Seismic optimization method of steel frame structure based on particle swarm optimization algorithm

Abstract

The application discloses a steel frame structure anti-seismic optimization method based on a particle swarm algorithm, and relates to the technical field of building anti-seismic, and comprises the following steps: acquiring geometric constraint data for a target beam; generating corresponding mechanical compensation parameters based on the geometric constraint data; managing state identifiers of a plurality of particle data in a particle swarm optimization engine through a coordination controller based on the mechanical compensation parameters; adjusting at least one of a constraint condition pool and a particle generation strategy of the particle swarm optimization engine based on a check result of the state identifiers, and outputting a steel frame structure anti-seismic optimization design scheme satisfying a strong-column weak-beam anti-seismic mechanism and geometric constraints; and the application ensures that only a scheme satisfying the mechanical compensation parameters can be adopted, thereby avoiding failure of the strong-column weak-beam mechanism caused by reduction of the section height of the target beam, and improving the safety of the target design scheme.

Description

Technical Field

[0001] This invention relates to the field of seismic resistance technology in buildings, specifically to a seismic optimization method for steel frame structures based on particle swarm optimization. Background Technology

[0002] In the current building structure design workflow, the architectural, structural, and mechanical and electrical (MEP) professionals need to work on the same BIM platform. The layout of MEP pipelines often imposes space clearance requirements on structural components. Designers will lock the cross-sectional height of structural beams in the BIM model. When performing seismic design, the structural professionals usually use particle swarm optimization to optimize the cross-section. The algorithm reads the component parameters from the BIM database, uses the locked beam cross-sectional height as a fixed condition, generates a design scheme, and performs mechanical index verification.

[0003] Existing optimization methods, after obtaining geometric constraint data for the target beam, only treat it as a geometric dimension constraint without assessing the impact of this constraint on the seismic resistance mechanism of the structure. When the cross-sectional height of the target beam is reduced due to space requirements, its bending bearing capacity decreases, which may lead to the failure of the strong column-weak beam mechanism. However, existing systems cannot identify this mechanism failure risk in real time during the optimization process, resulting in the output target design scheme meeting geometric requirements but having seismic safety hazards. Summary of the Invention

[0004] This invention aims to address the technical problem in existing seismic optimization methods for steel frame structures, which treat spatial geometric constraints as fixed boundary conditions but fail to assess their impact on the strong column-weak beam seismic resistance mechanism, leading to potential seismic safety hazards in the optimization results. To overcome the shortcomings of existing technologies, this invention provides a seismic optimization method for steel frame structures based on particle swarm optimization.

[0005] This invention provides the following technical solution: a seismic optimization method for steel frame structures based on particle swarm optimization, comprising the following steps: Obtain geometric constraint data for the target beam; Generate corresponding mechanical compensation parameters based on the geometric constraint data; Based on the aforementioned mechanical compensation parameters, a coordination controller manages the state identifiers of multiple particle data in the particle swarm optimization engine. Based on the verification results of the state identifier, at least one of the constraint pool and particle generation strategy of the particle swarm optimization engine is adjusted, and a seismic optimization design scheme for steel frame structure that satisfies the strong column-weak beam seismic resistance mechanism and geometric constraints is output.

[0006] As a preferred embodiment of the present invention, the step of generating corresponding mechanical compensation parameters based on the geometric constraint data includes: Extract the section height limit value from the geometric constraint data, and obtain the span value and load value of the target beam; Based on the cross-section height limit value, the span value, and the load value, a preset specification-stiffness mapping table is queried to generate the linear stiffness calculation value of the target beam. The product of the calculated linear stiffness value and the design coefficient of strong column-weak beam is calculated to generate a linear stiffness threshold, and the linear stiffness threshold is used as the mechanical compensation parameter.

[0007] As a preferred embodiment of the present invention, the step of managing the state identifiers of multiple particle data in the particle swarm optimization engine through a coordination controller includes: In the memory space of the particle swarm optimization engine, the state identifier of the generated particle data is initialized to a verification identifier; Based on the geometric constraint data, the particle data is geometrically verified. If the verification passes, the identifier to be verified is updated to a geometric compliance identifier in the memory space. Under the stated geometric compliance flag, a seismic mechanism safety verification is performed based on the stated mechanical compensation parameters; If the verification passes, the geometric compliance identifier will be migrated to a mechanism security identifier in the memory space. If the verification fails, the geometric compliance flag will be migrated to a mechanism failure flag, and particles with the mechanism failure flag will be prohibited from entering the next generation population.

[0008] As a preferred embodiment of the present invention, the step of generating the linear stiffness threshold further includes the following steps: Identify the floor location or area type of the target beam within the structural system; If the floor location identifier or the area type identifier corresponds to a preset structural reinforcement area, then the first correction weight is obtained; If the floor location identifier or the area type identifier does not correspond to the preset structural reinforcement area, then a second correction weight with a value less than the first correction weight is obtained; The linear stiffness threshold is adjusted by weighting using the first or the second correction weight to generate updated mechanical compensation parameters.

[0009] As a preferred embodiment of the present invention, adjusting at least one of the constraint pool and particle generation strategy of the particle swarm optimization engine based on the verification result of the state identifier includes: Based on the distribution region of particle data marked as failure indicators of the mechanism in the current multidimensional parameter space, a search weight adjustment instruction is generated; The search weight adjustment instruction is used to instruct the particle swarm optimization engine to reduce the search intensity of the distribution region in subsequent iterations.

[0010] As a preferred embodiment of the present invention, the method further includes a collision detection process, which includes: Calculate the number of mechanism failure marker particles that are generated consecutively for the same geometric constraint data; If the quantity reaches the preset quantity threshold, then the limit value of the cross-sectional parameters of the connected columns of the target beam is obtained; If the limit value of the cross-sectional parameter is determined to be insufficient for stiffness compensation, a conflict warning signal is generated and the current iteration task is terminated.

[0011] As a preferred embodiment of the present invention, after generating the conflict warning signal, the method further includes: Calculate the difference between the mechanical compensation parameter and the actual achievable stiffness value to generate the mechanical notch value; Summarize the geometric constraint data and the mechanical notch values ​​to generate a conflict diagnosis report; The conflict diagnosis report is sent to the corresponding electromechanical design terminal via a cooperative bus, so that the electromechanical design terminal can return a correction instruction for the geometric constraint data.

[0012] As a preferred embodiment of the present invention, the coordination controller ensures the consistency between mechanical compensation parameters and geometric constraint data through atomic transaction logic, including: Start a write transaction before updating the structure model database; The determination results of the geometric compliance identifier and the mechanism security identifier are obtained simultaneously; The write operation transaction is committed only if all the determination results are true; otherwise, a rollback operation is performed to maintain the original state of the structure model database.

[0013] As a preferred embodiment of the present invention, the method further includes an adaptive recovery process based on version snapshots, the adaptive recovery process comprising: When starting the iterative task of the particle swarm optimization engine, a data snapshot is generated for the current particle population and stored in the version repository; If the proportion of particles with the failure indicator of the mechanism exceeds the preset proportion threshold in N consecutive iterations, the data snapshot is called to restore the initial state of the population, and a risk warning indicator for the geometric constraint data is output to the user interface.

[0014] As a preferred embodiment of the present invention, the step of performing geometric verification on the particle data based on the geometric constraint data includes: Extract the actual cross-sectional height value of the target beam from the particle data; Calculate the difference between the actual cross-sectional height value and the cross-sectional height limit value; If the difference is less than or equal to zero, the particle is determined to meet the geometric compliance requirements.

[0015] The beneficial effects of this invention are: 1. In this invention, corresponding mechanical compensation parameters are generated based on the geometric constraint data, so that the optimization process considers not only geometric size constraints but also mechanical mechanism requirements. By coordinating the controller to manage the status flags of multiple particle data in the particle swarm optimization engine, it is ensured that only the scheme that meets the mechanical compensation parameters can be adopted, thereby avoiding the failure of the strong column-weak beam mechanism caused by the reduction of the target beam cross-section height and improving the safety of the target design scheme.

[0016] 2. In this invention, by initializing the state identifier as a verification identifier and migrating it to a geometric compliance identifier, a mechanism safety identifier, or a mechanism failure identifier based on the results of geometric verification and seismic mechanism safety verification, fine-grained state management of particle data is achieved. Particles with the mechanism failure identifier are prohibited from entering the next generation population, ensuring that all design schemes participating in the iteration meet the seismic mechanism safety requirements and reducing invalid calculations. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram illustrating the workflow of the seismic optimization method for steel frame structures according to the present invention. Figure 2 This is a schematic diagram of the status identifier flow and migration logic of the present invention; Figure 3 A schematic diagram of particle distribution evolution under the search weight adjustment command; Figure 4 This is a schematic diagram illustrating the quantity threshold triggering judgment in the conflict detection process. Detailed Implementation

[0018] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0019] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0020] like Figure 1 As shown, the seismic optimization method for steel frame structures based on particle swarm optimization includes the following steps: Obtain geometric constraint data for the target beam, which includes at least the identifier of the locked component (e.g., beam L1) and specific locking parameters (e.g., the section height limit is 500 mm). The purpose of generating corresponding mechanical compensation parameters based on geometric constraint data is to transform simple geometric size constraints into quantifiable mechanical requirements for the seismic resistance system of the structure. The essential meaning of these mechanical compensation parameters is: to ensure that the seismic resistance mechanism of the structure does not fail due to the weakening of the target beam, the minimum linear stiffness value that the associated columns connected to the target beam must achieve. Based on mechanical compensation parameters, the state identifiers of multiple particle data in the particle swarm optimization engine are managed by a coordination controller. At the beginning of the optimization task, the state identifier of each newly generated particle data is initialized as a verification pending identifier. Then, the coordination controller first performs geometric verification on each particle data based on the original geometric constraint data. If the verification passes, the verification pending identifier is updated to a geometric compliance identifier. For all particle data that have obtained a geometric compliance identifier, the coordination controller will further perform a seismic mechanism safety verification. If the seismic mechanism safety verification passes, the state identifier of the particle data is migrated from the geometric compliance identifier to the mechanism safety identifier.

[0021] Finally, the coordinating controller returns the verification results of all particle data, i.e. the set of state identifiers, to the particle swarm optimization engine. Based on the verification results of the state identifiers, it adjusts at least one of the constraint pool and particle generation strategy of the particle swarm optimization engine, and outputs a seismic optimization design scheme for steel frame structure that satisfies the strong column-weak beam seismic resistance mechanism and geometric constraints. Specifically, the engine will only select particle data whose status is marked as a mechanism safety marker as the parent generation to generate the next generation of particle population. This is equivalent to dynamically adjusting the constraint pool of the optimization engine and taking mechanical compensation parameters into consideration as new constraints.

[0022] It should be noted that in a typical implementation scenario of this invention, the system is deployed on a multi-disciplinary collaborative design BIM platform. The operator prepares to use the particle swarm optimization algorithm to perform seismic optimization on a steel frame structure in order to find a structural scheme with the least amount of steel and that meets various mechanical indicators. At the same time, a second operator performs integrated pipeline design on the same platform and proposes a space clearance requirement for a key frame beam (i.e., the target beam referred to in this method). The cross-sectional height attribute of the beam is locked in the BIM model, for example, setting its cross-sectional height not to exceed 500 mm. This operation generates the geometric constraint data for the target beam required by this invention at the system level.

[0023] The executing entity of this invention can be an independent coordination controller, which is deployed between the data layer and the particle swarm optimization engine of the BIM collaboration platform to manage and coordinate data and instructions from different disciplines.

[0024] Furthermore, based on the geometric constraint data, corresponding mechanical compensation parameters are generated, including: Extract the section height limit value from the geometric constraint data, and obtain the span value and load value of the target beam; Specifically, the coordinating controller first parses the constraint information related to the cross-sectional height of the target beam from the received geometric constraint data, i.e., the cross-sectional height constraint value. At the same time, in order to evaluate the impact of this height constraint on the structural mechanical performance, the controller needs to know the actual stress characteristics of the target beam in the current structural system. To this end, the controller sends a query request to the structural calculation model database to read the geometric and load properties of the target beam, thereby obtaining its span value (such as the distance between the nodes at both ends of the beam) and the load value it bears (including permanent loads and variable loads).

[0025] Based on the cross-section height limit value, span value, and load value, the preset specification-stiffness mapping table is queried to generate the linear stiffness calculation value of the target beam. The coordinating controller internally stores or accesses a specification-stiffness mapping table. This mapping table is a database table pre-calculated based on structural mechanics principles and commonly used steel specifications. It records the theoretical values ​​of linear stiffness of the components under different section heights, different section forms (such as H-beams, box sections, etc.), and different spans. (Specifically, the standard section specifications specified in the current national standards GB / T 11263-2017 "Hot-rolled H-beams and Split T-beams" and GB / T 6728-2017 "Cold-formed Hollow Steel for Structural Use" can be referenced, and commonly used steel series in engineering can be selected.) Linear stiffness is usually defined as the elastic modulus of the component multiplied by the moment of inertia of the section and then divided by the span. It reflects the component's ability to resist bending deformation. That is, the linear stiffness value K is calculated according to the structural mechanics formula K=EI / L, where E is the elastic modulus of the steel, I is the moment of inertia of the section, and L is the span of the component. The controller uses the obtained cross-section height limit value (e.g., 500 mm), span value (e.g., 6 m), and load value (optional, for more accurate matching of working conditions) as combined query conditions to retrieve data from the specification-stiffness mapping table. This mapping table can be indexed by fields such as cross-section height, cross-section form, and span as combined primary keys to quickly locate the corresponding linear stiffness record. The query result is the calculated linear stiffness value that the target beam can achieve under the given cross-section height limit. In addition, when there are variable wall thicknesses within the cross-sectional shape, detailed dimensions such as flange width and web thickness are required as auxiliary keys for retrieval. This information can be obtained from the original properties of the target beam in the BIM model mentioned above.

[0026] This calculated linear stiffness value is a quantitative indicator, representing the bending stiffness that the target beam can objectively provide after meeting the space clearance requirements.

[0027] To facilitate understanding, an example of a specification-stiffness mapping table is provided, as shown in Table 1 below:

[0028] Table 1 Specification-Stiffness Mapping Table The meanings of each field in the table are as follows: The section height represents the section height of the target beam, and the section width represents the flange width (for H-beams) or section width (for box sections) of the target beam. Cross-sectional shapes: including common types such as H-beams and box sections; different cross-sectional shapes correspond to different formulas for calculating the moment of inertia. Material strength grades, such as Q235 and Q345, correspond to different steel grades and are used to determine the elastic modulus E. The span is the span of the target beam, which corresponds to the span value read from the BIM model above; Linear stiffness value: The linear stiffness calculated according to structural mechanics formulas, in cm³, characterizes the ability of a component to resist bending deformation. It should be noted that the linear stiffness values ​​in the table use the engineering-standard unit cm³, and its value is I / L (moment of inertia / span), which is the elastic modulus of steel E (2.06 × 10⁻⁶). 5 MPa) is a simplified expression after being treated as a constant. If it needs to be converted to the standard unit N·m, it needs to be multiplied by the E value (E=2.06×10¹¹N / m²) and the unit conversion needs to be performed. The details will not be elaborated here.

[0029] The product of the calculated linear stiffness value and the design coefficient of strong column and weak beam is used to generate the linear stiffness threshold, and the linear stiffness threshold is used as a mechanical compensation parameter. To ensure that the structure can achieve the strong column-weak beam yielding mechanism under seismic loading, i.e., the plastic hinge first appears at the beam end rather than the column end, it is necessary to ensure that the column connected to the target beam has sufficient stiffness. Current seismic design codes usually specify the design requirements for strong column-weak beam and quantify the stiffness margin requirement of the column relative to the beam with a coefficient greater than 1 (i.e., the strong column-weak beam design coefficient). For example, this coefficient is often taken as 1.2.

[0030] The coordinating controller stores the strong column-weak beam design coefficient, which can be set in the system configuration according to the seismic code of the project location. The controller multiplies the linear stiffness calculation value generated in the previous step by the strong column-weak beam design coefficient, and the resulting product is the linear stiffness threshold. This linear stiffness threshold clearly defines the minimum linear stiffness requirement that the associated columns connected to the target beam must meet in order to ensure the overall seismic resistance mechanism from a mechanical perspective.

[0031] Finally, the controller encapsulates the calculated linear stiffness threshold as a mechanical compensation parameter. This mechanical compensation parameter is no longer a simple geometric constraint, but includes derived constraints required by the mechanical mechanism. It will be used together with the original geometric constraint data in the subsequent particle state management process as a basis for judging whether each candidate design scheme (i.e. each particle data) simultaneously meets both spatial and mechanical standards.

[0032] It should be noted that, according to Article 6.2.2 of the "Code for Seismic Design of Buildings" (GB / T 50011-2010) (2024 Edition), the verification formula for strong column-weak beam in frame structures is as follows: ; Where, ∑M c The sum of the bending moments at the upper and lower column ends of the node, ∑M b η is the sum of the bending moments at the beam ends on the left and right sides of the node. c This is the column end moment amplification factor.

[0033] In the flexible design phase, for frame beams primarily subjected to vertical loads, the beam end bending moments can be approximated as follows: ; Where q represents the equivalent uniformly distributed line load acting on the frame beam bearing the vertical load, and L represents the calculated span of the beam.

[0034] The flexural capacity of a cross section is directly proportional to the moment of inertia I of the cross section; therefore, ∑M c / ∑M b This can be approximated as the ratio of their linear stiffness: .

[0035] Where E represents the elastic modulus of steel, I c I is expressed as the moment of inertia of the cross section of the column connected to the node. b L is expressed as the moment of inertia of the beam section connected to the node. c L represents the calculated length of the column connected to the node (which typically corresponds to the floor height in a conventional frame). b EI represents the calculated span of the beam connected to the node. c / L c EI represents the calculated linear stiffness of the column. b / L b This is expressed as the calculated value of the beam's linear stiffness.

[0036] For conventional multi-story steel frame structures, the stress at the nodes is mainly bending moment. The linear stiffness ratio and the bending moment bearing capacity ratio have a high positive correlation. Engineering practice shows that using the linear stiffness ratio as a preliminary screening index for strong columns and weak beams can control the error within 15%, which is sufficient to meet the needs of rapid verification in the optimization process.

[0037] For special and complex nodes (such as transition layers, long-span beams, etc.), this invention adjusts the linear stiffness threshold by weighting it with a first correction weight to correct the error of the simplified model and ensure the safety of critical areas.

[0038] Furthermore, such as Figure 2 As shown, the state flags of multiple particle data in the particle swarm optimization engine are managed by the coordination controller, including: In the memory space of the particle swarm optimization engine, the state identifier of the generated particle data is initialized to the verification identifier; Specifically, after the particle swarm optimization engine completes the generation of a generation of particle data, it sends this batch of particle data to the coordinator. The coordinator assigns a status flag to each received particle data to record its verification status and sets the initial value of the status flag to the pending verification flag.

[0039] Geometric verification is performed on particle data based on geometric constraint data. If the verification passes, the identifier to be verified is updated to the geometric compliance identifier in the memory space. Subsequently, the coordination controller began to verify each piece of data in the pending verification category; The first step is geometric verification, which aims to check whether the design scheme represented by the particle data meets the space clearance requirements for the target beam proposed by the electromechanical engineering department. The controller extracts the actual cross-sectional height value of the target beam from the particle data and compares it with the cross-sectional height limit value in the geometric constraint data. If the actual cross-sectional height value is less than or equal to the cross-sectional height limit value, it means that the particle data meets the requirements in terms of geometric dimensions, and the geometric verification is passed. At this time, the coordinating controller updates the status identifier of the particle data from the pending verification identifier to the geometric compliance identifier. The geometric compliance identifier means that the particle data is qualified in terms of spatial geometry and can enter the next stage of verification. Conversely, if the actual cross-sectional height is greater than the cross-sectional height limit, the geometric verification will fail, the particle data will be marked as other violations, and subsequent processing will be terminated; it will not participate in the seismic mechanism safety verification.

[0040] Under the geometric compliance flag, the seismic mechanism safety verification is performed based on the mechanical compensation parameters; For all particle data that have obtained geometric compliance identification, the coordinating controller continues to perform seismic mechanism safety verification. The purpose of this verification step is to evaluate whether the design scheme maintains the key strong column-weak beam seismic mechanism of the structure while meeting the geometric requirements. The coordinating controller analyzes the cross-sectional parameters of the associated columns (such as columns C1 and C2) connected to the target beam in the particle data based on the mechanical compensation parameters. Specifically, the coordinating controller calculates the actual linear stiffness of the associated columns based on the cross-sectional dimensions and material properties in the particle data, and then compares the calculated actual linear stiffness with the linear stiffness threshold required by the mechanical compensation parameters. If the verification passes, the geometric compliance flag will be migrated to the mechanism safety flag in the memory space. This means that the actual linear stiffness of all associated columns is greater than or equal to the linear stiffness threshold required by the mechanical compensation parameters. This indicates that the design scheme effectively compensates for the stiffness loss caused by the reduction in beam height by strengthening the columns, thus ensuring the effectiveness of the strong column-weak beam mechanism.

[0041] If the verification fails, the geometry compliance flag will be migrated to the mechanism failure flag, and particles with the mechanism failure flag will be prohibited from entering the next generation population. Specifically, if the verification fails, it means that the actual linear stiffness of at least one associated column is less than the linear stiffness threshold required by the mechanical compensation parameter. This indicates that although the design scheme meets the space clearance, the stiffness of the column is insufficient to maintain the seismic resistance mechanism, which poses a safety hazard. In this case, the coordination controller will migrate the status identifier of the particle data from the geometric compliance identifier to the mechanism failure identifier. More importantly, the controller will prohibit all particles with mechanism failure indicators from entering the next generation population. This means that these unqualified designs will be removed from the evolutionary process and will not be used as parent generations to generate subsequent candidate designs, thus avoiding the continuous derivation of invalid designs.

[0042] Furthermore, the generation of the linear stiffness threshold also includes the following steps: Identify the floor location or area type of the target beam within the structural system; These location attributes include, but are not limited to, the floor number to which it belongs, whether it is located in the bottom reinforcement zone, whether it belongs to the transfer layer component, whether it is located in the large span area, etc. The coordination controller standardizes these location attributes and generates corresponding floor location identifiers (such as "F01" and "F02" to represent floors) or area type identifiers (such as "BASE-ZONE" to represent the bottom reinforcement zone). These identifiers are used to determine whether the target beam belongs to the critical or reinforcement area of ​​the structure.

[0043] If the floor location identifier or area type identifier corresponds to a preset structural reinforcement area, then the first correction weight is obtained; The coordination controller has a pre-configured structural importance mapping table, which associates different floor location identifiers or area type identifiers with corresponding correction weights. The coordination controller compares the floor location identifier or area type identifier of the identified target beam with the mapping table. If the identifier corresponds to a preset structural reinforcement area (e.g., identified as "BASE-ZONE"), the controller obtains the corresponding first correction weight from the mapping table. The first correction weight is usually a value greater than 1, such as 1.2, which is used to increase the linear stiffness threshold, thereby imposing higher stiffness requirements on the associated columns to achieve stronger protection for the critical area.

[0044] If the floor location identifier or area type identifier does not correspond to the preset structural reinforcement area, that is, the target beam is located in a general area (e.g., identified as "STANDARD-ZONE"), then a second correction weight with a value less than the first correction weight is obtained. In this case, the second correction weight can usually be set to 1.0, indicating that the benchmark requirements are followed and no additional reinforcement is performed.

[0045] Finally, the linear stiffness threshold is adjusted by weighting using the first or second correction weight to generate updated mechanical compensation parameters. Specifically, the system performs a multiplication operation between the linear stiffness threshold and the selected correction weight to generate updated mechanical compensation parameters. The updated mechanical compensation parameters are pushed to the verification logic of the coordination controller in real time. In the subsequent particle state transition process, the coordination controller will perform a more stringent seismic mechanism safety verification on the data of each particle located in the structural reinforcement area based on this regionally weighted adjustment index.

[0046] Furthermore, based on the verification results of the state identifier, at least one of the following is adjusted: the constraint pool of the particle swarm optimization engine and the particle generation strategy. This adjustment can be either an adjustment to the constraint pool (e.g., converting the failure region into a dynamic constraint), a direct adjustment to the particle generation strategy (e.g., adjusting the velocity and position update strategies), or a combination of both. The coordination controller can select one or more of these methods to execute based on the actual configuration and the interface capabilities of the optimization algorithm. Based on the distribution region of particle data marked as mechanism failure indicators in the current multidimensional parameter space, a search weight adjustment instruction is generated; Specifically, the coordinator collects all particle data that are marked as mechanism failures in the current iteration. Each particle data corresponds to a point in a multidimensional parameter space. The dimensions of this space are composed of all variable component parameters (such as the cross-sectional height and width of each associated column). The coordinator analyzes the parameters of these failed particles and identifies their clustering regions in the multidimensional parameter space.

[0047] In this embodiment, the system uses a multidimensional bounding box algorithm for region identification. Specifically, the coordinating controller traverses all mechanism failure marker particles in this generation and calculates the minimum and maximum values ​​of each component parameter dimension to construct a multidimensional orthogonal bounding box that defines the failure range.

[0048] For example, by statistically analyzing the cross-sectional dimensions of the associated columns of the failed particles, it may be found that when the cross-sectional height of a certain associated column (such as column C1) falls within the range of 400mm to 450mm, almost all particles involved in that column are marked as mechanism failure indicators. Based on this, the controller judges that the parameter range (i.e., the internal space of the bounding box) is a "high-risk" area.

[0049] Based on the above analysis, the coordinator generates a search weight adjustment instruction. This instruction is a structured data object containing at least two core pieces of information: first, the target parameter region for which the search intensity needs to be adjusted (i.e., the distribution region of the failed particles, which can be specifically represented as the upper and lower limits of the value range of each parameter dimension); and second, the adjustment method, i.e., the requirement to reduce the search intensity in this region. This instruction can be encapsulated in a message format agreed upon with the particle swarm optimization engine (e.g., containing specific penalty coefficient configurations).

[0050] Secondly, by using search weight adjustment instructions, the particle swarm optimization engine is instructed to reduce the search intensity of the distribution region in subsequent iterations; The coordinator sends the generated search weight adjustment instructions to the particle swarm optimization engine. After receiving the instructions, the particle swarm optimization engine adjusts its internal particle generation strategy accordingly in subsequent iterations. The specific methods of adjustment may include, but are not limited to, the following: Firstly, when updating the velocity and position of particles, for particles falling into the parameter area specified by the instruction, the velocity update magnitude is reduced, or their fitness value is lowered by setting a larger penalty term. In the specific program implementation, the optimization engine performs coordinate determination before calculating the particle's fitness function. If all parameters of the new particle fall within the upper and lower limits of the target parameter region (bounding box), its basic fitness value is directly multiplied by a preset penalty constant (e.g., a penalty coefficient of 100), causing its fitness evaluation result to deteriorate sharply, thereby reducing the probability that particles in this region will be selected as parents.

[0051] Secondly, when initializing the new generation of particles, the number of initial particles generated within the specified parameter region is reduced. In specific implementation, the system introduces an initialization range truncation judgment: After randomly generating the initial position parameters of new particles, if it is determined that the position falls within the target parameter region, the position is discarded and random sampling is retried, forcing more initial particles to be distributed in other parameter spaces, thus blocking the ineffective computing power consumption in known failure regions from the source.

[0052] Thirdly, in terms of global guidance, the guidance weights of the global optimum and individual optimum of the algorithm are adjusted. For example, for particles that fall into the target parameter region, the optimization engine forces the inertia weight in its iteration formula to be set to the minimum value allowed by the algorithm (such as 0.1), thereby greatly limiting its exploration step size and activity in this region, so that the overall movement trend of the particle swarm deviates from the failure region.

[0053] Through the above adjustments, the particle swarm optimization engine will spontaneously reduce its exploration efforts in parameter regions that have been proven invalid in subsequent iterations, thereby avoiding repeated attempts and wasting computational resources in the same infeasible region. The search focus of the algorithm will gradually shift to feasible parameter spaces that can generate mechanism safety marker particles, thereby accelerating convergence to the optimal solution that simultaneously satisfies geometric constraints and seismic safety mechanisms.

[0054] It should be noted that the constraint pool is a collection used to store the constraints that all design schemes must meet. These constraints include the original geometric constraint data and the mechanical compensation parameters generated by the coordination controller. When evaluating the fitness of each particle data, the optimization engine will check whether it meets all the requirements in the constraint pool one by one.

[0055] The particle generation strategy defines how the optimization engine generates the next generation of particle populations from the current generation of particle populations, including but not limited to parameters such as particle velocity update formula, position update rule, global optimal guiding weight, individual optimal guiding weight, and initialization range of new particles.

[0056] Specifically, such as Figure 3 As shown in the figure, using a two-dimensional parameter space as an example, the horizontal axis represents the cross-sectional height of the associated column (in millimeters), and the vertical axis represents the cross-sectional width of the associated column. After conversion based on the mechanical compensation parameters (linear stiffness threshold), the minimum cross-sectional size requirement for the associated column is obtained. The critical value of 520 mm, marked by a solid black line in the figure, represents the lower limit of this size. Particles located in the area below and to the left of this critical line (i.e., both the cross-sectional height and width are less than 520 mm) have associated column linear stiffness that does not meet the mechanical compensation requirements, and are therefore marked as mechanism failure marker particles (represented by the symbol "×"). Particles located in the area above and to the right of the critical line (i.e., both the cross-sectional height and width are not less than 520 mm) have associated column linear stiffness that meets the requirements, and are marked as mechanism safety marker particles (represented by the hollow circle "○").

[0057] As shown in the figure, the mechanism failure marker particles are clustered in the lower left corner (around 420 mm), forming a high-risk distribution area; while the mechanism safety marker particles are distributed in the upper right area (around 630 mm), constituting a safe search area. The coordinating controller generates search weight adjustment instructions through statistical analysis of the failure particle distribution area, instructing the particle swarm optimization engine to reduce the search intensity of the high-risk area in subsequent iterations. The evolution trend is illustrated by thick arrows in the figure: the arrows point from the high-risk area (starting coordinates 440, 440) to the safe area (ending coordinates 600, 600), indicating that under the action of the search weight adjustment instructions, the search center of the particle swarm gradually shifts towards the parameter space that meets the mechanical compensation requirements, thereby accelerating the convergence to a feasible solution.

[0058] Furthermore, after each iteration, or after the state marking of each generation of particle data is completed, the coordination controller executes a conflict detection process, which includes: First sub-step: Calculate the number of mechanism failure marker particles that are continuously generated for the same geometric constraint data; Specifically, the coordination controller maintains a counter internally, which is dedicated to tracking failures for specific geometric constraints (such as the 500mm height limit of beam L1). Each time the coordination controller completes the status marking of a generation of particle data, it counts the number of particles in the current generation that are marked as mechanism failures and are related to the current geometric constraints. If there are multiple consecutive generations, and the number of mechanism failure marker particles for the geometric constraints in each generation is not zero, the coordination controller will sum up the number of failures in each generation, or more precisely, record the number of generations that have not generated any mechanism safety marker particles or the total number of consecutively failed particles since the geometric constraints took effect. When the total number of consecutively generated mechanism failure marker particles reaches a preset threshold (this threshold can be set in the system configuration according to the project size and computing resources, such as 500 or 1000), the coordination controller triggers the next step.

[0059] Second sub-step: If the quantity reaches the preset quantity threshold, then obtain the limit values ​​of the cross-sectional parameters of the columns connected to the target beam; When the number of consecutively failing particles exceeds the threshold, the coordination controller determines that the current geometric constraints may be too stringent and further evaluation is needed to determine if there is a physically feasible solution. To this end, the coordinating controller sends a query request to the structural model database or material specification library to obtain the limit values ​​of the cross-sectional parameters of the associated columns (such as columns C1 and C2) connected to the target beam. These limit values ​​include, but are not limited to: The maximum allowable cross-sectional dimensions are limited by the building's outer contour or red line; the maximum steel profile specifications are based on market supply; and the upper limit of the cross-section corresponding to the maximum allowable axial compression ratio according to design specifications.

[0060] This data is pre-stored in the system's configuration database and can be updated according to the specific circumstances of the project.

[0061] Third sub-step: If the limit value of the cross-sectional parameter is determined to be insufficient for stiffness compensation, then generate a conflict warning signal and terminate the current iteration task; The coordination controller compares the obtained limit values ​​of the associated column section parameters (for example, the maximum possible section size of column C1 is 500mm×500mm, and the corresponding maximum linear stiffness is calculated to be 4000 cubic centimeters) with the previously generated mechanical compensation parameters (requiring the associated column linear stiffness to be no less than 4500 cubic centimeters). If even if the cross-sectional dimensions of the associated column are taken to their physical limits (such as cross-sectional dimensions, material strength, etc.), the maximum linear stiffness that can be achieved still cannot meet the requirements of the mechanical compensation parameters, then the coordination controller determines that there is an irreconcilable conflict between the current spatial geometric constraints and the seismic safety requirements of the structure.

[0062] At this point, the coordination controller immediately generates a conflict warning signal. This conflict warning signal is a standardized system event that includes at least the specific location of the conflict (such as project ID, floor, axis), the identification of the components involved (such as target beam L1, associated columns C1, C2), the cause of the conflict (such as "the column section has reached its limit, and the linear stiffness still cannot meet the compensation requirements"), and relevant numerical values ​​(such as the threshold required by the mechanical compensation parameters, the maximum linear stiffness that the associated column can actually achieve).

[0063] After generating a conflict warning signal, the coordination controller sends an instruction to the particle swarm optimization engine to forcibly terminate the currently ongoing iterative task and stop generating a new generation of particles.

[0064] Furthermore, after generating the conflict warning signal, it also includes: First sub-step: Calculate the difference between the mechanical compensation parameters and the actual achievable stiffness values ​​to generate mechanical notch values; Specifically, the coordination controller subtracts the mechanical compensation parameters (i.e., the linear stiffness threshold, 4500 cubic centimeters) previously generated based on the geometric constraint data from the actual achievable stiffness value (i.e., the maximum possible linear stiffness, 4000 cubic centimeters) that the associated column can achieve under its cross-sectional parameter limit value to obtain a mechanical gap value. In this scenario, the mechanical gap value is 500 cubic centimeters.

[0065] To ensure stable operation of the electromechanical terminal, the coordination controller, after obtaining the mechanical gap value, executes a reverse lookup algorithm based on a mapping table: Calculate the minimum target linear stiffness of the target beam required to satisfy the mechanism (i.e., the sum of the target beam linear stiffness under the current constrained state and the mechanical notch value); Then, in the preset specification-stiffness mapping table, the minimum standard steel specification that meets the target linear stiffness requirement is retrieved using the span and load of the target beam as filtering conditions. Extract the section height of this specification as the new recommended beam height value, and calculate the difference between this recommended value and the original section height limit value to generate geometric dimension correction recommendation data (e.g., recommending a 50 mm allowance for clearance).

[0066] The second sub-step: Summarize the geometric constraint data and mechanical notch values ​​to generate a conflict diagnosis report; The coordination controller will compile all relevant information that triggered the conflict to form a structured conflict diagnosis report. This report will include at least the following: the specific location of the conflict, the identification of the components involved, the original geometric constraint data, the mechanical compensation parameters, the limit values ​​of the cross-sectional parameters of the associated columns, the calculated mechanical gap values, the geometric dimension correction suggestions generated by the above transformation, and the possible causes of the conflict.

[0067] The third sub-step: Send the conflict diagnosis report to the corresponding electromechanical design terminal via the cooperation bus to drive the electromechanical design terminal to return correction instructions for the geometric constraint data; The coordination controller publishes the encapsulated conflict diagnosis report to the collaboration bus of the multi-discipline collaboration manager. The collaboration bus is a message middleware in the system responsible for cross-discipline and cross-module data communication. It can route messages to the corresponding electromechanical design terminal based on the target discipline identifier (e.g., "Mechatronics") in the report.

[0068] After receiving the conflict diagnosis report, the electromechanical design terminal will display the report content to the user in a visual way on its user interface. For example, the target beam will be highlighted in the BIM 3D view, and a suggestion box containing clear engineering instructions will pop up (such as: [Conflict Warning] The current height limit causes the structural seismic resistance mechanism to fail and the column stiffness cannot be compensated. It is recommended to move the elevation of the electromechanical pipeline here up to provide an additional clearance space of

[50] mm, so that the structural beam height can be widened to

[550] mm). This will make it easy for the user to directly refer to the geometric dimension correction suggestion data and directly input the correction instructions for the geometric constraint data in the user interface. After inputting the correction instructions, the electromechanical design terminal will send the instructions back to the coordination controller through the collaboration bus. After receiving the correction instructions, the coordination controller will use them as new geometric constraint data and re-trigger the above complete processing flow to start a new round of mechanical compensation parameter generation and optimization iteration.

[0069] Furthermore, the coordination controller ensures the consistency between mechanical compensation parameters and geometric constraint data through atomic transaction logic, including: Before updating the structural model database, a write operation transaction is started. This transaction incorporates a series of subsequent write operations on the database (including updating component section parameters, modifying model version information, etc.) into an indivisible execution unit. Simultaneously obtain the judgment results of geometric compliance mark and mechanism security mark; The write transaction is committed only if all the results are true; otherwise, a rollback is performed to preserve the original state of the structure model database. The coordination controller performs a logical AND operation on the Boolean values ​​of the synchronously acquired geometric compliance flag and mechanism safety flag. If both flags are true, it means that the particle data still meets the requirements of geometric constraints and seismic mechanism safety at the current moment, which is an effective design scheme. At this point, the controller executes the write operation transaction commit operation, formally writing the information such as component cross-section parameters contained in the particle data into the structural model database, and updating the relevant version records. After the commit, these modifications are visible globally.

[0070] Conversely, if any of the flags is not true (e.g., the geometry compliance flag is false, the mechanism security flag is false, or both are false), it means that the particle data has expired at the current moment. It may be due to subsequent constraint changes or other factors that it no longer meets the requirements. In this case, the controller does not perform a commit operation, but performs a transaction rollback operation to undo all temporary modifications to the database since the transaction started, so that the structural model database remains in its original state before the transaction started, without any impact.

[0071] Furthermore, in some cases, due to overly stringent geometric constraints or inherent design limitations in the structure itself, the proportion of particles that can pass both geometric verification and seismic mechanism safety verification (i.e., particles with mechanism safety identifiers) may be extremely low across multiple iterations. Therefore, the method also includes an adaptive recovery process based on version snapshots. This adaptive recovery process includes: When starting an iterative task of the particle swarm optimization engine, a data snapshot is generated for the current particle population and stored in the version repository. This data snapshot contains the starting state information of the current iterative task, such as the initial position, velocity, and individual optimal value of all particles. The version repository is a persistent storage system used to manage data snapshots at different points in time or for different tasks, so that they can be quickly restored when needed later.

[0072] If the proportion of particles with mechanism failure indicators exceeds a preset proportion threshold (e.g., 90%) for N consecutive generations (N is a preset generation threshold, such as 5 generations or 10 generations, which can be set in the system configuration according to the project scale and computing resources), the coordinator controller determines that the current optimization process has fallen into a state of continuous failure. That is, under the current geometric constraint data, the optimization algorithm is unable to find a feasible design solution. Then, it calls the data snapshot to restore the initial state of the population and outputs a risk warning indicator for the geometric constraint data to the user interface. The process of restoring the initial state of the population is as follows: The coordinator retrieves a snapshot of the data stored at the start of the iteration task from the version repository and loads the snapshot data back into the particle swarm optimization engine to replace the current particle population, thereby restoring the population to its initial state before the iteration began.

[0073] Specifically, such as Figure 4 As shown in the figure, the curve records the growth trend of the cumulative number of failed particles in each generation. In the early stage of iteration (generation 1 to generation 6), the number of failed particles continued to increase, indicating that the optimization algorithm had difficulty finding a feasible solution under the current constraints. When the cumulative number of failures reaches the preset threshold (1000 marked by the black dashed line in the figure), that is, when the cumulative number of failures in the 7th generation exceeds 1000, the coordination controller determines that the current geometric constraints may be too stringent. Subsequently, the controller queries the limit values ​​of the cross-sectional parameters of the associated columns and finds that even if the maximum allowable cross-section (such as 500 mm × 500 mm) is used, the actual achievable stiffness value still cannot meet the mechanical compensation parameter requirements (for example, the linear stiffness needs to be 4500 cubic centimeters, but can only provide 4000 cubic centimeters in reality). Therefore, the system determines that stiffness compensation cannot be performed, generates a conflict warning signal, and terminates the current iteration task.

[0074] The diagram shows "reaching the preset quantity threshold" and "generating a conflict warning signal" at generation 7. The cumulative number of particles remains unchanged at 1300 from generation 8 to generation 10, which shows that the number of failed particles no longer increases after the iterative task is terminated.

[0075] Furthermore, geometric verification of the particle data is performed based on geometric constraint data, including: Extract the actual cross-sectional height value of the target beam from the particle data. This value is the actual cross-sectional height value of the target beam L1 in the candidate design scheme, such as 450 mm, 480 mm, 520 mm, etc.

[0076] Calculate the difference between the actual cross-section height and the cross-section height limit. The formula for calculating this difference is: Difference = Actual cross-section height - Cross-section height limit. Taking a cross-section height limit of 500 mm as an example, if the actual cross-section height is 450 mm, the difference is -50 mm; if it is 500 mm, the difference is 0 mm; and if it is 520 mm, the difference is +20 mm.

[0077] If the difference is less than or equal to zero, the particle is determined to meet the geometric compliance requirements. According to the above process, the particle's status identifier will then be updated from the pending verification identifier to the geometric compliance identifier, and it will be allowed to enter the subsequent seismic mechanism safety verification. Conversely, if the difference is greater than zero, the particle is determined not to meet the geometric compliance requirements.

[0078] Finally, it should be noted that the above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A seismic optimization method for steel frame structures based on particle swarm optimization algorithm, characterized in that, Includes the following steps: Obtain geometric constraint data for the target beam; Generate corresponding mechanical compensation parameters based on the geometric constraint data; Based on the aforementioned mechanical compensation parameters, a coordination controller manages the state identifiers of multiple particle data in the particle swarm optimization engine. Based on the verification results of the state identifier, at least one of the constraint pool and particle generation strategy of the particle swarm optimization engine is adjusted, and a seismic optimization design scheme for steel frame structure that satisfies the strong column-weak beam seismic resistance mechanism and geometric constraints is output. The generation of corresponding mechanical compensation parameters based on the geometric constraint data includes: Extract the section height limit value from the geometric constraint data, and obtain the span value and load value of the target beam; Based on the cross-section height limit value, the span value, and the load value, a preset specification-stiffness mapping table is queried to generate the linear stiffness calculation value of the target beam. The product of the calculated linear stiffness value and the design coefficient of strong column and weak beam is calculated to generate a linear stiffness threshold, and the linear stiffness threshold is used as the mechanical compensation parameter. The process of managing the state identifiers of multiple particle data in the particle swarm optimization engine through a coordination controller includes: In the memory space of the particle swarm optimization engine, the state identifier of the generated particle data is initialized to a verification identifier; Based on the geometric constraint data, the particle data is geometrically verified. If the verification passes, the identifier to be verified is updated to a geometric compliance identifier in the memory space. Under the stated geometric compliance flag, a seismic mechanism safety verification is performed based on the stated mechanical compensation parameters; If the verification passes, the geometric compliance identifier will be migrated to a mechanism security identifier in the memory space. If the verification fails, the geometric compliance flag will be migrated to a mechanism failure flag, and particles with the mechanism failure flag will be prohibited from entering the next generation population. The adjustment of at least one of the constraint pool and particle generation strategy of the particle swarm optimization engine based on the verification result of the state identifier includes: Based on the distribution region of particle data marked as failure indicators of the mechanism in the current multidimensional parameter space, a search weight adjustment instruction is generated; The search weight adjustment instruction is used to instruct the particle swarm optimization engine to reduce the search intensity of the distribution region in subsequent iterations.

2. The seismic optimization method for steel frame structures based on particle swarm optimization algorithm according to claim 1, characterized in that, The process of generating the linear stiffness threshold also includes the following steps: Identify the floor location or area type of the target beam within the structural system; If the floor location identifier or the area type identifier corresponds to a preset structural reinforcement area, then the first correction weight is obtained; If the floor location identifier or the area type identifier does not correspond to the preset structural reinforcement area, then a second correction weight with a value less than the first correction weight is obtained; The linear stiffness threshold is adjusted by weighting using the first or the second correction weight to generate updated mechanical compensation parameters.

3. The seismic optimization method for steel frame structures based on particle swarm optimization algorithm according to claim 1, characterized in that, The method further includes a collision detection process, which includes: Calculate the number of mechanism failure marker particles that are generated consecutively for the same geometric constraint data; If the quantity reaches the preset quantity threshold, then the limit value of the cross-sectional parameters of the connected columns of the target beam is obtained; If the limit value of the cross-sectional parameter is determined to be insufficient for stiffness compensation, a conflict warning signal is generated and the current iteration task is terminated.

4. The seismic optimization method for steel frame structures based on particle swarm optimization algorithm according to claim 3, characterized in that, After generating the conflict warning signal, the method further includes: Calculate the difference between the mechanical compensation parameter and the actual achievable stiffness value to generate the mechanical notch value; Summarize the geometric constraint data and the mechanical notch values ​​to generate a conflict diagnosis report; The conflict diagnosis report is sent to the corresponding electromechanical design terminal via a cooperative bus, so that the electromechanical design terminal can return a correction instruction for the geometric constraint data.

5. The seismic optimization method for steel frame structures based on particle swarm optimization algorithm according to claim 1, characterized in that, The coordination controller ensures the consistency between mechanical compensation parameters and geometric constraint data through atomic transaction logic, including: Start a write transaction before updating the structure model database; The determination results of the geometric compliance identifier and the mechanism security identifier are obtained simultaneously; The write operation transaction is committed only if all the determination results are true; otherwise, a rollback operation is performed to maintain the original state of the structure model database.

6. The seismic optimization method for steel frame structures based on particle swarm optimization algorithm according to claim 1, characterized in that, The method also includes an adaptive recovery process based on version snapshots, the adaptive recovery process comprising: When starting the iterative task of the particle swarm optimization engine, a data snapshot is generated for the current particle population and stored in the version repository; If the proportion of particles with the failure indicator of the mechanism exceeds the preset proportion threshold in N consecutive iterations, the data snapshot is called to restore the initial state of the population, and a risk warning indicator for the geometric constraint data is output to the user interface.

7. The seismic optimization method for steel frame structures based on particle swarm optimization algorithm according to claim 1, characterized in that, The geometric verification of the particle data based on the geometric constraint data includes: Extract the actual cross-sectional height value of the target beam from the particle data; Calculate the difference between the actual cross-sectional height value and the cross-sectional height limit value; If the difference is less than or equal to zero, the particle is determined to meet the geometric compliance requirements.

Images