Steel reinforced concrete member full-period management method based on industrial big data

Abstract

The invention discloses a steel reinforced concrete member full-period management method based on industrial big data. The method comprises the steps that a member model is established through BIM modeling software, an easily-buckling area is screened in combination with finite element buckling analysis, a buckling monitoring point is arranged on the surface of an analysis plate body, and a plate body monitoring point is arranged on the splicing edge; extracting model coordinate data to generate processing parameters, and guiding plate cutting and monitoring point position marking; actual coordinates of all monitoring point positions in the profile steel component are obtained through laser scanning and compared with model coordinates, and reworking is triggered if errors exceed the limit; constructing a neural network model, inputting various parameters related to analysis of plate body buckling to obtain a predicted buckling monitoring point out-of-plane displacement value, and judging compliance; according to the method, the problems of low efficiency, error accumulation and lack of full-period closed-loop management of a traditional method are solved, the machining precision and the structural safety of the steel reinforced concrete member are remarkably improved, and the hardware cost and the detection time consumption are reduced.

Description

Technical Field

[0001] The present invention relates to the field of intelligent construction of buildings. More specifically, the present invention relates to a full-cycle management method for steel reinforced concrete components based on industrial big data. Background Art

[0002] A steel reinforced concrete component (SRC component) is a composite structure formed by using a steel section as the skeleton and outer layer, with concrete wrapped or embedded. In structures such as steel plate shear walls and precast steel columns, the structure with steel plates wrapping concrete has the advantages of both the flexural capacity of steel and the compressive performance of concrete, and the excellent mechanical property that the outer steel plates can work together with the concrete, and is widely used in key parts such as high-rise building cores and long-span bridges.

[0003] The outer steel plates of such components need to be processed into shape through processes such as cutting and welding, and be closely combined with the concrete to achieve collaborative stress. However, in the processing stage, due to factors such as cutting errors and welding thermal deformation, local initial displacements are likely to occur in the steel plates; in the on-site welding stage, uneven welding heat input or residual stress release will cause the steel plates to deviate from the preset positions. The above initial displacements result in gaps or uneven stress distribution at the contact surface between the steel plates and the concrete. During the subsequent stress process, premature buckling occurs due to the initial deformation of the steel plates, resulting in the steel plates being unable to work together with the concrete. Since the interface performance between the steel section and the concrete directly affects the overall structural bearing capacity, the accumulation of initial displacements may lead to the bearing capacity of the steel reinforced concrete structure being much lower than the design target, reducing the structural safety and durability. The difficulty in controlling and solving this problem lies in that it is difficult to quantify and check the deformation in real time during the processing and welding of steel section components. The traditional manual measurement and quality inspection method relies on experience for adjustment and lacks an accurate prediction and instant feedback mechanism based on theoretical models.

[0004] In existing building management, industrial big data and building information model technology have been widely applied to construction progress management, resource allocation and other links. For steel reinforced concrete structures, the existing technologies mostly focus on the flow management or model visualization display during the construction stage, and have not deeply intervened in the whole life cycle of steel section components, especially quality management. Specifically, although the component models established by using BIM modeling software in the design stage can export geometric data and can form a dynamic association with processing parameters; however, the processing error data in actual processing is not fed back to the design model to optimize subsequent production; the correlation between on-site welding deformation, concrete pouring parameters and steel section deformation during the construction stage has not been systematically analyzed.

[0005] In the quality monitoring of steel reinforced concrete structures, point cloud scanning technology is often used to obtain the three-dimensional coordinates of the component surface, and the processing errors are identified by comparing with the component model established by BIM modeling software. However, the point cloud data volume is huge, data processing depends on high-performance computing devices, and the coordinate matching algorithm is complex, resulting in a calculation time of up to several hours, which cannot meet the real-time monitoring requirements during processing or construction. The lag of the existing point cloud technology makes it difficult to effectively integrate into the actual production process.

[0006] Therefore, there is an urgent need to propose a new full-cycle management method for steel reinforced concrete components, which can effectively manage the entire process of steel reinforced concrete design, processing, and on-site construction. Summary of the Invention

[0007] An object of the present invention is to provide a full-cycle management method for steel reinforced concrete components based on industrial big data, which realizes the quality control of the entire process of steel reinforced concrete components by connecting each link through data flow.

[0008] In order to achieve these and other advantages of the present invention, according to one aspect of the present invention, there is provided a full-cycle management method for steel reinforced concrete components based on industrial big data, including the following steps: S1. Establish a component model in BIM modeling software, select each plate in contact with concrete on each side of the component model as the analysis plate, import it into finite element calculation software for buckling analysis, and select several monitoring points on the non-concrete contact surface of the analysis plate. The monitoring points include buckling monitoring points and plate monitoring points; S2. Extract the component contour data and the coordinates of the monitoring points from the component model to form processing parameters, and perform cutting of the steel plate profiles and marking of the monitoring points based on the processing parameters to complete the processing of the steel component; S3. After the steel component is processed, scan it to obtain the actual coordinates of each monitoring point on the steel component and compare them with the model coordinates of each monitoring point in the component model. If the error between the actual coordinates and the model coordinates exceeds the processing threshold, reprocess the steel component; S4. Establish a prediction model for the out-of-plane displacement of the buckling monitoring points, input the initial out-of-plane displacement values of the buckling monitoring points, plate geometric parameters, plate welding parameters, and the parameters of the infilled concrete of each analysis plate, and obtain the predicted out-of-plane displacement values of the buckling monitoring points after the concrete pouring is completed. If the predicted out-of-plane displacement values of the buckling monitoring points exceed the allowable buckling deformation amount, reprocess the steel component, and repeat steps S3 to S4 until each analysis plate meets the requirements; S5. The steel members are installed, welded and filled with internal concrete on site. After the concrete begins to set, the final out-of-plane displacement values of the buckling monitoring points on the steel members are collected by a laser scanner, and the final out-of-plane displacement values of the buckling monitoring points, the corresponding plate welding parameters and the internal concrete filling parameters are fed back to the out-of-plane displacement prediction model of the buckling monitoring points for training.

[0009] Preferably, step S1 includes the following steps: S11. Establish a three-dimensional component model in the BIM modeling software. Each plate in the component model needs to be modeled independently. After model checking, select the plate that contacts the concrete on one side and has a non-concrete contact surface on the other side as the analysis plate and transfer it to the finite element calculation software as the finite element model. Assign material properties to each calculation object in the finite element calculation software. S12. Set constraint conditions for the analysis plate in the finite element calculation software, apply the design load and perform a non-linear buckling analysis to obtain the prone-to-buckle area in the analysis plate. Select at least one point on the non-concrete contact surface of the plate in the prone-to-buckle area as the buckling monitoring point. S13. Mark the buckling monitoring points on each analysis plate in the component model and set several plate monitoring points at the splicing edges of each analysis plate and other plates.

[0010] Preferably, step S12 includes the following steps: S121. Take the state of the steel members after all on-site connection measures are completed according to the design requirements and before the concrete is poured on site as the basis for the constraint conditions of the edges of each analysis plate. Apply constraint conditions to the edges of the analysis plate in the finite element calculation software. The constraint conditions include unconstrained and six-degree-of-freedom constraints. S122. Apply the design load to the analysis plate in the finite element calculation software and extract the Mises stress data of each point in the first-order buckling mode of the analysis plate. , select the area as the prone-to-buckle area, where , is the Mises stress, is the ultimate buckling stress, K is the buckling coefficient, E is the material elastic modulus, v is the Poisson's ratio, t is the plate thickness, s is the length of the analysis plate in the force direction; S123. Select the point with the largest displacement in the prone-to-buckle area as the buckling monitoring point.

[0011] Preferably, step S3 includes the following steps: S31. Establish a component model coordinate system. In the component model coordinate system, define the model coordinate set of all monitoring points , and calculate the model distances between any two monitoring points to form a model distance set . Conduct laser scanning on the processed steel component to obtain the actual coordinate set of all monitoring points in the steel component , and calculate the actual distances between any two monitoring points to form an actual distance set . Among them, N is the number of monitoring points, is the model coordinate set, is the l model coordinate of the -th monitoring point, is the actual coordinate set, l is the actual coordinate of the -th monitoring point; S32. Traverse all multi-point combinations of monitoring points . Extract the corresponding model distance from the model distance set , extract the corresponding actual distance from the actual distance set , calculate the sum of the global distance errors of this multi-point combination . Select the multi-point combination that makes M the smallest as the optimal reference point group. Among them, a1, a2, aM is the number of monitoring points within the multi-point combination, ai and aj are the monitoring point indices within the multi-point combination, is the ai model distance between aj and , ai is the aj actual distance between S33. Align the actual optimal reference point group with the center point of the optimal reference point group in the component model coordinate system, translate the actual coordinates of the monitoring points to the component model coordinate system to obtain the translated corrected actual coordinates , calculate the three-dimensional deviation between it and the corresponding model coordinate . If is greater than the processing threshold, reprocess this steel component. Among them, is the corrected actual coordinate after the i -th monitoring point is translated to the component model coordinate system.

[0012] Preferably, step S4 includes the following steps: S41. Establish an independent coordinate system for each analysis plate body based on the positions of the plate body monitoring points on each steel section member, and obtain the initial out-of-plane displacement values of the buckling monitoring points on each analysis plate body; S42. Use the plate body geometric parameters, initial out-of-plane displacement values of the buckling monitoring points, plate body welding parameters, internal filled concrete parameters, and final out-of-plane displacement values of the buckling monitoring points in the historical processing data of each steel section member as sample set data to construct a prediction model for the out-of-plane displacement of the buckling monitoring points; S43. Input the plate body geometric parameters, plate body welding parameters, and internal filled concrete parameters of the current analysis plate body into the trained prediction model for the out-of-plane displacement of the buckling monitoring points, and output the predicted out-of-plane displacement value of the buckling monitoring points. The allowable buckling deformation quantity is calculated based on the plate and shell buckling theory, considering the critical displacement and the safety factor. If the predicted out-of-plane displacement value of the buckling monitoring points exceeds the allowable buckling deformation quantity, it is determined that the analysis plate body does not meet the requirements, and the steel section member is readjusted.

[0013] Preferably, in step S42, the plate body geometric parameters include the plate thickness, width-thickness ratio, height-thickness ratio, and root mean square error of each monitoring point position on the plate in the actual processing of the analysis plate body. The plate body welding parameters include the on-site welding line energy, total on-site weld length, designed distance between the buckling monitoring point and the nearest weld. The internal filled concrete parameters include the concrete compressive strength, concrete water-binder ratio, concrete elastic modulus, and concrete slump. Among them, a neural network model including an input layer, a hidden layer, and an output layer is adopted. The input layer of the neural network model includes multiple input neuron nodes. Multiple hidden neuron nodes are arranged on the hidden layer. Each input neuron node is respectively connected to each hidden neuron node. The hidden neuron nodes are all connected to the output layer. The plate body geometric parameters, initial out-of-plane displacement values of the buckling monitoring points, plate body welding parameters, and internal filled concrete parameters are used as the input values of the input neuron nodes, and the output result is compared with the corresponding final out-of-plane displacement value of the buckling monitoring point, and the neural network model is optimized by the algorithm to obtain the prediction model for the out-of-plane displacement of the buckling monitoring point.

[0014] Preferably, the allowable buckling deformation quantity , where f is the safety factor, with a value range of 2.0 to 4.0, D is the bending stiffness of the plate body, U a is the allowable buckling deformation quantity, s is the length of the analysis plate body in the force direction.

[0015] In a second aspect, the present invention provides a full-cycle management system for steel reinforced concrete components based on industrial big data, which applies the above-mentioned full-cycle management method for steel reinforced concrete components based on industrial big data, and includes: an industrial big data processing module, a point monitoring module, and a neural network prediction module that are connected by signals to each other; The industrial big data processing module includes a component processing sub-module and a data storage sub-module. The component processing sub-module includes BIM modeling software and finite element calculation software. The BIM modeling software is used to model each steel component, set buckling monitoring points and plate body monitoring points, and export sheet metal processing information. The data storage sub-module is used to collect and store the initial out-of-plane displacement values of the buckling monitoring points of each steel component after processing, the corresponding plate welding parameters, the in-filled concrete parameters, and the final out-of-plane displacement values of the buckling monitoring points during on-site construction; The point monitoring module includes two sets of laser scanners. One set of the two sets of laser scanners is set at the construction site, and the other set is set in the steel component processing factory. The laser scanner can collect the position information of the buckling monitoring points and the plate body monitoring points on the steel component; The neural network prediction module uses the data in the data storage sub-module to train and optimize the neural network model, and obtains the predicted out-of-plane displacement value of the buckling monitoring point after inputting the plate geometry parameters, the initial out-of-plane displacement value of the buckling monitoring point, the plate welding parameters, and the in-filled concrete parameters.

[0016] In a third aspect, the present invention provides a computer device, including a memory and a processor. Computer instructions are stored in the memory, and the processor executes the above-mentioned full-cycle management method for steel reinforced concrete components based on industrial big data by executing the computer instructions.

[0017] The present invention has at least the following beneficial effects: First, the present invention exports processing parameters through the component model made by the BIM modeling software, and obtains the actual coordinates by scanning after processing, and compares them with the component model in real time. If the error exceeds the limit, the rework process is triggered. The final deformation data collected during the construction stage is further used to update the neural network model, forming a data closed-loop of "design - processing - construction - optimization". Compared with the existing steel structure method that relies on manual inspection and empirical adjustment, this method penetrates each link through the data flow, significantly reduces the information lag, and realizes the full-process quality control of the steel reinforced concrete structure.

[0018] Second, the present invention constructs a prediction model for out-of-plane displacement of buckling monitoring points based on a neural network model. The input parameters include multiple data such as plate geometry parameters, plate welding parameters, and in-filled concrete parameters. Through training with historical data, the neural network model can dynamically adapt to different process conditions, and the output results are closer to the actual working conditions. This solution reduces the error of predicting the out-of-plane displacement value of buckling monitoring points by a large margin through multi-parameter coupling analysis, providing a reliable basis for process adjustment.

[0019] Third, the present invention proposes a coordinate alignment algorithm based on an optimal reference point group. It extracts the set of model distances between monitoring points from the component model, calculates the sum of global distance errors by traversing multiple combinations of actual scanned points, and selects the combination with the smallest error as the optimal reference point group for coordinate alignment. This method only needs to process the data of a small number of key monitoring points. Compared with the traditional matching technology after point cloud scanning, this technical solution greatly improves the detection efficiency on the premise of ensuring accuracy, significantly reduces the scanning cost, and is applicable to real-time quality monitoring in processing workshops and construction sites.

[0020] Fourth, in the finite element analysis stage, the present invention identifies the easily buckling areas of the plate through nonlinear buckling calculation, selects the point with the largest displacement in this area as the buckling monitoring point, and at the same time adds plate monitoring points at the splicing edges of the plate to track the cumulative errors in the processing and welding stages. Compared with the existing methods of randomly distributing points or evenly distributing points, this solution guides the positioning of monitoring points through mechanical analysis, making the monitoring data more directly reflect the weak parts of the component and avoiding potential risks caused by monitoring blind spots.

[0021] Fifth, the present invention can complete the coordinate acquisition of monitoring points using a conventional laser scanner, and reduces the dependence on hardware accuracy through an optimized algorithm. The algorithm for coordinate alignment with the optimal reference point group can automatically compensate for environmental vibrations or equipment pose deviations, ensuring data reliability under complex working conditions such as dust in the welding workshop and vibration during concrete pouring. Compared with the traditional quality inspection method that requires strict environmental control for point cloud scanning, this solution greatly reduces the hardware cost and is more adaptable to the flexibility and timeliness requirements of construction sites.

[0022] Other advantages, objectives, and features of the present invention will be partially reflected by the following description and partially understood by those skilled in the art through the research and practice of the present invention. Brief Description of the Drawings

[0023] Figure 1 It is the overall flowchart in a technical solution of the present invention; Figure 2 It is the schematic diagram of monitoring points of a steel section member in a technical solution of the present invention. Detailed Embodiment

[0024] The present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments, so that those skilled in the art can implement it with reference to the text of the specification.

[0025] It should be understood that the terms such as "having", "including" and "comprising" used herein do not exclude the presence or addition of one or more other elements or combinations thereof.

[0026] It should be noted that the experimental methods described in the following embodiments are all conventional methods unless otherwise specified, and the reagents and materials can be obtained from commercial channels unless otherwise specified; in the description of the present invention, it should be noted that unless otherwise clearly defined and limited, the terms "installed", "connected" and "set" should be understood in a broad sense. For example, they can be fixedly connected and set, or detachably connected and set, or integrally connected and set. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations. The orientation or positional relationship indicated by the terms "lateral", "longitudinal", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present invention.

[0027] As Figure 1-2 shown, the present invention provides a full-cycle management method for steel-concrete composite members based on industrial big data, including the following steps: S1. Establish a component model in BIM modeling software, select the plate in contact with concrete on each single side of the component model as the analysis plate, import it into finite element calculation software for buckling analysis, and select several monitoring points on the non-concrete contact surface of the analysis plate. The monitoring points include buckling monitoring points and plate monitoring points. Specifically, a three-dimensional component model of a steel member is established in BIM modeling software such as Revit and Tekla according to the design drawings. Since the buckling of the steel-concrete structure has the greatest impact on the overall force, and for the convenience of subsequent monitoring operations, the plate in contact with concrete on a single side is selected as the analysis plate, and the model of the analysis plate is imported into finite element calculation software. The finite element calculation software can adopt ANAYS, ABAQUS, etc. Apply the design load to the model of the analysis plate for nonlinear buckling analysis, extract the Mises stress distribution under the first-order buckling mode and screen out the easy-buckling area. Select the point with the largest displacement in the easy-buckling area as the buckling monitoring point, and add plate monitoring points at the splicing edges of the analysis plate and other plates. The plate monitoring points can be 10-20 mm away from the edge. Accurately locate the monitoring points through mechanical analysis to provide a data basis for subsequent processing and deformation monitoring.

[0028] S2. Extract the component contour data and the coordinates of the monitoring points from the component model to form processing parameters. Based on the processing parameters, cut the steel plate profiles and mark the monitoring points. After the section steel components are processed. Specifically, extract the contour data of the analysis plate body and the coordinates of the monitoring points from the component model to obtain a file such as DXF format that records the coordinate positions, and generate numerical control cutting codes based on this. Use a numerical control plasma cutting machine to cut the steel plate according to the contour, and simultaneously engrave the monitoring points on the surface of the plate body through a laser marking machine. The cut plate bodies are connected by welding and assembled into complete section steel components. During the processing, ensure that the splicing points at the edges of each plate body are aligned and the welds are smooth.

[0029] S3. After the section steel components are processed, scan them to obtain the actual coordinates of each monitoring point on the section steel components and compare them with the model coordinates of each monitoring point in the component model. If the error between the actual coordinates and the model coordinates exceeds the processing threshold, reprocess the section steel component. Specifically, after the section steel components are processed, quickly collect the actual coordinate data of all monitoring points on the surface of the section steel components through laser scanning technology and compare them with the preset model coordinates in the component model. Optionally, use a lightweight algorithm to screen the combination of reference points with the highest matching degree with the component model, calculate the sum of the global distance errors between the monitoring points, and determine the optimal coordinate system alignment scheme, so as to accurately calculate the three-dimensional position deviation of each monitoring point. The processing threshold can be set to ±2 mm with reference to the existing steel structure processing specifications. If the deviation of any monitoring point exceeds the processing threshold, a rework instruction will be generated and fed back to the processing link to readjust or reprocess the section steel component. Compared with the point cloud scanning technology, the scanning and calculation time of this solution is greatly reduced, and it does not require high-performance hardware support.

[0030] S4. Establish a prediction model for the out-of-plane displacement of the buckling monitoring points. Input the initial out-of-plane displacement values of the buckling monitoring points, the geometric parameters of the plate body, the welding parameters of the plate body, and the parameters of the infilled concrete of each analysis plate body to obtain the predicted out-of-plane displacement values of the buckling monitoring points after the concrete pouring is completed. If the predicted out-of-plane displacement value of the buckling monitoring point exceeds the allowable buckling deformation amount, reprocess the section steel component. Repeat steps S3 - S4 until each analysis plate body meets the requirements. Specifically, construct a neural network model. The neural network model can adopt a BP neural network model. The input parameters include parameters related to plate body buckling such as the geometric dimensions of the plate body, the welding line energy in subsequent on-site welding operations, and the elastic modulus of the concrete. The neural network model adopts a double hidden layer structure, and the output value is the predicted out-of-plane displacement value of the buckling monitoring point. The neural network model can optimize the network weights through the particle swarm algorithm and train the model using historical data.

[0031] S5. The steel section members are installed, welded, and internally filled with concrete on-site. After the concrete begins to set, the final out-of-plane displacement values of the buckling monitoring points on the steel section members are collected by a laser scanner. The final out-of-plane displacement values of the buckling monitoring points, the corresponding plate welding parameters, and the parameters of the infilled concrete are fed back to the out-of-plane displacement prediction model of the buckling monitoring points for training. Specifically, after the concrete begins to set, the laser scanner is used to collect the final position data of each monitoring point. The construction parameters and deformation data are normalized and then stored, and input into the neural network model for incremental training. Optionally, the mini-batch gradient descent method is used to update the weights. Through dynamic feedback, the model prediction error decreases as data accumulates.

[0032] In this technical solution, the systematic optimization of the full-cycle management of steel-concrete composite members is realized. First, based on BIM modeling and finite element buckling analysis, the monitoring points are accurately positioned to ensure that the monitoring data focuses on the weak parts of the structure. Then, through digital processing parameter transfer and automated cutting technology, each plate in the steel section member is processed and welded to form the steel section member. In the monitoring of the finished steel section member, a lightweight scanning device is used to replace the traditional point cloud scanning device for the surface monitoring of the steel section member. The position information of each monitoring point is compared with the component model established in the BIM modeling software. If the overall processing of the steel section member does not meet the requirements, rework is carried out. Then, further use the existing data, historical data, and design data to predict the buckling deformation of the component after construction is completed and decide whether to reprocess. This solution effectively monitors and manages the quality of steel section members throughout the whole cycle. The comprehensive cost and implementation difficulty are greatly reduced compared with the existing method of using point cloud scanning data for quality control, and the efficiency is greatly improved.

[0033] In another technical solution, step S1 includes the following steps: S11. Establish a three-dimensional component model in the BIM modeling software. Each plate in the component model needs to be modeled independently. After model checking, select the plate that contacts the concrete on one side and has a non-concrete contact surface on the other side as the analysis plate and transfer it to the finite element calculation software as a finite element model. In the finite element calculation software, material properties are assigned to each calculation object. Specifically, establish a three-dimensional component model of the steel section member in the BIM modeling software, ensure that each plate is modeled independently to avoid overlapping or gap problems between models. Then, screen out the plates that contact the concrete on one side and have a non-concrete contact surface on the other side, and export them as independent geometric files. The format can be SAT, STEP, etc., which can be imported into the finite element calculation software. In the finite element calculation software, assign material properties to the analysis plates according to the design requirements.

[0034] S12. Set the constraint conditions for the analysis plate body in the finite element calculation software, apply the design load, and perform non-linear buckling analysis to obtain the easily buckled area in the analysis plate body. Select at least one point on the non-concrete contact surface of the plate body in the easily buckled area as the buckling monitoring point. Specifically, apply the constraint conditions to the edge of the analysis plate body in the finite element calculation software, load the concrete side pressure according to the design specifications, and apply the line load on the stressed side. Perform non-linear buckling analysis, extract the Mises stress distribution of each point under the first-order buckling mode, and select the area where it exceeds the ultimate buckling stress as the easily buckled area. Select the point with the maximum displacement in the easily buckled area.

[0035] S13. Mark the buckling monitoring points on each analysis plate body in the component model and set several plate body monitoring points at the splicing edges of each analysis plate body and other plate bodies. At least one plate body monitoring point is set on each edge of the analysis plate body, and it is set at a distance of 10 mm to 20 mm from the edge line.

[0036] In another technical solution, step S12 includes the following steps: S121. Take the state of the steel member after all on-site connection measures are completed according to the design requirements and before the concrete is poured on-site as the basis for the constraint conditions of the edges of each analysis plate body. Apply the constraint conditions to the edges of the analysis plate body in the finite element calculation software. The constraint conditions include unconstrained and six-degree-of-freedom constraints. Specifically, before the finite element analysis, it is necessary to clarify the actual constraint state of the steel member before the concrete is poured. Based on this state, apply the constraint conditions to the edges of the analysis plate body in the finite element calculation software. The constraint types include two: unconstrained is used to simulate the state when the edge of the analysis plate body is not welded, and six-degree-of-freedom constraint is used to simulate the rigid connection after the edge of the analysis plate body is welded, so as to accurately simulate the boundary conditions in the actual working condition, ensure that the buckling analysis results conform to the construction reality, and avoid prediction distortion caused by deviation of constraint assumptions.

[0037] S122. Apply the design load to the analysis plate body in the finite element calculation software, and extract the Mises stress data of each point of the analysis plate body under the first-order buckling mode , select the area as the easily buckled area, where , is the Mises stress, is the ultimate buckling stress, K is the buckling coefficient, E is the material elastic modulus, v is the Poisson's ratio, t is the plate thickness, sTo analyze the length of the plate body in the force direction, specifically, first apply the design load to the plate body in the finite element calculation software to simulate the actual stress state; then perform a nonlinear buckling analysis to extract the Mises stress distribution data under the first-order buckling mode, which reflects the most likely deformation pattern when the plate body loses stability, where K is the buckling coefficient, and the reference value ranges from 0.425 to 4.0 according to relevant design codes, and the value is taken according to the edge restraint conditions of the analyzed plate body, where the buckling coefficient K is used to quantify the buckling resistance of the plate body under different boundary conditions. K The value is based on the classical plate and shell buckling theory and the Chinese national standard GB50017-2017 "Steel Structure Design Standard". In engineering practice, generally, the plate body in the profiled steel member is in a state of simply supported on four sides or simply supported on three sides and free on one side before pouring concrete. If the four sides of the analyzed plate body are welded to other structures, it is regarded as a simply supported plate body on four sides. K = 4. If three sides of the analyzed plate body are welded to other plate bodies and the other side is not welded and fixed, it is regarded as a plate body simply supported on three sides and free on one side. K = 0.425. For the mixed restraint forms in between, the buckling coefficient table in various specification guides can be referred to and the value can be matched according to the boundary conditions. K value.

[0038] S123. Select the point with the largest displacement in the easily buckling area as the buckling monitoring point. In the easily buckling area screened in step S122, extract the displacement data of each point and select the point with the largest displacement as the buckling monitoring point to ensure that the buckling monitoring point is set at the position with the most significant deformation.

[0039] In another technical solution, step S3 includes the following steps: S31. Establish the component model coordinate system. In the component model coordinate system, define the model coordinate set of all monitoring point positions, and calculate the model distances between any two monitoring point positions to form the model distance set . Perform a laser scan on the processed profiled steel member to obtain the actual coordinate set of all monitoring point positions in the profiled steel member, and calculate the actual distances between any two monitoring point positions to form the actual distance set , where is the model coordinate set, is the model coordinate of the l th monitoring point position, , is the actual coordinate set, is the actual coordinate of the l th monitoring point position. Specifically, first in the BIM modeling software, a three-dimensional rectangular coordinate system of the component model coordinate system is established with the geometric center or key splicing point of the component model as the origin. The model coordinate data of all monitoring points are exported from the BIM modeling software, where the number of monitoring points is N, and the monitoring point labels are l , and the model coordinates of each monitoring point . Next, it is necessary to calculate the model distance between any two monitoring points to form a model distance set , where ai and aj are indexes of different monitoring points; The actual coordinate data of the processed steel component is obtained through laser scanning technology. The actual coordinates of each monitoring point are After the scanning is completed, the actual coordinate set of all monitoring points is extracted , and the actual distance set is calculated, where the distance between any two monitoring points adopts the same calculation method.

[0040] S32. Traverse all multi-point combinations of all monitoring points , extract the corresponding model distance from the model distance set , extract the corresponding actual distance from the actual distance set , calculate the global distance error sum of this multi-point combination , and select the multi-point combination that makes the smallest as the optimal reference point group. Among them, M is the number of monitoring points in the multi-point combination, a1, a2, aM is the monitoring point index, ai and aj are the monitoring point indexes in the multi-point combination, is ai and aj the model distance between is ai and aj the actual distance between. Specifically, it is necessary to traverse all possible multi-point combinations, select one monitoring point from each analysis plate as a multi-point combination, where the multi-point combination contains at least three monitoring points, calculate the global distance error sum of the actual distance and the corresponding model distance of each multi-point combination, where , and screen out the smallest combination as the optimal reference point group.

[0041] S33. Align the actual optimal reference point group with the center point of the optimal reference point group in the component model coordinate system, and translate the actual coordinates of the monitoring points to the component model coordinate system to obtain the translated corrected actual coordinates , calculate its three-dimensional deviation from the corresponding model coordinates , if is greater than the processing threshold, reprocess the steel section member, where is the corrected actual coordinate after the i monitoring point is translated to the component model coordinate system . Specifically, the center point alignment is used to eliminate the overall position deviation between the actual scanned data and the theoretical model. By translating the coordinates of the actual monitoring points into the component model coordinate system, it can be ensured that both are under the same reference, so as to accurately identify the true local deviation. The processing threshold can refer to the steel structure processing specification.

[0042] In another technical solution, step S4 includes the following steps: S41. Establish an independent coordinate system for each analysis plate body based on the positions of the plate body monitoring points on each analysis plate body of the steel section member, and obtain the initial out-of-plane displacement values of the buckling monitoring points on each analysis plate body. Specifically, an independent local coordinate system is established for each analysis plate body to eliminate the interference of the overall structural deformation on the local buckling analysis. The geometric center of the plate body monitoring points of each analysis plate body is used as the origin of the coordinate system of this analysis plate body. The X-axis and Y-axis are defined according to the plane direction of the analysis plate body, and the Z-axis is perpendicular to the surface of the analysis plate body. The deviation value of the buckling monitoring point in the Z-axis direction in the actual data obtained by laser scanning is used as the initial out-of-plane displacement value of the buckling monitoring point.

[0043] S42. Use the plate body geometric parameters, initial out-of-plane displacement values of the buckling monitoring points, plate body welding parameters, internal filled concrete parameters and final out-of-plane displacement values of the buckling monitoring points in the historical processing data of each steel section member as sample set data to construct a prediction model for the out-of-plane displacement of the buckling monitoring points. Specifically, the neural network model uses the BP neural network model, and the input layer contains neurons corresponding to the parameters. Specifically, first, the data is normalized to eliminate the dimension difference; then, the neural network structure is designed, and the particle swarm optimization algorithm is used to optimize the network weights, and the prediction error is minimized through iterative training.

[0044] S43. Input the plate body geometric parameters, plate body welding parameters, and internal filled concrete parameters of the current analysis plate body into the trained prediction model for the out-of-plane displacement of the buckling monitoring points, and output the predicted out-of-plane displacement value of the buckling monitoring point. The allowable buckling deformation is calculated based on the plate shell buckling theory, considering the critical displacement and the safety factor. If the predicted out-of-plane displacement value of the buckling monitoring point exceeds the allowable buckling deformation, it is determined that this analysis plate body does not meet the requirements, and the steel section member is readjusted. Specifically, the allowable buckling deformation can be calculated according to the plate shell buckling theory to accurately quantify the maximum allowable deformation of the structure before instability, and the output result of the neural network model is compared with the allowable buckling deformation to determine whether to trigger the process adjustment.​

[0045] In another technical solution, in step S42, the geometric parameters of the plate body include plate thickness, width-thickness ratio, height-thickness ratio, and the root mean square error of each monitoring point on the plate during the actual processing of the analyzed plate body. The welding parameters of the plate body include the on-site welding line energy of the analyzed plate body, the total length of the on-site welds, the designed distance between the buckling monitoring point and the nearest weld. The parameters of the infilled concrete include concrete compressive strength, concrete water-binder ratio, concrete elastic modulus, and concrete slump. Among them, a neural network model including an input layer, a hidden layer, and an output layer is adopted. The input layer of the neural network model includes multiple input neuron nodes. Multiple hidden neuron nodes are arranged on the hidden layer. Each input neuron node is respectively connected to each hidden neuron node. The hidden neuron nodes are all connected to the output layer. The geometric parameters of the plate body, the initial out-of-plane displacement value of the buckling monitoring point, the welding parameters of the plate body, and the parameters of the infilled concrete are used as the input values of the input neuron nodes. The output result is compared with the final out-of-plane displacement value of the corresponding buckling monitoring point, and the algorithm optimizes the neural network model to obtain the out-of-plane displacement prediction model of the buckling monitoring point.

[0046] In this technical solution, the neural network model is a BP neural network model. The method for training and optimizing the neural network model includes the following steps; A1. Several factors such as the geometric parameters of the plate body, the initial out-of-plane displacement value of the buckling monitoring point, the welding parameters of the plate body, and the parameters of the infilled concrete are influencing factors. The total number of parameter types is used as the number of input neuron nodes m, and the predicted out-of-plane displacement value of the buckling monitoring point is used as the only output value, that is, the number of nodes in the output layer c = 1, and the number of hidden neuron nodes c in the hidden layer 1 is , where a is a random constant between 1 and 10; A2. Normalize the sample set data, and its mathematical expression is , where x i represents the sample data of the influencing factors, x min and x max are the minimum and maximum values in the sample data respectively, is the dimensionless processed data of the influencing factors; A3. Initialize the mapping relationship between the population particles and the weights and thresholds of the BP neural network, including particle dimension, initial velocity, population size, learning factor, and inertia weight; A4. Input the normalized input variables and output variables into the neural network model, calculate the fitness function value of the particles, and obtain the historical optimal fitness and global fitness of the particles. The fitness function value of the particles is the mean square error of the calculation result, and its function expression is , where represents the predicted value of the i-th sample, and y i is the true value of the i-th sample, and n is the total number of calculation results of the neural network model; A5. Iteratively calculate the fitness of the particles, and update the historical optimal fitness and the global fitness according to the preset update conditions until the preset iteration end condition is met; A6. Update the weights and thresholds of the neural network model to obtain a prediction model for the out-of-plane displacement of the buckling monitoring point.

[0047] In another technical solution, the allowable buckling deformation , where f is the safety factor, with a value range of 2.0 to 4.0, D is the bending stiffness of the plate body, U a is the allowable buckling deformation, s is the length of the plate body in the force direction for analysis.

[0048] A steel-concrete composite member full-cycle management system based on industrial big data, which applies the above-mentioned steel-concrete composite member full-cycle management method based on industrial big data, includes: an industrial big data processing module, a point monitoring module, and a neural network prediction module that are interconnected by signals; the industrial big data processing module includes a component processing sub-module and a data storage sub-module. The component processing sub-module includes BIM modeling software and finite element calculation software. The BIM modeling software is used to model each steel member, set buckling monitoring points and plate body monitoring points, and export sheet metal processing information. The data storage sub-module is used to collect and store the initial out-of-plane displacement values of the buckling monitoring points of each steel member after processing, the corresponding plate body welding parameters, the in-filled concrete parameters, and the final out-of-plane displacement values of the buckling monitoring points during on-site construction; the point monitoring module includes two groups of laser scanners. One group of the two groups of laser scanners is set at the construction site, and the other group is set in the steel member processing factory. The laser scanners can collect the position information of the buckling monitoring points and the plate body monitoring points on the steel members; the neural network prediction module uses the data in the data storage sub-module to train and optimize the neural network model, and inputs the plate body geometric parameters, the initial out-of-plane displacement values of the buckling monitoring points, the plate body welding parameters, and the in-filled concrete parameters to obtain the output predicted out-of-plane displacement values of the buckling monitoring points.

[0049] A computer device includes a memory and a processor. Computer instructions are stored in the memory, and the processor executes the computer instructions to perform the above-mentioned full-cycle management method for steel-concrete composite members based on industrial big data. This computer device can be any terminal device including mobile phones, laptop computers, desktop computers, tablet computers, PDAs (Personal Digital Assistants), POS (Point of Sales) terminals, in-vehicle computers, etc.

[0050] The computer instructions are stored in a readable storage medium, such as a computer floppy disk, USB flash drive, mobile hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disc, etc. The instructions include several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to perform the methods described in various embodiments of the present invention.

[0051] It should be noted that although the above steps are described in a specific order, it does not mean that the steps must be executed in the above specific order. In fact, some of these steps can be executed concurrently or even in a changed order as long as the required functions can be achieved. The number of devices and the processing scale described here are used to simplify the description of the present invention, and it is obvious to those skilled in the art for the application, modification, and variation of the present invention.

[0052] Although the embodiments of the present invention have been disclosed above, it is not limited to only the applications listed in the specification and embodiments. It can be fully applied to various fields suitable for the present invention. For those familiar with the field, additional modifications can be easily made. Therefore, without departing from the general concept defined by the claims and the equivalent scope, the present invention is not limited to the specific details and the illustrated and described examples here.

Claims

1. A full-cycle management method for steel-concrete components based on industrial big data, characterized in that: The following steps are involved: S1. Establish a component model in the BIM modeling software, select each plate in the component model that is in contact with concrete on one side as the analysis plate, import the finite element calculation software to perform buckling analysis, and select a number of monitoring points on the non-concrete contact surface of the analysis plate, wherein the monitoring points include buckling monitoring points and plate monitoring points; S2. Extracting component contour data and monitoring point coordinates from the component model to form processing parameters, cutting the steel plate profile and marking the monitoring points based on the processing parameters, and processing the steel component; S3, after the steel component is processed, scan it to obtain the actual coordinates of each monitoring point on the steel component and compare them with the model coordinates of each monitoring point in the component model. If the error between the actual coordinates and the model coordinates exceeds the processing threshold, reprocess the steel component; S4, establishing a buckling monitoring point out-of-plane displacement prediction model, inputting the initial out-of-plane displacement value of the buckling monitoring point of each analysis plate, plate geometry parameters, plate welding parameters, and internal filling concrete parameters, and obtaining the predicted out-of-plane displacement value of each buckling monitoring point after the concrete pouring is completed. If the predicted out-of-plane displacement value of the buckling monitoring point exceeds the allowable buckling deformation, reprocessing the steel member, and repeating steps S3-S4 until each analysis plate meets the requirements; S5. The steel structure is installed, welded and concrete is poured on site. After the initial setting of the concrete, the final out-of-plane displacement value of the buckling monitoring point on the steel structure is collected by a laser scanner. The final out-of-plane displacement value of the buckling monitoring point and the corresponding plate welding parameters and internal concrete filling parameters are fed back to the out-of-plane displacement prediction model of the buckling monitoring point for training.

2. The full-cycle management method of steel-concrete components based on industrial big data as described in claim 1 is characterized in that: Step S1 includes the following steps: S11, establishing a three-dimensional component model in the BIM modeling software, each plate in the component model needs to be independently modeled, and after the model is verified, a plate with one side in contact with concrete and the other side being a non-concrete contact surface is selected as an analysis plate and transferred into the finite element calculation software as a finite element model, and material properties are assigned to each calculation object in the finite element calculation software; S12, setting constraint conditions for the analysis plate in the finite element calculation software, applying the design load and performing nonlinear buckling analysis to obtain the buckling-prone area in the analysis plate, and selecting at least one point on the non-concrete contact surface of the plate in the buckling-prone area as a buckling monitoring point; S13. Mark buckling monitoring points on each analysis plate in the component model and set a number of plate monitoring points at the joint edges of each analysis plate and other plates.

3. The full-cycle management method for steel-concrete components based on industrial big data according to claim 2 is characterized in that: Step S12 includes the following steps: S121. The state of the steel structure after all the on-site connection measures are completed according to the design requirements and before the concrete is poured on-site is used as the basis for the constraint conditions of the edges of each analysis plate. The constraint conditions are imposed on the edges of the analysis plate in the finite element calculation software. The constraint conditions include no constraint and six-degree-of-freedom constraint. S122. Apply the design load to the analysis plate in the finite element calculation software and extract the Mises stress data of each point of the analysis plate under the first-order buckling mode. , select The area of ​​​​the flexible region is , is the Mises stress, is the ultimate buckling stress, K is the buckling coefficient, E is the elastic modulus of the material, v is Poisson's ratio, t is the plate thickness, s To analyze the length of the plate in the direction of force; S123. Select the point with the largest displacement in the buckling-prone area as the buckling monitoring point.

4. The full-cycle management method for steel-concrete components based on industrial big data according to claim 1 is characterized in that: Step S3 includes the following steps: S31. Establish a component model coordinate system, and define the model coordinate set of all monitoring points in the component model coordinate system. , and calculate the model distance between any two monitoring points to form a model distance set , perform laser scanning on the processed steel structure to obtain the actual coordinate set of all monitoring points in the steel structure , and calculate the actual distance between any two monitoring points to form an actual distance set ,in, N The number of monitoring points, is the model coordinate set, For the l Model coordinates of monitoring points , is the actual coordinate set, For the l The actual coordinates of the monitoring points ; S32, traverse the multi-point combination of all monitoring points , from the model distance set Extract the corresponding model distance , from the actual distance set Extract the corresponding actual distance , calculate the global distance error and , select The smallest multi-point combination is taken as the optimal reference point group, where M is the number of monitoring points in the multi-point combination, a1, a2, aM is the monitoring point index, ai and aj is the index of the monitoring point in the multi-point combination, for ai and aj The model distance between for ai and aj The actual distance between S33, aligning the actual optimal reference point group with the center point of the optimal reference point group in the component model coordinate system, and translating the actual coordinates of the monitoring point to the component model coordinate system to obtain the corrected actual coordinates after translation , calculate its corresponding model coordinates Three-dimensional deviation ,like If it is greater than the processing threshold, the steel member is reprocessed, where: For the i Corrected actual coordinates after the monitoring point is translated to the component model coordinate system .

5. The full-cycle management method for steel-concrete components based on industrial big data according to claim 1, characterized in that: Step S4 includes the following steps: S41, establishing an independent coordinate system for each analysis plate based on the position of the plate monitoring point on each analysis plate on the steel member, and obtaining the initial out-of-plane displacement value of the buckling monitoring point on each analysis plate; S42, using the plate geometric parameters of the analysis plate in each steel member in the historical processing data, the initial out-of-plane displacement value of the buckling monitoring point, the plate welding parameters, the inner filling concrete parameters and the final out-of-plane displacement value of the buckling monitoring point as sample set data to construct an out-of-plane displacement prediction model for the buckling monitoring point; S43. Use the plate geometry parameters, plate welding parameters, and internal concrete filling parameters of the current analysis plate to input the trained buckling monitoring point out-of-plane displacement prediction model, output the predicted out-of-plane displacement value of the buckling monitoring point, and calculate the allowable buckling deformation variable based on the plate-shell buckling theory, taking into account the critical displacement and safety factor. If the predicted out-of-plane displacement value of the buckling monitoring point exceeds the allowable buckling deformation variable, it is determined that the analysis plate does not meet the requirements and the steel structure is readjusted.

6. The full-cycle management method for steel-concrete components based on industrial big data according to claim 5 is characterized in that: In step S42, the plate body geometric parameters include plate thickness, width-to-thickness ratio, height-to-thickness ratio, and the root mean square error of each monitoring point on the plate during actual processing of the analyzed plate body; the plate body welding parameters include the on-site welding line energy of the analyzed plate body, the total length of the on-site weld, the design distance between the buckling monitoring point and the nearest weld, and the internal filling concrete parameters include the concrete compressive strength, concrete water-cement ratio, concrete elastic modulus, and concrete slump. A neural network model including an input layer, a hidden layer, and an output layer is adopted, wherein the input layer of the neural network model includes a plurality of input neuron nodes, and a plurality of hidden neuron nodes are arranged on the hidden layer, each of the input neuron nodes is respectively connected to each of the hidden neuron nodes, and the hidden neuron nodes are all connected to the output layer. The plate body geometric parameters, the initial out-of-plane displacement value of the buckling monitoring point, the plate body welding parameters, and the internal filling concrete parameters are used as input values ​​of the input neuron nodes, and the output results are compared with the corresponding final out-of-plane displacement values ​​of the buckling monitoring points. The algorithm optimizes the neural network model to obtain the out-of-plane displacement prediction model of the buckling monitoring point.

7. The full-cycle management method for steel-concrete components based on industrial big data according to claim 5 is characterized in that: The allowable buckling deformation ,in f is the safety factor, ranging from 2.0 to 4.0, D is the bending stiffness of the plate, U a To allow for the buckling deformation, s To analyze the length of the plate in the direction of force.

8. A full-cycle management system for steel-concrete components based on industrial big data, applied to the full-cycle management method for steel-concrete components based on industrial big data as claimed in any one of claims 1 to 7, characterized in that: include: Industrial big data processing module, point monitoring module, and neural network prediction module that are interconnected by signals; The industrial big data processing module includes a component processing submodule and a data storage submodule. The component processing submodule includes BIM modeling software and finite element calculation software. The BIM modeling software is used to model each steel component, set buckling monitoring points and plate monitoring points, and export plate processing information. The data storage submodule is used to collect and store the initial out-of-plane displacement value of the buckling monitoring point of each steel component after processing, the corresponding plate welding parameters during the on-site construction process, the internal filling concrete parameters, and the final out-of-plane displacement value of the buckling monitoring point; The point monitoring module includes two groups of laser scanners, one of which is set at the construction site and the other is set in the steel component processing plant. The laser scanners can collect the position information of the buckling monitoring points and the plate monitoring points on the steel component; The neural network prediction module uses the data in the data storage submodule to train and optimize the neural network model, and obtains the output predicted buckling monitoring point out-of-plane displacement value after inputting the plate geometric parameters, initial out-of-plane displacement value of the buckling monitoring point, plate welding parameters, and internal filling concrete parameters.

9. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores computer instructions, and the processor executes the full-cycle management method of steel-concrete components based on industrial big data as described in any one of claims 1 to 7 by executing the computer instructions.

Images