Cargo box floor structure strength design optimization method and system based on load prediction

Abstract

The application discloses a load prediction-based cargo box bottom plate structure strength design optimization method and system, belongs to the technical field of freight vehicle structure design, and through construction of a load prediction model of dynamic distribution states of goods in a cargo box, calculation of maximum equivalent stresses of each node in a bottom plate grid area and normalization processing, a stress intensity coefficient matrix is formed; a high-risk area is identified and a reinforcement target area set is constructed; initial parameters of a bottom plate structure are acquired, a finite element model is established, and adjustable structure parameters are set in the target area; a candidate structure design scheme set is generated, simulation calculation is executed, structure total mass, maximum stress value and first-order natural frequency of each scheme are extracted, and a performance evaluation vector is formed; an optimal design scheme is selected through a multi-objective optimization algorithm, and structure parameters thereof are output as a final design result; the application realizes structure optimization design of the cargo box bottom plate under complex load working conditions, and has high engineering adaptability and practical value.

Description

Technical Field

[0001] This invention relates to the field of freight vehicle structural design technology, specifically to a method and system for optimizing the structural strength design of cargo box floor based on load prediction. Background Technology

[0002] Currently, road freight vehicles frequently encounter non-standard load distributions when traversing unpaved road conditions such as urban-rural fringe areas and mountainous gravel roads. These include special load situations such as short-term concentrated loads caused by tipping loading and unloading, sudden off-center loading, and uneven particle accumulation. These conditions can lead to structural fatigue, cracks, and even perforations in the cargo box floor. Particularly when transporting certain special bulk materials (such as mineral powder, industrial waste, and construction backfill materials), the complex and variable loading conditions make it impossible to accurately recreate the load distribution using traditional rule-based loading models.

[0003] Existing cargo box floor structures are mostly designed using static uniformly distributed load models, assuming constant and stable loads. However, in actual operation, the floor structure is frequently subjected to localized, short-term, high-intensity impact loads, resulting in a severe lack of design margin. To improve structural reliability, some manufacturers choose to thicken the floor and increase the number of longitudinal and transverse beams. While this can temporarily delay failure, it significantly increases the overall vehicle weight, reduces transportation efficiency, and lowers economic viability. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for optimizing the structural strength design of cargo box floor plates based on load prediction, so as to overcome the shortcomings in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing the structural strength design of a cargo box floor based on load prediction, comprising:

[0006] S100. Collect operating condition data in the target transportation route and construct a load prediction model to describe the dynamic distribution of goods in the cargo container.

[0007] S200. Based on the load prediction model, extract the load spatial distribution function of the cargo box floor at different time periods, and calculate the maximum equivalent stress per unit area of ​​each node in the grid region of the cargo box floor.

[0008] S300: The maximum equivalent stress is normalized to form a regional stress intensity coefficient matrix, which is used to reflect the structural bearing risk level of each base plate location.

[0009] S400. Based on the base plate areas in the stress intensity coefficient matrix where the stress intensity coefficient is greater than the first set threshold, establish a set of base plate reinforcement target areas, the set including at least one high-risk node area.

[0010] S500: Obtain the initial design parameters of the base plate structure, including beam thickness, beam spacing, base plate thickness and material type, construct the structural finite element model, and mark the target set region for base plate reinforcement in the structural finite element model;

[0011] S600. In the structural finite element model, the local structural parameters of the target set region for base plate reinforcement are adjusted to generate a set of candidate structural design schemes, including several local reinforcement design sub-schemes.

[0012] S700: Perform simulation calculations under load prediction for each local strengthening design sub-scheme to obtain its total structural mass, maximum stress value and first natural frequency, and form a performance evaluation vector.

[0013] S800: Based on the preset objective function, perform multi-objective optimization selection on all performance evaluation vectors to determine the optimal design scheme, and use the corresponding structural parameters as the final design output of the cargo box floor.

[0014] Preferably, the operating condition data includes vehicle speed, road surface type, turning radius, braking intensity, and loading / unloading mode; the construction of a load prediction model to describe the dynamic distribution of goods in the cargo container includes: collecting video image data of the loading status of goods during transportation and acquiring the spatial stacking pattern of goods in the cargo container in real time; performing three-dimensional reconstruction on the image data to obtain a volume distribution cloud map of the goods stacking at different times, and dividing it into multiple regular sub-regions; combining the vehicle's acceleration / deceleration status and turning path data, using a multi-factor regression algorithm to train a motion trajectory function of the cargo's centroid changing over time; based on the motion trajectory function and the volume distribution cloud map, calculating the average load density per unit time in each sub-region, and constructing a prediction model for the change of cargo load in the spatiotemporal domain.

[0015] Preferably, the calculation of the maximum equivalent stress per unit area of ​​each node in the grid region of the cargo box floor includes:

[0016] The load prediction model is discretized into a sequence of two-dimensional load matrices with fixed time steps, each matrix corresponding to the load distribution of a time period;

[0017] A two-dimensional finite element mesh model of the cargo box floor is established. Shell elements are used to divide the mesh region, and each load matrix is ​​mapped to the floor mesh node at the corresponding time moment as the boundary load input.

[0018] The transient dynamic analysis method is used to solve the two-dimensional finite element analysis model F time step by time to obtain the equivalent stress response value of each grid node in all time steps.

[0019] The maximum equivalent stress of all nodes over the entire time series is statistically analyzed to obtain the maximum equivalent stress distribution map per unit area.

[0020] Preferably, the normalization of the maximum equivalent stress to form a regional stress intensity coefficient matrix includes:

[0021] Extract the maximum equivalent stress value of all nodes in the mesh of the cargo box bottom plate and construct a two-dimensional stress distribution matrix S, where each element S(i,j) represents the maximum equivalent stress at node (i,j);

[0022] Determine the yield strength of the structural material as a reference upper limit for normalization;

[0023] Normalize all stress values ​​in the two-dimensional stress distribution matrix S, calculate the normalized stress coefficients, and obtain the normalized stress intensity coefficient matrix R.

[0024] Preferably, the set of target areas for base plate reinforcement includes:

[0025] Traverse the stress intensity coefficient matrix R, extract all nodes whose stress intensity coefficient is greater than the first set threshold RH, and form a high-risk node set P;

[0026] Based on the spatial connectivity of the high-risk node set P, a connected component identification algorithm is used to cluster adjacent high-risk nodes to form multiple high-risk node regions.

[0027] For each high-risk node region, calculate its area, center coordinates, and surrounding average stress intensity coefficient as structural evaluation indicators;

[0028] High-risk areas that meet the preset area threshold and average stress intensity coefficient threshold are included in the target area set for base plate reinforcement.

[0029] Preferably, marking the target set region for base plate reinforcement in the structural finite element model includes:

[0030] Extract the initial structural design parameters of the target cargo box floor, including the cross-sectional thickness of the longitudinal and transverse beams, beam spacing, floor panel thickness, and material type of each component;

[0031] Based on the design parameters, a three-dimensional finite element model including the base plate, supporting beams and frame components was constructed using shell-beam hybrid modeling technology.

[0032] Map the set of reinforcement target regions Rtarget to the base plate node region of the model, and set variable parameter identifiers in this region;

[0033] Define boundary constraints and load input interfaces.

[0034] Preferably, the step of adjusting the local structural parameters of the target set region for base plate reinforcement in the structural finite element model includes:

[0035] For the set of target reinforcement areas Rtarget, multiple adjustable design parameters are set, including panel thickness increment, local material replacement options, and beam spacing fine-tuning values;

[0036] Based on each set of parameter combinations, a corresponding local structural reinforcement configuration is generated and embedded into the finite element model to form a set of local reinforcement design sub-schemes.

[0037] A parameter-driven modeling script is used to automatically generate multiple candidate structural models with different local structural configurations in batches.

[0038] All sub-scheme models are saved as a candidate structural design scheme set D.

[0039] Preferably, the formation of the performance evaluation vector includes:

[0040] Finite element analysis was performed on each sub-scheme model in the candidate structural design scheme set, and time-varying surface loads corresponding to the load prediction model were applied.

[0041] Perform transient structural simulation analysis and extract the maximum equivalent stress value σmax and corresponding node coordinates for each model in the entire time domain;

[0042] The total structural mass m of the scheme is automatically calculated based on the material and structural element properties.

[0043] The first natural frequency f1 of each sub-scheme is extracted by modal analysis, and σmax, m and f1 are combined to form the performance evaluation vector V=[σmax,m,f1].

[0044] Preferably, the step of performing multi-objective optimization selection on all performance evaluation vectors according to a preset objective function includes:

[0045] Construct a set of objective functions G, which includes at least three performance objectives: minimizing the total mass of the structure, minimizing the maximum equivalent stress, and maximizing the first-order natural frequency.

[0046] The performance evaluation vector V of each scheme in the candidate structural design scheme set D is normalized.

[0047] The Pareto front optimization strategy is adopted to screen all solutions in the normalized objective space for non-dominated solutions;

[0048] The optimal solution Dopt with the best overall performance is selected from the Pareto optimal solution set, and its structural parameters are extracted as the final optimized design output of the cargo box floor.

[0049] This invention also provides a load prediction-based structural strength design optimization system for cargo box floor plates, comprising:

[0050] Load modeling module: Collects operating condition data in the target transportation route and constructs a load prediction model to describe the dynamic distribution of goods in the cargo container;

[0051] Stress calculation module: Based on the load prediction model, extract the load spatial distribution function of the cargo box floor at different time periods, and calculate the maximum equivalent stress per unit area of ​​each node in the grid region of the cargo box floor.

[0052] Structural risk identification module: Normalizes the maximum equivalent stress to form a regional stress intensity coefficient matrix, which is used to reflect the structural bearing risk level at each base plate location;

[0053] Reinforcement Area Extraction Module: Based on the base plate areas in the stress intensity coefficient matrix where the stress intensity coefficient is greater than the first set threshold, establish a set of base plate reinforcement target areas, which includes at least one high-risk node area;

[0054] Parameter mapping module: Obtains the initial design parameters of the base plate structure, including beam thickness, beam spacing, base plate thickness and material type, constructs the structural finite element model, and marks the target set region for base plate reinforcement in the structural finite element model;

[0055] Scheme generation module: Adjusts local structural parameters of the target set region for base plate reinforcement in the structural finite element model to generate a set of candidate structural design schemes, including several local reinforcement design sub-schemes;

[0056] Simulation module: Performs simulation calculations under load prediction for each local strengthening design sub-scheme to obtain its total structural mass, maximum stress value and first natural frequency, forming a performance evaluation vector;

[0057] Design output module: Based on the preset objective function, perform multi-objective optimization selection on all performance evaluation vectors, determine the optimal design scheme, and use the corresponding structural parameters as the final design output of the cargo box floor.

[0058] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0059] 1. This invention effectively solves the problem of inaccurate dynamic load distribution in traditional cargo box floor structure design by constructing a load prediction model based on actual operating conditions. By performing spatiotemporal mapping of the load distribution and combining it with finite element simulation analysis, it can accurately identify weak areas of the structure, realize quantitative assessment and targeted reinforcement of high-risk areas of the floor, and significantly improve the pertinence and scientific nature of structural strength matching.

[0060] 2. This invention introduces a multi-objective optimization strategy, incorporating the total structural mass, maximum equivalent stress, and first-order natural frequency into a unified evaluation system. A candidate scheme set is established, and the optimal design is selected based on the Pareto front, balancing structural lightweighting and safety requirements. This method achieves an effective balance between structural performance and manufacturing cost, improves design automation and engineering applicability, and is suitable for structural optimization design scenarios in various commercial vehicles or heavy-duty equipment. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0062] Figure 1 This is a flowchart of the method of the present invention.

[0063] Figure 2 This is a flowchart of the system modules of the present invention. Detailed Implementation

[0064] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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, 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.

[0065] Example 1, please refer to Figure 1 As shown in this embodiment, the cargo box floor structure strength design optimization method based on load prediction includes:

[0066] S100. Collect operating condition data in the target transportation route and construct a load prediction model to describe the dynamic distribution of goods in the cargo container.

[0067] S200. Based on the load prediction model, extract the load spatial distribution function of the cargo box floor at different time periods, and calculate the maximum equivalent stress per unit area of ​​each node in the grid region of the cargo box floor.

[0068] S300: The maximum equivalent stress is normalized to form a regional stress intensity coefficient matrix, which is used to reflect the structural bearing risk level of each base plate location.

[0069] S400. Based on the base plate areas in the stress intensity coefficient matrix where the stress intensity coefficient is greater than the first set threshold, establish a set of base plate reinforcement target areas, the set including at least one high-risk node area.

[0070] S500: Obtain the initial design parameters of the base plate structure, including beam thickness, beam spacing, base plate thickness and material type, construct the structural finite element model, and mark the target set region for base plate reinforcement in the structural finite element model;

[0071] S600. In the structural finite element model, the local structural parameters of the target set region for base plate reinforcement are adjusted to generate a set of candidate structural design schemes, including several local reinforcement design sub-schemes.

[0072] S700: Perform simulation calculations under load prediction for each local strengthening design sub-scheme to obtain its total structural mass, maximum stress value and first natural frequency, and form a performance evaluation vector.

[0073] S800: Based on the preset objective function, perform multi-objective optimization selection on all performance evaluation vectors to determine the optimal design scheme, and use the corresponding structural parameters as the final design output of the cargo box floor.

[0074] In this invention, constructing a dynamic load prediction model is a crucial step in accurately describing the spatial distribution of goods inside the cargo container over time. This model is used for subsequent calculations of local stress in the base plate and strength optimization. Its specific construction process includes:

[0075] First, industrial-grade vehicle-mounted cameras and LiDAR (Light Detection and Ranging) devices are installed on the top or side of the vehicle's cargo box to collect video image data and point cloud data of the cargo stacking status during typical transportation processes. The cameras provide two-dimensional visible light images, while the LiDAR acquires three-dimensional depth information.

[0076] To ensure complete coverage of the cargo container area, the camera frame rate was set to 30 frames per second, and the LiDAR point cloud sampling frequency was no less than 10Hz to guarantee temporal and spatial continuity. The collected data was timestamped using a synchronization mechanism to ensure consistency between image and depth information.

[0077] The acquired images and point cloud data were fused together, and a structured light registration algorithm and a depth projection algorithm were used to complete the 3D reconstruction of the cargo stacking morphology. The 3D model was represented using a voxel grid, and the reconstruction results were exported as a volume distribution cloud map.

[0078] The volume distribution cloud map is divided into several regular sub-regions (e.g., each region is a 0.5m × 0.5m × 0.5m cube) to refine the spatial load analysis. Each sub-region is identified by its number (i,j,k), and its volume percentage and material density estimate are calculated to infer the cargo mass distribution in that region.

[0079] By combining vehicle driving status data, including real-time acceleration, speed, turning radius, and loading / unloading action records, the displacement of the cargo's center of gravity at different points in time is tracked.

[0080] By combining vehicle dynamics data with the aforementioned three-dimensional mass distribution data, a multi-factor linear regression algorithm is used to establish the time-varying trajectory function G(t) of the cargo centroid, which is in the form: G(t) = [x(t), y(t), z(t)]; where x(t), y(t), and z(t) represent the displacement functions of the cargo centroid along the three spatial axes with time t, respectively. The training dataset is fitted using historical transportation data, with the goal of minimizing the sum of squared errors between the predicted centroid trajectory and the actual observed trajectory.

[0081] Based on the construction of the three-dimensional distribution and motion trajectory function G(t), the average load density ρ(i,j,t) of each sub-region at any time t is calculated using the principles of mass conservation and momentum distribution. The dynamic load prediction model L(x,y,t) is further defined as follows: ;in, This indicates the sub-region number containing the location (x, y), and A is the area of ​​the region on the horizontal projection plane.

[0082] Finally, L(x,y,t) represents the instantaneous load distribution function of the cargo at the horizontal position (x,y) of the cargo box at any time t.

[0083] In this invention, considering the complex loading conditions and dynamically uneven stress changes of the cargo box floor during transportation, a stress calculation method based on a predictive load model is designed to obtain the maximum equivalent stress per unit area of ​​each grid node. The specific implementation process includes:

[0084] The previously constructed dynamic load prediction model L(x,y,t) is discretized on the time axis, and a uniform time step Δt (e.g., Δt=0.5 seconds) is set, dividing the entire transportation cycle T into n time segments.

[0085] Every point in time (k=1 to n) corresponds to a two-dimensional load matrix This matrix describes the time of the cargo box floor. The load per unit area at each location at any given time (unit: kN / m²). The resulting load sequence is: .

[0086] This discretization process preserves the temporal characteristics of the model and facilitates gradual loading and processing during subsequent finite element analysis.

[0087] A two-dimensional finite element analysis model F of the cargo box floor is constructed, using shell elements to model the plate to account for plate thickness, bending stiffness, and in-plane stress characteristics. The floor model is divided into equally spaced mesh nodes in the horizontal direction (x,y), for example, each element is a 0.1 m × 0.1 m square shell element.

[0088] For each moment The corresponding load matrix The surface nodes of the two-dimensional finite element analysis model F are mapped to their spatial locations and applied as instantaneous surface loads to the center nodes of each element. The load units remain constant. It is automatically converted into nodal concentrated force input.

[0089] After the loads are applied at each time step, transient dynamic analysis is performed based on the two-dimensional finite element analysis model F to solve the structural response of all mesh nodes at each time point step by step.

[0090] Using Von Mises equivalent stress as the stress index, the equivalent stress value of each node at each time step is calculated. , where i,j represent node numbers and k represents time step numbers.

[0091] Post-processing is performed on the stress response of all nodes throughout the entire time series to extract the maximum equivalent stress experienced by each node over all time steps, defined as: ; where σ_max(i,j) represents the maximum equivalent stress per unit area borne by node (i,j) during the entire transportation process, in MPa.

[0092] All σ_max(i,j) are combined into a two-dimensional distribution map, denoted as the σ_max distribution map. This map can be used to identify high-risk areas of stress concentration, providing a basis for the next stage of structural reinforcement design and optimization.

[0093] To accurately identify high-risk areas in the cargo box floor during structural optimization design and avoid material compatibility errors caused by relying solely on absolute stress values, this invention introduces a normalization process. This process transforms the maximum equivalent stress into a regional stress intensity coefficient matrix, and based on this matrix, delineates structural load-bearing risk level zones. The process includes:

[0094] After completing the finite element analysis and extracting the maximum equivalent stress of all nodes, a two-dimensional stress distribution matrix S is constructed. This matrix consists of the stress values ​​of all nodes in the base plate mesh, and each element S(i,j) represents the maximum equivalent stress (unit: MPa) borne by the (i,j)th node in the entire time domain.

[0095] The dimensions of the two-dimensional stress distribution matrix S are consistent with those of the base plate mesh model, which can comprehensively reflect the stress response distribution law of the base plate structure under actual load conditions.

[0096] To achieve normalization, the material's yield strength σy needs to be introduced as a standardized benchmark value. The yield strength refers to the maximum stress a material can withstand before plastic deformation, and its unit is MPa.

[0097] For example, if the base plate is made of Q345 high-strength steel, its nominal yield strength can be taken as σy = 345 MPa; if it is made of aluminum alloy 6061-T6, then σy = 276 MPa. This value can be determined through material technical standards or measured mechanical properties. The yield strength, as a standard upper limit, helps to convert stress values ​​under different material conditions into a unified risk assessment dimension.

[0098] Normalizing each element S(i,j) in the two-dimensional stress distribution matrix S yields the stress intensity coefficient matrix R, which is calculated as follows: After normalization, R(i,j) represents the stress intensity coefficient at node (i,j), typically ranging from 0 to 1. If the coefficient at a node exceeds 1, it indicates that the maximum stress has exceeded the yield strength of the material, placing it in a critical region.

[0099] The stress intensity coefficient matrix R can be used to intuitively measure the load utilization rate at various locations of the base plate structure, and is the basis for the next step of risk area identification.

[0100] To effectively strengthen the local structure of the cargo box floor, this invention proposes a target region identification method based on a stress intensity coefficient matrix. This method can accurately extract areas with concentrated load-bearing risks, forming a set of input regions for optimized design. The specific steps include:

[0101] After obtaining the normalized stress intensity coefficient matrix R, iterate through it to extract all nodes (i,j) that satisfy the following condition: R(i,j)≥RH; where RH is a set first risk threshold, typically ranging from 0.85 to 0.95, with the specific value determined based on material properties and structural safety margin. Nodes satisfying this condition are grouped into a set P, called the high-risk node set, which reflects the distribution of areas with a high probability of potential failure in the structure.

[0102] Based on the two-dimensional spatial relationships of nodes in the high-risk node set P, an 8-neighborhood connection method is used to construct a node graph, and a connected component labeling algorithm is applied to cluster adjacent nodes to form several spatially continuous high-risk regions: Each It is an independent high-risk node region containing multiple connected high-risk nodes.

[0103] For each high-risk node region Ck, its key structural characteristic indicators are further calculated, including:

[0104] Region area Ak: is the number of nodes in the region multiplied by the unit area;

[0105] Region centroid coordinates : The arithmetic mean of the coordinates of all nodes within the region;

[0106] mean stress intensity coefficient : The average value of R(i,j) for all nodes in the region.

[0107] The above indicators are used to measure the importance of the region and its impact on the overall performance of the structure, and serve as the basis for subsequent screening.

[0108] Set an area threshold Amin (e.g., 0.05 m²) and an average stress intensity threshold Ravg (e.g., 0.90) to filter all high-risk node regions Ck, retaining regions that simultaneously meet the following condition: Ak ≥ Amin; All regions that meet the conditions are aggregated to form the target region set Rtarget for base plate reinforcement. This set contains at least one spatially connected high-risk node region, which is the direct input region for local structural parameter optimization.

[0109] To achieve structural strength optimization based on local load response, this invention further constructs a finite element simulation model of the base plate based on initial structural parameters and integrates the identified high-risk areas into the model. Specifically, this includes:

[0110] Extract the main structural parameters of the cargo box floor from design drawings, CAD files, or manufacturing databases, including but not limited to: longitudinal beam thickness tv and transverse beam thickness th, in mm; longitudinal beam spacing Lv and transverse beam spacing Lh, in mm or m; floor panel thickness tp, in mm; and material type Mx, including material information for the floor and beams (such as Q345, 6061-T6, etc.), corresponding to different material properties such as elastic modulus, density, and yield strength.

[0111] Based on the initial parameters mentioned above, a three-dimensional structural finite element model F0 is established using a modeling approach combining shell elements (for the base plate and panel) and beam elements (for the supporting structure). The specific modeling steps are as follows:

[0112] The base plate area is divided into two-dimensional meshes using shell elements such as SHELL181 or S4R.

[0113] The longitudinal and transverse beams are modeled using BEAM188 or B31 elements, arranged below the shell elements and rigidly connected to their nodes.

[0114] All beams and shells achieve structural integrity through shared nodes or rigid connection units;

[0115] Material properties are loaded from the material database, and a corresponding material model is assigned to each type of structural component.

[0116] The previously identified target areas for base plate reinforcement, Rtarget, are mapped onto the model's base plate shell element mesh, and their spatial positions in the model are located by coordinate matching or node number matching.

[0117] Within each mesh element corresponding to a reinforced region, adjustable design variable identifiers are set to indicate that this region will participate in subsequent design variable optimization calculations. Adjustable variables include: local shell element thickness variable Δtp; local material strength grade variable ΔM; and local support beam connection status variables (such as whether or not stiffeners are introduced). This annotation process establishes a mapping relationship between "region - structural parameters - optimization variables," serving as crucial input preparation for local structural adjustments.

[0118] To achieve high-precision analysis and optimization simulation, the three-dimensional structural finite element model F0 needs to be further configured with a complete analysis input environment, including: boundary condition settings: such as fixed support, simply supported state, or connection stiffness simulation of the bottom plate edge; load interface settings: configuring the input binding relationship with the time series dynamic load prediction model L(x,y,t) so that the predicted load can be automatically applied to the shell element surface; material failure criterion settings: such as yield strength, maximum deformation limit, etc., used to evaluate whether the local response exceeds the limit.

[0119] After completing the finite element modeling of the base plate structure and annotating the reinforcement target area Rtarget, in order to optimize the mass and stiffness performance of the structure as much as possible while meeting the strength requirements, this invention adjusts the local structural parameters within the Rtarget area and generates a series of candidate structural design schemes through multi-variable combinations, providing a foundation for subsequent multi-objective optimization. This process includes:

[0120] For the target area Rtarget for base plate reinforcement, several adjustable local structural design parameters are defined, including but not limited to: base plate thickness increment Δtp: such as increasing by 1 mm, 2 mm, or 3 mm; beam spacing adjustment ΔL: such as reducing the longitudinal beam / transverse beam spacing by 10% and 15% respectively; material replacement option M′: such as replacing the original Q235 steel with Q345 high-strength steel or 6061-T6 aluminum alloy; local support reinforcement structure type T: such as whether to add reinforcing ribs or transverse auxiliary beams. These parameters are set as a set of discrete variables and combined into multiple parameter configurations using a Cartesian product to construct different reinforcement schemes.

[0121] For each set of structural parameter combinations Pk, the original structural properties are replaced in the Rtarget region of the three-dimensional structural finite element model F0 to generate a new local configuration model.

[0122] For example, when a certain sub-scheme The combination is: Δtp = 2 mm, If the material is Q345 and reinforcing ribs are added, then in this area, the thickness of the bottom shell element increases by 2 mm from the original value, the support beams are rearranged with a denser spacing, and the material properties are changed. Simultaneously, transverse stiffening elements are added at the center. Each Pk configuration corresponds to a structural sub-scheme, denoted as Dk, and can form a total of [number missing] sub-schemes. to Multiple reinforced design sub-models.

[0123] To improve efficiency, this invention automatically performs modeling and local parameter modification using CAE parameter modeling scripts (such as Python + ABAQUS or APDL scripts). The process is as follows: import the original 3D structural finite element model F0; read the structural parameter combination Pk; perform parameter replacement in the Rtarget region; output and save the modified finite element model Dk.

[0124] All generated local enhancement sub-schemes The candidate structural design schemes are uniformly archived and formed into a set D, which will be used for subsequent performance simulation and multi-objective optimization analysis.

[0125] Each Dk model contains complete structural definitions, boundary constraints, and load interface settings, and can be directly used for batch calculations in the finite element analysis platform.

[0126] After generating the candidate structural design scheme set D, to achieve multi-dimensional performance evaluation of each local strengthening sub-scheme under real loads, this invention proposes a finite element simulation calculation method based on a load prediction model. This method can simultaneously extract the total structural mass, maximum equivalent stress, and first-order natural frequency to form a structural performance evaluation vector V. Specifically, it includes:

[0127] Each sub-scheme Dk in the candidate structural design scheme set is imported into the finite element analysis platform (such as ABAQUS, ANSYS, etc.) as an input model, and the load sequence corresponding to the aforementioned load prediction model L(x,y,t) is called.

[0128] The specific operation is as follows: Discretize the dynamic load prediction model L(x,y,t) into multiple time-series surface loads. In the Dk model, Lk is applied to the corresponding node region, and the time step Δt and total simulation duration T are set. Material models, boundary constraints, and contact conditions are set, and transient dynamic simulation (Transient Structural Analysis) is performed. This process can obtain the time-varying response of the structure during actual transportation and accurately simulate the evolution of load propagation and local stress concentration.

[0129] After the simulation calculation is completed, post-processing analysis is performed on the structural stress distribution at each time step within the entire analysis time domain. The Von Mises equivalent stress criterion is used to extract the stress value of each node (i,j) at each time step tk. The calculation method is as follows: In the formula, This represents the normal stress in the x-axis direction. This represents the normal stress in the y-axis direction. This represents the normal stress in the z-axis direction. This represents the shear stress in the xy plane. This represents the shear stress in the yz plane. The shear stress on the zx plane is represented by the maximum value selected from all nodes and all time steps, denoted as: The coordinates of the nodes where the stress occurs are recorded for analysis of structural weak points.

[0130] Based on the material density ρ, geometric parameters (such as thickness and cross-sectional area), and element volume of the shell and beam elements defined in the Dk model, perform mass statistics to calculate the total mass m of the entire structural model, in kg.

[0131] This process can be completed by calling the mass summary function in the finite element platform, or by automatically extracting parameters and calculating them in the modeling script: , where i is the element number, ρi is the material density of the i-th element, and Vi is the element volume or equivalent volume. This index is used to measure the impact of strengthening schemes on the structural lightweighting target.

[0132] Modal analysis was performed on the Dk model to extract its natural frequency sequence f1, f2, ..., fn under no-load conditions, where f1 is the first natural frequency (in Hz), reflecting the principal mode stiffness characteristics of the structure under free vibration. The extracted key performance parameters—σmax: upper limit of structural strength (in MPa); m: total structural mass (in kg); f1: first natural frequency (in Hz)—were combined to form a structural performance evaluation vector V, denoted as V = [σmax, m, f1]. This performance vector will serve as input for subsequent multi-objective optimization algorithms (such as lightweight-strength-stiffness), supporting automated scheme selection and optimal design recommendation.

[0133] After completing the simulation analysis of the candidate structural design scheme set D and extracting the corresponding performance evaluation vector V=[σmax,m,f1], this invention proposes an optimal structure screening method based on a multi-objective optimization algorithm to achieve quantitative comparison and selection of schemes. Through performance normalization and Pareto analysis, it comprehensively balances structural lightweighting, strength and safety, and dynamic performance to determine the final design output. Specifically, it includes:

[0134] A predefined set of objective functions G is used to evaluate the overall performance of each candidate solution, and includes at least the following three objectives: Minimize the total structural mass m to achieve the goal of lightweighting the chassis; Minimize the maximum equivalent stress σmax to ensure that the structural strength does not exceed the limit; The objective function is to maximize the first natural frequency f1 to improve vibration resistance and structural stiffness. Therefore, the set of objective functions can be expressed as: The above objectives are interdependent, and the global optimum cannot be obtained through a single optimization function; therefore, a multi-objective optimization method is required.

[0135] Since the dimensions of different objective items are inconsistent, the performance evaluation vector V=[σmax,m,f1] of all candidate schemes needs to be normalized to ensure the dimensionlessness and comparability of the optimization calculation.

[0136] The normalization method is as follows: For minimizing the objective (m and σmax), use: To maximize the objective (f1), use: Where Vmin and Vmax represent the minimum and maximum values ​​of the same objective term among all candidate solutions, respectively. The V vectors of all solutions are normalized to V′ and formed into a dimensionless performance matrix, which serves as the optimization input.

[0137] The normalized performance evaluation matrix is ​​input into the optimization module, and one of the following optimization strategies is adopted: multi-objective evolutionary algorithm (MOEA), such as NSGA-II or MOEA / D; or Pareto frontier analysis algorithm.

[0138] During the optimization process, the system searches for the non-dominated solution set of all schemes, that is, the solution set where no other scheme is simultaneously better on all objectives, denoted as the Pareto optimal solution set M. This solution set represents the compromise design results for different performance preferences.

[0139] From the Pareto optimal solution set M, the sub-solution with the best performance is selected according to the design weight or comprehensive scoring rules, denoted as: Dot∈M;

[0140] The scoring rules can adopt a weighted evaluation function or a decision algorithm such as TOPSIS, which comprehensively considers the relative importance of the three objectives (such as strength > stiffness > mass).

[0141] The structural design parameters contained in Dopt, including the bottom plate thickness, material type, beam arrangement, etc., are used as the final optimization result and output as the final structural design of the cargo box bottom plate.

[0142] Example 2, please refer to Figure 2 As shown in this embodiment, the cargo box floor structure strength design optimization system based on load prediction includes:

[0143] Load modeling module: Collects operating condition data in the target transportation route and constructs a load prediction model to describe the dynamic distribution of goods in the cargo container;

[0144] Stress calculation module: Based on the load prediction model, extract the load spatial distribution function of the cargo box floor at different time periods, and calculate the maximum equivalent stress per unit area of ​​each node in the grid region of the cargo box floor.

[0145] Structural risk identification module: Normalizes the maximum equivalent stress to form a regional stress intensity coefficient matrix, which is used to reflect the structural bearing risk level at each base plate location;

[0146] Reinforcement Area Extraction Module: Based on the base plate areas in the stress intensity coefficient matrix where the stress intensity coefficient is greater than the first set threshold, establish a set of base plate reinforcement target areas, which includes at least one high-risk node area;

[0147] Parameter mapping module: Obtains the initial design parameters of the base plate structure, including beam thickness, beam spacing, base plate thickness and material type, constructs the structural finite element model, and marks the target set region for base plate reinforcement in the structural finite element model;

[0148] Scheme generation module: Adjusts local structural parameters of the target set region for base plate reinforcement in the structural finite element model to generate a set of candidate structural design schemes, including several local reinforcement design sub-schemes;

[0149] Simulation module: Performs simulation calculations under load prediction for each local strengthening design sub-scheme to obtain its total structural mass, maximum stress value and first natural frequency, forming a performance evaluation vector;

[0150] Design output module: Based on the preset objective function, perform multi-objective optimization selection on all performance evaluation vectors, determine the optimal design scheme, and use the corresponding structural parameters as the final design output of the cargo box floor.

[0151] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for optimizing the structural strength design of cargo box floor plates based on load prediction, characterized in that: include: S100. Collect operating condition data in the target transportation route and construct a load prediction model to describe the dynamic distribution of goods in the cargo container. The operational data includes vehicle speed, road surface type, turning radius, braking intensity, and loading / unloading mode. The construction of a load prediction model to describe the dynamic distribution of goods within the cargo container includes: collecting video image data of the cargo loading status during transportation and acquiring the spatial stacking pattern of the goods within the cargo container in real time; performing three-dimensional reconstruction on the image data to obtain a volume distribution cloud map of the cargo stacking at different times, and dividing it into multiple regular sub-regions; combining the vehicle's acceleration / deceleration status and turning path data, using a multi-factor regression algorithm to train a motion trajectory function of the cargo's centroid changing over time; based on the motion trajectory function and the volume distribution cloud map, calculating the average load density per unit time in each sub-region, and constructing a prediction model for the change of cargo load in the spatiotemporal domain. S200. Based on the load prediction model, extract the load spatial distribution function of the cargo box floor at different time periods, and calculate the maximum equivalent stress per unit area of ​​each node in the cargo box floor grid region, including: The load prediction model is discretized into a sequence of two-dimensional load matrices with fixed time steps, each matrix corresponding to the load distribution of a time period; A two-dimensional finite element mesh model of the cargo box floor is established. Shell elements are used to divide the mesh region, and each load matrix is ​​mapped to the floor mesh node at the corresponding time moment as the boundary load input. The transient dynamic analysis method is used to solve the two-dimensional finite element analysis model F time step by time to obtain the equivalent stress response value of each grid node in all time steps. The maximum equivalent stress of all nodes over the entire time series is statistically analyzed to obtain the maximum equivalent stress distribution map per unit area; S300: The maximum equivalent stress is normalized to form a regional stress intensity coefficient matrix, which is used to reflect the structural bearing risk level of each base plate location. S400. Based on the base plate areas in the stress intensity coefficient matrix where the stress intensity coefficient is greater than the first set threshold, establish a set of base plate reinforcement target areas, the set including at least one high-risk node area. S500: Obtain the initial design parameters of the base plate structure, including beam thickness, beam spacing, base plate thickness and material type, construct the structural finite element model, and mark the target set region for base plate reinforcement in the structural finite element model; S600. In the structural finite element model, the local structural parameters of the target set region for base plate reinforcement are adjusted to generate a set of candidate structural design schemes, including several local reinforcement design sub-schemes. S700: Perform simulation calculations under load prediction for each local strengthening design sub-scheme to obtain its total structural mass, maximum stress value and first natural frequency, and form a performance evaluation vector. S800: Based on the preset objective function, perform multi-objective optimization selection on all performance evaluation vectors to determine the optimal design scheme, and use the corresponding structural parameters as the final design output of the cargo box floor.

2. The method for optimizing the structural strength design of cargo box floor based on load prediction according to claim 1, characterized in that: The normalization process for the maximum equivalent stress to form a regional stress intensity coefficient matrix includes: Extract the maximum equivalent stress value of all nodes in the mesh of the cargo box bottom plate and construct a two-dimensional stress distribution matrix S, where each element S(i,j) represents the maximum equivalent stress at node (i,j); Determine the yield strength of the structural material as a reference upper limit for normalization; Normalize all stress values ​​in the two-dimensional stress distribution matrix S, calculate the normalized stress coefficients, and obtain the normalized stress intensity coefficient matrix R.

3. The method for optimizing the structural strength design of cargo box floor based on load prediction according to claim 2, characterized in that: The set of target areas for base plate reinforcement includes: Traverse the stress intensity coefficient matrix R, extract all nodes whose stress intensity coefficient is greater than the first set threshold RH, and form a high-risk node set P; Based on the spatial connectivity of the high-risk node set P, a connected component identification algorithm is used to cluster adjacent high-risk nodes to form multiple high-risk node regions. For each high-risk node region, calculate its area, center coordinates, and surrounding average stress intensity coefficient as structural evaluation indicators; High-risk areas that meet the preset area threshold and average stress intensity coefficient threshold are included in the target area set for base plate reinforcement.

4. The method for optimizing the structural strength design of cargo box floor based on load prediction according to claim 3, characterized in that: The step of marking the target set region for base plate reinforcement in the structural finite element model includes: Extract the initial structural design parameters of the target cargo box floor, including the cross-sectional thickness of the longitudinal and transverse beams, beam spacing, floor panel thickness, and material type of each component; Based on the design parameters, a three-dimensional finite element model including the base plate, supporting beams and frame components was constructed using shell-beam hybrid modeling technology. Map the set of reinforcement target regions Rtarget to the base plate node region of the model, and set variable parameter identifiers in this region; Define boundary constraints and load input interfaces.

5. The method for optimizing the structural strength design of cargo box floor based on load prediction according to claim 4, characterized in that: The adjustment of local structural parameters for the target set region of the base plate reinforcement in the structural finite element model includes: For the set of target reinforcement areas Rtarget, multiple adjustable design parameters are set, including panel thickness increment, local material replacement options, and beam spacing fine-tuning values; Based on each set of parameter combinations, a corresponding local structural reinforcement configuration is generated and embedded into the finite element model to form a set of local reinforcement design sub-schemes. A parameter-driven modeling script is used to automatically generate multiple candidate structural models with different local structural configurations in batches. All sub-scheme models are saved as a candidate structural design scheme set D.

6. The method for optimizing the structural strength design of cargo box floor based on load prediction according to claim 1, characterized in that: The formation of the performance evaluation vector includes: Finite element analysis was performed on each sub-scheme model in the candidate structural design scheme set, and time-varying surface loads corresponding to the load prediction model were applied. Perform transient structural simulation analysis and extract the maximum equivalent stress value σmax and corresponding node coordinates for each model in the entire time domain; The total structural mass m of the scheme is automatically calculated based on the material and structural element properties. The first natural frequency f1 of each sub-scheme is extracted by modal analysis, and σmax, m and f1 are combined to form the performance evaluation vector V=[σmax, m, f1].

7. The method for optimizing the structural strength design of cargo box floor based on load prediction according to claim 1, characterized in that: The step of performing multi-objective optimization selection on all performance evaluation vectors according to a preset objective function includes: Construct a set of objective functions G, which includes at least three performance objectives: minimizing the total mass of the structure, minimizing the maximum equivalent stress, and maximizing the first-order natural frequency. The performance evaluation vector V of each scheme in the candidate structural design scheme set D is normalized. The Pareto front optimization strategy is adopted to screen all solutions in the normalized objective space for non-dominated solutions; The optimal solution Dopt with the best overall performance is selected from the Pareto optimal solution set, and its structural parameters are extracted as the final optimized design output of the cargo box floor.

8. A load-prediction-based structural strength design optimization system for cargo box floor plates, used to implement the load-prediction-based structural strength design optimization method for cargo box floor plates as described in any one of claims 1-7, characterized in that: include: Load modeling module: Collects operating condition data in the target transportation route and constructs a load prediction model to describe the dynamic distribution of goods in the cargo container; Stress calculation module: Based on the load prediction model, extract the load spatial distribution function of the cargo box floor at different time periods, and calculate the maximum equivalent stress per unit area of ​​each node in the grid region of the cargo box floor. Structural risk identification module: Normalizes the maximum equivalent stress to form a regional stress intensity coefficient matrix, which is used to reflect the structural bearing risk level at each base plate location; Reinforcement Area Extraction Module: Based on the base plate areas in the stress intensity coefficient matrix where the stress intensity coefficient is greater than the first set threshold, establish a set of base plate reinforcement target areas, which includes at least one high-risk node area; Parameter mapping module: Obtains the initial design parameters of the base plate structure, including beam thickness, beam spacing, base plate thickness and material type, constructs the structural finite element model, and marks the target set region for base plate reinforcement in the structural finite element model; Scheme generation module: Adjusts local structural parameters of the target set region for base plate reinforcement in the structural finite element model to generate a set of candidate structural design schemes, including several local reinforcement design sub-schemes; Simulation module: Performs simulation calculations under load prediction for each local strengthening design sub-scheme to obtain its total structural mass, maximum stress value and first natural frequency, forming a performance evaluation vector; Design output module: Based on the preset objective function, perform multi-objective optimization selection on all performance evaluation vectors, determine the optimal design scheme, and use the corresponding structural parameters as the final design output of the cargo box floor.

Images