a Bayesian optimization algorithm, the fiber angle deviation, resin viscoelasticity data, and layup efficiency are comprehensively considered to dynamically correct the layup path of the composite material and generate corresponding update instructions to ensure the accuracy and efficiency of the fiber laying process.

[0104] Bayesian optimization is a global optimization algorithm, particularly suitable for expensive and uncertain function optimization problems. In the process of composite fiber placement, due to the complexity of fiber angle deviations and resin viscoelasticity, Bayesian optimization can effectively find the optimal placement path.

[0105] As mentioned earlier, this data is obtained in real time through a polarizing laser polarimeter, reflecting the deviation between the actual and ideal fiber orientation.

[0106] Resin viscoelasticity data refers to the viscosity and elastic behavior of resins under specific temperatures and pressures, which has a significant impact on the molding quality and performance of composite materials. Update instructions are generated based on the results of a Bayesian optimization algorithm and are used to adjust the motion trajectory, tension, and speed parameters of the fiber placement head to optimize the layup path.

[0107] The three-dimensional position compensation vector is a vector that includes position correction, attitude adjustment, and velocity compensation, used to precisely control the movement of the wire placement head.

[0108] Tension compensation coefficient is a coefficient used to adjust fiber tension during the fiber laying process to ensure fiber flatness and tightness.

[0109] Step 51: Construct a three-objective optimization function; wherein the three-objective optimization function includes minimizing fiber angle deviation, optimizing resin viscoelastic data, and maximizing layup efficiency.

[0110] The three-objective optimization function is a function that comprehensively considers minimizing fiber angle deviation, optimizing resin viscoelastic data, and maximizing layup efficiency. This function is the foundation of the Bayesian optimization algorithm, guiding the algorithm to find the optimal solution in the search space.

[0111] Minimize fiber angle deviation: The goal is to make the actual fiber orientation as close as possible to the ideal orientation in order to reduce inter-layer errors and improve the performance of composite materials.

[0112] Resin viscoelasticity data optimization: The goal is to maintain optimal viscosity and elastic behavior of the resin during layup to ensure the molding quality and performance of the composite material.

[0113] Maximizing layup efficiency: The goal is to improve the efficiency of the fiber laying process and reduce production time and costs.

[0114] Step 52: Model the three-objective optimization function based on the preset Gaussian process model, and introduce Pareto front analysis to generate a Bayesian optimization algorithm; wherein, the Pareto front analysis is used to find the optimal solution set in a multi-objective optimization problem.

[0115] Gaussian process models are non-parametric probabilistic models that can be used to model and predict complex functions. In Bayesian optimization, Gaussian process models are used to model three-objective optimization functions to predict function values ​​under different parameters.

[0116] In multi-objective optimization problems, due to potential conflicts between objectives, it is impossible to find a solution that simultaneously satisfies the optimality of all objectives. Pareto front analysis is a method used to find the optimal set of solutions in multi-objective optimization problems. It finds a set of optimal solutions in a non-dominated sense by analyzing the trade-offs between different objectives.

[0117] Based on the Gaussian process model and Pareto front analysis, a Bayesian optimization algorithm is generated. This algorithm can efficiently find the optimal solution satisfying the three-objective optimization function in the search space.

[0118] Step 53: Generate the three-dimensional position compensation vector and tension compensation coefficient based on the Bayesian optimization algorithm; wherein the compensation vector includes position correction amount, attitude adjustment amount and velocity compensation amount.

[0119] The Bayesian optimization algorithm generates a three-dimensional position compensation vector based on the current position, attitude, and velocity information of the fiber placement head, as well as fiber angle deviation and resin viscoelasticity data. This vector includes position correction, attitude adjustment, and velocity compensation, used to precisely control the movement of the fiber placement head.

[0120] Simultaneously, the Bayesian optimization algorithm generates a tension compensation coefficient based on fiber tension and layup requirements. This coefficient is used to adjust fiber tension during the layup process to ensure fiber flatness and tightness.

[0121] Step 54: Determine the update command based on the three-dimensional position compensation vector and tension compensation coefficient.

[0122] Based on the three-dimensional position compensation vector and tension compensation coefficient generated by the Bayesian optimization algorithm, the update command is determined. This command contains the specific parameters and values ​​that the filament placement head needs to be adjusted to dynamically correct the placement path.

[0123] Update commands are applied to the fiber placement control system to adjust the movement trajectory, tension, and speed parameters of the fiber placement head, thereby optimizing the placement path. By continuously applying update commands, the optimal placement path is gradually approached, improving the molding quality and performance of the composite material.

[0124] In a specific example:

[0125] A polarization laser polarimeter and resin viscoelasticity measurement equipment were configured to ensure real-time acquisition of fiber angle deviation and resin viscoelasticity data. A three-objective optimization function was constructed, including minimizing fiber angle deviation, optimizing resin viscoelasticity data, and maximizing layup efficiency. The three-objective optimization function was modeled based on a Gaussian process model, and Pareto front analysis was introduced to generate a Bayesian optimization algorithm. Fiber angle deviation and resin viscoelasticity data were acquired in real time. The acquired data was input into the Bayesian optimization algorithm for processing and analysis. Based on the Bayesian optimization algorithm, a three-dimensional position compensation vector and tension compensation coefficient were generated. Update instructions were determined based on the generated three-dimensional position compensation vector and tension compensation coefficient. The update instructions were applied to the fiber layup control system to adjust the movement trajectory, tension, and speed parameters of the fiber layup head. The fiber layup process was monitored in real time, and the update instructions were continuously adjusted and optimized based on feedback data.

[0126] Step 6: Adjust the spatial position and tension of the filament placement head according to the update command.

[0127] Based on the update instructions generated by the Bayesian optimization algorithm, the spatial position (i.e., position and orientation) and tension of the fiber placement head are precisely adjusted to ensure the accuracy and stability of the composite material fiber placement process.

[0128] The update command is generated by processing fiber angle deviation and resin viscoelasticity data using a Bayesian optimization algorithm. It includes a three-dimensional position compensation vector for adjusting the spatial pose of the filament placement head and a tension compensation coefficient for adjusting the tension.

[0129] Spatial position refers to the position and orientation of the fiber placement head in three-dimensional space. It is a key parameter in the fiber placement process of composite materials and directly affects the quality and performance of the layup.

[0130] Tension is the pulling force on the fiber during the fiber laying process, which has an important impact on the smoothness and tightness of the fiber.

[0131] Step 61: Based on the preset robot inverse kinematics model, the three-dimensional position compensation vector is converted into a joint space trajectory, and the joint motion trajectory is optimized based on the preset weighted least squares method; wherein, the weighted least squares method is used to solve the optimal joint motion trajectory under multiple constraints.

[0132] Inverse kinematics (IK) is a mathematical model used to convert the motion of an end effector (such as a filament-laying head) in Cartesian space (i.e., three-dimensional space) into motion in the robot's joint space. Through IK, the three-dimensional position compensation vector is converted into the angles or displacements required for each joint of the robot to rotate.

[0133] In practice, the target angle or displacement that each joint of the robot needs to achieve is first calculated using the inverse kinematics model based on the current position and target position of the filament-laying head (determined by the three-dimensional position compensation vector).

[0134] Joint space trajectory refers to the motion trajectory of each joint of a robot over time. By optimizing the joint space trajectory, we can ensure smooth and accurate robot movement and avoid vibration or impact.

[0135] Weighted least squares is a mathematical optimization method used to find the optimal solution under multiple constraints. In robot motion trajectory optimization, weighted least squares considers multiple constraints such as joint velocity, acceleration, and torque to find the optimal joint motion trajectory.

[0136] In practice, the objective function of weighted least squares is first constructed based on the robot's dynamics model and kinematic constraints. Then, numerical optimization algorithms (such as gradient descent and Newton's method) are used to solve for the optimal solution of the objective function, i.e., the optimal joint motion trajectory.

[0137] Step 62: Adjust the tension output based on the tension compensation coefficient.

[0138] The tension compensation coefficient, as mentioned above, is generated using a Bayesian optimization algorithm and is used to adjust the fiber tension during the fiber laying process.

[0139] Adjusting the tension output: Based on the tension compensation coefficient, adjust the tension control system in the fiber laying machine to change the tension on the fiber.

[0140] In practice, the target tension value is first calculated based on the tension compensation coefficient. Then, the tension of the fibers is monitored and adjusted in real time through a tension control system (such as a tension sensor or tension controller) to keep it near the target tension value. This ensures that the fibers remain flat and tightly bonded during the laying process, improving the molding quality and performance of the composite material.

[0141] In a specific example, taking the dynamic optimization of the automated fiber placement path in composite materials as an example, the specific implementation steps are as follows: Configure a polarized laser polarimeter and a resin viscoelasticity measurement device to ensure real-time acquisition of fiber angle deviation and resin viscoelasticity data. Construct a three-objective optimization function, including minimizing fiber angle deviation, optimizing resin viscoelasticity data, and maximizing placement efficiency. Model the three-objective optimization function based on a Gaussian process model and introduce Pareto front analysis to generate a Bayesian optimization algorithm. Establish a robot inverse kinematics model and a weighted least squares optimization method. Collect fiber angle deviation and resin viscoelasticity data in real time. Input the collected data into the Bayesian optimization algorithm for processing and analysis, generating update instructions (including three-dimensional position compensation vectors and tension compensation coefficients). Based on the robot inverse kinematics model, convert the three-dimensional position compensation vectors into joint space trajectories. Optimize the joint motion trajectory using the weighted least squares method to ensure smooth and accurate robot movement. Adjust the spatial position of the fiber placement head according to the optimized joint motion trajectory. Calculate the target tension value based on the tension compensation coefficient. Monitor and adjust the fiber tension in real time through a tension control system to keep it near the target tension value. During the fiber placement process, fiber angle deviation, resin viscoelasticity data, and the spatial position and tension of the fiber placement head are monitored in real time. Based on the monitoring data, instructions are continuously adjusted, optimized, and updated to ensure the accuracy and stability of the fiber placement process.

[0142] This application also includes the following methods:

[0143] A1. Construct a process database; wherein, the process knowledge base includes historical process parameters, material performance data, and environmental data; the historical process parameters include tension values, layup speeds, and temperature compensation coefficients under different material systems.

[0144] The process database is an integrated information system used to store and manage various types of data related to composite material fiber placement processes. It includes historical process parameters, material performance data, and environmental data, providing comprehensive data support for process optimization.

[0145] Historical process parameters refer to the records of process parameters used in past wire laying processes for different material systems and component geometries. These parameters include, but are not limited to, tension values, laying speed, and temperature compensation coefficients, and serve as important references for process optimization.

[0146] Tension value refers to the amount of tension exerted on the fiber during the fiber laying process. Different material systems have different tension requirements, and proper tension settings can ensure the flatness and tightness of the fibers.

[0147] Laying speed refers to the speed at which the filament placement head moves across the surface of the component. The selection of laying speed requires comprehensive consideration of material properties, equipment capabilities, and production efficiency.

[0148] Temperature compensation coefficient is used to adjust the temperature settings during the fiber placement process because changes in ambient temperature can affect material properties and fiber placement.

[0149] Material performance data includes information on the mechanical properties, thermal properties, and chemical stability of composite materials. This data is crucial for selecting process parameters and developing fiber placement strategies.

[0150] This refers to the environmental conditions during the fiber placement process, such as temperature, humidity, and air pressure. Environmental data has a certain impact on the fiber placement effect, so it needs to be considered in the process database.

[0151] In practice, by collecting and analyzing past wire laying cases, relevant process parameters, material performance data and environmental data are extracted, and then organized and stored according to a certain data structure and format to form a process database.

[0152] A2. Based on the material system and geometric features of the component to be laid, compare it with the process knowledge base to determine the search results, and sort the search results based on a preset similarity algorithm.

[0153] The material system and geometric characteristics of the component to be laid refer to the type of material, number of layers, layup sequence, shape, and size characteristics of the component. These characteristics are the key basis for selecting process parameters and formulating fiber placement strategies.

[0154] The material system and geometric characteristics of the components to be laid are compared with historical cases in the process database to find similar cases for reference.

[0155] This was achieved by comparing a set of similar historical cases obtained from a process knowledge base. These cases included material systems, geometric features, and corresponding process parameters similar to the components to be laid.

[0156] This is an algorithm used to assess the similarity between two objects. In this step, the similarity algorithm is used to rank the search results, placing the most similar cases first, in order to facilitate the subsequent selection and optimization of process parameters.

[0157] In practice, the process begins by retrieving similar historical cases from the process database based on the material system and geometric characteristics of the components to be laid. Then, a pre-defined similarity algorithm is used to evaluate and rank the search results. This similarity algorithm considers the similarity in multiple aspects, including material system, geometric characteristics, and process parameters, and ranks the search results by calculating a similarity score.

[0158] A3. Generate a pre-placement strategy based on the sorting results.

[0159] The ranking results refer to the set of cases obtained after sorting the search results using a similarity algorithm, which includes historical cases that are most similar to the component to be laid.

[0160] Pre-layout strategy refers to the process parameters and fiber placement strategy selected based on the sorting results, used to guide the actual fiber placement process. The pre-layout strategy includes specific parameters such as tension settings, layup speed, and temperature compensation coefficient, as well as fiber placement path and layup sequence strategies.

[0161] In practice, the most similar cases in the ranking results are first analyzed to extract their process parameters and fiber placement strategies. Then, considering the specific conditions and actual needs of the component to be placed, the extracted process parameters and fiber placement strategies are adjusted and optimized to generate a pre-placement strategy suitable for the component. Finally, the pre-placement strategy is input into the fiber placement equipment to guide the actual fiber placement process.

[0162] The above are embodiments of the method proposed in this application. Based on the same inventive concept, embodiments of this application also provide a dynamic optimization device for automatic fiber placement path of composite materials, the structure of which is as follows: Figure 2 As shown.

[0163] Figure 2 This is a schematic diagram of the internal structure of an automated fiber placement path dynamic optimization device for composite materials, provided as an embodiment of this application. Figure 2 As shown, the device includes:

[0164] At least one processor 201;

[0165] And a memory 202 that is communicatively connected to at least one processor;

[0166] The memory 202 stores instructions executable by at least one processor, which are executed by at least one processor 201 to enable at least one processor 201 to:

[0167] A three-dimensional model of the component to be laid is obtained, and the surface curvature distribution characteristics and fiber laying angle constraints are obtained based on the three-dimensional model. The fiber laying angle constraints are dynamically corrected based on the environmental and material data of the component to be laid. The environmental data includes temperature and humidity data, and the material data includes resin viscoelasticity data and fiber wetting state data. The surface curvature distribution characteristics are processed based on a preset path planning algorithm to generate an initial fiber laying path. Based on a preset tension mapping table and temperature compensation coefficient, the fiber laying head is driven to perform reverse pretension control. The reverse pretension value is positively correlated with the square of the real-time laying speed of the fiber laying head. Actual fiber orientation data is obtained using a polarization laser polarimeter according to a preset sampling period, and the fiber angle deviation of the actual fiber orientation data is extracted. The fiber angle deviation and resin viscoelasticity data are processed based on a preset Bayesian optimization algorithm to dynamically correct the laying path and generate an update instruction. The update instruction includes a three-dimensional position compensation vector and a tension compensation coefficient. The spatial position and tension of the fiber laying head are adjusted according to the update instruction.

[0168] Some embodiments of this application provide corresponding to Figure 1 A non-volatile computer storage medium for dynamic optimization of automatic fiber placement path in composite materials, storing computer-executable instructions, wherein the computer-executable instructions are configured as follows:

[0169] A three-dimensional model of the component to be laid is obtained, and the surface curvature distribution characteristics and fiber laying angle constraints are obtained based on the three-dimensional model. The fiber laying angle constraints are dynamically corrected based on the environmental and material data of the component to be laid. The environmental data includes temperature and humidity data, and the material data includes resin viscoelasticity data and fiber wetting state data. The surface curvature distribution characteristics are processed based on a preset path planning algorithm to generate an initial fiber laying path. Based on a preset tension mapping table and temperature compensation coefficient, the fiber laying head is driven to perform reverse pretension control. The reverse pretension value is positively correlated with the square of the real-time laying speed of the fiber laying head. Actual fiber orientation data is obtained using a polarization laser polarimeter according to a preset sampling period, and the fiber angle deviation of the actual fiber orientation data is extracted. The fiber angle deviation and resin viscoelasticity data are processed based on a preset Bayesian optimization algorithm to dynamically correct the laying path and generate an update instruction. The update instruction includes a three-dimensional position compensation vector and a tension compensation coefficient. The spatial position and tension of the fiber laying head are adjusted according to the update instruction.

[0170] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments for IoT devices and media are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0171] The systems, media, and methods provided in this application are one-to-one correspondences. Therefore, the systems and media also have similar beneficial technical effects as their corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the systems and media will not be repeated here.

[0172] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage) containing computer-usable program code.

[0173] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which are executable by the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0174] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0175] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0176] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0177] Memory may include non-persistent storage in computer-readable media, random access memory (RAM), and / or non-volatile memory forms such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0178] Computer-readable media include both permanent and non-permanent, removable and non-removable media that store information using any method or technology. Information is computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessed by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0179] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0180] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations will be apparent to those skilled in the art. Any modifications, substitutions, or improvements made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for dynamic optimization of automated fiber placement path in composite materials, applied to a fiber placement robot, the fiber placement robot comprising a fiber placement head and a polarizing laser polarimeter, characterized in that, The method includes: A three-dimensional model of the component to be laid is obtained, and the surface curvature distribution characteristics and fiber laying angle constraints are obtained based on the three-dimensional model; wherein, the fiber laying angle constraints are dynamically corrected according to the environmental data and material data of the component to be laid, the environmental data includes temperature data and humidity data, and the material data includes resin viscoelasticity data and fiber impregnation state data; The surface curvature distribution characteristics are processed based on a preset path planning algorithm to generate an initial filament laying path; Based on a preset tension mapping table and temperature compensation coefficient, the filament placement head is driven to perform reverse pretension control; wherein, the reverse pretension value is positively correlated with the square of the real-time placement speed of the filament placement head; The polarization laser polarizer acquires actual fiber orientation data according to a preset sampling period, and extracts the fiber angle deviation of the actual fiber orientation data. The fiber angle deviation and resin viscoelasticity data are processed based on a preset Bayesian optimization algorithm to dynamically correct the layup path and generate update instructions; wherein, the update instructions include a three-dimensional position compensation vector and a tension compensation coefficient. Adjust the spatial position and tension of the filament placement head according to the update instructions.

2. The method for dynamic optimization of automatic fiber placement path of composite materials according to claim 1, characterized in that, Obtain a 3D model of the component to be laid, and based on the 3D model, obtain the surface curvature distribution characteristics and fiber layup angle constraints, specifically including: The component to be laid is scanned using a pre-set CT scanner to obtain CT scan data; The CT scan data and the CAD model of the component to be laid are processed based on a multi-source heterogeneous data fusion algorithm to construct a three-dimensional model; Based on a preset local surface fitting algorithm, Gaussian curvature, mean curvature, and principal curvature directions are extracted from the three-dimensional model to determine the surface curvature distribution characteristics. The allowable range of fiber laying angle is corrected based on the temperature and humidity data in the environmental data to determine the fiber laying angle constraint; wherein, the allowable range is non-linearly related to the extreme value region in the curvature feature map.

3. The method for dynamic optimization of automatic fiber placement path of composite materials according to claim 1, characterized in that, The surface curvature distribution features are processed based on a preset path planning algorithm to generate an initial filament laying path, specifically including: Based on a pre-defined random sampling algorithm, path exploration is performed on the surface curvature distribution characteristics to generate candidate paths that meet the constraints of fiber continuity and layability. A cost function is constructed based on the resin viscoelastic parameters in the material data; wherein, the cost function includes a path length optimization objective, a curvature change rate optimization objective, and a wetting quality optimization objective; The weight coefficients of multiple optimization objectives in the cost function are adjusted based on the feedback from the immersion state data. The initial fiber placement path is determined based on the magnitude of the weighting coefficient.

4. The method for dynamic optimization of automatic fiber placement path of composite materials according to claim 1, characterized in that, Based on a preset tension mapping table and temperature compensation coefficient, the filament placement head is driven to perform reverse pretension control, specifically including: Construct a four-dimensional mapping table; wherein the four-dimensional mapping table includes tension, velocity, temperature, and viscoelasticity; The real-time placement speed of the filament placement head and the temperature data in the environmental data are imported into a preset neural network model to predict the temperature compensation coefficient. The reverse pretension value is determined based on the four-dimensional mapping table and the temperature compensation coefficient. The filament placement head is driven based on the reverse pretension value.

5. The method for dynamic optimization of automatic fiber placement path of composite materials according to claim 1, characterized in that, Based on the polarization laser polarimeter acquiring actual fiber orientation data according to a preset sampling period, and extracting the fiber angle deviation of the actual fiber orientation data, specifically including: The azimuth and elevation angles of the fiber orientation are calculated based on a preset polarized light interference fringe analysis algorithm. The azimuth and elevation angle data are subjected to spatiotemporal joint noise reduction based on a preset extended Kalman filter algorithm to determine the fiber angle deviation; wherein, the extended Kalman filter algorithm includes state prediction, covariance update and measurement update.

6. The method for dynamic optimization of automatic fiber placement path of composite materials according to claim 1, characterized in that, The fiber angle deviation and resin viscoelasticity data are processed based on a preset Bayesian optimization algorithm to dynamically correct the layup path and generate update instructions, specifically including: A three-objective optimization function is constructed; wherein, the three-objective optimization function includes minimizing fiber angle deviation, optimizing resin viscoelastic data, and maximizing layup efficiency; The three-objective optimization function is modeled based on a pre-defined Gaussian process model, and Pareto front analysis is introduced to generate a Bayesian optimization algorithm; wherein, the Pareto front analysis is used to find the optimal solution set in a multi-objective optimization problem; The three-dimensional position compensation vector and tension compensation coefficient are generated based on the Bayesian optimization algorithm; wherein, the compensation vector includes position correction, attitude adjustment and velocity compensation. The update command is determined based on the three-dimensional position compensation vector and tension compensation coefficient.

7. The method for dynamic optimization of automatic fiber placement path of composite materials according to claim 1, characterized in that, Adjusting the spatial position and tension of the filament placement head according to the update command specifically includes: The three-dimensional position compensation vector is converted into a joint motion trajectory based on a preset robot inverse kinematics model, and the joint motion trajectory is optimized based on a preset weighted least squares method; wherein, the weighted least squares method is used to solve the optimal joint motion trajectory under multiple constraints. The tension output is adjusted based on the aforementioned tension compensation coefficient.

8. The method for dynamic optimization of automatic fiber placement path of composite materials according to claim 1, characterized in that, The method further includes: Construct a process database; wherein the process database includes historical process parameters, material performance data, and environmental data; the historical process parameters include tension values, layup speeds, and temperature compensation coefficients under different material systems; Based on the material system and geometric features of the component to be laid, the process database is compared to determine the search results, and the search results are sorted based on a preset similarity algorithm; Generate a pre-placement strategy based on the sorting results.

9. A dynamic optimization device for automatic fiber placement path of composite materials, characterized in that, The device includes: At least one processor; And, a memory communicatively connected to the at least one processor; The memory stores instructions executable by the at least one processor, which, when executed by the at least one processor, enable the at least one processor to: A three-dimensional model of the component to be laid is obtained, and the surface curvature distribution characteristics and fiber laying angle constraints are obtained based on the three-dimensional model; wherein, the fiber laying angle constraints are dynamically corrected according to the environmental data and material data of the component to be laid, the environmental data includes temperature data and humidity data, and the material data includes resin viscoelasticity data and fiber impregnation state data; The surface curvature distribution characteristics are processed based on a preset path planning algorithm to generate an initial filament laying path; Based on a preset tension mapping table and temperature compensation coefficient, the filament placement head is driven to perform reverse pretension control; wherein, the reverse pretension value is positively correlated with the square of the real-time placement speed of the filament placement head; The actual fiber orientation data is obtained by a polarizing laser polarizer according to a preset sampling period, and the fiber angle deviation of the actual fiber orientation data is extracted. The fiber angle deviation and resin viscoelasticity data are processed based on a preset Bayesian optimization algorithm to dynamically correct the layup path and generate update instructions; wherein, the update instructions include a three-dimensional position compensation vector and a tension compensation coefficient. Adjust the spatial position and tension of the filament placement head according to the update instructions.

10. A non-volatile computer storage medium for dynamic optimization of automatic fiber placement path in composite materials, storing computer-executable instructions, characterized in that, The computer-executable instructions are set as follows: A three-dimensional model of the component to be laid is obtained, and the surface curvature distribution characteristics and fiber laying angle constraints are obtained based on the three-dimensional model; wherein, the fiber laying angle constraints are dynamically corrected according to the environmental data and material data of the component to be laid, the environmental data includes temperature data and humidity data, and the material data includes resin viscoelasticity data and fiber impregnation state data; The surface curvature distribution characteristics are processed based on a preset path planning algorithm to generate an initial filament laying path; Based on a preset tension mapping table and temperature compensation coefficient, the filament placement head is driven to perform reverse pretension control; wherein, the reverse pretension value is positively correlated with the square of the real-time placement speed of the filament placement head; The actual fiber orientation data is obtained by a polarizing laser polarizer according to a preset sampling period, and the fiber angle deviation of the actual fiber orientation data is extracted. The fiber angle deviation and resin viscoelasticity data are processed based on a preset Bayesian optimization algorithm to dynamically correct the layup path and generate update instructions; wherein, the update instructions include a three-dimensional position compensation vector and a tension compensation coefficient. Adjust the spatial position and tension of the filament placement head according to the update instructions.

Images