52 results about "Scale structure" patented technology

The invention discloses a rock mechanical parameter determination method based on a nanoindentation experiment, which comprises the following steps: S1, randomly selecting a small part of a target rock to carry out the nanoindentation experiment, and testing the elastic modulus, hardness and fracture toughness micromechanical parameters of the target rock; s2, constructing a multi-scale structure model of the target rock by considering internal cracks, pores, organic matters and proportion factors of each mineral component of the target rock based on the micromechanics parameters obtained by the nanoindentation experiment; s3, on the basis of the multi-scale structure model of the target rock, performing multi-scale transformation by utilizing a reformed group theory to obtain a simplified structure model of the target rock, and calculating mechanical parameters of the simplified structure model; and S4, carrying out a macromechanics experiment on the target rock, calculating mechanical parameters, and then checking the reliability of the method for calculating the mechanical parameters of the target rock based on the nanoindentation experiment. According to the method, the efficiency and precision of calculating the macromechanical parameters of the rock can be improved.

The invention relates to a large-scale structure dynamic response prediction method based on a hybrid deep learning model. The method comprises the following steps: establishing a finite element model of a large-scale structure; a vibration exciter is used for measuring modal parameters of the large-scale structure, and the finite element model of the large-scale structure is corrected according to the modal parameters of the large-scale structure; simulating the dynamic response of the large-scale structure under the dynamic load by using the corrected finite element model; meanwhile, a sensor is used for actually collecting the dynamic response of the large-scale structure under the dynamic load; constructing a training set through analog simulation and actually acquired dynamic response of the large-scale structure under the dynamic load; training the hybrid deep learning model according to the constructed training set; and inputting the dynamic load acting on the large-scale structure into the trained deep learning model to predict the dynamic response of the large-scale structure. According to the method, the prediction efficiency and accuracy are improved, the generalization ability of the model is enhanced, and the method can adapt to various different environments and conditions.

The present invention provides a large-scale structure spatial posture online compensation system based on multiple laser trackers, including a numerical control positioner, a laser tracker, a reflective target ball, a ball joint and a large structure; one laser tracker corresponds to one reflective target ball and is installed close to the transmitting target ball; the numerical control positioner is connected to the large structure through a ball joint; the reflective target balls are provided in multiple numbers and are dispersedly installed on the large structure; wherein the numerical control positioner and the laser tracker are provided with at least 3 groups. At the same time, a method of the above system is also provided, which can avoid the low efficiency caused by single-point measurement and iterative posture adjustment in the traditional large-scale structure posture adjustment and positioning method, and at the same time, the posture adjustment trajectory can be detected and compensated in real time, thereby ensuring the positioning accuracy of the large structure.

The invention provides a high-ductility cement-based composite material freeze-thaw damage prediction method and system fused with surface multi-scale structure recognition. The method comprises the following steps: S1, obtaining ECC surface damage information after a freeze-thaw cycle effect; s2, constructing an optimized ECC freeze-thaw damage strength prediction model based on a machine learning method; s3, according to the parameters obtained in the step S1, the surface damage coefficient of the ECC is determined through the prediction model in the step S2; and S4, predicting the damage strength of the ECC after freezing and thawing according to the surface damage coefficient. The invention provides an ECC freeze-thaw damage prediction model based on surface recognition, and the damage strength of a material can be speculated by analyzing surface damage picture information. According to the model, image information such as fiber exposure and surface peeling is directly associated with damage strength by using an image feature extraction and machine learning algorithm, and efficient and accurate prediction of the performance of the ECC material after freeze-thaw damage is realized.

The invention belongs to the technical field related to structure optimization design, and discloses a macro-micro dual-scale structure topological optimization method and device capable of controlling the minimum size of a microstructure, and the method comprises the following steps: (1), carrying out the interpolation of an elastic matrix of an actual microstructure in an offline database and the corresponding density through employing radial basis function interpolation, and obtaining an elastic matrix of the actual microstructure; constructing a mapping model; (2) a flexibility minimization optimization model is constructed, design variables of the optimization model are cutting heights and topological variables of the two virtual microstructures in each unit, and constraint functions are volume fractions and topological variable constraints of the overall structure; (3) respectively calculating the sensitivity of the target function and the constraint function to the design variable, and further updating the design variable; (4) when the updated flexibility error of the dual-scale structure meets an optimization termination condition, completing optimization and obtaining the dual-scale structure meeting the requirement; otherwise, turning to the step (3). According to the method, the problems of low calculation efficiency and poor connectivity are solved.

The invention relates to the technical field of displacement change monitoring, in particular to a large-scale structure displacement change tracking monitoring method and system based on digital images, and the method comprises the following steps: manufacturing a target with a known shape and size; collecting a global digital image of the large-scale structure when no target is arranged, defining the global digital image as a first digital image, marking a fixed point and a displacement tracking point to be measured in the first digital image based on prior knowledge of the large-scale structure, and arranging targets at the fixed point and the displacement tracking point to be measured; acquiring a digital image of the large-scale structure after the target is arranged, and defining the digital image as a second digital image; based on the target in the second digital image, target structure displacement analysis is carried out, target theoretical center point position coordinates corresponding to the displacement tracking points to be measured are obtained, and displacement changes of the large-scale structure are analyzed. According to the invention, the image identification monitoring precision of the to-be-measured displacement tracking point and the accuracy and robustness of change tracking of the to-be-measured displacement tracking point can be obviously improved.

The invention belongs to the related technical field of dual-scale structure design, and discloses a dual-scale structure optimization design method and system based on substructure data driving, and the method comprises the steps: an offline stage: for cell elements of a dual-scale structure, carrying out the cutting operation of a level set function, obtaining a virtual microstructure, and carrying out the cutting operation of the level set function; determining the density of the actual microstructure according to the density of the virtual microstructure; performing static condensation through a substructure method to obtain a stiffness matrix of the super-unit, and constructing a mapping model of the cutting height parameter, the stiffness matrix of the super-unit and the density of the actual microstructure; in the online stage, a data driving model with the cutting height parameter as the input quantity and the stiffness matrix of the super-unit and the density of the actual microstructure as the output quantity is obtained; and applying the data driving model to a finite element optimization process of the dual-scale structure, and finally obtaining an optimized dual-scale structure. According to the invention, an off-line data driving method is adopted to carry out dual-scale structure optimization, and the efficiency can be greatly improved under the condition that the precision is ensured.

The invention belongs to the technical field of microstructure and multi-scale structure design, and particularly relates to a microstructure rapid design and filling method and system based on macroscopic strain. Constructing a microstructure topological optimization model based on the macroscopic strain, sampling the macroscopic strain and the volume fraction to generate a corresponding optimized microstructure, and establishing a microstructure training data sample library with labels; carrying out equivalence analysis on a strain energy expression, solving the problem of variable inconsistency by utilizing an intermediate variable, and converting the strain in the training sample into the intermediate variable; establishing a deep neural network with an intermediate variable and a volume fraction as input and a microstructure topological configuration as output, and training the deep neural network by using a data sample to obtain a microstructure fast reverse design network; and carrying out staged multi-scale structure filling design by using the microstructure fast reverse design network. According to the method, the online calculation cost of microstructure and multi-scale structure design can be reduced, and the design efficiency is improved.

The present application belongs to the technical field of structure optimization design, and discloses a macro-micro double-scale structure topology optimization method and equipment capable of controlling the minimum size of microstructure, comprising the following steps: (1) interpolating the elastic matrix and the corresponding density of the actual microstructure in the offline database by using radial basis function interpolation, and then constructing a mapping model; (2) constructing a compliance minimization optimization model, wherein the design variables of the optimization model are the cutting height of two virtual microstructures in each unit and the topology variable, and the constraint functions are the volume fraction of the overall structure and the topology variable constraint; (3) calculating the sensitivity of the objective function and the constraint function to the design variable, and then updating the design variable; (4) when the compliance error of the updated double-scale structure meets the optimization termination condition, the optimization is completed and the double-scale structure meeting the requirements is obtained; otherwise, go to step (3). The present application solves the problems of low calculation efficiency and poor connectivity.

The invention relates to the technical field of structure fatigue evaluation, in particular to a pipe truss structure fatigue analysis parameter optimization method and system based on deep learning, and the method comprises the following steps: constructing a deep learning proxy model fusing a multi-scale structure diagram encoder, a working condition parameter encoder and a damage dimension curve decoder; training the model by using a high-fidelity training data set; for the target truss structure and the load working condition, predicting a damage dimension relation curve through the model; determining the optimal dimension of a rain flow counting matrix by analyzing the convergence characteristic of the curve; and predicting a multi-working-condition damage value through a model, and screening out a dominant working condition based on a damage contribution percentage. According to the method, the deep learning agent model is used for replacing traditional iterative simulation, rapid optimization of key parameters in fatigue analysis is achieved, the efficiency and accuracy of fatigue evaluation of the large truss structure are remarkably improved, and the generalization ability of the model is improved by fusing the physical effect of a multi-scale structure chart encoder to capture key hotspots.

The invention discloses a cross-scale structure-material rheology detection system, which belongs to the technical field of biomedical engineering, and realizes cross-scale geometric imaging by combining optical coherence tomography and laser scanning confocal or fluorescence imaging common optical path design. Geometric-mechanical synchronous observation and analysis of in-situ real-time monitoring and cross-scale synchronous monitoring of stress and strain of a multi-scale structure are supported; the biological 3D printing based on the six-axis robot supports in-situ printing on a rheology detection platform, so that a standardized rheology detection model of a bionic tissue inherent structure is constructed, and automatic and high-throughput bio-ink geometry-mechanics synchronous observation is supported; a multi-physics field coupling measurement model in a dynamic microenvironment is constructed, and a printing structure-material mechanics coupling mechanism is analyzed based on a rheology-structure data quantitative mapping model of deep learning. The system is based on rheology detection of the structure, and the theoretical and technical limitations of multi-scale structural mechanical property monitoring are solved.

The invention discloses a snakelike arm structure based on a negative Poisson's ratio metamaterial. A snakelike mechanical arm is obtained by assembling a plurality of mechanical metamaterial structural bodies. The mechanical metamaterial structural body is composed of an external frame unit capable of elastically deforming and bionic scales; the elastically-deformable outer frame unit is formed by arranging four connecting rods between a vertical rod A and a vertical rod B which are arranged in parallel. The elastically deformable outer frame unit has an elastic deformation characteristic of the flexible mechanism. The mechanical metamaterial structural body is divided into a first plane and a second plane from the space structure, the second plane tilts upwards by a certain angle alpha relative to the first plane, the alpha forms a micro-convex bionic scale structure with a certain backward angle, the highly asymmetric pawl characteristic with one sharp side and the other gentle side is shown, and the mechanical metamaterial structural body is formed. Therefore, the mechanical metamaterial structural body forms friction anisotropy in the scale blade direction, namely, different friction forces are shown in different movement directions.

The utility model relates to the technical field of wagon balance structures, in particular to a wagon balance connecting mechanism. According to the technical scheme, the scale comprises a first scale body, a second scale body, a square cap and a connecting rod, the first scale body is provided with a first connecting piece, the second scale body is provided with a second connecting piece, the first connecting piece and the second connecting piece are respectively provided with protrusions and grooves which are distributed in a staggered mode, the protrusions and the grooves of the first connecting piece are matched with the protrusions and the grooves of the second connecting piece, and the square cap is connected with the connecting rod. The connecting rod penetrates through the protrusions and the grooves, square caps are arranged at the two ends of the connecting rod in a threaded mode, and the square caps are limited by the first limiting assembly and the second limiting assembly and do not spin. According to the utility model, the two wagon balances are connected in a splicing manner through the cooperation of the structures, so that the installation stability of the wagon balances can be improved, and the hidden dangers that the wagon balances turn over and the table surfaces are uneven can be eliminated.

The invention discloses a weak wake turbine guide vane based on a trailing edge discrete multi-scale structure. The weak wake turbine guide vane comprises a turbine guide vane body and the discrete multi-scale structure. The discrete multi-scale structure is arranged on the tail edge of the turbine guide vane body and used for enhancing the interaction of vortexes of different scales in the wake and the interaction of the wake and the main flow so as to weaken the wake of the turbine guide vane and reduce the pneumatic excitation amplitude of the wake. Structures of different scales are discretely arranged on the tail edge of the turbine guide vane body, the interaction of vortexes of different scales in the wake and the interaction of the wake and the main flow are enhanced, and therefore the wake of the turbine guide vane is weakened, the excitation amplitude of the wake of the turbine guide vane is reduced, and the design of the weak wake excitation turbine guide vane is achieved.

The application provides a complex design domain multi-scale isometric topology optimization method, device, equipment and medium, relates to the isometric topology optimization technical field, the method constructs macroscopic structure design domain and microscopic structure design domain based on optimization design parameter information, and constructs a material interpolation model based on the equivalent elastic tensor of the microscopic structure obtained according to the microscopic structure density and a preset equivalent elastic tensor prediction model, constructs a preset isometric topology optimization model, and updates the design variable based on the macroscopic sensitivity and the microscopic sensitivity obtained by calculation, obtains the target macroscopic design variable and the target microscopic design variable, generates the target macroscopic structure and the target microscopic structure respectively, carries out smoothing treatment on the rough optimization multi-scale structure formed by filling the target microscopic structure into the target macroscopic structure, and obtains a target complex design domain multi-scale structure model, which breaks through the bottleneck that the existing isometric concurrent topology optimization is difficult to optimize a complex multi-scale model.

The application provides a life prediction method and system for airtight multi-cell steel structure, and relates to the technical field of structure finite element analysis, which comprises the following steps: taking the airtight multi-cell steel structure as a prototype, determining a scale reduction ratio, and establishing a scaled structure of the airtight multi-cell steel structure; performing laboratory accelerated environmental testing and mechanical testing on the scaled structure under the equivalent service environment of the airtight multi-cell steel structure, obtaining mechanical response test data of the scaled structure, and determining test targets of the scaled structure at different times; taking scaled plate material data as input to establish a simulation model of the scaled structure and obtain a transfer coefficient; taking original plate material data as input to establish a full-size structure simulation model of the airtight multi-cell steel structure; and using the full-size structure simulation model to output crack propagation size and residual life of a target welded structure at different times under the full-size structure. The application effectively improves the accuracy of life prediction of the airtight multi-cell steel structure.

The invention relates to a high-ice-grade controllable-pitch propeller hub assembly test model design method, which comprises the steps of S1, designing a scaling test model structure based on structural characteristics of a high-ice-grade controllable-pitch propeller hub assembly prototype; s2, determining a similarity criterion of the prototype machine and the test model, and calculating a similarity coefficient on the basis of geometric dimensions, material performance and load parameters of the prototype machine and the test model; s3, scaling the structural size of the prototype machine according to the similarity coefficient, and determining geometric parameters of each part of the test model; s4, selecting a test model material meeting similar criteria with the prototype material performance, and ensuring that the stress-strain characteristics of the test model are consistent with those of the prototype; s5, reproducing an anti-loose structure of the prototype machine in the test model, wherein the anti-loose structure comprises an arrangement mode of anti-loose stainless steel wires and anti-loose strips; and S6, based on the zoomed structure and the selected material, assembling a test model, carrying out a static strength and fatigue life reliability test, and verifying the mechanical similarity between the test model and a prototype.

The invention relates to the technical field of multi-dimensional multi-scale structure modeling and design, in particular to a modeling and design method for a curved-surface thin-wall multi-scale structure. And an implicit modeling method is adopted to calculate a meso-structure symbolic distance function. And calculating a curved surface two-dimensional parameter domain based on a conformal mapping method. On the basis of the periodicity characteristic of a multi-scale structure, Fourier transform is adopted in a two-dimensional parameter domain, and the irregular two-dimensional multi-scale structure meeting the specified design parameter requirement is calculated according to the relative density, the relative angle and the mesostructure size parameter of the multi-scale structure. And performing interpolation calculation on the single-layer grid model of the three-dimensional curved surface multi-scale structure, and providing a multi-scale structure thickening method combining a normal vector offset method and an interpolation method to obtain a curved surface thin-wall multi-scale structure geometric model. The method is suitable for a complex microstructure expressed by a two-dimensional symbol distance function, the irregular multi-scale structure meets specified relative density, relative angle and microstructure size parameters, the method is suitable for a complex curved surface design domain, and connectivity of the microstructure is guaranteed.

The utility model relates to a movable scaleplate structure which comprises a forklift carriage, a forklift gantry, a fixed scaleplate, a movable scaleplate assembly and a linkage mechanism, the forklift carriage is installed on the outer side of the forklift gantry, the forklift carriage is in sliding connection with the forklift gantry, and the forklift gantry comprises a fixed outer gantry and a telescopic inner gantry; the telescopic inner door frame is directly embedded in the fixed outer door frame, the movable scale assembly is installed on the lower portion of the forklift sliding frame and the lower portion of the telescopic inner door frame, the movement of the forklift sliding frame and the telescopic inner door frame drives the movable scale assembly to move through the linkage mechanism, and the linkage mechanism is located on the movable scale assembly. The linkage mechanism is detachably connected with the movable scale assembly, and the fixed scale is installed on the fixed outer door frame and detachably connected with the fixed outer door frame. The goods shelf has the technical effect that the specific goods shelf layer height can be indicated layer by layer.

The application provides a method for measuring damping ratio of a building by using solid sound waves, which comprises calculating an attenuation coefficient α of sound waves in the building; obtaining an undamped natural angular frequency ω n of the building according to physical parameters of the building; and calculating a damping ratio ζ by bringing the attenuation coefficient α and the undamped natural angular frequency ω n into a formula. The method has the advantages of simple testing device, easy operation, simple operation process, greatly reduced testing cost and time, and suitability for on-site testing and large-scale structure application, especially for structures difficult to test in a laboratory environment. Through reasonable data analysis method, the damping ratio of the building can be relatively accurately measured.

The application relates to the technical field of data processing, and discloses a surrounding rock structure surface model generation method and system based on three-dimensional machine vision, aiming to solve the problems of single scale, poor spatial continuity and disconnection of engineering semantics of the existing method, and mainly comprising the following steps: multi-source point cloud collection and fusion, bottom-up aggregation path, top-down segmentation path, double-path feature fusion and structure surface parameterization output. Through the hierarchical processing framework of the bottom-up aggregation and top-down segmentation double paths, the multi-scale geometric features in the point cloud data are extracted, full-scale accurate identification from large-scale structure surfaces to microstructure surfaces is automatically realized, the problems of single scale, poor spatial continuity and disconnection of engineering semantics in the prior art are solved, the surrounding rock structure surface model generation efficiency and accuracy are improved, and the application is suitable for rock mass stability analysis.

The invention discloses a large-scale structure reanalysis method and system based on a separation-reconstruction framework, and the method comprises the steps: dividing a truss structure into a plurality of sub-domains, and constructing an extraction matrix; according to the total stiffness matrix and the extraction matrix, constructing a flexibility matrix corresponding to the sub-domain, and according to structural design modification, correcting the stiffness of the corresponding sub-domain, and constructing a maintenance reduction positive equation of the sub-domain; carrying out iterative solution on the descending maintenance positive equation of the subdomains according to a stable biconjugate gradient method to obtain displacement correction of each subdomain; mapping the displacement correction of each sub-domain to the overall structure, and reconstructing a displacement response vector of the overall structure; and calculating internal and external force residual vectors of the displacement response vector, judging whether the internal and external force residual vectors meet a preset convergence condition, and if so, outputting a structure reanalysis result. A local correction structure reanalysis method based on dimension reduction iteration is combined with a high-dimensional model characterization method and a residual error correction mechanism so as to improve the reanalysis efficiency of a large-scale or overall modification structure.

The application discloses a kind of topological optimization result geometric feature size control method based on image processing, comprising: step one: constructing double-scale topological optimization model;Step two: update double-scale level set function and then drive cross-scale structure configuration update;Step three: the boundary number and the coordinate point of boundary of cross-scale structure are obtained;Step four: based on the coordinate point of boundary of cross-scale structure, macroscopic configuration and microstructure are quantified respectively, and the geometric feature specific value of cross-scale structure is obtained;Step five: regulate the extremely thin component feature of cross-scale structure;Step six: regulate the hole feature of cross-scale structure;Step seven: regulate the special angle feature of cross-scale structure.The application realizes the effective regulation of accurate identification of various geometric features based on image processing technology under the level set framework, promotes design and manufacturing integration, and helps to improve design efficiency and shorten design cycle.

The invention discloses a photovoltaic glass calender upper roller bearing pedestal moving scale structure, and relates to the technical field of photovoltaic glass calenders, the photovoltaic glass calender upper roller bearing pedestal moving scale structure comprises a scale structure, the scale structure is fixedly installed at one end of an upper roller bearing support, the scale structure comprises a first metal substrate, and the front end of the first metal substrate is provided with a scale groove and a number groove; the upper roller bearing support is movably installed on the bearing seat supporting frame, an arrow indicating board is fixedly installed at one end of the bearing seat supporting frame, the arrow indicating board is located below the scale structure and comprises a second metal base plate, and an arrow groove is formed in the front end of the second metal base plate. By arranging the scale structure on the upper roller bearing support and arranging the arrow indication board on the bearing seat supporting frame, an operator can directly read the moving distance of the upper roller bearing seat without depending on experience or auxiliary tools, the adjusting precision and the production efficiency are remarkably improved, and the machine adjusting time and personal errors are reduced.

The present application discloses a calibration-free laser vision fusion measurement method and system for large-scale structure deformation, wherein the method sets a visual image acquisition instrument, a laser emitter emits a laser line on the inner wall of the target structure and images it on the visual image acquisition instrument, and determines the target offset of the pixel points between the laser line and the reference line based on the target image formed on the acquisition instrument. Prior to this, by controlling the actual displacement of the measured object and the pixel displacement corresponding to the displacement process imaged by the visual image acquisition instrument, the corresponding relationship between the pixel displacement and the proportional coefficient is fitted to obtain a predetermined relationship table. After obtaining the target offset corresponding to the target image, the proportional coefficient corresponding to the target offset is obtained by querying the predetermined relationship table; based on the proportional coefficient, the target offset is converted into the actual physical displacement caused by the deformation of the target structure. The actual displacement can be obtained only through pixel displacement, without the need for calibration in a complex on-site working environment.

The invention discloses a scale structure and a snakelike robot. The scale structure comprises a first scale unit and a second scale unit which are parallel to each other and connected in sequence. A first rotating section is arranged on the side, close to the second scale unit, of the first scale unit, the first scale unit is connected with the second scale unit through the first rotating section, and a first hollow area is arranged at the position, located on one side of the first rotating section, of the first scale unit; a second rotating section in the length direction of the second scale unit is arranged at the position, close to the first scale unit, of the second scale unit, and a second hollow area is arranged at the position, between the second rotating section and the first rotating section, of the second scale unit; deformation of the scale structure can be directly and passively triggered by ground acting force without depending on additional control signal input, natural feedback from the environment to the robot structure is achieved, the overall complexity of the system is remarkably reduced, and adaptability and interaction efficiency of the robot to the complex environment are enhanced.

The invention provides a multi-material multi-scale structure design method and system based on deep learning. The method comprises the following steps: establishing a multi-material heterogeneous microstructure parameterized model; a data set is generated based on the constructed model, equidistant and uniform valuing is carried out on each cutting height at a set interval, each height variable can obtain a corresponding value, and then random sampling is carried out on all heights to obtain a data set of a corresponding sample scale; constructing an agent model, wherein the agent model is used for constructing a mapping relation between the height parameters of the microstructure and the corresponding performance parameters; based on neural network re-parameterization as an optimizer, performing optimization design by constructing constraints in a loss function and combining an agent model; and reconstructing the multi-material multi-scale configuration based on the optimized parameters.

The invention discloses a calibration-free large-scale structure deformation laser vision fusion measurement method and system, and the method comprises the steps: setting a visual image collection instrument, enabling a laser transmitter to transmit laser rays on the inner wall of a target structure, and enabling the laser rays to be imaged on the visual image collection instrument; and determining a target offset of a pixel point between the laser line and the reference line based on a target image formed on the acquisition instrument. And before this, fitting the corresponding relation between the pixel displacement and the proportionality coefficient by controlling the actual displacement of the measured object and the pixel displacement corresponding to the imaging of the visual image acquisition instrument in the displacement process to obtain a predetermined relation table. After a target offset corresponding to the target image is obtained, obtaining a proportionality coefficient corresponding to the target offset by querying a predetermined relation table; and converting the target offset into an actual physical displacement generated by deformation of the target structure based on the proportionality coefficient. The actual displacement can be obtained only through the pixel displacement, and calibration does not need to be carried out in an on-site complex working condition environment.

The present application relates to the technical field of structural fatigue evaluation, in particular to a pipe truss structure fatigue analysis parameter optimization method and system based on deep learning, the steps are as follows: a deep learning agent model is constructed by fusing a multi-scale structure graph encoder, a working condition parameter encoder and a damage dimension curve decoder; the model is trained by using a high-fidelity training data set; for a target truss structure and a load working condition, the damage dimension relationship curve of the target truss structure is predicted by the model; the optimal dimension of the rain flow counting matrix is determined by analyzing the convergence characteristics of the curve; and the multi-working condition damage values are predicted by the model, and the dominant working condition is screened out based on the damage contribution percentage. The present application replaces the traditional iterative simulation with the deep learning agent model, realizes the rapid optimization of the key parameters in the fatigue analysis, significantly improves the efficiency and accuracy of the large truss structure fatigue evaluation, and improves the generalization ability of the model by fusing the multi-scale structure graph encoder to capture the physical effects of the key hot spots.

The invention relates to a compression device (1) intended to be arranged around a limb (M) of a living being, said compression device (1) comprising a sleeve (10) extending lengthwise around a longitudinal axis (X1) and having several compression rings (A) juxtaposed along said longitudinal axis, each compression ring (A) comprising a strip (2) of material developed at least partially around the longitudinal axis (X1), characterized in that the compression rings (A) are connected to each other by a first connection (L1) along a first direction (X2) parallel to the longitudinal axis, and in that the strips (2) of material are connected to each other, along a second direction (X3) parallel to the longitudinal axis, distinct from the first direction (X2), by a second sliding connection (L2) so as to form a scale structure. Figure to be published with the abstract: Figure 2