Image generation method and system based on visual analysis

Abstract

The application relates to the field of image processing and computer-aided technology, and discloses an image generation method and system based on visual analysis. The method can accurately identify the geometric shape of a component by performing deep visual analysis on a non-standardized two-dimensional design input image, and can further deduce macroscopic physical property description information, function positioning, use environment and performance requirements of the component, and deduce accurate physical parameters on the basis, and finally generate a three-dimensional entity model containing rich physical properties. Therefore, the application can effectively solve the problems of non-standardized image processing difficulty and physical property information loss in the prior art, and significantly improves the application value of the three-dimensional model in virtual simulation and the efficiency of engineering design.

Description

Technical Field

[0001] This invention relates to the fields of image processing and computer-aided design, and more specifically, to an image generation method and system based on visual analysis. Background Technology

[0002] In the field of mechanical product design and manufacturing, it has become common practice to use computer systems to analyze two-dimensional design input images and generate three-dimensional solid models. These systems can receive standardized planar drawings provided by engineers, such as front, side, and sectional views of parts. By identifying geometric features, they automatically generate corresponding three-dimensional solid models, which are then imported into a virtual simulation environment for assembly checks, motion analysis, or preliminary strength assessments, significantly improving the efficiency of concept verification in the early stages of design.

[0003] However, in actual design work, especially in the conceptual design phase or when rapid modifications to existing products are required, engineers often provide non-standardized, low-precision 2D design input images, such as hand-drawn sketches, scanned old blueprints, or printouts with arbitrary annotations. These non-standardized visual inputs may have problems such as unclear or discontinuous lines, typos, stains, creases, and non-standard or unclear dimensions. When existing systems process such images, their geometric feature recognition capabilities face challenges, potentially leading to misjudgments or failure to correctly connect geometric elements. This results in 3D models with geometric deviations or even failing to form a valid entity, forcing engineers to spend a significant amount of time on manual corrections, ultimately reducing overall efficiency.

[0004] Even if the system solves the basic geometric shape recognition problem through image preprocessing and deep learning techniques, and can extract relatively accurate geometric information from blurry and incomplete visual input to generate roughly correct 3D geometries, these 2D drawings or sketches, even with relatively complete geometric information, rarely contain detailed information about key physical properties such as material properties, surface treatment, manufacturing tolerances, or assembly fits. For example, a sketch may clearly depict the geometry of a gear, but it won't directly state its material, hardness requirements, or fit precision. When generating a 3D model, the system can only obtain a pure geometry, lacking these physical parameters that are crucial for subsequent mechanical simulations, thermal analysis, fatigue life prediction, or assembly clearance analysis. This lack of information at the physical property level significantly reduces the application value of the generated 3D model in a virtual simulation environment. Because without specific physical properties, the simulation results will be generalized and unable to reflect the performance of the real product, thus failing to provide engineers with the refined and quantitative analysis results they require.

[0005] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0006] This invention provides an image generation method and system based on visual analysis, aiming to solve the problems in the prior art where it is difficult to accurately identify geometric features and generate effective three-dimensional models when processing non-standardized two-dimensional design input images, and where the generated three-dimensional models lack key physical attribute information, thus limiting their application value in virtual simulation environments.

[0007] The technical solution of this application is as follows: In a first aspect, this application discloses an image generation method based on visual analysis, comprising the following steps: Acquire non-standardized two-dimensional design input images; Image preprocessing is performed on non-standardized two-dimensional design input images, and two-dimensional geometric contour features are extracted from the preprocessed images; Based on the two-dimensional geometric contour features, identify the geometric shape of the component and the macroscopic physical property description information of the component; Based on the geometric shape, construct the initial three-dimensional geometric model of the component; Based on the description information of geometric shape and macroscopic physical properties, determine the functional positioning and usage environment of the component; Based on functional positioning and usage environment, obtain the performance requirements of components and rank the performance requirements by importance; Based on functional positioning, usage environment, and performance requirements, the physical parameters of the components are derived. Verify the consistency and engineering rationality between physical parameters, and adjust the physical parameters based on the verification results; The adjusted physical parameters are integrated into the initial 3D geometric model to generate a 3D solid model for virtual simulation, and the 3D solid model is used as input for virtual simulation.

[0008] This technical solution can effectively process non-standardized two-dimensional design input images. It can not only accurately identify geometric features and construct an initial three-dimensional model, but also further derive and integrate key physical parameters to generate a three-dimensional solid model with physical properties. This significantly improves the accuracy and application value of virtual simulation and solves the problem of lack of physical property information in existing three-dimensional models.

[0009] Furthermore, this application proposes that the steps for deriving the physical parameters of a component based on its functional positioning, usage environment, and performance requirements include: Based on macroscopic physical property description information, functional positioning, usage environment and performance requirements, the initial mechanical parameters, initial physical parameters and initial surface residual compressive stress of the components are extracted from the engineering database. The working cycle conditions of the component under the service environment are analyzed, including the amplitude of the cyclic load, the frequency of the cyclic load, the range of the periodic temperature fluctuation, and the frequency of the periodic temperature fluctuation. Based on the analyzed cyclic load and periodic temperature fluctuations, the relaxation process of residual compressive stress inside the component with the number of working cycles is predicted, and the evolution result of residual stress relaxation is obtained. Based on the results of residual stress relaxation evolution, the fatigue limit of the component is adjusted to reflect the decrease in the component's fatigue resistance during long-term operation. Verify whether the adjusted fatigue limit meets the performance requirements of the component.

[0010] This technical solution enables a more accurate derivation of the physical parameters of components by comprehensively considering multiple factors. In particular, by predicting the residual stress relaxation process and adjusting the fatigue limit, the assessment of the fatigue resistance of components during long-term operation becomes more accurate, thereby improving the reliability and engineering applicability of the physical parameter derivation.

[0011] More specifically, in some implementation schemes, the steps for verifying whether the adjusted fatigue limit meets the performance requirements of the component include: Set a safe threshold for the fatigue limit; When the adjusted fatigue limit is lower than the safety threshold, a fatigue risk warning is issued; Based on fatigue risk warning, the current damage of the component is calculated, and the failure time or failure location of the component is predicted based on the current damage.

[0012] This technical solution enables the timely detection of potential fatigue failure risks by setting safety thresholds and providing fatigue risk warnings. It also allows for the prediction of failure time and location, providing crucial decision-making support for component maintenance and replacement and effectively preventing sudden failures.

[0013] Based on the above, this application further proposes that the steps for verifying whether the adjusted fatigue limit meets the performance requirements of the component include: Obtain the adjusted fatigue limit, initial mechanical parameters, and discrete distribution information of initial physical parameters; Based on discrete distribution information, generate multiple parameter combination samples of the component; For each parameter combination sample in the multiple parameter combination samples, verify whether the fatigue limit, mechanical parameters and physical parameters in the parameter combination sample meet the performance requirements of the component. Based on the verification results of all parameter combination samples, calculate the probability that the component meets the performance requirements and output the probability value.

[0014] This technical solution enables the quantitative evaluation of component performance satisfaction through probabilistic statistical methods, thereby providing more robust decision support for design and manufacturing and reducing the risks associated with uncertainty.

[0015] Preferably, the steps of extracting the initial mechanical parameters, initial physical parameters, and initial surface residual compressive stress of a component from the engineering database based on macroscopic physical property description information, functional positioning, usage environment, and performance requirements include: Identify the type of manufacturing process used for the component; Based on the type of manufacturing process, identify the characteristics of the impact of the manufacturing process on the surface or near-surface structure of the material. Based on the influence characteristics, the initial surface residual compressive stress distribution of the component is derived; Based on the influence of manufacturing processes on the microstructure of materials, the initial mechanical and physical parameters are corrected.

[0016] This technical solution allows for more accurate acquisition of the initial mechanical parameters, physical parameters, and surface residual compressive stress of components by considering the impact of manufacturing processes on material properties, thereby improving the accuracy of subsequent physical parameter derivation.

[0017] In one embodiment, the step of deriving the initial surface residual compressive stress distribution of a component based on its influencing characteristics includes: Identify the geometric features of the components; Based on the identified geometric features, the surface regions of the component are divided, where the surface regions include geometric feature regions and non-geometric feature regions. For geometric feature regions, the initial surface residual compressive stress concentration distribution of the geometric feature regions is derived based on the manufacturing process type and local geometric parameters of the geometric features; For non-geometric feature regions, the initial surface residual compressive stress gradient distribution of the non-geometric feature regions is derived based on the manufacturing process type and the macroscopic geometric parameters of the non-geometric feature regions. By integrating the initial surface residual compressive stress concentration distribution in geometric feature regions with the initial surface residual compressive stress gradient distribution in non-geometric feature regions, a non-uniform distribution of initial surface residual compressive stress is obtained for the component.

[0018] This technical solution can distinguish between geometric feature regions and non-geometric feature regions, and derive their residual compressive stress distributions separately. Finally, it can be fused to obtain the non-uniform distribution of the initial surface residual compressive stress of the component, thus reflecting the actual situation of the residual stress on the component surface more accurately.

[0019] In another embodiment, the steps of predicting the relaxation process of residual compressive stress inside the component with the number of working cycles based on the analyzed cyclic load and periodic temperature fluctuations, and obtaining the residual stress relaxation evolution result, include: Identify the types, sizes, spatial distributions, and orientation distribution characteristics of micro-defects within the component; Based on the identified micro-defect types, sizes, spatial distributions, and grain orientation characteristics, the disturbance effect of micro-defects or grain orientation inhomogeneities on the local stress field of the component is determined. Based on the perturbation effect, the constitutive response of the material in the local area of ​​the component is modified to deduce the accelerating or inhibiting effect of micro-defects or grain orientation inhomogeneity on the relaxation rate of local residual compressive stress in the component. By integrating the accelerating or inhibiting effects into the relaxation prediction of residual compressive stress inside the component, a residual compressive stress relaxation process that takes into account microscopic inhomogeneities is obtained.

[0020] This technical solution can more accurately predict the relaxation process of residual compressive stress by taking into account the microscopic non-uniformity inside the component, thereby improving the accuracy of fatigue life prediction.

[0021] As an optional approach, the steps for adjusting the fatigue limit of the component based on the residual stress relaxation evolution results include: Acquire information on temperature changes and load frequency changes in the operating environment of the component during operation; The residual stress relaxation rate of the component is adjusted based on temperature change information and load frequency change information. Based on the corrected residual stress relaxation rate, the current residual stress state of the component is recalculated; Based on the current residual stress state and the material fatigue damage accumulation characteristics of the component, the fatigue limit of the component is dynamically adjusted.

[0022] This technical solution enables the real-time acquisition of working environment parameters and dynamic adjustment of fatigue limits, making the assessment of component fatigue performance closer to actual operating conditions, thereby improving the accuracy and reliability of predictions.

[0023] To enhance functionality, the steps for dynamically adjusting the fatigue limit of a component based on its current residual stress state and the cumulative fatigue damage characteristics of the component's material include: The real-time service environment parameters of the component during operation are obtained, including temperature, humidity, concentration of corrosive medium and load spectrum changes. Assess the current environmental damage level of component materials based on real-time service environment parameters; Based on the current level of environmental damage, modify the material fatigue damage accumulation characteristics of the component; Based on the modified material fatigue damage accumulation characteristics and combined with the current residual stress state, the fatigue limit of the component is dynamically adjusted.

[0024] This technical solution can further improve the accuracy of dynamic adjustment of fatigue limit by taking into account the influence of real-time service environment parameters on the cumulative characteristics of material fatigue damage, thus making the prediction of component fatigue life more precise.

[0025] Secondly, this application also discloses an image generation system based on visual analysis, comprising: The image acquisition module is used to acquire non-standardized two-dimensional design input images; The image preprocessing module is used to preprocess non-standardized two-dimensional design input images and extract two-dimensional geometric contour features from the preprocessed images. The geometric recognition module is used to identify the geometric shape of a component and its macroscopic physical property description information based on two-dimensional geometric contour features. The 3D model building module is used to construct the initial 3D geometric model of the component based on its geometric shape. The functional environment determination module is used to determine the functional positioning and usage environment of a component based on its geometric shape and macroscopic physical property description information. The performance requirement processing module is used to obtain the performance requirements of components based on their functional positioning and usage environment, and to prioritize the performance requirements. The physical parameter derivation module is used to derive the physical parameters of components based on functional positioning, usage environment, and performance requirements. The parameter verification and adjustment module is used to verify the consistency and engineering rationality between physical parameters, and to adjust the physical parameters based on the verification results. The model generation output module is used to integrate the adjusted physical parameters into the initial three-dimensional geometric model to generate a three-dimensional solid model for virtual simulation, and to use the three-dimensional solid model as input for virtual simulation.

[0026] This technical solution provides an integrated system solution that, through modular design, enables fully automated generation of a three-dimensional solid model with physical properties from non-standardized two-dimensional design input images, significantly improving design efficiency and the accuracy of virtual simulation. Beneficial effects

[0027] The image generation method and system based on visual analysis disclosed in this application can effectively address the challenges posed by low-precision and non-standard images such as hand-drawn sketches and scanned old blueprints by acquiring non-standard two-dimensional design input images, preprocessing them, and extracting features. It solves the problems of difficulty in geometric feature recognition and easy deviation in existing systems when processing such images.

[0028] Building upon this foundation, this application further constructs an initial three-dimensional geometric model of the component based on the identified geometric shape and macroscopic physical property description information. More importantly, by determining the component's functional positioning and usage environment, performance requirements are obtained and prioritized, thereby deriving the component's physical parameters. These physical parameters, after consistency and engineering rationality verification and adjustment, are integrated into the initial three-dimensional geometric model, ultimately generating a three-dimensional solid model for virtual simulation. This three-dimensional solid model not only possesses accurate geometric information but also includes key physical parameters such as material properties and surface treatments, overcoming the shortcomings of existing technologies where the generated three-dimensional models lack physical property information, significantly reducing their application value in virtual simulation environments.

[0029] Therefore, the method of this application can provide a three-dimensional solid model with complete geometric and physical properties, which greatly improves the accuracy and practicality of virtual simulation in mechanical simulation, thermal analysis, fatigue life prediction, etc., and provides engineers with refined and quantitative analysis results, significantly improving the efficiency and quality of design verification. Attached Figure Description

[0030] Figure 1 This is a flowchart illustrating an image generation method based on visual analysis provided in an embodiment of the present invention; Figure 2 This is a flowchart of a method for deriving the physical parameters of a component according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an image generation system based on visual analysis provided in an embodiment of the present invention. Detailed Implementation

[0031] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0032] Reference Figure 1 , Figure 1 This is a flowchart illustrating an image generation method based on visual analysis provided in an embodiment of the present invention, including: S11, Obtain non-standardized two-dimensional design input image; S12, perform image preprocessing on the non-standard two-dimensional design input image, and extract two-dimensional geometric contour features from the preprocessed image; S13, Based on the two-dimensional geometric contour features, identify the geometric shape of the component and the macroscopic physical property description information of the component; S14, Based on the geometric shape, construct the initial three-dimensional geometric model of the component; S15, Based on the geometric shape and the macroscopic physical property description information, determine the functional positioning and usage environment of the component; S16, Based on the functional positioning and the usage environment, obtain the performance requirements of the component, and rank the performance requirements by importance; S17. Based on the functional positioning, the usage environment, and the performance requirements, derive the physical parameters of the component; S18, verify the consistency and engineering rationality between the physical parameters, and adjust the physical parameters according to the verification results; S19, the adjusted physical parameters are integrated into the initial three-dimensional geometric model to generate a three-dimensional solid model for virtual simulation, and the three-dimensional solid model is used as the input for virtual simulation.

[0033] This application utilizes deep visual analysis of non-standardized two-dimensional design input images to accurately identify the geometric shape of components. Furthermore, it derives macroscopic physical property descriptions, functional positioning, usage environment, and performance requirements, and on this basis, derives precise physical parameters, ultimately generating a three-dimensional solid model containing rich physical properties. Therefore, this application effectively solves the problems of difficult non-standardized image processing and missing physical property information in existing technologies, significantly improving the application value of three-dimensional models in virtual simulation and the efficiency of engineering design.

[0034] In order to better understand the technical solutions proposed in this application, it is necessary to explain some key terms involved therein.

[0035] "Non-standard 2D design input images" refer to images that do not conform to traditional engineering drawing specifications, such as hand-drawn sketches, scanned old drawings, and printouts with arbitrary annotations. These images may have problems such as unclear or discontinuous lines, non-standard or blurry dimensioning.

[0036] "Two-dimensional geometric contour features" refer to the basic geometric elements extracted from two-dimensional images that are used to describe the shape and size of components, such as straight lines, arcs, curves, and corner points.

[0037] "The geometric form of a component" refers to the overall shape, size proportions, and relative positional relationships between its parts.

[0038] "Macrophysical property description information" refers to non-quantitative or semi-quantitative information related to the physical properties of a component, identified from an image. Examples include descriptive terms such as "high strength," "wear resistance," and "lightweight," or information implied through symbols or text labels, such as material type and surface treatment method.

[0039] "Functional positioning" refers to the role and main function of a component in the whole product or system, such as "support component", "transmission component", "sealing component", etc.

[0040] "Usage environment" refers to the external conditions under which a component operates, such as temperature, humidity, load type, and corrosive media.

[0041] "Performance requirements" refer to the technical indicators that a component needs to meet under specific functional positioning and usage environment, such as strength, stiffness, fatigue life, corrosion resistance, etc.

[0042] "Physical parameters" refer to specific numerical values ​​used to quantitatively describe the physical properties of a component, such as the elastic modulus, yield strength, density, coefficient of thermal expansion, and surface roughness of a material.

[0043] "Initial 3D geometric model" refers to a 3D model that only contains geometric shape information and has not yet integrated detailed physical parameters.

[0044] A "3D solid model" refers to a complete 3D model that integrates geometric information and physical parameters and can be directly used for virtual simulation.

[0045] The image generation method based on visual analysis proposed in this application is essentially about transforming a non-standardized two-dimensional design input image into a three-dimensional solid model with complete physical properties through a series of refined processing steps.

[0046] First, it is necessary to acquire non-standardized two-dimensional design input images. This can be achieved in several ways. For example, users can digitize hand-drawn sketches or old blueprints using a scanner to generate bitmap or PDF image files; alternatively, the system can directly receive digital sketches drawn in real time by engineers from devices such as digital drawing tablets or tablets; or the system can load existing non-standardized image files from network storage or the local file system.

[0047] Secondly, the acquired non-standardized two-dimensional design input image undergoes image preprocessing, and two-dimensional geometric contour features are extracted from the preprocessed image. Image preprocessing aims to eliminate interference such as noise, blur, creases, and stains in the image, improving image quality for subsequent feature extraction. For example, Gaussian filtering and median filtering can be used for noise reduction; image enhancement techniques, such as histogram equalization, can be used to improve image contrast; and geometric correction can be performed on tilted or deformed images. After preprocessing, edge detection algorithms (such as the Canny operator and Sobel operator) can be used to identify edges in the image, and straight lines and arcs can be detected using methods such as Hough transform, thereby extracting the two-dimensional geometric contour features of the components. For example, for a hand-drawn gear sketch, preprocessing can make its lines clearer, and then edge detection and contour tracking can accurately extract the gear's tooth shape, addendum circle, and dedendum circle.

[0048] Next, based on the extracted two-dimensional geometric contour features, the geometric shape of the component and its macroscopic physical property description information are identified. Geometric shape identification can be achieved by matching the extracted contour features with a preset geometric template, or by using deep learning models (such as convolutional neural networks) to classify and identify the contour features, thereby determining whether the component is circular, square, polygonal, and its specific dimensional proportions. For example, the component might be identified as a circular flange with a diameter of 50mm and six evenly distributed bolt holes with a diameter of 10mm. The identification of macroscopic physical property description information requires a combination of natural language processing (NLP) and image recognition technologies. For example, the system can identify text labels in an image, such as "high-strength steel" or "corrosion-resistant coating," and parse them into macroscopic physical property description information; or, by identifying specific symbols, colors, or textures in an image, the system can infer the physical properties that the component may possess; for example, red might indicate a high-temperature component, and a rough texture might indicate a casting.

[0049] Then, based on the identified geometric shape, an initial three-dimensional geometric model of the component is constructed. This step transforms two-dimensional geometric information into a three-dimensional representation. For example, if the component is identified as a circular outline, the system can generate a cylinder or disk based on its diameter and a preset thickness through an extrusion operation. If multiple interrelated two-dimensional outlines are identified, the system can construct more complex three-dimensional geometries through operations such as rotation, sweeping, and lofting. For example, for an identified flange outline, it can be extruded into a three-dimensional flange model with a certain thickness, and corresponding holes can be generated on the three-dimensional model based on the two-dimensional position and diameter of the bolt holes.

[0050] Subsequently, based on the identified geometric shape and macroscopic physical property descriptions, the functional positioning and usage environment of the component are determined. Functional positioning can be determined by reasoning based on the component's geometric features and macroscopic physical property descriptions. For example, a cylindrical component with a threaded hole, whose macroscopic physical property description includes the word "connection," might be positioned as a "connecting component"; a component with a heat sink structure might be positioned as a "heat dissipation component." The usage environment can be determined by inference based on the macroscopic physical property descriptions. For example, if the macroscopic physical property descriptions include terms such as "high temperature resistance" or "corrosion resistance," it can be inferred that its usage environment may involve high-temperature or corrosive media.

[0051] Next, based on the defined functional positioning and usage environment, the performance requirements of the components are obtained and prioritized. Performance requirements can be obtained by querying a pre-defined engineering specification database or expert knowledge base. For example, for a component positioned as a "transmission component" and used in a "high-speed operation" environment, its performance requirements might include "high strength," "high wear resistance," and "low vibration." These performance requirements can be prioritized according to their criticality to the component's functional realization and safe operation; for example, in some applications, strength may be more important than wear resistance.

[0052] Furthermore, based on functional positioning, usage environment, and performance requirements, the physical parameters of the component are derived. This step is one of the key innovations of this application. For example, if the component's functional positioning is "load-bearing component," its usage environment is "high temperature and high pressure," and its performance requirements are "high strength and long life," then the system can use this information to select materials that meet the conditions from the material database and derive their specific physical parameters such as elastic modulus, yield strength, fatigue limit, and coefficient of thermal expansion. The derivation process can utilize machine learning models, learning the relationship between component attributes and physical parameters from a large number of engineering cases, to achieve the transformation from qualitative description to quantitative parameters.

[0053] Then, the consistency and engineering rationality of the derived physical parameters are verified, and the physical parameters are adjusted according to the verification results. For example, if the derived material density and strength parameters do not match, or if some parameters exceed the range of conventional engineering materials, adjustments are required. Verification can be achieved by comparing with engineering specifications, material handbooks, or rules of thumb. If inconsistencies or unreasonableness are found, the system can correct the physical parameters according to preset adjustment strategies, such as prioritizing the satisfaction of key performance requirements or iterative optimization within a certain range, until the consistency and rationality requirements are met.

[0054] Finally, the adjusted physical parameters are integrated into the initial 3D geometric model to generate a 3D solid model for virtual simulation, which is then used as input for the virtual simulation. This step fuses geometric and physical property information to form a complete 3D model suitable for advanced engineering analysis. For example, parameters such as material properties, surface roughness, and thermal conductivity can be attached to the corresponding parts of the 3D geometric model as metadata or attribute tags. The generated 3D solid model can be directly imported into finite element analysis (FEA), computational fluid dynamics (CFD), or multibody dynamics (MBD) software for more accurate mechanical analysis, thermal analysis, fatigue life prediction, or assembly simulation.

[0055] The image generation method based on visual analysis proposed in this application can transform non-standardized two-dimensional design input images into three-dimensional solid models with complete physical properties through the aforementioned series of steps. Compared with existing technologies, the innovation of this application lies in its ability to process non-standardized two-dimensional design input images and extract not only geometric information from these images, but also to deeply mine and deduce the macroscopic physical property description information, functional positioning, usage environment, performance requirements, and specific physical parameters of the components.

[0056] refer to Figure 2 , Figure 2 This is a flowchart of a method for deriving the physical parameters of a component according to an embodiment of the present invention. S17 includes: S171, Based on the macroscopic physical property description information, the functional positioning, the usage environment and the performance requirements, extract the initial mechanical parameters, initial physical parameters and initial surface residual compressive stress of the component from the engineering database; S172, Analyze the working cycle conditions of the component in the usage environment, wherein the working cycle conditions include the amplitude of the cyclic load, the frequency of the cyclic load, the range of the periodic temperature fluctuation, and the frequency of the periodic temperature fluctuation. S173, Based on the analyzed cyclic load and the periodic temperature fluctuation, predict the relaxation process of the residual compressive stress inside the component with the number of working cycles, and obtain the residual stress relaxation evolution result; S174, Based on the residual stress relaxation evolution results, adjust the fatigue limit of the component to reflect the decrease in the fatigue resistance of the component during long-term operation; S175, Verify whether the adjusted fatigue limit meets the performance requirements of the component.

[0057] Extracting the initial mechanical parameters, initial physical parameters, and initial surface residual compressive stress of components from the engineering database involves retrieving and acquiring basic mechanical property data (such as yield strength, tensile strength, and elastic modulus), physical property data (such as density, coefficient of thermal expansion, and thermal conductivity), and initial surface residual compressive stress distribution information that may form during manufacturing from a pre-established engineering database, based on the identified macroscopic physical property description information of the component, its functional positioning, expected usage environment, and set performance requirements. These initial parameters provide the foundational data for subsequent fatigue analysis and physical parameter derivation.

[0058] Furthermore, analyzing the operational cycle conditions of a component under the stated service environment refers to clarifying the dynamic loads and thermal environment characteristics it experiences by analyzing the component's actual or anticipated service conditions. Specifically, operational cycle conditions can include the amplitude of cyclic loads (e.g., the difference between the maximum and minimum values ​​of stress or strain), the frequency of cyclic loads (i.e., the rate of load change), the range of periodic temperature fluctuations (e.g., the difference between the highest and lowest temperatures), and the frequency of periodic temperature fluctuations. These parameters are key inputs for evaluating the component's fatigue behavior and residual stress relaxation process.

[0059] Based on this, and using the analyzed cyclic loads and periodic temperature fluctuations, the relaxation process of residual compressive stress inside the component with the number of working cycles is predicted, yielding the residual stress relaxation evolution results. This typically involves using material creep, plastic deformation, and thermodynamic models, combined with numerical simulation methods such as finite element analysis, to simulate the dynamic process of the gradual decrease or redistribution of residual compressive stress inside the component with time or the number of working cycles under cyclic loads and temperature fluctuations. The residual stress relaxation evolution results reflect the long-term changing trend of the stress state inside the component.

[0060] Based on the residual stress relaxation evolution results, the fatigue limit of the component is adjusted to reflect the decrease in its fatigue resistance during long-term operation. Residual compressive stress can generally improve the fatigue strength of materials, but its relaxation under cyclic loading and high temperature will weaken this beneficial effect, thus reducing the actual fatigue limit of the component. Therefore, it is necessary to dynamically correct the fatigue limit of the component according to the predicted degree of residual stress relaxation, so that it more accurately reflects the true fatigue resistance of the component under long-term service conditions.

[0061] Finally, verifying whether the adjusted fatigue limit meets the performance requirements of the component involves comparing the fatigue limit after correction for residual stress relaxation with the fatigue performance requirements set during the component design. This step aims to ensure that the component still meets its expected service life and reliability standards even after considering the decline in fatigue performance caused by residual stress relaxation during long-term operation.

[0062] This application systematically integrates the dynamic characteristics of components under actual service conditions, particularly the impact of residual stress relaxation on fatigue performance, into the derivation of physical parameters. Through this technical solution, the accuracy and reliability of component physical parameter derivation are significantly improved, making it more reflective of the long-term performance of components under actual service conditions. By considering the dynamic influence of residual stress relaxation on fatigue limit, the fatigue life and potential failure risks of components can be predicted more accurately, effectively overcoming the shortcomings of traditional methods in assessing long-term service performance. Furthermore, this solution provides more realistic and comprehensive input for subsequent 3D solid model generation and virtual simulation, thereby enhancing the engineering rationality and safety of the entire design and analysis process. It helps to identify and mitigate potential fatigue failure risks early in the design phase, guiding design optimization and extending the service life of components.

[0063] This application further proposes a more refined verification method, which aims to achieve a comprehensive assessment and forward-looking management of component fatigue state by setting safety thresholds, issuing risk warnings, and predicting failure time or location.

[0064] The steps for verifying whether the adjusted fatigue limit meets the performance requirements of the component include: Set the safety threshold for the fatigue limit; When the adjusted fatigue limit is lower than the safety threshold, a fatigue risk warning is issued; Based on the fatigue risk warning, the current damage of the component is calculated, and the failure time or failure location of the component is predicted based on the current damage.

[0065] Specifically, the fatigue limit safety threshold refers to a minimum acceptable fatigue limit value set in engineering practice to ensure the safe and reliable operation of a component within its expected service life. This threshold typically considers factors such as the inherent fatigue characteristics of the material, the design margin of the component, the severity of the service environment, and the severity of the failure consequences, and can be determined based on industry standards, design specifications, or empirical data. Its purpose is to provide a clear judgment criterion to promptly identify components with potential fatigue performance risks.

[0066] Specifically, when the adjusted fatigue limit falls below the safety threshold, the system will issue a fatigue risk warning. This warning can be understood as a signal or notification, designed to alert operators or maintenance personnel that the component's fatigue state has entered a potentially risky area. The warning may take the form of, but is not limited to, audible and visual alarms, system message prompts, and report generation. Its purpose is to prompt relevant personnel to conduct further inspections, assessments, or take appropriate intervention measures to prevent potential failures.

[0067] In practical applications, based on the fatigue risk warning, the current damage of the component can be calculated, and the failure time or location of the component can be predicted based on the current damage. The current damage refers to the degree of material property degradation caused by fatigue accumulation after the component has undergone a certain number of working cycles. This damage can be calculated using various fatigue damage accumulation models (such as Miner's rule, linear damage accumulation theory, or nonlinear damage accumulation theory). Based on the current damage, combined with the component's remaining life prediction model, it is possible to further predict whether the component may fail at a future point in time, or whether failure may occur in a specific area of ​​the component (such as stress concentration areas or defect areas). The purpose is to provide a quantitative basis for preventative maintenance, component replacement planning, and structural optimization.

[0068] This application's solution introduces a safety threshold for fatigue limits, providing a quantified safety boundary for evaluating the fatigue performance of components. When the adjusted fatigue limit falls below this threshold, the system can promptly issue a fatigue risk warning, thus transforming traditional passive failure detection into proactive risk identification. It is precisely this warning mechanism that makes subsequent damage calculation and failure prediction possible. By quantitatively assessing the current damage of the component and combining it with a failure prediction model, the remaining life and potential failure modes of the component can be analyzed in depth. This method not only determines whether the component meets performance requirements, but more importantly, it reveals the evolution trend of the component's fatigue state and provides specific predictions of failure time and location, thereby providing more comprehensive and forward-looking information for engineering decisions.

[0069] This application further proposes the following steps for verifying whether the adjusted fatigue limit meets the performance requirements of the component: Obtain the adjusted fatigue limit, the initial mechanical parameters, and the discrete distribution information of the initial physical parameters; Based on the discrete distribution information, generate multiple parameter combination samples of the component; For each of the multiple parameter combination samples, verify whether the fatigue limit, mechanical parameters, and physical parameters in the parameter combination sample meet the performance requirements of the component; Based on the verification results of all parameter combination samples, calculate the probability that the component meets the performance requirements and output the probability value.

[0070] Specifically, obtaining the discrete distribution information of the adjusted fatigue limit, initial mechanical parameters, and initial physical parameters refers to collecting and quantifying the variability of these key parameters through experimental testing, historical data analysis, expert experience, or statistical modeling. This discrete distribution information can be represented as a probability density function, a cumulative distribution function, or a series of discrete values ​​and their corresponding probability weights. The purpose is to comprehensively capture the uncertainty range and distribution characteristics of these parameters in practical applications.

[0071] In this process, generating multiple parameter combination samples of the component based on the discrete distribution information can be understood as using random sampling techniques such as Monte Carlo simulation, Latin hypercube sampling, and importance sampling to extract a large number of parameter combinations from the acquired discrete distribution information. Each parameter combination sample represents the state of the component under a specific parameter configuration. By generating a sufficient number of samples, the parameter space can be effectively covered, simulating the behavior of the component under various possible parameter combinations.

[0072] In practical applications, verifying whether the fatigue limit, mechanical parameters, and physical parameters of each parameter combination sample meet the performance requirements of the component involves substituting each generated parameter combination sample into a preset performance evaluation model or criterion. This verification process may involve numerical simulation (such as finite element analysis), empirical formula calculations, or judgments based on engineering specifications to determine whether the fatigue limit, mechanical parameters, and physical parameters of the component all meet its design performance requirements under that specific parameter combination.

[0073] Furthermore, based on the verification results of all parameter combination samples, the probability that the component meets the performance requirements is calculated and output as a probability value. This means that after verifying all parameter combination samples, the number of samples that meet the performance requirements is counted, and the ratio of this number to the total number of samples is taken as the probability that the component meets the performance requirements. This probability value is a quantitative indicator that intuitively reflects the reliability level of the component considering parameter uncertainties.

[0074] This application's solution overcomes the limitation of traditional deterministic verification methods, which cannot effectively assess the impact of parameter uncertainty on component performance, by introducing discrete distribution information of parameters and generating multiple parameter combination samples for individual verification. Through the above technical solution, this application provides a more comprehensive and reliable method for verifying the fatigue limit of components. This method considers the inherent uncertainty of key physical parameters of components, avoiding the risks that may arise from relying solely on nominal values. By calculating the probability that a component meets performance requirements, it provides designers and engineers with quantifiable reliability indicators, thereby better assessing the long-term service performance of components, optimizing design schemes, reducing potential failure risks, and improving the overall reliability and safety of products. This probabilistic assessment method is particularly suitable for critical components with high reliability requirements, significantly improving the scientific nature and accuracy of engineering decisions.

[0075] The step of extracting the initial mechanical parameters, initial physical parameters, and initial surface residual compressive stress of the component from the engineering database based on the macroscopic physical property description information, the functional positioning, the usage environment, and the performance requirements includes: Identify the type of manufacturing process used for the component; Based on the manufacturing process type, identify the characteristics of the impact of the manufacturing process on the surface or near-surface structure of the material. Based on the aforementioned influence characteristics, the initial surface residual compressive stress distribution of the component is derived; Based on the influence of the manufacturing process on the microstructure of the material, the initial mechanical parameters and the initial physical parameters are corrected.

[0076] Specifically, identifying the manufacturing process type used for a component refers to analyzing design drawings, production records, or using image recognition technology to analyze the surface features of the component to determine the manufacturing methods it underwent during production, such as casting, forging, welding, heat treatment, and surface strengthening treatments (e.g., shot peening, laser shock peening). The purpose is to provide basic information for subsequent parameter adjustments.

[0077] Specifically, identifying the impact characteristics of the manufacturing process on the surface or near-surface structure of the material, based on the manufacturing process type, can be understood as querying a pre-defined process-impact database for the identified manufacturing process to obtain the influence patterns of the process on the microstructure and mechanical properties of the material, such as surface roughness, hardness gradient, grain size, phase transformation, dislocation density, and residual stress distribution. For example, shot peening introduces high-pressure stress into the material surface and may lead to grain refinement and work hardening. The purpose is to quantify the specific changes in component performance caused by the manufacturing process.

[0078] In practical applications, based on the aforementioned influencing characteristics, the initial surface residual compressive stress distribution of the component is derived. Specifically, this refers to predicting the magnitude, depth, and distribution pattern of the initial residual compressive stress on the component surface based on the identified manufacturing process influence characteristics, combined with the principles of materials mechanics and solid mechanics, through numerical simulation (such as finite element analysis) or empirical formulas. For example, for components that have undergone surface strengthening treatment, a region with high residual compressive stress typically forms on its surface. The purpose is to obtain the true stress state of the component before service.

[0079] Furthermore, modifying the initial mechanical and physical parameters based on the influence of the manufacturing process on the material's microstructure refers to adjusting the initial mechanical parameters (such as yield strength, tensile strength, elastic modulus, fatigue strength, etc.) and initial physical parameters (such as density, coefficient of thermal expansion, thermal conductivity, etc.) extracted from the engineering database based on the changes in the material's microstructure (such as grain size, grain boundary characteristics, phase composition, etc.) caused by the manufacturing process. For example, heat treatment may alter the grain structure of a material, thereby affecting its strength and toughness. The aim is to make these parameters more accurately reflect the actual performance of the component under a specific manufacturing process.

[0080] This application's solution, by deeply identifying the manufacturing process type of the component and its impact on the surface or near-surface structure of the material, can more accurately derive the initial surface residual compressive stress distribution of the component and correct the initial mechanical and physical parameters. Through the above technical solution, this application can significantly improve the accuracy and reliability of extracting the initial mechanical parameters, initial physical parameters, and initial surface residual compressive stress of the component. By considering the unique influence of the manufacturing process on the material's microstructure and surface condition, simulation errors caused by using general parameters can be avoided, allowing the subsequently derived physical parameters to more realistically reflect the actual performance of the component. This not only helps to more accurately predict the component's service behavior and lifespan but also provides more reliable data support for design optimization and failure analysis, thereby improving the accuracy and effectiveness of the entire virtual simulation process.

[0081] This application further proposes the following steps for deriving the initial surface residual compressive stress distribution of a component based on the aforementioned influence characteristics: Identify the geometric features of the component; Based on the identified geometric features, the surface region of the component is divided, wherein the surface region includes geometric feature regions and non-geometric feature regions; For the geometric feature region, based on the manufacturing process type and the local geometric parameters of the geometric feature, the initial surface residual compressive stress concentration distribution of the geometric feature region is derived; The initial surface residual compressive stress gradient distribution of the non-geometric feature region is derived based on the manufacturing process type and the macroscopic geometric parameters of the non-geometric feature region. By integrating the initial surface residual compressive stress concentration distribution in the geometric feature region with the initial surface residual compressive stress gradient distribution in the non-geometric feature region, a non-uniform distribution of the initial surface residual compressive stress of the component is obtained.

[0082] Specifically, identifying the geometric features of a component refers to using image processing and pattern recognition techniques to identify various geometric shapes and structures on the component's surface from a pre-processed image. These include features such as holes, slots, chamfers, bosses, grooves, and threads. These geometric features are typically areas of stress concentration or significant stress gradient changes. Based on the identified geometric features, the surface area of ​​the component is divided into two main categories: geometric feature areas and non-geometric feature areas. Geometric feature areas are those with obvious geometric discontinuities or shape changes, such as the edges of holes and chamfers mentioned above. Non-geometric feature areas are relatively flat areas on the component's surface with minimal geometric shape changes. This division aims to use different stress derivation models for different areas to improve the accuracy of the derivation.

[0083] In practical applications, the derivation of the initial surface residual compressive stress concentration distribution for geometrically characteristic regions is based on the manufacturing process type (e.g., shot peening, laser shock peening, rolling, etc.) and the local geometric parameters of the geometric features (e.g., the diameter of holes, the radius of chamfers, the depth and width of grooves, etc.). These local geometric parameters have a decisive influence on the concentration effect of residual stress. For example, after shot peening, higher residual compressive stress concentration may occur at small-radius chamfers. Meanwhile, the derivation of the initial surface residual compressive stress gradient distribution for non-geometrically characteristic regions is based on the manufacturing process type and the macroscopic geometric parameters of the non-geometrically characteristic regions (e.g., the length, width, radius of curvature, etc. of the region). These macroscopic parameters mainly affect the gradient change of residual stress over a large range, rather than local concentration.

[0084] For example, after rolling, the residual stress in flat areas may exhibit a gradient distribution that gradually decreases from the surface to the interior. Ultimately, by fusing the initial surface residual compressive stress concentration distribution in geometrically characteristic areas with the initial surface residual compressive stress gradient distribution in non-geometrically characteristic areas, the aim is to superimpose or smoothly transition these two different types of stress distributions using mathematical models or numerical methods, thereby obtaining a non-uniform distribution of the initial surface residual compressive stress of the component. This ensures an accurate and continuous description of residual stress across the entire component surface, regardless of whether it is a stress concentration area or a stress gradient area.

[0085] This application's solution addresses the limitation of traditional methods in deriving initial surface residual compressive stress distribution by meticulously identifying the geometric features of components and dividing the component surface into geometrically characteristic regions and non-geometrically characteristic regions. This approach may overlook local stress concentrations or gradient changes when deriving the initial surface residual compressive stress distribution. Through this technical solution, this application overcomes the shortcomings of existing technologies in handling the residual compressive stress distribution on complex geometrical components, significantly improving the accuracy and precision of deriving the initial surface residual compressive stress distribution. This regionalized and differentiated derivation strategy effectively reflects the stress concentration effect in geometrically characteristic regions and the stress gradient changes in non-geometrically characteristic regions, thus avoiding evaluation errors caused by simplification.

[0086] This application further proposes considering the influence of microscopic inhomogeneities within the component when predicting the relaxation process of residual compressive stress with the number of working cycles, in order to obtain more accurate results on the evolution of residual stress relaxation. The steps described above for predicting the relaxation process of residual compressive stress within the component with the number of working cycles based on the analyzed cyclic load and the periodic temperature fluctuations, and obtaining the results on the evolution of residual stress relaxation, include: Identify the types, sizes, spatial distributions, and orientation distribution characteristics of micro-defects within the component; Based on the identified microdefect type, microdefect size, microdefect spatial distribution, and grain orientation distribution characteristics, the disturbance effect of the microdefect or grain orientation inhomogeneity on the local stress field of the component is determined. Based on the perturbation effect, the material constitutive response of the local region of the component is modified to deduce the accelerating or inhibiting effect of the micro-defects or the non-uniformity of grain orientation on the relaxation rate of the local residual compressive stress of the component. By integrating the accelerating or inhibiting effect into the relaxation prediction of the residual compressive stress inside the component, a residual compressive stress relaxation process that takes into account microscopic inhomogeneities is obtained.

[0087] Specifically, identifying the types, sizes, spatial distributions, and grain orientation characteristics of microscopic defects within a component refers to obtaining information about its internal microstructure using non-destructive testing techniques (such as X-ray diffraction, ultrasonic testing, and electron microscopy) or metallographic analysis. Microscopic defect types can include pores, inclusions, and microcracks; defect size refers to their geometric dimensions; spatial distribution describes their location and density within the component; and grain orientation characteristics describe the orientation and extent of crystal arrangement within the material. This information is crucial for understanding the local mechanical behavior of materials.

[0088] Specifically, based on the identified micro-defect types, sizes, spatial distributions, and grain orientation characteristics, the disturbance effect of micro-defects or grain orientation inhomogeneities on the local stress field of a component is determined. This can be understood as analyzing how these microstructural features alter the stress distribution within the component. For example, pores or microcracks can cause stress concentration, while grains with different orientations may lead to differences in local elastic modulus or yield strength, thus generating local stress inhomogeneities under macroscopic loads.

[0089] In practical applications, based on the aforementioned perturbation effect, modifying the material constitutive response of local regions of the component to deduce the accelerating or inhibiting effects of microscopic defects or grain orientation inhomogeneities on the local residual compressive stress relaxation rate of the component refers to introducing microstructural parameters into the material constitutive model so that it can reflect the influence of local stress field changes on the plastic deformation and creep behavior of the material. For example, materials in stress concentration regions are more prone to plastic deformation and creep, thereby accelerating residual stress relaxation; while certain specific grain orientations may have higher creep resistance, thereby inhibiting residual stress relaxation.

[0090] Furthermore, integrating the accelerating or inhibiting effects into the relaxation prediction of residual compressive stress within the component yields a residual compressive stress relaxation process that considers microscopic inhomogeneities. This refers to applying the modified material constitutive response to the component's residual stress relaxation calculation in numerical simulations or analytical models. This can be achieved through finite element analysis (FEA) combined with multi-scale modeling, where microstructural information is mapped into a macroscopic model to more accurately predict the evolution of residual stress.

[0091] This application's approach first identifies microscopic defects and grain orientation inhomogeneities within the component, then quantifies the perturbation effects of these microscopic features on the local stress field. Because these perturbation effects alter the constitutive response of the material in local regions of the component, the relaxation rate of residual compressive stress exhibits either an accelerating or inhibiting trend in different microscopic regions. By integrating the accelerating or inhibiting effects caused by these microscopic inhomogeneities into the residual compressive stress relaxation prediction model, this application can more accurately simulate the evolution of residual stress in the component under working cycle conditions, thereby overcoming the prediction bias caused by neglecting the influence of microstructure in traditional methods.

[0092] This application further proposes steps for adjusting the fatigue limit of a component based on the residual stress relaxation evolution results, including: Acquire information on temperature changes and load frequency changes in the operating environment of the component during operation; Based on the temperature change information and the load frequency change information, the residual stress relaxation rate of the component is corrected; Based on the corrected residual stress relaxation rate, the current residual stress state of the component is recalculated; Based on the current residual stress state and the material fatigue damage accumulation characteristics of the component, the fatigue limit of the component is dynamically adjusted.

[0093] Specifically, acquiring information on temperature and load frequency changes in the operating environment of a component during operation refers to using various sensors or monitoring systems deployed at the component's service site to collect real-time or periodic temperature data and load frequency data of the component's environment. For example, thermocouples and infrared thermometers can be used to obtain temperature information, while strain gauges, accelerometers, or vibration sensors can be used to obtain load frequency information. This information forms the basis for dynamically correcting the residual stress relaxation rate.

[0094] The process of correcting the residual stress relaxation rate of the component based on the temperature and load frequency changes can be understood as using an established material residual stress relaxation model, combined with real-time acquired temperature and load frequency data, to adjust the parameters in the model to more accurately predict the actual rate of residual stress relaxation under the current working environment. For example, the residual stress relaxation rate typically accelerates under high temperature or high frequency loads, and may slow down under lower temperatures.

[0095] In practical applications, based on the corrected residual stress relaxation rate, the current residual stress state of the component is recalculated. Specifically, the corrected relaxation rate is substituted into the residual stress evolution equation, and combined with the initial residual stress distribution of the component and the number of service cycles already completed, the residual stress distribution and magnitude within the component at the current moment are calculated. The purpose is to obtain the most accurate residual stress level of the component under its current service condition.

[0096] Furthermore, based on the current residual stress state and the material fatigue damage accumulation characteristics of the component, the fatigue limit of the component is dynamically adjusted. This means that after obtaining the current residual stress state of the component, it is combined with the inherent fatigue damage accumulation law of the component material. For example, the presence of residual compressive stress usually increases the fatigue limit of the material, but this gain gradually weakens as the residual compressive stress relaxes. At the same time, the material accumulates fatigue damage during service, which also leads to a decrease in its fatigue limit. By comprehensively considering these two factors, the fatigue limit of the component can be updated in real time or periodically, making it closer to the actual fatigue resistance.

[0097] This application's solution dynamically captures key external factors influencing residual stress relaxation by acquiring information on temperature and load frequency changes in the operating environment during component operation. Given that temperature and load frequency are crucial parameters affecting material creep and plastic deformation behavior, their changes directly determine the dynamics of residual stress relaxation. Therefore, by correcting the residual stress relaxation rate based on this dynamic information, the residual stress evolution model can better reflect the actual service conditions of the component. Based on this, the current residual stress state of the component is recalculated, ensuring a real-time and accurate assessment of the internal stress distribution. Since the residual stress state significantly impacts the fatigue performance of materials, combining this with the material's fatigue damage accumulation characteristics and dynamically adjusting the component's fatigue limit allows for a more accurate reflection of the component's true fatigue resistance under different service stages and environmental conditions, effectively compensating for the limitations of traditional static assessment methods.

[0098] The above-mentioned steps for dynamically adjusting the fatigue limit of the component based on the current residual stress state and the material fatigue damage accumulation characteristics of the component include: The real-time service environment parameters of the component during operation are obtained, wherein the real-time service environment parameters include temperature, humidity, concentration of corrosive medium and load spectrum changes; The current environmental damage level of the component material is assessed based on the real-time service environment parameters. Based on the current environmental damage level, the material fatigue damage accumulation characteristics of the component are corrected; Based on the corrected material fatigue damage accumulation characteristics and combined with the current residual stress state, the fatigue limit of the component is dynamically adjusted.

[0099] Real-time service environment parameters refer to the various environmental conditions that affect the material properties and damage accumulation of components during actual operation. Specifically, temperature can affect the mechanical properties, corrosion rate, and fatigue crack propagation rate of materials; humidity, especially in high-humidity environments, may accelerate the corrosion fatigue process of some materials; the concentration of corrosive media directly affects the corrosion fatigue and stress corrosion cracking behavior of materials; and changes in the load spectrum reflect the dynamic and random nature of the actual loads borne by the component, which has a significant impact on the accumulation mode of fatigue damage. These parameters can be collected in real time by deploying various sensors (such as temperature sensors, humidity sensors, gas analyzers, strain gauges, etc.) on the component or in its service environment, or obtained through data interaction with environmental monitoring systems and historical operation databases.

[0100] Furthermore, assessing the current environmental damage level of the component material based on the real-time service environment parameters refers to quantifying the degree of performance degradation or damage accumulation caused by environmental factors (such as corrosion, oxidation, wear, radiation, etc.) based on acquired real-time environmental data. This assessment process can utilize pre-established physical models, empirical formulas, data-driven models (e.g., damage prediction models based on machine learning), or finite element analysis. For example, the corrosion depth or corrosion rate of the material can be predicted using a corrosion kinetic model based on temperature, humidity, and corrosive medium concentration, thereby obtaining the current degree of corrosion damage.

[0101] Based on this, and according to the current degree of environmental damage, the material fatigue damage accumulation characteristics of the component are corrected. This refers to adjusting the parameters or models describing the damage accumulation law of the material under cyclic loading. Material fatigue damage accumulation characteristics are typically characterized by SN curves (stress-life curves), ε-N curves (strain-life curves), or various fatigue damage accumulation rules (such as Miner's rule). Environmental damage, such as corrosion, can lead to defects and stress concentration on the material surface, thereby reducing the material's fatigue strength, fatigue limit, and shortening its fatigue life. Therefore, the correction process involves adjusting these fatigue characteristic curves or parameters according to the assessed degree of environmental damage, so that they can more accurately reflect the actual fatigue behavior of the material under the current environmental damage state. For example, corrosion damage can reduce the effective load-bearing cross-section of the material, thus lowering its fatigue limit.

[0102] Finally, based on the corrected material fatigue damage accumulation characteristics and the current residual stress state, the fatigue limit of the component is dynamically adjusted. This dynamic adjustment process comprehensively considers the beneficial effects of residual stress on fatigue performance (e.g., residual compressive stress can increase the fatigue limit) and the adverse effects of environmental damage on fatigue performance (e.g., environmental damage can decrease the fatigue limit). By coupling the corrected fatigue damage accumulation characteristics with the current residual stress state in the calculation, a more comprehensive and accurate fatigue limit value that reflects the actual fatigue resistance of the component under the current complex service environment and stress state can be obtained. This dynamic adjustment makes the fatigue limit no longer a static material property, but a parameter that changes in real time with the service conditions of the component.

[0103] This application's solution, by introducing real-time service environment parameters, can more comprehensively capture the complex environmental impacts faced by components during actual operation. Through this technical solution, this application can achieve more precise and adaptive dynamic adjustment of the component's fatigue limit. Specifically, by considering real-time service environment parameters and their impact on the degree of material environmental damage, it can effectively compensate for the shortcomings of adjusting the fatigue limit solely based on residual stress state, avoiding the neglect of fatigue performance degradation caused by environmental factors. Therefore, the adjusted fatigue limit can more accurately reflect the component's true fatigue resistance under complex and variable service environments, significantly improving the accuracy and reliability of fatigue life prediction. This provides more solid data support for component health monitoring, maintenance strategy formulation, and remaining life assessment, thereby effectively reducing the risk of unexpected component failure.

[0104] refer to Figure 3 , Figure 3 This is a schematic diagram of the structure of an image generation system based on visual analysis provided in an embodiment of the present invention, comprising: The image acquisition module is used to acquire non-standardized two-dimensional design input images; The image preprocessing module is used to perform image preprocessing on the non-standardized two-dimensional design input image and extract two-dimensional geometric contour features from the preprocessed image; A geometric recognition module is used to identify the geometric shape of a component and the macroscopic physical property description information of the component based on the two-dimensional geometric contour features. A 3D model building module is used to build an initial 3D geometric model of the component based on the geometric shape. The functional environment determination module is used to determine the functional positioning and usage environment of the component based on the geometric shape and the macroscopic physical property description information. The performance requirement processing module is used to obtain the performance requirements of the component based on the functional positioning and the usage environment, and to rank the performance requirements by importance. The physical parameter derivation module is used to derive the physical parameters of the component based on the functional positioning, the usage environment, and the performance requirements. The parameter verification and adjustment module is used to verify the consistency and engineering rationality between the physical parameters, and to adjust the physical parameters according to the verification results; The model generation output module is used to integrate the adjusted physical parameters into the initial three-dimensional geometric model to generate a three-dimensional solid model for virtual simulation, and to use the three-dimensional solid model as input for virtual simulation.

[0105] The image generation system based on visual analysis proposed in this application effectively solves the problems of difficulty in processing non-standardized two-dimensional design input images and the lack of key physical attribute information in the generated three-dimensional models in existing technologies through the collaborative work of its various functional modules. The system receives non-standardized images through an image acquisition module, extracts geometric contours and macroscopic physical attribute descriptions through an image preprocessing module and a geometry recognition module, and then generates an initial three-dimensional geometric model through a three-dimensional model construction module. Based on this, a functional environment determination module, a performance requirement processing module, and a physical parameter derivation module progressively refine the functional positioning, usage environment, performance requirements, and physical parameters of the components. Finally, a parameter verification and adjustment module ensures the engineering rationality of the parameters, and a model generation and output module integrates all information to generate a three-dimensional solid model that can be directly used for virtual simulation. This systematic processing flow ensures the transformation from non-standardized visual input to a high-fidelity three-dimensional solid model, greatly improving design efficiency and simulation accuracy.

[0106] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. An image generation method based on visual analysis, characterized in that, include: Acquire non-standardized two-dimensional design input images; The non-standard two-dimensional design input image is preprocessed, and two-dimensional geometric contour features are extracted from the preprocessed image. Based on the two-dimensional geometric contour features, the geometric shape of the component and the macroscopic physical property description information of the component are identified; Based on the described geometric shape, construct an initial three-dimensional geometric model of the component; Based on the described geometric shape and macroscopic physical properties, the functional positioning and usage environment of the component are determined; Based on the functional positioning and the usage environment, the performance requirements of the component are obtained, and the performance requirements are ranked by importance. Based on the functional positioning, the usage environment, and the performance requirements, the physical parameters of the component are derived. Verify the consistency and engineering rationality among the physical parameters, and adjust the physical parameters according to the verification results; The adjusted physical parameters are integrated into the initial three-dimensional geometric model to generate a three-dimensional solid model for virtual simulation, and the three-dimensional solid model is used as input for virtual simulation.

2. The image generation method based on visual analysis according to claim 1, characterized in that, The step of deriving the physical parameters of the component based on the functional positioning, the usage environment, and the performance requirements includes: Based on the macroscopic physical property description information, the functional positioning, the usage environment, and the performance requirements, the initial mechanical parameters, initial physical parameters, and initial surface residual compressive stress of the component are extracted from the engineering database. The working cycle conditions of the component under the operating environment are analyzed, wherein the working cycle conditions include the amplitude of the cyclic load, the frequency of the cyclic load, the range of the periodic temperature fluctuation, and the frequency of the periodic temperature fluctuation. Based on the analyzed cyclic load and the periodic temperature fluctuation, the relaxation process of the residual compressive stress inside the component with the number of working cycles is predicted, and the residual stress relaxation evolution result is obtained. Based on the residual stress relaxation evolution results, the fatigue limit of the component is adjusted to reflect the decrease in the component's fatigue resistance during long-term operation; Verify whether the adjusted fatigue limit meets the performance requirements of the component.

3. The image generation method based on visual analysis according to claim 2, characterized in that, The step of verifying whether the adjusted fatigue limit meets the performance requirements of the component includes: Set the safety threshold for the fatigue limit; When the adjusted fatigue limit is lower than the safety threshold, a fatigue risk warning is issued; Based on the fatigue risk warning, the current damage of the component is calculated, and the failure time or failure location of the component is predicted based on the current damage.

4. The image generation method based on visual analysis according to claim 2, characterized in that, The step of verifying whether the adjusted fatigue limit meets the performance requirements of the component includes: Obtain the adjusted fatigue limit, the initial mechanical parameters, and the discrete distribution information of the initial physical parameters; Based on the discrete distribution information, generate multiple parameter combination samples of the component; For each of the multiple parameter combination samples, verify whether the fatigue limit, mechanical parameters, and physical parameters in the parameter combination sample meet the performance requirements of the component; Based on the verification results of all parameter combination samples, calculate the probability that the component meets the performance requirements and output the probability value.

5. The image generation method based on visual analysis according to claim 2, characterized in that, The step of extracting the initial mechanical parameters, initial physical parameters, and initial surface residual compressive stress of the component from the engineering database based on the macroscopic physical property description information, the functional positioning, the usage environment, and the performance requirements includes: Identify the type of manufacturing process used for the component; Based on the manufacturing process type, identify the characteristics of the impact of the manufacturing process on the surface or near-surface structure of the material. Based on the aforementioned influence characteristics, the initial surface residual compressive stress distribution of the component is derived; Based on the influence of the manufacturing process on the microstructure of the material, the initial mechanical parameters and the initial physical parameters are corrected.

6. The image generation method based on visual analysis according to claim 5, characterized in that, The step of deriving the initial surface residual compressive stress distribution of the component based on the influence characteristics includes: Identify the geometric features of the component; Based on the identified geometric features, the surface region of the component is divided, wherein the surface region includes geometric feature regions and non-geometric feature regions; For the geometric feature region, based on the manufacturing process type and the local geometric parameters of the geometric feature, the initial surface residual compressive stress concentration distribution of the geometric feature region is derived; For the non-geometric feature region, the initial surface residual compressive stress gradient distribution of the non-geometric feature region is derived based on the manufacturing process type and the macroscopic geometric parameters of the non-geometric feature region; By integrating the initial surface residual compressive stress concentration distribution in the geometric feature region with the initial surface residual compressive stress gradient distribution in the non-geometric feature region, a non-uniform distribution of the initial surface residual compressive stress of the component is obtained.

7. The image generation method based on visual analysis according to claim 2, characterized in that, The step of predicting the relaxation process of the residual compressive stress inside the component with the number of working cycles based on the analyzed cyclic load and the periodic temperature fluctuation, and obtaining the residual stress relaxation evolution result, includes: Identify the types, sizes, spatial distributions, and orientation distribution characteristics of micro-defects within the component; Based on the identified microdefect type, microdefect size, microdefect spatial distribution, and grain orientation distribution characteristics, the disturbance effect of the microdefect or grain orientation inhomogeneity on the local stress field of the component is determined. Based on the perturbation effect, the material constitutive response of the local region of the component is modified to deduce the accelerating or inhibiting effect of the micro-defects or the non-uniformity of grain orientation on the relaxation rate of the local residual compressive stress of the component. By integrating the accelerating or inhibiting effect into the relaxation prediction of the residual compressive stress inside the component, a residual compressive stress relaxation process that takes into account microscopic inhomogeneities is obtained.

8. The image generation method based on visual analysis according to claim 2, characterized in that, The step of adjusting the fatigue limit of the component based on the residual stress relaxation evolution result includes: Acquire information on temperature changes and load frequency changes in the working environment of the component during operation; Based on the temperature change information and the load frequency change information, the residual stress relaxation rate of the component is corrected; Based on the corrected residual stress relaxation rate, the current residual stress state of the component is recalculated; Based on the current residual stress state and the material fatigue damage accumulation characteristics of the component, the fatigue limit of the component is dynamically adjusted.

9. The image generation method based on visual analysis according to claim 8, characterized in that, The step of dynamically adjusting the fatigue limit of the component based on the current residual stress state and the material fatigue damage accumulation characteristics of the component includes: The real-time service environment parameters of the component during operation are obtained, wherein the real-time service environment parameters include temperature, humidity, concentration of corrosive medium and load spectrum changes; The current environmental damage level of the component material is assessed based on the real-time service environment parameters. Based on the current environmental damage level, the material fatigue damage accumulation characteristics of the component are corrected; Based on the corrected material fatigue damage accumulation characteristics and combined with the current residual stress state, the fatigue limit of the component is dynamically adjusted.

10. An image generation system based on visual analysis, characterized in that, include: The image acquisition module is used to acquire non-standardized two-dimensional design input images; The image preprocessing module is used to perform image preprocessing on the non-standardized two-dimensional design input image and extract two-dimensional geometric contour features from the preprocessed image; A geometric recognition module is used to identify the geometric shape of a component and the macroscopic physical property description information of the component based on the two-dimensional geometric contour features. A 3D model building module is used to build an initial 3D geometric model of the component based on the geometric shape. The functional environment determination module is used to determine the functional positioning and usage environment of the component based on the geometric shape and the macroscopic physical property description information. The performance requirement processing module is used to obtain the performance requirements of the component based on the functional positioning and the usage environment, and to rank the performance requirements by importance. The physical parameter derivation module is used to derive the physical parameters of the component based on the functional positioning, the usage environment, and the performance requirements. The parameter verification and adjustment module is used to verify the consistency and engineering rationality between the physical parameters, and to adjust the physical parameters according to the verification results; The model generation output module is used to integrate the adjusted physical parameters into the initial three-dimensional geometric model to generate a three-dimensional solid model for virtual simulation, and to use the three-dimensional solid model as input for virtual simulation.

Images