Method and system for optimizing stability of a crossing operation based on mechanical modeling

Abstract

This invention discloses a method and system for optimizing the stability of scaffolding operations based on mechanical modeling, relating to the field of engineering design technology. The method involves disassembling the structure and generating mechanical property feature vectors for each part using a graph convolutional neural network. Using the minimization of overall mass and the maximization of global stability as the fitness, and combining attention weight adjustment and adaptive step-size iteration, the dimensions and materials of sub-parts are optimized. When the fitness converges or the maximum number of iterations is reached, the dimensions, materials, and mechanical characteristics of each sub-part are output, forming a solution that balances structural stability and economy. This method not only optimizes quality during the design process but also achieves a superior structural state through scientific analysis. While ensuring safety and stability, it enables lightweight design, improves design efficiency and optimization accuracy, and provides reliable technical support for engineering construction.

Description

Technical Field

[0001] This invention relates to the field of engineering design technology, and in particular to a method and system for optimizing the stability of cross-passage operations based on mechanical modeling. Background Technology

[0002] Cross-beam scaffolds typically need to be erected according to site conditions. Their height, span, support locations, member arrangements, and connection methods will all change with the construction environment. Their structural stability is a key concern in site design and construction management.

[0003] Furthermore, the truss structure is not supported by a single component, but rather by a system of multiple rods, nodes, and support units that work together to form a load-bearing system. A change in any dimension or material may alter the stability of the entire structure.

[0004] Currently, the design and selection of bridging frames largely rely on specifications, standard configurations, and the engineering experience of technical personnel. When there are many optional components, various material types, and complex on-site constraints, it is difficult for manual methods to fully reflect the cascading effects of changes in local components on overall stability. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a method for optimizing the operational stability of a gantry based on mechanical modeling, which includes obtaining the geometric structure of the gantry to be optimized;

[0008] After the geometric structure is solidified, it is disassembled to obtain the sub-parts of the crossing frame;

[0009] Each of the sub-parts is bound to a corresponding physical parameter, serving as an individual to be optimized;

[0010] After adjusting the physical parameters of all the individuals to be optimized, the overall operational stability of the crossing frame is used as the fitness to iteratively optimize the physical parameters of the individuals.

[0011] When the fitness converges or the maximum number of iterations is reached, the optimization results of the global rational parameters are output.

[0012] As a preferred embodiment of the bridging frame operation stability optimization method based on mechanical modeling described in this invention, the geometric structure includes the spatial form constituting the overall skeleton of the bridging frame.

[0013] As a preferred embodiment of the gantry operation stability optimization method based on mechanical modeling described in this invention, the sub-parts of the gantry include basic component units that are disassembled according to the physical connection relationship in the geometric structure, each unit having independent degrees of freedom of motion and structurally indivisible.

[0014] As a preferred embodiment of the stability optimization method for cross-passage operation based on mechanical modeling described in this invention, the physical parameters include the specific dimensions and material types of each sub-part.

[0015] Each type of material is of uniform density.

[0016] As a preferred embodiment of the stability optimization method for bridging operations based on mechanical modeling described in this invention, the adjustment of physical parameters for all individuals to be optimized includes:

[0017] A heterogeneous knowledge graph is constructed with size specifications, material types, and mechanical properties as nodes, and a graph convolutional neural network is used to extract features and vectorize the graph.

[0018] Using the size specifications and material type as input, the mechanical property feature vector of the sub-part is generated by utilizing the message passing mechanism of the graph convolutional neural network;

[0019] The mechanical property feature vector is the encoded result characterizing bending stiffness, compressive strength and damping characteristics;

[0020] For each material type of each sub-part, the corresponding geometric structure size constraint range is input into the graph convolutional neural network, and the adjustable space of the mechanical property feature vector of the sub-part is output.

[0021] Furthermore, by sampling each dimension at equal intervals within the constrained range of geometric structure dimensions, and inputting the dimension specifications of the sampling points into the graph convolutional neural network, the mechanical property feature vector of each sampling point is obtained;

[0022] The mechanical property feature vectors are sorted according to the sequence of sampling points. By smoothing the feature values ​​of each dimension of the mechanical property feature vectors, the fitting results of the mechanical properties with respect to size specifications are obtained for each type of material.

[0023] Integration is due to the discreteness of the adjustable space of mechanical properties of different types of materials, forming a feasible domain space of mechanical property eigenvectors that is discontinuous.

[0024] By simultaneously driving each sub-part to change its size and specifications in random directions within its corresponding constraint range, each individual component can simultaneously traverse its corresponding feasible domain space, thereby achieving optimization of the overall operational stability of the gantry under physical reality constraints.

[0025] As a preferred embodiment of the bridging frame operation stability optimization method based on mechanical modeling described in this invention, the fitness of the bridging frame is calculated globally using a multi-objective optimization model.

[0026] In each sub-part, the volume is calculated according to the size specifications, and then multiplied by the density corresponding to the material type to obtain the mass of the sub-part; the mass of each sub-part in the whole is accumulated to obtain the global mass data, and the objective function 1 is constructed with the minimum of the mass data;

[0027] The mechanical property feature vectors of all sub-parts are projected into the pre-trained global stability coupling space to generate the global stability of the straddle.

[0028] If the global stability is within the preset constraint range, then the maximum of the global stability is taken as the objective function 2; if the relative position or relative local curvature exceeds the constraint range, then an infinitely large cost is imposed on the objective function 2.

[0029] The coupling space for global stability is a high-dimensional mechanical logic field constructed using a deep neural network. By performing deep learning on a training set containing force transmission and geometric nonlinear responses between sub-parts, the physical coupling causal chain between local variations and overall responses within the strut is solidified. By capturing the interference and superposition effects of the mechanical feature vectors of the sub-parts during the projection process, key mechanical features characterizing the overall stiffness distribution and modal stability of the structure are extracted. After the mechanical features of each sub-part are nonlinearly interfered and superimposed in the coupling space, a measure of global stability is obtained.

[0030] For objective function 1 and objective function 2, the weighted sum is calculated according to the preset attention weights and used as the fitness. The optimization is then performed with the minimum fitness.

[0031] Specifically, when the objective function 2 is greater than the preset target structure stability threshold, the attention adjustment strategy is activated; otherwise, the attention weights are reset and measured according to the preset attention weights.

[0032] The attention adjustment strategy includes calculating the target function 2 using a monotonically increasing function to obtain a scaling factor, then scaling down the attention weights of the target function 2 using the scaling factor, and finally normalizing the attention to achieve attention adjustment and update.

[0033] As a preferred embodiment of the bridging frame operation stability optimization method based on mechanical modeling described in this invention, in the process of optimizing the overall operation stability of the bridging frame, the rate of change of the mechanical property feature vector with the change of the size specification is extracted according to the fitting result between the size specification and the mechanical property feature vector.

[0034] When each individual iterates by changing its size specifications, the standard step size is multiplied by an adjustment factor to obtain the actual step size of each individual in the current iteration stage;

[0035] Specifically, the size specification of the individual that has not been changed is obtained, and the change rate value v of the current size specification position is matched in the fitting result; the adjustment coefficient T is calculated as: T=θ / (ε+v); θ represents a constant used for proportional adjustment; ε represents a constant term to prevent the denominator from being 0.

[0036] Secondly, the present invention provides a gantry operation stability optimization system based on mechanical modeling, comprising a data acquisition unit for acquiring the geometry of the gantry to be optimized;

[0037] The decomposition unit, after solidifying the geometric structure, disassembles the geometric structure to obtain the sub-parts of the crossing frame;

[0038] The binding unit binds each of the sub-parts to the corresponding physical parameters, treating them as individual components to be optimized;

[0039] The iterative unit, after adjusting the physical parameters of all the individuals to be optimized, uses the overall operational stability of the crossing frame as the fitness to iteratively optimize the physical parameters of the individuals.

[0040] The output unit outputs the optimization results of the global rational parameters when the fitness converges or the maximum number of iterations is reached.

[0041] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein the computer program, when executed by the processor, implements any step of the method for optimizing the stability of gantry operations based on mechanical modeling as described in the first aspect of the present invention.

[0042] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the method for optimizing the stability of gantry operations based on mechanical modeling as described in the first aspect of the present invention.

[0043] The beneficial effects of this invention are as follows: it significantly improves the overall stability and safety of the bridging structure, evenly distributes structural stress, and reduces the risk of local instability, buckling, or overturning. While ensuring safety, it effectively controls the overall mass, achieving lightweight design, reducing material consumption and construction costs, and improving construction feasibility. Through reasonable coordination of the performance of sub-components, it can balance the strength of local components with the stability of the overall structure, ensuring that adjustments to each component have a positive impact on overall performance. The optimization process can adapt to different construction environments and load conditions, ensuring the structure maintains reliability and robustness in actual use. Attached Figure Description

[0044] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is a flowchart of a method for optimizing the stability of a gantry operation based on mechanical modeling.

[0046] Figure 2 This is a computer equipment diagram for a method to optimize the stability of gantry operations based on mechanical modeling. Detailed Implementation

[0047] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0048] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0050] Reference Figure 1 and Figure 2 As one embodiment of the present invention, this embodiment provides a method for optimizing the stability of a gantry operation based on mechanical modeling, comprising the following steps:

[0051] S1: Obtain the geometry of the gantry to be optimized.

[0052] The geometric structure includes the spatial form that constitutes the overall skeleton of the bridging frame.

[0053] Specifically, this geometric description can be the spatial arrangement of the strut's columns, main beams, crossbeams, and diagonal braces. This includes the length, relative position, and overall three-dimensional spatial contour characteristics of each component, so as to accurately reflect the geometry of the strut in structural stability analysis.

[0054] S2: After the geometric structure is solidified, the geometric structure is disassembled to obtain the sub-parts of the crossing frame.

[0055] The solidified geometry serves to provide a stable foundational geometric object for subsequent analysis and optimization. By fixing the spatial layout, component lengths, relative positions, and topological relationships of the truss, the disassembly standards of the sub-components can be ensured to be consistent, giving each sub-component specific degrees of freedom and positional constraints in the mechanical calculation model.

[0056] Each sub-part is a basic component unit divided according to the physical connection relationship between components. It has independent degrees of freedom of motion in the mechanical calculation model, and its structural properties cannot be further divided.

[0057] It can transform a complex overall structure into the smallest unit that can be analyzed independently, allowing each sub-part to be individually bound to dimensions, materials, and mechanical properties, while also reflecting the impact of local component changes on overall stability in subsequent analyses.

[0058] S3: Bind each of the sub-parts to the corresponding physical parameters, as each individual to be optimized.

[0059] The purpose of binding each sub-part to its corresponding physical parameters is to constrain its dimensions during the optimization iteration process. These dimensional constraints are predetermined in the design phase, typically taking into account engineering requirements and construction feasibility, and limiting the length, cross-section, and layout range of each component.

[0060] This binding ensures that each sub-part can only change within the allowed size range during iterative adjustments, guaranteeing the rationality of structural parameters, overall stability, and the reliability of stress analysis, while avoiding changes beyond the feasible range.

[0061] Furthermore, the physical parameters include the specific dimensions and material types of each sub-part.

[0062] In this embodiment, the size specification refers to the specific geometric parameters of the sub-component, including length, cross-sectional dimensions, width-to-height ratio, etc., which can accurately describe the spatial size and shape of the component.

[0063] Material type refers to the type of material used in the component, such as steel, aluminum, or other engineering materials. Furthermore, to simplify calculations, each material is assumed to be of uniform density in the analysis, meaning the material density remains consistent throughout the entire component.

[0064] In mechanical analysis and mass statistics, volume and mass can be directly calculated (by multiplying the density corresponding to the material type by the volume). At the same time, it is convenient to generate mechanical property feature vectors, ensuring the efficiency and reliability of calculations in iterative optimization and stability assessment.

[0065] S4: After adjusting the physical parameters of all the individuals to be optimized, the overall operational stability of the crossing frame is used as the fitness, and the individual physical parameters are iteratively optimized.

[0066] Adjusting the physical parameters for all the individuals to be optimized includes:

[0067] A heterogeneous knowledge graph is constructed with size specifications, material types, and mechanical properties as nodes, and a graph convolutional neural network is used to extract features and vectorize the graph.

[0068] Using the dimensions and material type as input, the mechanical property feature vectors of the sub-parts are generated using the message passing mechanism of the graph convolutional neural network.

[0069] The mechanical property feature vector is the encoded result characterizing bending stiffness, compressive strength, and damping properties.

[0070] Each material has inherent mechanical properties, such as elastic modulus, compressive strength, bending stiffness and damping characteristics. The size and specifications of the component determine the specific manifestation of these material properties in the sub-parts. For example, the size and length of the cross-section directly affect the stiffness and load-bearing capacity.

[0071] By constructing a heterogeneous knowledge graph with dimensions, material types, and mechanical properties as nodes, and using graph convolutional neural networks for feature extraction and vectorization mapping, the combined influence of dimensions and materials on mechanical properties can be systematically analyzed.

[0072] The neural network uses a message passing mechanism to fuse the dimensional information of each sub-part with its inherent material properties and the correlation information of surrounding components, thereby generating a mechanical property feature vector that can accurately characterize bending stiffness, compressive strength and damping characteristics.

[0073] For each material type of each sub-part, the corresponding geometric structure size constraint interval is input into the graph convolutional neural network, and the adjustable space of the mechanical property feature vector of the sub-part is output; due to the discreteness of the adjustable space of mechanical properties of different material types, a feasible domain space of mechanical property feature vector is formed.

[0074] By simultaneously driving each sub-part to change its size and specifications in random directions within its corresponding constraint range, each individual component can simultaneously traverse its corresponding feasible domain space, thereby achieving optimization of the overall operational stability of the gantry under physical reality constraints.

[0075] In this embodiment, the "geometric dimension constraint range" refers to the pre-determined dimension range of each sub-part during the design phase, including length, cross-sectional dimensions, etc., used to limit the space allowed for adjustment of the component during the optimization iteration process. These constraint ranges are pre-set according to engineering design requirements, construction feasibility, and mechanical rationality to ensure that the dimensions generated during the optimization process do not exceed the actual constructable or mechanically safe range. Since the mechanical properties of the sub-part exhibit an approximately linear relationship with the dimension change, when the constraint range is input to the graph convolutional neural network, it actually allows the network to identify the mechanical properties corresponding to the endpoints of the dimension range, thereby deriving the range of mechanical property changes within that range. Considering that different materials have different mechanical properties, the mechanical performance range exhibited by each material at the same size is also different, and the adjustable space of the mechanical property feature vector formed by the combination of multiple materials may have discontinuities. By integrating the adjustable ranges of each material, the feasible domain space of the discontinuous mechanical properties of the entire sub-part can be obtained, providing an accurate input boundary for subsequent stability assessment and optimization.

[0076] By sampling each dimension at equal intervals within the constrained range of geometric dimensions, and inputting the dimensional specifications of the sampling points into the graph convolutional neural network, a mechanical property feature vector is obtained for each sampling point. The mechanical property feature vectors are then sorted according to the sequence of sampling points. By smoothing the eigenvalues ​​of the mechanical property feature vectors in each dimension, the fitting result of the mechanical properties for each material type, considering both the dimensional specifications and the mechanical property feature vectors, is obtained.

[0077] "Sampling at equal intervals for each dimension" means dividing the length, width, or cross-sectional dimensions within the allowable dimensional range of each sub-part into several uniform points for observation. This is analogous to marking several points evenly on a line segment, each point corresponding to a specific dimension. This allows us to understand the trend of dimensional changes on mechanical properties throughout the adjustable range, rather than just looking at the maximum or minimum values. By organizing the mechanical characteristics obtained from these sampling points in sequence and performing a smoothing process, we can obtain a continuous relationship curve or surface that shows the relationship between dimensional changes and mechanical properties. This fitting is not intended to replace calculations, but rather to make it easier to predict performance changes at arbitrary dimensions during subsequent analysis or optimization, improving efficiency while maintaining reasonableness.

[0078] Furthermore, the fitness of the traverse frame is calculated globally using a multi-objective optimization model:

[0079] Within each sub-section, the volume is calculated based on the dimensions and specifications, then multiplied by the density corresponding to the material type to obtain the sub-section's mass. The masses of all sub-sections are then summed to obtain the global mass data. The objective function is constructed by minimizing this global mass data. By minimizing the overall mass as the objective function, the total weight of the bridging frame can be minimized while ensuring stability and mechanical performance, thereby reducing material usage, transportation difficulty, and construction costs. This approach incorporates weight control into a multi-objective optimization system, considering it simultaneously with stability, stiffness, and other indicators, ensuring that the optimization results balance economy and safety, rather than simply pursuing mechanical strength or stability.

[0080] To evaluate the impact of the combination of each sub-component on the stability of the bridging frame at the overall level, the mechanical property feature vectors of all sub-components are projected into a pre-trained coupling space of global stability to generate the global stability of the bridging frame. If the global stability is within a preset constraint range, the maximum of the global stability is used as the objective function 2; if the relative position or relative local curvature exceeds the constraint range, an infinitely large cost is imposed on objective function 2 (at which point objective function 2 decreases to infinitesimal). Maximizing global stability as the objective function ensures that the structure remains within a safe and stable range during the iteration process.

[0081] The coupling space for global stability is a high-dimensional mechanical logic field constructed using a deep neural network. By performing deep learning on a training set containing force transmission and geometric nonlinear responses between sub-parts, the physical coupling causal chain between local variations and overall response within the strut is solidified. By capturing the interference and superposition effects of the sub-part mechanical feature vectors during projection, key mechanical features characterizing the overall stiffness distribution and modal stability of the structure are extracted. The mechanical features of each sub-part undergo nonlinear interference and superposition mapping in the coupling space (not a simple numerical addition, but a comprehensive processing of the mechanical properties of each sub-part according to its position and action path within the structure: local stiffness, strength, and damping characteristics influence each other, which is reflected through network mapping, ultimately generating a comprehensive global stability metric. Simply put, it considers the interactions and nonlinear superposition effects between sub-parts, converting local mechanical characteristics into an overall stability evaluation index, ensuring that the optimization results reflect both local parameter changes and the overall structural safety and reliability). This process yields the global stability metric.

[0082] It's important to note that the overall stability of the bridging frame depends not only on the size and material of individual sub-components, but also on the force transmission between sub-components, nodal constraints, and geometric nonlinear responses. The coupling space for global stability is constructed using deep neural networks: by performing deep learning on a training set containing this information, the neural network can "learn" the influence of changes in each sub-component on the overall force and deformation, forming a high-dimensional mechanical logic field that links local component parameters with global stability. This allows for rapid and reasonable prediction of overall stability during the optimization process.

[0083] For objective function 1 and objective function 2, the weighted sum is calculated according to the preset attention weights and used as the fitness. The optimization is performed with the minimum fitness (when performing the weighted summation of objective function 2, the weight coefficient of the objective function is the negative value of the attention weight).

[0084] Specifically, when the objective function 2 is greater than a preset target structure stability threshold, an attention adjustment strategy is initiated; otherwise, the attention weights are reset and measured according to preset attention weights. The attention adjustment strategy includes calculating a scaling factor for the objective function 2 using a monotonically increasing function (a preset linear function), proportionally reducing the attention weights of the objective function 2 using this scaling factor, normalizing the attention, updating the attention, and calculating the objective function according to the updated attention weights.

[0085] It's important to note that the attention weighting mechanism allows the optimization process to dynamically allocate attention among multiple objectives, thus balancing structural lightweighting and overall stability. The fitness is obtained by weighting and summing objective function 1 (minimizing quality) and objective function 2 (global stability) according to preset weights, simultaneously considering both economy and safety. During iteration, when objective function 2 exceeds the preset structural stability threshold, it indicates that structural stability is sufficient or excessive. The optimization process can then appropriately reduce the weight of the stability objective, allocating more attention to quality or other objectives. Conversely, if stability is insufficient, the attention weights are adjusted to enhance the influence of stability on fitness, prioritizing safety improvements in the optimization process.

[0086] The preset target structural stability threshold is typically determined based on engineering design specifications or historical analysis results. For example, the threshold can be set according to the maximum allowable stress, minimum buckling coefficient, overturning margin, or safety factor to ensure that the structural parameters generated during the optimization process always meet the design safety requirements. Furthermore, the attention adjustment strategy can automatically adjust the weights of each objective based on the current stability of the structure, making the iteration process smoother; thus, while ensuring safety, it effectively balances structural lightweighting and construction economy.

[0087] Furthermore, during the optimization of the overall operational stability of the gantry, the rate of change of the mechanical property feature vector with respect to the size specification is extracted based on the fitting result between the size specification and the mechanical property feature vector. When each individual iterates by changing its size specification, the standard step size is multiplied by an adjustment coefficient to obtain the actual step size of each individual in the current iteration stage. It should be noted that during the optimization of the overall operational stability of the gantry, special handling is required during iteration because the mechanical property feature vectors of sub-parts may have discontinuous boundaries in the feasible region space (jumps in the mechanical properties of different materials). A jump mechanism is used (when jumping from one position to another discontinuous feasible region, the point closest to the current position is found in the target feasible region to achieve the jump; at this time, the jump step size is automatically equivalent to the actual step size of the current individual) to directly cross infeasible intervals and adjust the size to the next feasible region. This jump ensures that the optimization process is not blocked by unadjustable gaps or discrete points, and at the same time, it can explore the mechanical property intervals corresponding to different material or size combinations, ensuring that each iteration is always within the legal and feasible range. In this way, even if there are discontinuities in the mechanical properties, iterative optimization can be successfully completed, and a globally optimal structural scheme that balances stability and economy can be obtained.

[0088] To avoid excessive jumps or entering infeasible regions when dimensional changes significantly impact mechanical properties, a smaller step size is used, ensuring that each adjustment does not compromise structural stability. Conversely, when mechanical properties are less sensitive to dimensional changes, the step size can be appropriately increased to improve optimization efficiency. The unchanged dimensional specifications of the individual components are obtained, and the rate of change v of the current dimensional specification position is matched in the fitting results. The adjustment coefficient T is calculated as: T = θ / (ε + v); θ represents a constant used for proportional adjustment; ε represents a constant term to prevent the denominator from being zero. This adjustment coefficient balances search accuracy and convergence speed, ensuring smooth optimization within complex, nonlinear, and discontinuous mechanically feasible regions, while reducing the risk of skipping the optimal solution or entering unstable configurations.

[0089] S5: When the fitness converges or the maximum number of iterations is reached, output the optimization results of the global rational parameters.

[0090] Furthermore, the output results include the dimensions, materials, and mechanical characteristics of each sub-part.

[0091] Fitness convergence indicates that if further adjustments to sub-part dimensions or materials offer limited improvement to overall performance, the optimization is close to its optimal state. The maximum number of iterations is set to ensure the optimization process does not loop indefinitely.

[0092] By combining these two termination conditions, not only is the reliability and feasibility of the optimization results guaranteed, but the optimization process is also ensured to be completed efficiently with limited computing resources, providing an operable final solution that balances stability and economy for engineering design.

[0093] This embodiment also provides a system for optimizing the operational stability of a bridging frame based on mechanical modeling, including: a data acquisition unit for acquiring the geometry of the bridging frame to be optimized.

[0094] The decomposition unit, after solidifying the geometric structure, disassembles the geometric structure to obtain the sub-parts of the crossing frame.

[0095] The binding unit binds each of the sub-parts to the corresponding physical parameters, treating them as individual components to be optimized.

[0096] The iterative unit adjusts the physical parameters of all the individuals to be optimized, and uses the overall operational stability of the crossing frame as the fitness to iteratively optimize the physical parameters of the individuals.

[0097] The output unit outputs the optimization results of the global rational parameters when the fitness converges or the maximum number of iterations is reached. This embodiment also provides a computer device suitable for the stability optimization method of gantry operation based on mechanical modeling, comprising: a memory and a processor; the memory stores computer-executable instructions, and the processor executes the computer-executable instructions to implement the stability optimization method of gantry operation based on mechanical modeling as proposed in the above embodiment.

[0098] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0099] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the stability optimization method for gantry operations based on mechanical modeling as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0100] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for optimizing the stability of gantry operations based on mechanical modeling, characterized in that: This includes obtaining the geometry of the gantry to be optimized; After the geometric structure is solidified, it is disassembled to obtain the sub-parts of the crossing frame; Each of the sub-parts is bound to a corresponding physical parameter, serving as an individual to be optimized; After adjusting the physical parameters of all the individuals to be optimized, the overall operational stability of the crossing frame is used as the fitness to iteratively optimize the physical parameters of the individuals. When the fitness converges or the maximum number of iterations is reached, the optimization results of the global rational parameters are output.

2. The method for optimizing the stability of gantry operations based on mechanical modeling as described in claim 1, characterized in that: The geometric structure includes the spatial form that constitutes the overall skeleton of the bridging frame.

3. The method for optimizing the stability of gantry operations based on mechanical modeling as described in claim 2, characterized in that: The sub-parts of the bridging frame include basic component units that are disassembled according to the physical connection relationships in the geometric structure, each possessing independent degrees of freedom of movement and structurally indivisible.

4. The method for optimizing the stability of gantry operations based on mechanical modeling as described in claim 3, characterized in that: The physical parameters include the specific dimensions and material types of each sub-part; Each type of material is of uniform density.

5. The method for optimizing the stability of a gantry operation based on mechanical modeling as described in claim 4, characterized in that: Adjusting the physical parameters for all the individuals to be optimized includes: A heterogeneous knowledge graph is constructed with size specifications, material types, and mechanical properties as nodes, and a graph convolutional neural network is used to extract features and vectorize the graph. Using the size specifications and material type as input, the mechanical property feature vector of the sub-part is generated by utilizing the message passing mechanism of the graph convolutional neural network; The mechanical property feature vector is the encoded result characterizing bending stiffness, compressive strength and damping characteristics; For each material type of each sub-part, the corresponding geometric structure size constraint range is input into the graph convolutional neural network, and the adjustable space of the mechanical property feature vector of the sub-part is output. Furthermore, by sampling each dimension at equal intervals within the constrained range of geometric structure dimensions, and inputting the dimension specifications of the sampling points into the graph convolutional neural network, the mechanical property feature vector of each sampling point is obtained; The mechanical property feature vectors are sorted according to the sequence of sampling points. By smoothing the feature values ​​of each dimension of the mechanical property feature vectors, the fitting results of the mechanical properties with respect to size specifications are obtained for each type of material. Integration is due to the discreteness of the adjustable space of mechanical properties of different types of materials, forming a feasible domain space of mechanical property eigenvectors that is discontinuous. By simultaneously driving each sub-part to change its size and specifications in random directions within its corresponding constraint range, each individual component can simultaneously traverse its corresponding feasible domain space, thereby achieving optimization of the overall operational stability of the gantry under physical reality constraints.

6. The method for optimizing the stability of a gantry operation based on mechanical modeling as described in claim 5, characterized in that: The fitness of the bridging frame is calculated globally using a multi-objective optimization model: In each sub-part, the volume is calculated according to the size specifications, and then multiplied by the density corresponding to the material type to obtain the mass of the sub-part; the mass of each sub-part in the whole is accumulated to obtain the global mass data, and the objective function 1 is constructed with the minimum of the mass data; The mechanical property feature vectors of all sub-parts are projected into the pre-trained global stability coupling space to generate the global stability of the straddle. If the global stability is within the preset constraint range, then the maximum of the global stability is taken as the objective function 2; if the relative position or relative local curvature exceeds the constraint range, then an infinitely large cost is imposed on the objective function 2. The coupling space for global stability is a high-dimensional mechanical logic field constructed using a deep neural network. By performing deep learning on a training set containing force transmission and geometric nonlinear responses between sub-parts, the physical coupling causal chain between local variations and overall responses within the strut is solidified. By capturing the interference and superposition effects of the mechanical feature vectors of the sub-parts during the projection process, key mechanical features characterizing the overall stiffness distribution and modal stability of the structure are extracted. After the mechanical features of each sub-part are nonlinearly interfered and superimposed in the coupling space, a measure of global stability is obtained. For objective function 1 and objective function 2, the weighted sum is calculated according to the preset attention weights and used as the fitness. The optimization is then performed with the minimum fitness. Specifically, when the objective function 2 is greater than the preset target structure stability threshold, the attention adjustment strategy is activated; Conversely, the attention weights are reset and measured according to the preset attention weights; The attention adjustment strategy includes calculating the target function 2 using a monotonically increasing function to obtain a scaling factor, then scaling down the attention weights of the target function 2 using the scaling factor, and finally normalizing the attention to achieve attention adjustment and update.

7. The method for optimizing the stability of a gantry operation based on mechanical modeling as described in claim 6, characterized in that: In the process of optimizing the overall operational stability of the crossing frame, the rate of change of the mechanical property feature vector with the change of the size specification is extracted based on the fitting result between the size specification and the mechanical property feature vector. When each individual iterates by changing its size specifications, the standard step size is multiplied by an adjustment factor to obtain the actual step size of each individual in the current iteration stage; Specifically, the size specification of the individual that has not been changed is obtained, and the change rate value v of the current size specification position is matched in the fitting result; the adjustment coefficient T is calculated as: T=θ / (ε+v); θ represents a constant used for proportional adjustment; ε represents a constant term to prevent the denominator from being 0.

8. A stability optimization system for bridging operations based on mechanical modeling, based on the stability optimization method for bridging operations based on mechanical modeling as described in any one of claims 1 to 7, characterized in that: Includes a data acquisition unit to acquire the geometry of the gantry to be optimized; The decomposition unit, after solidifying the geometric structure, disassembles the geometric structure to obtain the sub-parts of the crossing frame; The binding unit binds each of the sub-parts to the corresponding physical parameters, treating them as individual components to be optimized; The iterative unit, after adjusting the physical parameters of all the individuals to be optimized, uses the overall operational stability of the crossing frame as the fitness to iteratively optimize the physical parameters of the individuals. The output unit outputs the optimization results of the global rational parameters when the fitness converges or the maximum number of iterations is reached.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the method for optimizing the stability of cross-train operations based on mechanical modeling as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the method for optimizing the stability of cross-train operations based on mechanical modeling as described in any one of claims 1 to 7.

Images