A method and system for analyzing abnormal stress of automobile structure based on deep learning
By constructing a three-dimensional automotive structural model and multi-layer perceptron analysis, the problems of inaccurate stress abnormality analysis of automobile structures and unclear threat analysis of impact deformation in traditional methods are solved, and more accurate stress abnormality identification and structural optimization are achieved. Improves the safety and durability of the vehicle.
Patent Information
- Application Number
- CN202410749224.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-12
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2044-06-12
AI Technical Summary
The traditional deep learning-based analysis method for automobile structural stress anomalies has problems such as inaccurate analysis of stress anomalies and unclear analysis of the threat of structural deformation to personnel during a car collision.
By obtaining automotive structural design data, a three-dimensional automotive structural model is constructed, the connection point stress bearing analysis, structural stress distribution simulation, and collision deformation simulation are carried out, the abnormal deformation angle of the body structure is identified, the deformation angle threat is calculated, and the multi-layer perceptron is based on the multi-layer perceptron Build an abnormal structure analysis model, conduct vehicle structural stress abnormality analysis and optimization adjustment.
提高了汽车结构受力异常分析的精确度,明确了冲撞过程中结构形变对人员的威胁程度,确保了车辆在使用过程中的安全性和结构的耐久性。
Smart Images

Figure CN118313070B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of automobile structure analysis, and in particular to a method and system for analyzing abnormal stress of automobile structure based on deep learning. Background Art
[0002] Abnormal stress in automobile structures is caused by design defects, material problems, poor manufacturing or fatigue factors caused by long-term use. These abnormalities can cause automobile failures or even safety hazards during use. Deep learning is a machine learning method based on artificial neural networks, which has powerful pattern recognition and feature extraction capabilities. The abnormal stress analysis method of automobile structures based on deep learning uses a large amount of structural stress data to learn and understand the normal and abnormal stress modes of automobile structures by training deep neural network models. First, a large amount of automobile structure stress data is collected and annotated, including stress information under normal and abnormal conditions. Then, the deep learning model is trained using these data to enable it to automatically learn and identify the associations and characteristics between different stress modes. Finally, by predicting and detecting new automobile structure stress data, potential abnormal stress conditions can be discovered in a timely manner. However, the traditional abnormal stress analysis method of automobile structures based on deep learning has the problems of inaccurate analysis of abnormal stress in automobile structures and unclear analysis of the threat to personnel caused by structural deformation of automobile structures during collisions. Summary of the invention
[0003] Based on this, it is necessary to provide a method and system for analyzing abnormal stress of automobile structures based on deep learning to solve at least one of the above technical problems.
[0004] To achieve the above purpose, a method for analyzing abnormal stress of automobile structure based on deep learning is provided, and the method comprises the following steps:
[0005] Step S1: Acquire automobile structure design data; wherein the automobile structure design data includes: design parameters and geometric information of the vehicle body, chassis, door, and engine compartment components; construct a three-dimensional automobile structure model based on the automobile structure design data to obtain a three-dimensional automobile structure model;
[0006] Step S2: performing a connection point force bearing analysis according to the three-dimensional automobile structure model to obtain connection point force bearing data; performing automobile structure force distribution simulation according to the connection point force bearing data to obtain structural force distribution simulation data; performing force numerical component analysis of different connection points on the connection point force bearing data according to the structural force distribution simulation data to obtain connection point force numerical component data;
[0007] Step S3: Carry out a collision deformation simulation of the vehicle structure according to the numerical component data of the force at the connection point to obtain structural collision deformation simulation data; identify the abnormal deformation angle of the vehicle body structure according to the structural collision deformation simulation data to obtain abnormal deformation angle data of the vehicle body structure; calculate the deformation angle threat degree of the abnormal deformation angle data of the vehicle body structure to obtain deformation angle threat degree data; construct a vehicle body abnormal structure analysis model based on the deformation angle threat degree data based on a multi-layer perceptron to obtain a vehicle body abnormal structure analysis model;
[0008] Step S4: performing abnormal stress analysis on the vehicle structure based on the abnormal structural analysis model of the vehicle body to obtain abnormal stress data on the vehicle structure; performing optimized stress adjustment on the vehicle structure according to the abnormal stress data on the vehicle structure to obtain optimized stress adjustment data on the vehicle structure.
[0009] The present invention obtains automobile structural design data and converts it into a three-dimensional automobile structural model, so that designers and engineers can more clearly visualize the entire automobile structure, including the appearance and internal components. By constructing a three-dimensional model, the design team can better evaluate the structural design of the automobile, discover potential design defects or optimization space, and make adjustments and improvements in advance. Analyzing the stress conditions of the connection points is helpful to evaluate the strength and stability of the automobile structure in actual use and collision conditions. Through the stress bearing analysis, existing fatigue or overload problems can be identified, providing an important basis for ensuring the safety of the automobile. The stress data of key connection points or components found in the analysis can be used to guide the optimization and improvement of the structure to improve the overall performance and durability. By simulating the stress distribution of the automobile structure, the stress conditions of various parts of the whole vehicle can be fully understood, and potential weaknesses or high-risk areas can be identified. The analysis results can be used to predict the performance of the automobile under different working conditions, including vibration, stress and deformation. By analyzing the numerical components of the stress at the connection points, the load conditions of various parts of the structure can be deeply understood, providing a basis for local design adjustments. By simulating the deformation of the vehicle under collision conditions, it is helpful to evaluate the safety performance of the vehicle when an accident occurs, including The simulation data can be used to identify the area that causes abnormal deformation of the body structure, thereby guiding the optimization and improvement of the structure and improving the protection ability of the vehicle in the event of a collision. By identifying the abnormal deformation angle of the body structure, the abnormal deformation of the vehicle in the accident can be discovered in time, which is helpful to take repair measures or adjust the design in advance. The analysis of the abnormal deformation angle data of the body structure can provide basic data support for the subsequent calculation of the threat degree of the deformation angle. The calculation of the threat degree of the deformation angle helps to assess the potential threat degree of abnormal deformation to the vehicle structure and occupant safety, and provide a basis for risk management. The threat degree data can be used to guide the optimal design of the body structure, reduce the threat of deformation angle, and improve the safety performance of the vehicle in accidents. The abnormal structure analysis model built based on the multi-layer perceptron can be used as an early warning system for vehicle structural abnormalities, timely monitor and respond to changes in the body structure, and ensure the safety of the vehicle during use. Analysis of abnormal structural force data helps to monitor the health of the vehicle structure, timely discover and solve potential force problems, and improve the durability and stability of the structure. According to the abnormal analysis results, the vehicle structure can be optimized and adjusted to reduce the risk of abnormal structural force and improve the reliability and safety of the vehicle. Therefore, the present invention optimizes the traditional method of analyzing abnormal stress of automobile structure based on deep learning, solves the problems of inaccurate analysis of abnormal stress of automobile structure and unclear analysis of threat to personnel caused by deformation of automobile structure during collision, improves the accuracy of analysis of abnormal stress of automobile structure and clarifies the problem of threat degree of deformation of automobile structure to personnel during collision.
[0010] Preferably, step S1 comprises the following steps:
[0011] Step S11: Acquire automobile structural design data; wherein the automobile structural design data includes: design parameters and geometric information of the vehicle body, chassis, door, and engine compartment components;
[0012] Step S12: performing structural grid division on the automobile structure design data to obtain automobile structure grid data;
[0013] Step S13: constructing a three-dimensional automobile structure model for the automobile structure design data according to the automobile structure meshing data to obtain a three-dimensional automobile structure model.
[0014] The present invention ensures that comprehensive automobile structural design data is obtained, including design parameters and geometric information of components such as the body, chassis, doors, and engine compartment, and ensures that the acquired data is accurate to ensure the accuracy of subsequent model construction and analysis. Meshing the automobile structural design data helps to maintain the accuracy and detail of the structure when establishing a three-dimensional model. Dividing the structural grid can improve the calculation efficiency of subsequent analysis, making simulation analysis of large automobile structures more efficient. Constructing a three-dimensional automobile structural model based on structural gridded data can more intuitively present the appearance and internal structure of the automobile, which is convenient for the design team to perform visual design evaluation. The three-dimensional model provides a basis for subsequent simulation analysis, including force analysis, collision simulation, etc., and provides the necessary tools and platform for automobile design and performance evaluation.
[0015] Preferably, step S2 comprises the following steps:
[0016] Step S21: performing structural connection unit analysis according to the three-dimensional automobile structure model to obtain structural connection unit data; performing structural connection point extraction on the structural connection unit data to obtain structural connection point data;
[0017] Step S22: performing connection point force bearing analysis on the structural connection point data according to the three-dimensional automobile structure model to obtain connection point force bearing data;
[0018] Step S23: simulating the force distribution of the vehicle structure according to the three-dimensional vehicle structure model and the force bearing data of the connection points to obtain structural force distribution simulation data;
[0019] Step S24: performing force numerical component analysis of different connection points on the connection point force bearing data according to the structural force distribution simulation data to obtain the force numerical component data of the connection points.
[0020] By analyzing the structural connection units, the present invention can obtain relevant data about each connection unit in the automobile structure, including material properties, connection methods, etc., to provide basic data for subsequent connection point extraction and force analysis, and ensure that each connection unit of the automobile structure is comprehensively analyzed to maintain the integrity and stability of the structure; analyzing the force bearing conditions of the connection points is helpful to evaluate the strength and stability of the automobile structure under actual use, and discover potential force concentration or overload problems. According to the force bearing data of the connection points, it can be verified whether the design of the connection points meets safety standards and requirements, and ensure the safety performance of the vehicle under different working conditions. By simulating the force distribution of the automobile structure, the force conditions of various parts of the vehicle body under different working conditions can be fully understood, and potential high-stress areas or structural weaknesses can be identified. The structural force distribution simulation data can be used to evaluate the performance of the automobile under different conditions, including vibration, stress and deformation. Analyzing the numerical components of the force at the connection points helps to deeply understand the force conditions of local parts of the structure, and provides a basis for the optimization and improvement of the local design. By analyzing the numerical components of the force at the connection points, it can be verified whether the design meets engineering standards and safety requirements, and the stability and reliability of the structure can be ensured.
[0021] Preferably, step S23 includes the following steps:
[0022] Step S231: performing force diffusion orientation analysis on the force bearing data of the connection point to obtain the force diffusion orientation data of the connection point; performing structural section stress analysis based on the force diffusion orientation data of the connection point and the force bearing data of the connection point to obtain structural section stress data;
[0023] Step S232: performing a vehicle structure stress simulation according to the three-dimensional vehicle structure model and the structural section stress data to obtain structural stress simulation data; performing a local structure stress change analysis on the structural section stress data according to the vehicle structure stress simulation data to obtain local structure stress change data;
[0024] Step S233: performing instantaneous force diffusion topological structure analysis on the force diffusion orientation data of the connection points according to the local structure force change data to obtain instantaneous force diffusion topological structure data;
[0025] Step S234: Carry out vehicle structural force distribution simulation according to the structural force simulation data and the instantaneous force diffusion topological structure data to obtain structural force distribution simulation data.
[0026] The present invention determines the force diffusion orientation by analyzing the force bearing data of the connection point, which is helpful to understand the transmission path and influence range of the force in the structure. The stress analysis of the structural section is performed according to the force diffusion orientation data, and the stress distribution of each section of the structure can be evaluated to find potential stress concentration or abnormal stress conditions. The force simulation is performed based on the three-dimensional automobile structure model and the structural section stress data, and the force conditions of the vehicle structure under different working conditions can be fully understood. By analyzing the force simulation data, the change of the local structure force, including the area where the force increases or decreases, is determined, which provides a reference for subsequent optimization and adjustment. By analyzing the local structure force change data, the topological structure of the instantaneous force diffusion, that is, the propagation path and influence range of the force in the structure, can be determined, and the existing abnormal force propagation path can be identified, which helps to find the weaknesses or potential force concentration areas in the structure. The structural force distribution simulation is performed according to the force simulation data and the instantaneous force diffusion topological structure data, and the force conditions of the vehicle structure under different working conditions can be fully evaluated. The structural force distribution simulation data is provided, which provides a basis for structural optimization and adjustment, so as to improve the safety and stability of the vehicle structure.
[0027] Preferably, step S24 includes the following steps:
[0028] Step S241: extracting force distribution data of different connection points from the force bearing data of the connection points according to the structural collision force distribution simulation data, and obtaining the force distribution data of the connection points;
[0029] Step S242: performing connection point force vector analysis according to the connection point force distribution data to obtain the connection point force vector; performing decomposed force calculation on the connection point force vector to obtain connection point decomposed force data;
[0030] Step S243: performing calculation of adjacent connection point decomposition force variance on the connection point decomposition force data to obtain adjacent connection point decomposition force variance values; performing non-uniform variance decomposition force data sampling on the connection point decomposition force data according to the adjacent connection point decomposition force variance values to obtain non-uniform variance decomposition force sampling data;
[0031] Step S244: performing non-uniform numerical approximation processing according to the non-uniform variance decomposition force sampling data to obtain non-uniform decomposition force numerical approximation data;
[0032] Step S245: performing order-limited piecewise linear interpolation processing on the numerical approximation data of the non-uniform decomposition force to obtain decomposition force order-limited interpolation data;
[0033] Step S246: performing force numerical component analysis of different connection points on the connection point decomposition force data according to the non-uniform decomposition force numerical approximation data and the decomposition force order limited interpolation data to obtain the force numerical component data of the connection points.
[0034] The present invention extracts force distribution data of different connection points according to structural collision force distribution simulation data, which is helpful to understand the force conditions of different connection points under collision conditions and provide basic data for subsequent analysis. The force distribution data of the connection points are analyzed to obtain the force vector of the connection points, so that the force direction and magnitude of the connection points can be understood in detail. The force vector is decomposed into force calculation to obtain the decomposed force data of the connection points, which is helpful to understand the force conditions in various directions. The decomposed force variance values of adjacent connection points are calculated to evaluate the non-uniformity of the force of the connection points. The non-uniform variance decomposed force data is sampled according to the variance value to more comprehensively understand the distribution of the force of the connection points. The non-uniform decomposed force sampling data is numerically approximated, which is helpful to simplify the data and reduce the calculation complexity while retaining key information. According to the non-uniform decomposed force numerical approximation data, the order-limited piecewise linear interpolation processing is performed to obtain a smoother data curve to accurately reflect the force conditions. According to the processed data, the numerical component analysis of the force of the connection points is performed to obtain the numerical component data of the force of each connection point, which provides a basis and guidance for structural optimization.
[0035] Preferably, performing order-limited piecewise linear interpolation processing on the numerical approximation data of the non-uniformity decomposition force comprises the following steps:
[0036] Perform local feature clustering analysis on the numerical approximation data of non-uniform decomposition force to obtain local clustering data of decomposition force;
[0037] According to the local clustering data of decomposition force, the data of numerical approximation of non-uniform decomposition force is clustered and divided to obtain the clustering data of numerical approximation of decomposition force;
[0038] Based on the numerical approximation of decomposition power, clustering data is divided to perform order matching restrictions on different data categories to obtain order matching restricted data;
[0039] According to the order matching limited data, the numerical approximation data of the non-uniform decomposition force are subjected to order limited piecewise linear interpolation processing to obtain the decomposition force order limited interpolation data.
[0040] The present invention can extract local features from non-uniform decomposition force numerical approximation data through local feature clustering analysis, such as clustering features of different force modes or force areas, and group and cluster the data, which is helpful to reduce the complexity and noise of the data and highlight the main features of the data. According to the local feature clustering data, the non-uniform decomposition force numerical approximation data is reasonably divided so that data of different categories can be clearly distinguished and described. On the basis of the decomposition force numerical approximation clustering data division, order matching limitation is performed, that is, according to the characteristics of different data categories, a suitable interpolation order is determined to ensure the matching degree of the interpolation result with the original data. Through matching limitation, the consistency and rationality of the interpolation result and the original data in each category are guaranteed. Based on the order matching limitation data, piecewise linear interpolation processing is performed, and data of different categories are interpolated and fitted to obtain decomposition force order-limited interpolation data. The data obtained after the interpolation processing can not only accurately reflect the characteristics of the original data, but also maintain the smoothness and continuity of the interpolation result.
[0041] Preferably, step S3 comprises the following steps:
[0042] Step S31: performing a vehicle structure collision deformation simulation according to the connection point force numerical component data, the structure collision force distribution simulation data and the connection point force bearing data to obtain the structure collision deformation simulation data;
[0043] Step S32: performing deformation time series analysis on the structural collision deformation simulation data to obtain structural collision deformation time series data; performing time series deformation angle analysis on the structural collision deformation time series data to obtain structural collision deformation angle data;
[0044] Step S33: identifying abnormal deformation angles of the vehicle body structure according to the structural collision time-series deformation angle data and the structural collision force distribution simulation data, and obtaining abnormal deformation angle data of the vehicle body structure;
[0045] Step S34: using a deformation structure angle threat degree algorithm to calculate the deformation angle threat degree of the abnormal deformation angle data of the vehicle body structure to obtain deformation angle threat degree data;
[0046] Step S35: constructing a vehicle body abnormal structure analysis model based on the deformation angle threat degree data and the vehicle body structure abnormal deformation angle data to obtain a vehicle body abnormal structure analysis model.
[0047] The present invention simulates the numerical component data of the force on the connection point, the simulation data of the force distribution of the structure collision and the force bearing data of the connection point, so as to obtain the simulation data of the deformation of the automobile structure under the collision situation. The effect of this step is to provide the simulation data of the structural collision deformation, which can help understand the deformation of the automobile under different collision conditions and provide a reliable basis for subsequent analysis. The structural collision deformation simulation data is subjected to time series analysis to obtain the time series data of the structural collision deformation, which is further processed to obtain the time series deformation angle data of the structural collision. This step is helpful to identify the time series change trend of the automobile structure deformation and provides a data basis for the subsequent abnormal deformation analysis. Based on the structural collision time series deformation angle data and the structural collision force distribution simulation data, the abnormal deformation angle of the vehicle body structure is identified, that is, the abnormal deformation angle of the vehicle body structure that is different from the conventional collision situation is identified. Different abnormal deformation situations are detected. This step helps to timely detect abnormal deformation of the vehicle body structure, thereby improving the efficiency of vehicle collision safety and structural health status monitoring. The abnormal deformation angle data of the vehicle body structure are calculated using the deformation structure angle threat degree algorithm to obtain the deformation angle threat degree data, that is, the potential threat assessment of the deformation degree to the safety of the vehicle body structure. This step helps to assess the impact of vehicle structure deformation on safety, provides quantitative indicators for vehicle collision safety, and provides a basis for subsequent safety improvements. Based on a multi-layer perceptron, the deformation angle threat degree data and the abnormal deformation angle data of the vehicle body structure are comprehensively analyzed and modeled, and a vehicle body abnormal structure analysis model is constructed. The model has a certain predictive ability and can identify potential abnormalities in the vehicle body structure through deformation data, providing a basis for automobile safety assessment and improvement.
[0048] Preferably, step S33 includes the following steps:
[0049] Step S331: performing a collision energy time series change evaluation based on the structural collision time series deformation angle data to obtain collision energy time series change data; performing energy distribution simulation on the structural collision force distribution simulation data according to the collision energy time series change data to obtain collision energy structure distribution data;
[0050] Step S332: analyzing the fragile area of the vehicle body structure based on the structural collision force distribution simulation data to obtain the fragile area data of the vehicle body structure; performing a vehicle body structure dislocation simulation on the fragile area data of the vehicle body structure according to the collision energy structure distribution data to obtain the vehicle body structure dislocation simulation data;
[0051] Step S333: performing geometric collision interlacing influence analysis on the vehicle body structure misalignment simulation data to obtain geometric collision interlacing influence data; performing vehicle body structure collapse orientation simulation on the vehicle body structure misalignment simulation data according to the geometric collision interlacing influence data to obtain vehicle body structure collapse orientation simulation data;
[0052] Step S334: Calculate the vehicle body seat collapse orientation probability based on the vehicle body structure collapse orientation simulation data to obtain vehicle body seat collapse orientation probability data; identify the abnormal deformation angle of the vehicle body structure based on the vehicle body structure collapse orientation simulation data and the vehicle body seat collapse orientation probability data to obtain the abnormal deformation angle data of the vehicle body structure.
[0053] The present invention obtains the change of collision energy over time, i.e., collision energy time-series change data, by evaluating based on structural collision time-series deformation angle data. Based on the collision energy time-series change data, energy distribution simulation is performed on the structural collision force distribution simulation data, thereby obtaining the distribution of collision energy on the vehicle body structure, i.e., collision energy structural distribution data. This step helps to understand the change trend of collision energy over time and its distribution on the vehicle body structure, and provides basic data for subsequent structural fragile area analysis. Based on the structural collision force distribution simulation data, a fragile area analysis is performed on the vehicle body structure to determine the area that is subject to a greater impact, and obtain vehicle body structure fragile area data. According to the collision energy structural distribution data, a dislocation simulation is performed on the fragile area, i.e., the displacement and deformation of these areas in the case of a collision are analyzed to obtain vehicle body structure dislocation simulation data. This step helps to identify the area in the vehicle body structure that is most vulnerable to damage, and provides key information for further analysis. Geometric collision interleaving influence analysis is performed on the vehicle body structure dislocation simulation data. Analysis, that is, analyzing the impact of the misalignment and deformation of different parts on other parts, obtaining the geometric collision staggered influence data, and simulating the collapse direction of the body structure for the body structure misalignment simulation data according to the geometric collision staggered influence data, that is, analyzing the structural collapse that occurs in the case of a collision, and obtaining the body structure collapse direction simulation data. This step helps to evaluate the degree of damage to the body structure caused by the misalignment and deformation between different areas, and provides important information for safety assessment; analyzing the body structure collapse direction simulation data, calculating the probability of the body seat collapsing in different directions, and obtaining the body seat collapse direction probability data. This step helps to evaluate the safety of the vehicle passenger seat under different collision conditions, and provides a basis for vehicle design and safety system improvement. Based on the body structure collapse direction simulation data and the body seat collapse direction probability data, the abnormal deformation angle of the body structure is identified, that is, the abnormal structural deformation is identified. This step helps to timely discover the abnormal deformation of the body structure, and provides important support for vehicle structure safety monitoring.
[0054] Preferably, the deformation structure angle threat degree algorithm in step S34 is as follows:
[0055] ;
[0056] In the formula, The result value representing the threat degree of deformation angle, Indicates the number of angles that need to be calculated. Represented as an index, Represented as the first Angle value, Indicates the phase position of the abnormal deformation angle and the driving position, The overall angle difference representing the abnormal deformation angle, Represents the overall strength coefficient of the vehicle body structure, represents a natural constant, represents the material elastic modulus of the vehicle body structure, Indicates the error correction value of the deformable structure angle threat algorithm.
[0057] The present invention constructs an algorithm for the angle threat of a deformable structure. The algorithm evaluates the angle threat of a deformable structure by comprehensively considering factors such as the number of angles, angle value, phase position of the abnormal deformation angle and the driver's seat, the overall angle difference of the abnormal deformation angle, the overall strength coefficient of the vehicle body structure, and the elastic modulus of the material. The adjustment of each parameter has an impact on the final threat result. The algorithm fully considers the number of angles that need to be calculated. ,Increase The value of will increase the number of angles calculated, making the result more accurate. By considering more angles, the threat level of the deformed structure can be more comprehensively assessed. Angle value The size and change of each angle have an impact on the threat level of the deformed structure. By considering each angle value, the deformation of the structure can be evaluated more accurately; the phase position of the abnormal deformation angle and the driving position , which reflects the threat level of abnormal deformation to the driver. The value indicates that the greater the phase difference between the abnormal deformation angle and the driving position, the smaller the threat to the driver; the overall angle difference of the abnormal deformation angle , this parameter is used to measure the degree of abnormal deformation, the larger the The value indicates that the overall difference in deformation angle is large, indicating that the deformation of the structure is more serious; the overall strength coefficient of the body structure , this parameter reflects the resistance of the vehicle body structure, the larger The value indicates that the body structure is stronger and can withstand greater deformation without threatening the safety of the occupants; the natural constant , which is approximately equal to 2.71828, and appears in the formula for the material elastic modulus In the exponential function of , adjust the influence of the material elastic modulus on the threat result; the material elastic modulus of the body structure The elastic modulus of a material reflects its stiffness and resistance to deformation. The value indicates that the material is more rigid and has a stronger resistance to deformation; the error correction value of the deformation structure angle threat algorithm , this value is used to correct the formula to improve the accuracy of the calculation results.
[0058] Preferably, the present invention further provides a deep learning-based automobile structure stress anomaly analysis system, which is used to execute the deep learning-based automobile structure stress anomaly analysis method as described above. The deep learning-based automobile structure stress anomaly analysis method includes:
[0059] A three-dimensional automobile structure model building module is used to obtain automobile structure design data; construct a three-dimensional automobile structure model based on the automobile structure design data to obtain a three-dimensional automobile structure model;
[0060] The connection point force numerical component analysis module is used to perform connection point force bearing analysis based on the three-dimensional automobile structure model to obtain connection point force bearing data; perform automobile structure collision force distribution simulation based on the connection point force bearing data to obtain structure collision force distribution simulation data; perform different connection point force numerical component analysis on the connection point force bearing data based on the structure collision force distribution simulation data to obtain connection point force numerical component data;
[0061] The vehicle body abnormal structure analysis model construction module is used to simulate the collision deformation of the vehicle structure according to the numerical component data of the force at the connection point to obtain the structural collision deformation simulation data; identify the abnormal deformation angle of the vehicle body structure according to the structural collision deformation simulation data to obtain the abnormal deformation angle data of the vehicle body structure; calculate the deformation angle threat degree of the abnormal deformation angle data of the vehicle body structure to obtain the deformation angle threat degree data; construct the vehicle body abnormal structure analysis model based on the deformation angle threat degree data based on the multi-layer perceptron to obtain the vehicle body abnormal structure analysis model;
[0062] The vehicle structure force optimization and adjustment module is used to perform vehicle structure force abnormality analysis based on the vehicle body abnormal structure analysis model to obtain vehicle structure force abnormality data; perform vehicle structure force optimization and adjustment based on the vehicle structure force abnormality data to obtain vehicle structure force optimization and adjustment data.
[0063] The invention has the beneficial effect that, by acquiring the automobile structure design data and converting it into a three-dimensional automobile structure model, designers and engineers can more clearly visualize the entire automobile structure, including the exterior and internal components. By constructing a three-dimensional model, the design team can better evaluate the structural design of the automobile, discover potential design defects or optimization space, and thus make adjustments and improvements in advance. Analyzing the stress conditions of the connection points is helpful to evaluate the strength and stability of the automobile structure in actual use and collision conditions. Through the stress bearing analysis, existing fatigue or overload problems can be identified, providing an important basis for ensuring the safety of the automobile. The stress data of the key connection points or components found in the analysis can be used to guide the optimization and improvement of the structure to improve the overall performance and durability. By simulating the stress distribution of the automobile structure, the stress conditions of various parts of the vehicle can be fully understood, and potential weaknesses or high-risk areas can be identified. The analysis results can be used to predict the performance of the automobile under different working conditions, including vibration, stress and deformation. By analyzing the numerical components of the stress at the connection points, the load conditions of various parts of the structure can be deeply understood, providing a basis for local design adjustments. By simulating the deformation of the vehicle under collision conditions, it is helpful to evaluate the safety performance of the vehicle when an accident occurs. , including the deformation degree of the body and the stability of the structure. The simulation data can be used to identify the areas that cause abnormal deformation of the body structure, thereby guiding the optimization and improvement of the structure and improving the protection ability of the vehicle in the event of a collision. By identifying the abnormal deformation angle of the body structure, the abnormal deformation of the vehicle in the accident can be discovered in time, which helps to take repair measures or adjust the design in advance. The analysis of the abnormal deformation angle data of the body structure can provide basic data support for the subsequent calculation of the threat degree of the deformation angle. The calculation of the threat degree of the deformation angle helps to assess the potential threat degree of abnormal deformation to the vehicle structure and occupant safety, and provide a basis for risk management. The threat degree data can be used to guide the optimal design of the body structure, reduce the threat of deformation angles, and improve the safety performance of the vehicle in accidents. The abnormal structure analysis model built based on the multi-layer perceptron can be used as an early warning system for vehicle structural abnormalities, timely monitor and respond to changes in the body structure, and ensure the safety of the vehicle during use. Analyzing the abnormal structural force data helps to monitor the health of the vehicle structure, timely discover and solve potential force problems, and improve the durability and stability of the structure. According to the abnormal analysis results, the vehicle structure is optimized and adjusted to reduce the risk of abnormal structural force and improve the reliability and safety of the vehicle. Therefore, the present invention optimizes the traditional method of analyzing abnormal stress of automobile structure based on deep learning, solves the problems of inaccurate analysis of abnormal stress of automobile structure and unclear analysis of threat to personnel caused by deformation of automobile structure during collision, improves the accuracy of analysis of abnormal stress of automobile structure and clarifies the problem of threat degree of deformation of automobile structure to personnel during collision. BRIEF DESCRIPTION OF THE DRAWINGS
[0064] Figure 1 A schematic diagram of the steps of a method for analyzing abnormal stress of automobile structure based on deep learning;
[0065] Figure 2 for Figure 1 Detailed implementation steps of step S2 in the flowchart;
[0066] Figure 3 for Figure 1 Detailed implementation steps of step S3 in FIG.
[0067] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0068] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.
[0069] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.
[0070] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.
[0071] To achieve this, please refer to Figures 1 to 3 , a method for analyzing abnormal stress of automobile structure based on deep learning, the method comprising the following steps:
[0072] Step S1: Acquire automobile structure design data; wherein the automobile structure design data includes: design parameters and geometric information of the vehicle body, chassis, door, and engine compartment components; construct a three-dimensional automobile structure model based on the automobile structure design data to obtain a three-dimensional automobile structure model;
[0073] Step S2: performing a connection point force bearing analysis according to the three-dimensional automobile structure model to obtain connection point force bearing data; performing automobile structure force distribution simulation according to the connection point force bearing data to obtain structural force distribution simulation data; performing force numerical component analysis of different connection points on the connection point force bearing data according to the structural force distribution simulation data to obtain connection point force numerical component data;
[0074] Step S3: Carry out a collision deformation simulation of the vehicle structure according to the numerical component data of the force at the connection point to obtain structural collision deformation simulation data; identify the abnormal deformation angle of the vehicle body structure according to the structural collision deformation simulation data to obtain abnormal deformation angle data of the vehicle body structure; calculate the deformation angle threat degree of the abnormal deformation angle data of the vehicle body structure to obtain deformation angle threat degree data; construct a vehicle body abnormal structure analysis model based on the deformation angle threat degree data based on a multi-layer perceptron to obtain a vehicle body abnormal structure analysis model;
[0075] Step S4: performing abnormal stress analysis on the vehicle structure based on the abnormal structural analysis model of the vehicle body to obtain abnormal stress data on the vehicle structure; performing optimized stress adjustment on the vehicle structure according to the abnormal stress data on the vehicle structure to obtain optimized stress adjustment data on the vehicle structure.
[0076] In the embodiment of the present invention, reference Figure 1 The above is a schematic diagram of the steps of a method for analyzing abnormal stress of an automobile structure based on deep learning of the present invention. In this example, the method for analyzing abnormal stress of an automobile structure based on deep learning includes the following steps:
[0077] Step S1: Acquire automobile structure design data; wherein the automobile structure design data includes: design parameters and geometric information of the vehicle body, chassis, door, and engine compartment components; construct a three-dimensional automobile structure model based on the automobile structure design data to obtain a three-dimensional automobile structure model;
[0078] In the embodiment of the present invention, the structural design data of the automobile, including design drawings and CAD files, are obtained from the automobile manufacturer or design team, and a three-dimensional automobile structural model is constructed based on the automobile structural design data using computer-aided design (CAD) software. For example, the main structure, body, wheels and other parts of the automobile are drawn based on the dimensions and geometric shapes in the design drawings, and the accuracy and completeness of the model are ensured.
[0079] Step S2: performing a connection point force bearing analysis according to the three-dimensional automobile structure model to obtain connection point force bearing data; performing automobile structure force distribution simulation according to the connection point force bearing data to obtain structural force distribution simulation data; performing force numerical component analysis of different connection points on the connection point force bearing data according to the structural force distribution simulation data to obtain connection point force numerical component data;
[0080] In an embodiment of the present invention, an engineering simulation software is used to perform a stress analysis of the connection points of a three-dimensional automobile structure model. This involves determining the stress conditions of connection points (such as bolts and welding points) under different working conditions, including the magnitude and direction of the stress, and simulating the stress distribution of the automobile structure based on the stress bearing data of the connection points. This includes simulating the distribution of stress on the entire automobile structure to determine which areas are subjected to greater forces and which areas are the main load-bearing areas of the structure. According to the structural stress distribution simulation data, the stress bearing data of the connection points is further analyzed to obtain the numerical component data of the stress at the connection points. This involves decomposing the stress into numerical components of the magnitude of the stress in different directions and at different connection points in order to better understand the stress conditions of the connection points, that is, there is a total force that is evenly distributed by the numerical amount of the stress at each connection point, that is, the numerical components of the magnitude of the stress in different directions and at different connection points of each connection point, and the magnitude of the stress at different connection points is expressed numerically.
[0081] Step S3: Carry out a collision deformation simulation of the vehicle structure according to the numerical component data of the force at the connection point to obtain structural collision deformation simulation data; identify the abnormal deformation angle of the vehicle body structure according to the structural collision deformation simulation data to obtain abnormal deformation angle data of the vehicle body structure; calculate the deformation angle threat degree of the abnormal deformation angle data of the vehicle body structure to obtain deformation angle threat degree data; construct a vehicle body abnormal structure analysis model based on the deformation angle threat degree data based on a multi-layer perceptron to obtain a vehicle body abnormal structure analysis model;
[0082] In an embodiment of the present invention, simulation software is used to take the numerical component data of the force at the connection point as input to simulate the deformation of the automobile structure under collision conditions. For example, the deformation of the vehicle under different collision scenarios is simulated, including deformations such as bending and twisting of the vehicle body, and the abnormal deformation angle of the vehicle body structure is identified by analyzing the structural collision deformation simulation data. Including detecting abnormal deformations such as twisting and bending that occur in the vehicle during the collision, expressing them in the form of angles, and calculating the threat degree of the deformation angle according to the abnormal deformation angle data of the vehicle body structure. This involves determining the degree of influence of different deformation angles on the safety of the vehicle structure to quantify the degree of threat of deformation to the safety of the vehicle, and using machine learning methods such as multi-layer perceptrons to build a vehicle body abnormal structure analysis model based on the deformation angle threat degree data. Including training a model to identify the degree of abnormality of the vehicle body structure under different deformation conditions to help evaluate the safety of the vehicle during a collision.
[0083] Step S4: performing abnormal stress analysis on the vehicle structure based on the abnormal structural analysis model of the vehicle body to obtain abnormal stress data on the vehicle structure; performing optimized stress adjustment on the vehicle structure according to the abnormal stress data on the vehicle structure to obtain optimized stress adjustment data on the vehicle structure.
[0084] In the embodiment of the present invention, based on the abnormal structural analysis model of the vehicle body, the abnormal stress of the vehicle structure is analyzed. This involves detecting whether the stress distribution of the structure in a collision situation meets the design requirements, and whether there are abnormal stress concentration areas. According to the abnormal stress data of the vehicle structure, the structural stress is optimized and adjusted. This includes adjusting the design of the connection points, adding or adjusting structural parts to reduce the stress concentration, and thus improving the safety of the vehicle in a collision situation.
[0085] The present invention obtains automobile structural design data and converts it into a three-dimensional automobile structural model, so that designers and engineers can more clearly visualize the entire automobile structure, including the appearance and internal components. By constructing a three-dimensional model, the design team can better evaluate the structural design of the automobile, discover potential design defects or optimization space, and make adjustments and improvements in advance. Analyzing the stress conditions of the connection points is helpful to evaluate the strength and stability of the automobile structure in actual use and collision conditions. Through the stress bearing analysis, existing fatigue or overload problems can be identified, providing an important basis for ensuring the safety of the automobile. The stress data of key connection points or components found in the analysis can be used to guide the optimization and improvement of the structure to improve the overall performance and durability. By simulating the stress distribution of the automobile structure, the stress conditions of various parts of the whole vehicle can be fully understood, and potential weaknesses or high-risk areas can be identified. The analysis results can be used to predict the performance of the automobile under different working conditions, including vibration, stress and deformation. By analyzing the numerical components of the stress at the connection points, the load conditions of various parts of the structure can be deeply understood, providing a basis for local design adjustments. By simulating the deformation of the vehicle under collision conditions, it is helpful to evaluate the safety performance of the vehicle when an accident occurs, including The simulation data can be used to identify the area that causes abnormal deformation of the body structure, thereby guiding the optimization and improvement of the structure and improving the protection ability of the vehicle in the event of a collision. By identifying the abnormal deformation angle of the body structure, the abnormal deformation of the vehicle in the accident can be discovered in time, which is helpful to take repair measures or adjust the design in advance. The analysis of the abnormal deformation angle data of the body structure can provide basic data support for the subsequent calculation of the threat degree of the deformation angle. The calculation of the threat degree of the deformation angle helps to assess the potential threat degree of abnormal deformation to the vehicle structure and occupant safety, and provide a basis for risk management. The threat degree data can be used to guide the optimal design of the body structure, reduce the threat of deformation angle, and improve the safety performance of the vehicle in accidents. The abnormal structure analysis model built based on the multi-layer perceptron can be used as an early warning system for vehicle structural abnormalities, timely monitor and respond to changes in the body structure, and ensure the safety of the vehicle during use. Analysis of abnormal structural force data helps to monitor the health of the vehicle structure, timely discover and solve potential force problems, and improve the durability and stability of the structure. According to the abnormal analysis results, the vehicle structure can be optimized and adjusted to reduce the risk of abnormal structural force and improve the reliability and safety of the vehicle. Therefore, the present invention optimizes the traditional method of analyzing abnormal stress of automobile structure based on deep learning, solves the problems of inaccurate analysis of abnormal stress of automobile structure and unclear analysis of threat to personnel caused by deformation of automobile structure during collision, improves the accuracy of analysis of abnormal stress of automobile structure and clarifies the problem of threat degree of deformation of automobile structure to personnel during collision.
[0086] Preferably, step S1 comprises the following steps:
[0087] Step S11: Acquire automobile structural design data; wherein the automobile structural design data includes: design parameters and geometric information of the vehicle body, chassis, door, and engine compartment components;
[0088] Step S12: performing structural grid division on the automobile structure design data to obtain automobile structure grid data;
[0089] Step S13: constructing a three-dimensional automobile structure model for the automobile structure design data according to the automobile structure meshing data to obtain a three-dimensional automobile structure model.
[0090] In an embodiment of the present invention, automobile structure design data, including design drawings, CAD files or other related documents, are obtained from automobile design departments or related fields. Finite element analysis software is used to grid the automobile structure according to a grid division algorithm, and the body structure is divided into multiple grid units for simulation analysis. Regarding the gridded data, computer-aided design software and other tools are used to construct a complete three-dimensional automobile structure model for subsequent simulation analysis and design optimization, including using CAD software to construct a three-dimensional structure model of a new type of automobile based on the gridded data, including components such as the body, chassis, and doors, for subsequent structural analysis and design evaluation.
[0091] The present invention ensures that comprehensive automobile structural design data is obtained, including design parameters and geometric information of components such as the body, chassis, doors, and engine compartment, and ensures that the acquired data is accurate to ensure the accuracy of subsequent model construction and analysis. Meshing the automobile structural design data helps to maintain the accuracy and detail of the structure when establishing a three-dimensional model. Dividing the structural grid can improve the calculation efficiency of subsequent analysis, making simulation analysis of large automobile structures more efficient. Constructing a three-dimensional automobile structural model based on structural gridded data can more intuitively present the appearance and internal structure of the automobile, which is convenient for the design team to perform visual design evaluation. The three-dimensional model provides a basis for subsequent simulation analysis, including force analysis, collision simulation, etc., and provides the necessary tools and platform for automobile design and performance evaluation.
[0092] Preferably, step S2 comprises the following steps:
[0093] Step S21: performing structural connection unit analysis according to the three-dimensional automobile structure model to obtain structural connection unit data; performing structural connection point extraction on the structural connection unit data to obtain structural connection point data;
[0094] Step S22: performing connection point force bearing analysis on the structural connection point data according to the three-dimensional automobile structure model to obtain connection point force bearing data;
[0095] Step S23: simulating the force distribution of the vehicle structure according to the three-dimensional vehicle structure model and the force bearing data of the connection points to obtain structural force distribution simulation data;
[0096] Step S24: performing force numerical component analysis of different connection points on the connection point force bearing data according to the structural force distribution simulation data to obtain the force numerical component data of the connection points.
[0097] As an example of the present invention, refer to Figure 2 As shown, in this example, step S2 includes:
[0098] Step S21: performing structural connection unit analysis according to the three-dimensional automobile structure model to obtain structural connection unit data; performing structural connection point extraction on the structural connection unit data to obtain structural connection point data;
[0099] In the embodiment of the present invention, a connection unit analysis is performed on a three-dimensional automobile structure model using professional structural analysis software. This includes identifying various connection units in the automobile structure, such as bolts and welding points, extracting relevant data of the structural connection units from the three-dimensional model, such as position, type, and material, analyzing the structural connection units, and determining their role and influence in the automobile structure, including using finite element analysis software to analyze the automobile structure model, identify the position and type of each connection unit, such as body welding points, bolt connection points, etc., and extract relevant data.
[0100] Step S22: performing connection point force bearing analysis on the structural connection point data according to the three-dimensional automobile structure model to obtain connection point force bearing data;
[0101] In the embodiment of the present invention, a connection point force bearing analysis is performed based on a three-dimensional automobile structure model and structural connection point data. This includes determining the force, direction, and magnitude of the connection point, using professional simulation software to perform force bearing analysis, considering the force conditions under various working conditions, such as static loads and dynamic loads, including using finite element analysis software to perform force bearing analysis on the structural connection point, determining the force on the connection point in collision, driving, etc., and obtaining the connection point force bearing data.
[0102] Step S23: simulating the force distribution of the vehicle structure according to the three-dimensional vehicle structure model and the force bearing data of the connection points to obtain structural force distribution simulation data;
[0103] In an embodiment of the present invention, professional structural analysis software is used to simulate the force distribution of the automobile structure in combination with a three-dimensional automobile structure model and connection point force bearing data. The force conditions under various working conditions, such as collision, acceleration, braking, etc., are considered in the simulation to obtain a comprehensive force distribution condition, including using structural analysis software to simulate the force distribution of the vehicle under different conditions, such as front-end collision, side collision, etc., based on the automobile structure model and the actual measured connection point force data, to obtain structural force distribution simulation data.
[0104] Step S24: performing force numerical component analysis of different connection points on the connection point force bearing data according to the structural force distribution simulation data to obtain the force numerical component data of the connection points.
[0105] In an embodiment of the present invention, based on the structural force distribution simulation data, the connection point force bearing data is analyzed to obtain the numerical components of the forces at different connection points, and the force components in different directions are considered to better understand the force conditions of the connection points and determine whether there are abnormal or excessive force conditions. By analyzing the data, it is determined which connection points are subjected to greater forces and the impact of these forces on the structure. The method includes analyzing the connection point force bearing data according to the structural force distribution simulation data to determine the force components of each connection point in different directions, such as tension, pressure, etc., to obtain the connection point force numerical component data, wherein the analysis of the numerical components of the forces at different connection points means that a total force is evenly distributed by the numerical amount of the force at each connection point, that is, the numerical components of the forces at different directions and at different connection points of each connection point, and the forces at different connection points are represented by numerical values.
[0106] By analyzing the structural connection units, the present invention can obtain relevant data about each connection unit in the automobile structure, including material properties, connection methods, etc., to provide basic data for subsequent connection point extraction and force analysis, and ensure that each connection unit of the automobile structure is comprehensively analyzed to maintain the integrity and stability of the structure; analyzing the force bearing conditions of the connection points is helpful to evaluate the strength and stability of the automobile structure under actual use, and discover potential force concentration or overload problems. According to the force bearing data of the connection points, it can be verified whether the design of the connection points meets safety standards and requirements, and ensure the safety performance of the vehicle under different working conditions. By simulating the force distribution of the automobile structure, the force conditions of various parts of the vehicle body under different working conditions can be fully understood, and potential high-stress areas or structural weaknesses can be identified. The structural force distribution simulation data can be used to evaluate the performance of the automobile under different conditions, including vibration, stress and deformation. Analyzing the numerical components of the force at the connection points helps to deeply understand the force conditions of local parts of the structure, and provides a basis for the optimization and improvement of the local design. By analyzing the numerical components of the force at the connection points, it can be verified whether the design meets engineering standards and safety requirements, and the stability and reliability of the structure can be ensured.
[0107] Preferably, step S23 includes the following steps:
[0108] Step S231: performing force diffusion orientation analysis on the force bearing data of the connection point to obtain the force diffusion orientation data of the connection point; performing structural section stress analysis based on the force diffusion orientation data of the connection point and the force bearing data of the connection point to obtain structural section stress data;
[0109] Step S232: performing a vehicle structure stress simulation according to the three-dimensional vehicle structure model and the structural section stress data to obtain structural stress simulation data; performing a local structure stress change analysis on the structural section stress data according to the vehicle structure stress simulation data to obtain local structure stress change data;
[0110] Step S233: performing instantaneous force diffusion topological structure analysis on the force diffusion orientation data of the connection points according to the local structure force change data to obtain instantaneous force diffusion topological structure data;
[0111] Step S234: Carry out vehicle structural force distribution simulation according to the structural force simulation data and the instantaneous force diffusion topological structure data to obtain structural force distribution simulation data.
[0112] In an embodiment of the present invention, the force bearing data at the connection point, including the magnitude and direction of the force, are collected and sorted, and a professional force analysis software or tool is used to perform a diffusion azimuth analysis on the force at the connection point. This step is intended to determine the direction and path of the force diffusion from the connection point to the surrounding structure. According to the diffusion azimuth analysis results, the force diffusion azimuth data of the connection point is obtained. These data describe the specific direction and degree of force propagation from the connection point to the outside. Combined with the force diffusion azimuth data of the connection point and the force bearing data of the connection point, a stress analysis of the structural section is performed. This step uses mechanical principles and structural analysis methods to calculate the stress distribution on the structural section. After stress analysis, stress data at each position on the structural section is obtained. These data reflect the stress state of the structural section when it is subjected to external forces, establish or import a three-dimensional automobile structure model, including components such as the body, chassis, and doors, ensure that the model accurately describes the geometric shape and material properties of the automobile structure, combine the structural section stress data with the three-dimensional automobile structure model, and perform force simulation of the automobile structure on a simulation software or platform. This step simulates the stress conditions of the automobile during actual use and operation by applying methods such as engineering mechanics and finite element analysis. After completing the structural stress simulation, the stress simulation data of the automobile structure under different working conditions are obtained. These data include information such as stress and deformation of various components of the automobile under stress. According to the stress simulation data of the automobile structure, the local structural stress change analysis is performed on the structural section stress data. This step focuses on the local stress changes of various components of the automobile, as well as the stress concentration or deformation, and finally obtains the local structural stress change data. These data provide detailed information on the local stress changes of the automobile structure under different working conditions and stress conditions, which helps to optimize the structural design and improve the structural strength; prepare the local structural stress change data obtained in the previous step, including local stress concentration, deformation and other information, and use professional topological analysis tools to perform instantaneous force diffusion topological structure analysis on the connection point force diffusion orientation data based on the local structural force change data. This step aims to identify the instantaneous path and structure of the force transmission of the connection point, and obtain the instantaneous path and structural data of the force diffusion of the connection point based on the topological structure analysis results. These data describe the transmission path and connection relationship of the force in the structure. Prepare the structural force simulation data obtained in the previous step, including the force conditions and stress distribution of each component of the car. Combine the structural force simulation data and the instantaneous force diffusion topological structure data to simulate the force distribution of the car structure on the relevant simulation platform or software. This step simulates the force distribution of the car structure under different working conditions through numerical calculation methods. After completing the force distribution simulation, the force distribution simulation data of the car structure at different parts are obtained. These data reflect the stress distribution of various parts of the car structure under force, which helps to analyze the strength and stability of the structure.
[0113] The present invention determines the force diffusion orientation by analyzing the force bearing data of the connection point, which is helpful to understand the transmission path and influence range of the force in the structure. The stress analysis of the structural section is performed according to the force diffusion orientation data, and the stress distribution of each section of the structure can be evaluated to find potential stress concentration or abnormal stress conditions. The force simulation is performed based on the three-dimensional automobile structure model and the structural section stress data, and the force conditions of the vehicle structure under different working conditions can be fully understood. By analyzing the force simulation data, the change of the local structure force, including the area where the force increases or decreases, is determined, which provides a reference for subsequent optimization and adjustment. By analyzing the local structure force change data, the topological structure of the instantaneous force diffusion, that is, the propagation path and influence range of the force in the structure, can be determined, and the existing abnormal force propagation path can be identified, which helps to find the weaknesses or potential force concentration areas in the structure. The structural force distribution simulation is performed according to the force simulation data and the instantaneous force diffusion topological structure data, and the force conditions of the vehicle structure under different working conditions can be fully evaluated. The structural force distribution simulation data is provided, which provides a basis for structural optimization and adjustment, so as to improve the safety and stability of the vehicle structure.
[0114] Preferably, step S24 includes the following steps:
[0115] Step S241: extracting force distribution data of different connection points from the force bearing data of the connection points according to the structural collision force distribution simulation data, and obtaining the force distribution data of the connection points;
[0116] Step S242: performing connection point force vector analysis according to the connection point force distribution data to obtain the connection point force vector; performing decomposed force calculation on the connection point force vector to obtain connection point decomposed force data;
[0117] Step S243: performing calculation of adjacent connection point decomposition force variance on the connection point decomposition force data to obtain adjacent connection point decomposition force variance values; performing non-uniform variance decomposition force data sampling on the connection point decomposition force data according to the adjacent connection point decomposition force variance values to obtain non-uniform variance decomposition force sampling data;
[0118] Step S244: performing non-uniform numerical approximation processing according to the non-uniform variance decomposition force sampling data to obtain non-uniform decomposition force numerical approximation data;
[0119] Step S245: performing order-limited piecewise linear interpolation processing on the numerical approximation data of the non-uniform decomposition force to obtain decomposition force order-limited interpolation data;
[0120] Step S246: performing force numerical component analysis of different connection points on the connection point decomposition force data according to the non-uniform decomposition force numerical approximation data and the decomposition force order limited interpolation data to obtain the force numerical component data of the connection points.
[0121] In an embodiment of the present invention, force distribution simulation data of the structure under collision conditions are prepared. These data describe the distribution of forces received by various parts in a collision event. According to the structure collision force distribution simulation data, corresponding connection point force bearing data are extracted for different connection points. These data reflect the distribution of forces received by the connection points under collision conditions. According to the extracted connection point force bearing data, the distribution data of the connection point force is obtained, including information such as the magnitude and direction of the force. According to the connection point force distribution data, the connection point force vector is analyzed to determine the magnitude and direction of the force received by each connection point, and the force vector of each connection point is decomposed to calculate its decomposed force in different directions. This step helps to understand the force conditions in various directions, obtain the data of the connection point decomposed force, including the magnitude of the force in various directions, analyze the connection point decomposed force data, and calculate the variance value of the decomposed force between adjacent connection points. This step is used to evaluate the non-uniformity of the connection point force. According to the variance value of the decomposed force of adjacent connection points, the connection point decomposed force data is sampled to extract data representing the variance of the non-uniformity. These data reflect the degree of non-uniformity of the connection point force. Prepare the non-uniform decomposition force sampling data obtained in the previous step as the basic data for processing, use numerical approximation methods such as polynomial fitting, curve fitting, etc. to process the non-uniform decomposition force sampling data to obtain more accurate non-uniform decomposition force numerical approximation data. After completing the approximation processing, obtain the numerical approximation data of the non-uniform decomposition force. These data better describe the non-uniformity of the force at the connection point. Prepare the non-uniform decomposition force numerical approximation data obtained in the previous step as the basic data for piecewise linear interpolation processing, and perform piecewise linear interpolation on the non-uniform decomposition force numerical approximation data. According to the order limit, the data is segmented and linearly interpolated to obtain smoother and continuous interpolation data. After the interpolation is completed, the order-limited interpolation data of the decomposed force are obtained. These data are more suitable for subsequent analysis and processing. According to the numerical approximation data of the non-uniform decomposed force and the order-limited interpolation data of the decomposed force, the numerical component analysis of the connection point force data is performed to decompose the force into numerical components in various directions. After completing the numerical component analysis, the numerical component data of the connection point force in various directions are obtained. These data reflect the specific numerical distribution of the connection point force.
[0122] The present invention extracts force distribution data of different connection points according to structural collision force distribution simulation data, which is helpful to understand the force conditions of different connection points under collision conditions and provide basic data for subsequent analysis. The force distribution data of the connection points are analyzed to obtain the force vector of the connection points, so that the force direction and magnitude of the connection points can be understood in detail. The force vector is decomposed into force calculation to obtain the decomposed force data of the connection points, which is helpful to understand the force conditions in various directions. The decomposed force variance values of adjacent connection points are calculated to evaluate the non-uniformity of the force of the connection points. The non-uniform variance decomposed force data is sampled according to the variance value to more comprehensively understand the distribution of the force of the connection points. The non-uniform decomposed force sampling data is numerically approximated, which is helpful to simplify the data and reduce the calculation complexity while retaining key information. According to the non-uniform decomposed force numerical approximation data, the order-limited piecewise linear interpolation processing is performed to obtain a smoother data curve to accurately reflect the force conditions. According to the processed data, the numerical component analysis of the force of the connection points is performed to obtain the numerical component data of the force of each connection point, which provides a basis and guidance for structural optimization.
[0123] Preferably, performing order-limited piecewise linear interpolation processing on the numerical approximation data of the non-uniformity decomposition force comprises the following steps:
[0124] Perform local feature clustering analysis on the numerical approximation data of non-uniform decomposition force to obtain local clustering data of decomposition force;
[0125] According to the local clustering data of decomposition force, the data of numerical approximation of non-uniform decomposition force is clustered and divided to obtain the clustering data of numerical approximation of decomposition force;
[0126] Based on the numerical approximation of decomposition power, clustering data is divided to perform order matching restrictions on different data categories to obtain order matching restricted data;
[0127] According to the order matching limited data, the numerical approximation data of the non-uniform decomposition force are subjected to order limited piecewise linear interpolation processing to obtain the decomposition force order limited interpolation data.
[0128] In an embodiment of the present invention, a cluster analysis method (such as K-means clustering, density clustering, etc.) is used to perform local feature clustering analysis on the non-uniform decomposition force numerical approximation data to determine an appropriate number of clusters so as to divide the data into local clusters with similar features. According to the results of the local feature clustering analysis, the non-uniform decomposition force numerical approximation data is clustered and divided into different cluster categories, each cluster category represents a group of data points with similar features. Based on the decomposition force local clustering data, different data categories are limited by order matching, and the interpolation order to be used for each data category is determined to ensure that the interpolation result can effectively reflect the characteristics of the data. According to the order matching limit data, the non-uniform decomposition force numerical approximation data is subjected to order-limited piecewise linear interpolation processing, the data is segmented according to different cluster categories and orders, and each segment is interpolated using a linear interpolation method to ensure that the interpolation result is smooth and continuous between the segments and can accurately reflect the trends and characteristics of the original data.
[0129] The present invention can extract local features from non-uniform decomposition force numerical approximation data through local feature clustering analysis, such as clustering features of different force modes or force areas, and group and cluster the data, which is helpful to reduce the complexity and noise of the data and highlight the main features of the data. According to the local feature clustering data, the non-uniform decomposition force numerical approximation data is reasonably divided so that data of different categories can be clearly distinguished and described. On the basis of the decomposition force numerical approximation clustering data division, order matching limitation is performed, that is, according to the characteristics of different data categories, a suitable interpolation order is determined to ensure the matching degree of the interpolation result with the original data. Through matching limitation, the consistency and rationality of the interpolation result and the original data in each category are guaranteed. Based on the order matching limitation data, piecewise linear interpolation processing is performed, and data of different categories are interpolated and fitted to obtain decomposition force order-limited interpolation data. The data obtained after the interpolation processing can not only accurately reflect the characteristics of the original data, but also maintain the smoothness and continuity of the interpolation result.
[0130] Preferably, step S3 comprises the following steps:
[0131] Step S31: performing a vehicle structure collision deformation simulation according to the connection point force numerical component data, the structure collision force distribution simulation data and the connection point force bearing data to obtain the structure collision deformation simulation data;
[0132] Step S32: performing deformation time series analysis on the structural collision deformation simulation data to obtain structural collision deformation time series data; performing time series deformation angle analysis on the structural collision deformation time series data to obtain structural collision deformation angle data;
[0133] Step S33: identifying abnormal deformation angles of the vehicle body structure according to the structural collision time-series deformation angle data and the structural collision force distribution simulation data, and obtaining abnormal deformation angle data of the vehicle body structure;
[0134] Step S34: using a deformation structure angle threat degree algorithm to calculate the deformation angle threat degree of the abnormal deformation angle data of the vehicle body structure to obtain deformation angle threat degree data;
[0135] Step S35: constructing a vehicle body abnormal structure analysis model based on the deformation angle threat degree data and the vehicle body structure abnormal deformation angle data to obtain a vehicle body abnormal structure analysis model.
[0136] As an example of the present invention, refer to Figure 3 As shown, in this example, step S3 includes:
[0137] Step S31: performing a vehicle structure collision deformation simulation according to the connection point force numerical component data, the structure collision force distribution simulation data and the connection point force bearing data to obtain the structure collision deformation simulation data;
[0138] In an embodiment of the present invention, numerical component data of connection point forces, structural collision force distribution simulation data, and connection point force bearing data are prepared as input data for simulation, and the model and parameters required for the simulation are determined, including the vehicle structure model, collision model, and material parameters. Based on the input data and model, a vehicle structure collision deformation simulation is performed on a corresponding simulation platform or software. According to the connection point force data and the collision force distribution, the deformation of the vehicle during the collision is simulated, including the degree of deformation and stress distribution. After the simulation is completed, the structural collision deformation simulation data is obtained, including information such as the deformation of the vehicle structure at different time points and stress distribution.
[0139] Step S32: performing deformation time series analysis on the structural collision deformation simulation data to obtain structural collision deformation time series data; performing time series deformation angle analysis on the structural collision deformation time series data to obtain structural collision deformation angle data;
[0140] In an embodiment of the present invention, a time series analysis is performed on the structural collision deformation simulation data, and the deformation of the automobile structure at different time points is recorded and analyzed. Based on the deformation time series data, the structural collision time series deformation angle is analyzed, and the deformation angle of the automobile structure during the collision is calculated, including the angle changes such as rotation and bending. The deformation angle changes of each part are analyzed to understand the deformation characteristics and degree of the automobile structure in the collision.
[0141] Step S33: identifying abnormal deformation angles of the vehicle body structure according to the structural collision time-series deformation angle data and the structural collision force distribution simulation data, and obtaining abnormal deformation angle data of the vehicle body structure;
[0142] In an embodiment of the present invention, structural collision sequence deformation angle data and structural collision force distribution simulation data are prepared as input data for abnormal deformation angle identification, and the structural collision sequence deformation angle data are analyzed. In combination with the structural collision force distribution simulation data, existing abnormal deformation angles, such as excessive deformation angles or abnormal deformation modes, are identified. After the abnormal deformation angles are identified, the abnormal deformation angle data of the vehicle body structure are obtained. These data describe the abnormal deformation of the vehicle body structure under collision or force conditions.
[0143] Step S34: using a deformation structure angle threat degree algorithm to calculate the deformation angle threat degree of the abnormal deformation angle data of the vehicle body structure to obtain deformation angle threat degree data;
[0144] In an embodiment of the present invention, abnormal deformation angle data of the vehicle body structure are prepared as input data for deformation angle threat degree calculation. Based on the deformation structure angle threat degree algorithm, the abnormal deformation angle data of the vehicle body structure are calculated. According to the design of the algorithm, the threat degree of each abnormal deformation angle is calculated to evaluate its influence on the safety of the vehicle body structure. After completing the threat degree calculation, the deformation angle threat degree data are obtained, which reflect the threat degree of each abnormal deformation angle to the safety of the vehicle body structure.
[0145] Step S35: constructing a vehicle body abnormal structure analysis model based on the deformation angle threat degree data and the vehicle body structure abnormal deformation angle data to obtain a vehicle body abnormal structure analysis model.
[0146] In an embodiment of the present invention, deformation angle threat degree data and abnormal deformation angle data of vehicle body structure are prepared as input data for constructing an abnormal structure analysis model of the vehicle body. The deformation angle threat degree data and abnormal deformation angle data of the vehicle body structure are modeled based on a multi-layer perceptron or other appropriate machine learning algorithms. The model is trained to identify abnormal conditions of the vehicle body structure and predict their impact on vehicle safety. After the model is built, an abnormal structure analysis model of the vehicle body is obtained, which can perform analysis based on the input abnormal deformation data and provide predictions and evaluations of abnormal conditions of the vehicle body structure.
[0147] The present invention simulates the numerical component data of the force on the connection point, the simulation data of the force distribution of the structure collision and the force bearing data of the connection point, so as to obtain the simulation data of the deformation of the automobile structure under the collision situation. The effect of this step is to provide the simulation data of the structural collision deformation, which can help understand the deformation of the automobile under different collision conditions and provide a reliable basis for subsequent analysis. The structural collision deformation simulation data is subjected to time series analysis to obtain the time series data of the structural collision deformation, which is further processed to obtain the time series deformation angle data of the structural collision. This step is helpful to identify the time series change trend of the automobile structure deformation and provides a data basis for the subsequent abnormal deformation analysis. Based on the structural collision time series deformation angle data and the structural collision force distribution simulation data, the abnormal deformation angle of the vehicle body structure is identified, that is, the abnormal deformation angle of the vehicle body structure that is different from the conventional collision situation is identified. Different abnormal deformation situations are detected. This step helps to timely detect abnormal deformation of the vehicle body structure, thereby improving the efficiency of vehicle collision safety and structural health status monitoring. The abnormal deformation angle data of the vehicle body structure are calculated using the deformation structure angle threat degree algorithm to obtain the deformation angle threat degree data, that is, the potential threat assessment of the deformation degree to the safety of the vehicle body structure. This step helps to assess the impact of vehicle structure deformation on safety, provides quantitative indicators for vehicle collision safety, and provides a basis for subsequent safety improvements. Based on a multi-layer perceptron, the deformation angle threat degree data and the abnormal deformation angle data of the vehicle body structure are comprehensively analyzed and modeled, and a vehicle body abnormal structure analysis model is constructed. The model has a certain predictive ability and can identify potential abnormalities in the vehicle body structure through deformation data, providing a basis for automobile safety assessment and improvement.
[0148] Preferably, step S33 includes the following steps:
[0149] Step S331: performing a collision energy time series change evaluation based on the structural collision time series deformation angle data to obtain collision energy time series change data; performing energy distribution simulation on the structural collision force distribution simulation data according to the collision energy time series change data to obtain collision energy structure distribution data;
[0150] Step S332: analyzing the fragile area of the vehicle body structure based on the structural collision force distribution simulation data to obtain the fragile area data of the vehicle body structure; performing a vehicle body structure dislocation simulation on the fragile area data of the vehicle body structure according to the collision energy structure distribution data to obtain the vehicle body structure dislocation simulation data;
[0151] Step S333: performing geometric collision interlacing influence analysis on the vehicle body structure misalignment simulation data to obtain geometric collision interlacing influence data; performing vehicle body structure collapse orientation simulation on the vehicle body structure misalignment simulation data according to the geometric collision interlacing influence data to obtain vehicle body structure collapse orientation simulation data;
[0152] Step S334: Calculate the vehicle body seat collapse orientation probability based on the vehicle body structure collapse orientation simulation data to obtain vehicle body seat collapse orientation probability data; identify the abnormal deformation angle of the vehicle body structure based on the vehicle body structure collapse orientation simulation data and the vehicle body seat collapse orientation probability data to obtain the abnormal deformation angle data of the vehicle body structure.
[0153] In the embodiment of the present invention, the structural collision time-series deformation angle data are collected and sorted to ensure the accuracy and completeness of the data, and the time interval and sampling frequency of the data are determined for subsequent analysis and processing. The structural collision time-series deformation angle data are used to calculate the collision energy at each time point. The kinetic energy formula in physics or the energy calculation algorithm designed according to the specific situation can be used to calculate the collision energy. The calculated collision energy data are subjected to time-series change analysis, the change trend of the collision energy over time is analyzed, the peak value, fluctuation and the existence of mutation points of the energy are identified, and the collision energy time-series change data are visualized for more intuitive understanding and analysis. The collision energy time-series change data are evaluated, and the characteristics and The energy distribution pattern is obtained by processing and analyzing the time series data of deformation angles to determine the energy change trend and peak value during the collision process. According to the collision energy time series change data, the energy distribution simulation is performed on the structural collision force distribution simulation data. The energy distribution of each part during the collision is described using the energy distribution simulation data. After completing the energy distribution simulation, the collision energy structural distribution data is obtained. These data describe the energy distribution of the vehicle structure during the collision process. Based on the structural collision force distribution simulation data, the fragile area of the vehicle body structure is analyzed to analyze the areas with concentrated force, high stress or other structural damage, and the fragile area of the vehicle body structure is determined. According to the collision energy structural distribution data, the fragile area data of the vehicle body structure is analyzed. Carry out body structure dislocation simulation to simulate the dislocation or deformation of the body structure caused by energy during the collision. After completing the dislocation simulation, obtain the body structure dislocation simulation data, which describe the dislocation or deformation of the body structure during the collision. Perform geometric collision staggered influence analysis on the body structure dislocation simulation data to analyze the dislocation of each part in the dislocation simulation data, consider the geometric collision staggered influence between different parts, such as overlap, extrusion, etc., and perform body structure collapse orientation simulation on the body structure dislocation simulation data according to the geometric collision staggered influence data to simulate the collapse or collapse orientation of the body structure under the influence of collision or staggering, including the part and shape. After completing the collapse orientation simulation, obtain the body structure. The body structure collapse orientation simulation data describes the body structure collapse orientation under specific collision or impact conditions, and the body seat collapse orientation probability is calculated for the body structure collapse orientation simulation data. The occurrence probabilities of different collapse orientations, the structural characteristics of the body structure, material strength and other factors are comprehensively considered to calculate the collapse probability of each orientation. The abnormal deformation angle of the body structure is identified based on the body structure collapse orientation simulation data and the body seat collapse orientation probability data, and the abnormal deformation angle of the body structure is analyzed under different collapse orientations and their probabilities, including rotation and twisting. After the abnormal deformation angle is identified, the abnormal deformation angle data of the body structure is obtained, which describes the abnormal deformation of the body structure under different collision or collapse conditions.
[0154] The present invention obtains the change of collision energy over time, i.e., collision energy time-series change data, by evaluating based on structural collision time-series deformation angle data. Based on the collision energy time-series change data, energy distribution simulation is performed on the structural collision force distribution simulation data, thereby obtaining the distribution of collision energy on the vehicle body structure, i.e., collision energy structural distribution data. This step helps to understand the change trend of collision energy over time and its distribution on the vehicle body structure, and provides basic data for subsequent structural fragile area analysis. Based on the structural collision force distribution simulation data, a fragile area analysis is performed on the vehicle body structure to determine the area that is subject to a greater impact, and obtain vehicle body structure fragile area data. According to the collision energy structural distribution data, a dislocation simulation is performed on the fragile area, i.e., the displacement and deformation of these areas in the case of a collision are analyzed to obtain vehicle body structure dislocation simulation data. This step helps to identify the area in the vehicle body structure that is most vulnerable to damage, and provides key information for further analysis. Geometric collision interleaving influence analysis is performed on the vehicle body structure dislocation simulation data. Analysis, that is, analyzing the impact of the misalignment and deformation of different parts on other parts, obtaining the geometric collision staggered influence data, and simulating the collapse direction of the body structure for the body structure misalignment simulation data according to the geometric collision staggered influence data, that is, analyzing the structural collapse that occurs in the case of a collision, and obtaining the body structure collapse direction simulation data. This step helps to evaluate the degree of damage to the body structure caused by the misalignment and deformation between different areas, and provides important information for safety assessment; analyzing the body structure collapse direction simulation data, calculating the probability of the body seat collapsing in different directions, and obtaining the body seat collapse direction probability data. This step helps to evaluate the safety of the vehicle passenger seat under different collision conditions, and provides a basis for vehicle design and safety system improvement. Based on the body structure collapse direction simulation data and the body seat collapse direction probability data, the abnormal deformation angle of the body structure is identified, that is, the abnormal structural deformation is identified. This step helps to timely discover the abnormal deformation of the body structure, and provides important support for vehicle structure safety monitoring.
[0155] Preferably, the deformation structure angle threat degree algorithm in step S34 is as follows:
[0156] ;
[0157] In the formula, The result value representing the threat degree of deformation angle, Indicates the number of angles that need to be calculated. Represented as an index, Represented as the first Angle value, Indicates the phase position of the abnormal deformation angle and the driving position, The overall angle difference representing the abnormal deformation angle, Represents the overall strength coefficient of the vehicle body structure, represents a natural constant, represents the material elastic modulus of the vehicle body structure, Indicates the error correction value of the deformable structure angle threat algorithm.
[0158] The present invention constructs an algorithm for the angle threat of a deformable structure. The algorithm evaluates the angle threat of a deformable structure by comprehensively considering factors such as the number of angles, angle value, phase position of the abnormal deformation angle and the driver's seat, the overall angle difference of the abnormal deformation angle, the overall strength coefficient of the vehicle body structure, and the elastic modulus of the material. The adjustment of each parameter has an impact on the final threat result. The algorithm fully considers the number of angles that need to be calculated. ,Increase The value of will increase the number of angles calculated, making the result more accurate. By considering more angles, the threat level of the deformed structure can be more comprehensively assessed. Angle value The size and change of each angle have an impact on the threat level of the deformed structure. By considering each angle value, the deformation of the structure can be evaluated more accurately; the phase position of the abnormal deformation angle and the driving position , which reflects the threat level of abnormal deformation to the driver. The value indicates that the greater the phase difference between the abnormal deformation angle and the driving position, the smaller the threat to the driver; the overall angle difference of the abnormal deformation angle , this parameter is used to measure the degree of abnormal deformation, the larger the The value indicates that the overall difference in deformation angle is large, indicating that the deformation of the structure is more serious; the overall strength coefficient of the body structure , this parameter reflects the resistance of the vehicle body structure, the larger The value indicates that the body structure is stronger and can withstand greater deformation without threatening the safety of the occupants; the natural constant , which is approximately equal to 2.71828, and appears in the formula for the material elastic modulus In the exponential function of , adjust the influence of the material elastic modulus on the threat result; the material elastic modulus of the body structure The elastic modulus of a material reflects its stiffness and resistance to deformation. The value indicates that the material is more rigid and has a stronger resistance to deformation; the error correction value of the deformation structure angle threat algorithm , this value is used to correct the formula to improve the accuracy of the calculation results.
[0159] Preferably, the present invention further provides a deep learning-based automobile structure stress anomaly analysis system, which is used to execute the deep learning-based automobile structure stress anomaly analysis method as described above. The deep learning-based automobile structure stress anomaly analysis method includes:
[0160] A three-dimensional automobile structure model building module is used to obtain automobile structure design data; construct a three-dimensional automobile structure model based on the automobile structure design data to obtain a three-dimensional automobile structure model;
[0161] The connection point force numerical component analysis module is used to perform connection point force bearing analysis based on the three-dimensional automobile structure model to obtain connection point force bearing data; perform automobile structure collision force distribution simulation based on the connection point force bearing data to obtain structure collision force distribution simulation data; perform different connection point force numerical component analysis on the connection point force bearing data based on the structure collision force distribution simulation data to obtain connection point force numerical component data;
[0162] The vehicle body abnormal structure analysis model construction module is used to simulate the collision deformation of the vehicle structure according to the numerical component data of the force at the connection point to obtain the structural collision deformation simulation data; identify the abnormal deformation angle of the vehicle body structure according to the structural collision deformation simulation data to obtain the abnormal deformation angle data of the vehicle body structure; calculate the deformation angle threat degree of the abnormal deformation angle data of the vehicle body structure to obtain the deformation angle threat degree data; construct the vehicle body abnormal structure analysis model based on the deformation angle threat degree data based on the multi-layer perceptron to obtain the vehicle body abnormal structure analysis model;
[0163] The vehicle structure force optimization and adjustment module is used to perform vehicle structure force abnormality analysis based on the vehicle body abnormal structure analysis model to obtain vehicle structure force abnormality data; perform vehicle structure force optimization and adjustment based on the vehicle structure force abnormality data to obtain vehicle structure force optimization and adjustment data.
[0164] The present invention obtains automobile structural design data and converts it into a three-dimensional automobile structural model, so that designers and engineers can more clearly visualize the entire automobile structure, including the appearance and internal components. By constructing a three-dimensional model, the design team can better evaluate the structural design of the automobile, discover potential design defects or optimization space, and make adjustments and improvements in advance. Analyzing the stress conditions of the connection points is helpful to evaluate the strength and stability of the automobile structure in actual use and collision conditions. Through the stress bearing analysis, existing fatigue or overload problems can be identified, providing an important basis for ensuring the safety of the automobile. The stress data of key connection points or components found in the analysis can be used to guide the optimization and improvement of the structure to improve the overall performance and durability. By simulating the stress distribution of the automobile structure, the stress conditions of various parts of the whole vehicle can be fully understood, and potential weaknesses or high-risk areas can be identified. The analysis results can be used to predict the performance of the automobile under different working conditions, including vibration, stress and deformation. By analyzing the numerical components of the stress at the connection points, the load conditions of various parts of the structure can be deeply understood, providing a basis for local design adjustments. By simulating the deformation of the vehicle under collision conditions, it is helpful to evaluate the safety performance of the vehicle when an accident occurs, including The simulation data can be used to identify the area that causes abnormal deformation of the body structure, thereby guiding the optimization and improvement of the structure and improving the protection ability of the vehicle in the event of a collision. By identifying the abnormal deformation angle of the body structure, the abnormal deformation of the vehicle in the accident can be discovered in time, which is helpful to take repair measures or adjust the design in advance. The analysis of the abnormal deformation angle data of the body structure can provide basic data support for the subsequent calculation of the threat degree of the deformation angle. The calculation of the threat degree of the deformation angle helps to assess the potential threat degree of abnormal deformation to the vehicle structure and occupant safety, and provide a basis for risk management. The threat degree data can be used to guide the optimal design of the body structure, reduce the threat of deformation angle, and improve the safety performance of the vehicle in accidents. The abnormal structure analysis model built based on the multi-layer perceptron can be used as an early warning system for vehicle structural abnormalities, timely monitor and respond to changes in the body structure, and ensure the safety of the vehicle during use. Analysis of abnormal structural force data helps to monitor the health of the vehicle structure, timely discover and solve potential force problems, and improve the durability and stability of the structure. According to the abnormal analysis results, the vehicle structure can be optimized and adjusted to reduce the risk of abnormal structural force and improve the reliability and safety of the vehicle. Therefore, the present invention optimizes the traditional method of analyzing abnormal stress of automobile structure based on deep learning, solves the problems of inaccurate analysis of abnormal stress of automobile structure and unclear analysis of threat to personnel caused by deformation of automobile structure during collision, improves the accuracy of analysis of abnormal stress of automobile structure and clarifies the problem of threat degree of deformation of automobile structure to personnel during collision.
[0165] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.
[0166] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.
Claims
1. A method for analyzing abnormal stress of automobile structure based on deep learning, characterized in that: The following steps are involved: Step S1: Acquire automobile structure design data; The automobile structure design data includes: design parameters and geometric information of the body, chassis, doors, and engine compartment components; constructing a three-dimensional automobile structure model based on the automobile structure design data to obtain a three-dimensional automobile structure model; Step S2: performing a connection point force bearing analysis according to the three-dimensional automobile structure model to obtain connection point force bearing data; performing a vehicle structure force distribution simulation according to the connection point force bearing data to obtain structure collision force distribution simulation data; performing a force numerical component analysis of different connection points on the connection point force bearing data according to the structure collision force distribution simulation data to obtain connection point force numerical component data; wherein step S2 includes: Step S21: performing structural connection unit analysis according to the three-dimensional automobile structure model to obtain structural connection unit data; performing structural connection point extraction on the structural connection unit data to obtain structural connection point data; Step S22: performing connection point force bearing analysis on the structural connection point data according to the three-dimensional automobile structure model to obtain connection point force bearing data; Step S23: simulating the force distribution of the vehicle structure according to the three-dimensional vehicle structure model and the force bearing data of the connection points to obtain the structural collision force distribution simulation data; wherein step S23 includes: Step S231: performing force diffusion orientation analysis on the force bearing data of the connection point to obtain the force diffusion orientation data of the connection point; performing structural section stress analysis based on the force diffusion orientation data of the connection point and the force bearing data of the connection point to obtain structural section stress data; Step S232: performing a vehicle structure stress simulation according to the three-dimensional vehicle structure model and the structural section stress data to obtain structural stress simulation data; performing a local structure stress change analysis on the structural section stress data according to the vehicle structure stress simulation data to obtain local structure stress change data; Step S233: performing instantaneous force diffusion topological structure analysis on the force diffusion orientation data of the connection points according to the local structure force change data to obtain instantaneous force diffusion topological structure data; Step S234: simulating the force distribution of the automobile structure according to the structural force simulation data and the instantaneous force diffusion topological structure data to obtain structural collision force distribution simulation data; Step S24: performing force numerical component analysis of different connection points on the connection point force bearing data according to the structural collision force distribution simulation data to obtain the connection point force numerical component data; Step S3: Carry out a collision deformation simulation of the vehicle structure according to the numerical component data of the force at the connection point to obtain structural collision deformation simulation data; identify the abnormal deformation angle of the vehicle body structure according to the structural collision deformation simulation data to obtain abnormal deformation angle data of the vehicle body structure; calculate the deformation angle threat degree of the abnormal deformation angle data of the vehicle body structure to obtain deformation angle threat degree data; construct a vehicle body abnormal structure analysis model based on the deformation angle threat degree data based on a multi-layer perceptron to obtain a vehicle body abnormal structure analysis model; Step S4: performing abnormal stress analysis on the vehicle structure based on the abnormal structural analysis model of the vehicle body to obtain abnormal stress data on the vehicle structure; performing optimized stress adjustment on the vehicle structure according to the abnormal stress data on the vehicle structure to obtain optimized stress adjustment data on the vehicle structure.
2. The method for analyzing abnormal stress of automobile structure based on deep learning according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: Acquire automobile structural design data; wherein the automobile structural design data includes: design parameters and geometric information of the vehicle body, chassis, door, and engine compartment components; Step S12: performing structural grid division on the automobile structure design data to obtain automobile structure grid data; Step S13: constructing a three-dimensional automobile structure model for the automobile structure design data according to the automobile structure meshing data to obtain a three-dimensional automobile structure model.
3. The method for analyzing abnormal stress of automobile structure based on deep learning according to claim 2 is characterized in that: Step S24 includes the following steps: Step S241: extracting force distribution data of different connection points from the force bearing data of the connection points according to the structural collision force distribution simulation data, and obtaining the force distribution data of the connection points; Step S242: performing connection point force vector analysis according to the connection point force distribution data to obtain the connection point force vector; performing decomposed force calculation on the connection point force vector to obtain connection point decomposed force data; Step S243: performing calculation of adjacent connection point decomposition force variance on the connection point decomposition force data to obtain adjacent connection point decomposition force variance values; performing non-uniform variance decomposition force data sampling on the connection point decomposition force data according to the adjacent connection point decomposition force variance values to obtain non-uniform variance decomposition force sampling data; Step S244: performing non-uniform numerical approximation processing according to the non-uniform variance decomposition force sampling data to obtain non-uniform decomposition force numerical approximation data; Step S245: performing order-limited piecewise linear interpolation processing on the numerical approximation data of the non-uniform decomposition force to obtain decomposition force order-limited interpolation data; Step S246: performing force numerical component analysis of different connection points on the connection point decomposition force data according to the non-uniform decomposition force numerical approximation data and the decomposition force order limited interpolation data to obtain the force numerical component data of the connection points.
4. The method for analyzing abnormal stress of automobile structure based on deep learning according to claim 3 is characterized in that: The order-limited piecewise linear interpolation processing of the numerical approximation data of the non-uniform decomposition force includes the following steps: Perform local feature clustering analysis on the numerical approximation data of non-uniform decomposition force to obtain local clustering data of decomposition force; According to the local clustering data of decomposition force, the data of numerical approximation of non-uniform decomposition force is clustered and divided to obtain the clustering data of numerical approximation of decomposition force; Based on the numerical approximation of decomposition power, clustering data is divided to perform order matching restrictions on different data categories to obtain order matching restricted data; According to the order matching limited data, the numerical approximation data of the non-uniform decomposition force are subjected to order limited piecewise linear interpolation processing to obtain the decomposition force order limited interpolation data.
5. The method for analyzing abnormal stress of automobile structure based on deep learning according to claim 4 is characterized in that: Step S3 includes the following steps: Step S31: performing a vehicle structure collision deformation simulation according to the connection point force numerical component data, the structure collision force distribution simulation data and the connection point force bearing data to obtain the structure collision deformation simulation data; Step S32: performing deformation time series analysis on the structural collision deformation simulation data to obtain structural collision deformation time series data; performing time series deformation angle analysis on the structural collision deformation time series data to obtain structural collision deformation angle data; Step S33: identifying abnormal deformation angles of the vehicle body structure according to the structural collision time-series deformation angle data and the structural collision force distribution simulation data, and obtaining abnormal deformation angle data of the vehicle body structure; Step S34: using a deformation structure angle threat degree algorithm to calculate the deformation angle threat degree of the abnormal deformation angle data of the vehicle body structure to obtain deformation angle threat degree data; Step S35: constructing a vehicle body abnormal structure analysis model based on the deformation angle threat degree data and the vehicle body structure abnormal deformation angle data to obtain a vehicle body abnormal structure analysis model.
6. The method for analyzing abnormal stress of automobile structure based on deep learning according to claim 5 is characterized in that: Step S33 includes the following steps: Step S331: performing a collision energy time series change evaluation based on the structural collision time series deformation angle data to obtain collision energy time series change data; performing energy distribution simulation on the structural collision force distribution simulation data according to the collision energy time series change data to obtain collision energy structure distribution data; Step S332: analyzing the fragile area of the vehicle body structure based on the structural collision force distribution simulation data to obtain the fragile area data of the vehicle body structure; performing a vehicle body structure dislocation simulation on the fragile area data of the vehicle body structure according to the collision energy structure distribution data to obtain the vehicle body structure dislocation simulation data; Step S333: performing geometric collision interlacing influence analysis on the vehicle body structure misalignment simulation data to obtain geometric collision interlacing influence data; performing vehicle body structure collapse orientation simulation on the vehicle body structure misalignment simulation data according to the geometric collision interlacing influence data to obtain vehicle body structure collapse orientation simulation data; Step S334: Calculate the vehicle body seat collapse orientation probability based on the vehicle body structure collapse orientation simulation data to obtain vehicle body seat collapse orientation probability data; identify the abnormal deformation angle of the vehicle body structure based on the vehicle body structure collapse orientation simulation data and the vehicle body seat collapse orientation probability data to obtain the abnormal deformation angle data of the vehicle body structure.
7. The method for analyzing abnormal stress of automobile structure based on deep learning according to claim 6 is characterized in that: The algorithm for the threat degree of the deformable structure angle in step S34 is as follows: ; In the formula, The result value representing the threat degree of deformation angle, Indicates the number of angles that need to be calculated. Represented as an index, Represented as the first Angle value, Indicates the phase position of the abnormal deformation angle and the driving position, The overall angle difference representing the abnormal deformation angle, Represents the overall strength coefficient of the vehicle body structure, represents a natural constant, represents the material elastic modulus of the vehicle body structure, Indicates the error correction value of the deformable structure angle threat algorithm.
8. A deep learning-based automobile structure stress anomaly analysis system, used to execute the deep learning-based automobile structure stress anomaly analysis method according to claim 7, characterized in that: The deep learning-based automobile structure stress anomaly analysis system includes: A three-dimensional automobile structure model building module is used to obtain automobile structure design data; construct a three-dimensional automobile structure model based on the automobile structure design data to obtain a three-dimensional automobile structure model; The connection point force numerical component analysis module is used to perform connection point force bearing analysis based on the three-dimensional automobile structure model to obtain connection point force bearing data; perform automobile structure collision force distribution simulation based on the connection point force bearing data to obtain structure collision force distribution simulation data; perform different connection point force numerical component analysis on the connection point force bearing data based on the structure collision force distribution simulation data to obtain connection point force numerical component data; The vehicle body abnormal structure analysis model construction module is used to simulate the collision deformation of the vehicle structure according to the numerical component data of the force at the connection point to obtain the structural collision deformation simulation data; identify the abnormal deformation angle of the vehicle body structure according to the structural collision deformation simulation data to obtain the abnormal deformation angle data of the vehicle body structure; calculate the deformation angle threat degree of the abnormal deformation angle data of the vehicle body structure to obtain the deformation angle threat degree data; construct the vehicle body abnormal structure analysis model based on the deformation angle threat degree data based on the multi-layer perceptron to obtain the vehicle body abnormal structure analysis model; The vehicle structure force optimization and adjustment module is used to perform vehicle structure force abnormality analysis based on the vehicle body abnormal structure analysis model to obtain vehicle structure force abnormality data; perform vehicle structure force optimization and adjustment based on the vehicle structure force abnormality data to obtain vehicle structure force optimization and adjustment data.
