A method for designing a steel reinforcement member

Abstract

The application discloses a design method for steel reinforcement components. In the application, a machine learning model can quickly process and analyze a large amount of historical design data, automatically identify key factors affecting the calculation of axial force reinforcement, and improve the calculation efficiency. The machine learning model automatically completes the calculation through an algorithm, reducing errors in the calculation process. A fully trained machine learning model can accurately predict the reinforcement area based on the input parameters, improving the accuracy of the design. The machine learning model can discover hidden relationships between data by analyzing historical data, which is particularly important for the calculation of reinforcement caused by axial force, as axial force is related to various factors such as structural size, material properties, etc. Adaptability and generalization ability: the application of machine learning to the calculation process of reinforcement caused by axial force can significantly improve the intelligent level of design, improve the efficiency and quality of design, and provide strong technical support for cost optimization.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of steel reinforcement member design, and particularly relates to a design method for a steel reinforcement member. BACKGROUND

[0002] In the modern construction field, steel reinforcement members play a crucial role as the backbone of building structures, providing solid support for various types of buildings. Today, let's delve into the world of steel reinforcement members and understand their unique charm and the importance of design. Steel reinforcement members, as the "powerful" in construction, with their high strength and good ductility, ensure the stability and safety of buildings. Steel reinforcement members have good corrosion resistance, ensuring the long service life of buildings even in harsh environments. Standardized production of steel reinforcement members facilitates rapid assembly on construction sites, improving construction efficiency and shortening construction period. Steel reinforcement member design is not only a technology but also an art. It combines knowledge from multiple disciplines such as mechanics, materials science, and structural science, injecting infinite innovation vitality into the construction industry. With the continuous development of construction technology, steel reinforcement member design will play an increasingly important role in future construction.

[0003] However, common traditional manual calculation methods are prone to errors due to the involvement of multiple variables and complex relationships in axial force reinforcement calculation, especially when dealing with large amounts of data and iterative design. At the same time, traditional reinforcement calculation methods often rely on fixed calculation formulas and empirical rules, which may not be flexible enough for special or complex design situations, and may not provide the optimal reinforcement recommendations. SUMMARY

[0004] The purpose of the present application is to solve the above-mentioned problems and provide a design method for steel reinforcement members.

[0005] The technical solution adopted by the present application is as follows: a design method for steel reinforcement members, the method comprising the following steps:

[0006] S1: preliminary preparation, determine the design task, understand the engineering requirements, and clearly define the purpose, size, shape, and material requirements of the steel reinforcement member;

[0007] S2: collect data, collect relevant design specifications, standards, atlases, and material performance data, and analyze the geological, hydrological, and climatic conditions of the project site to provide a basis for subsequent design;

[0008] S3: preliminary design, determine the type of member according to the engineering requirements, select the appropriate member type, including beams, slabs, columns, and walls; determine the member size according to the stress analysis, construction requirements, and construction convenience to determine the cross-sectional size of the member;

[0009] S4: Perform stress analysis on the component to determine internal forces, including bending moment, shear force, and axial force. According to the results of the stress analysis, select appropriate steel and concrete grade;

[0010] S5: Perform reinforcement calculation. According to the internal force calculation results, determine the steel cross-sectional area. Collect historical structural design case data, including structure type, material properties, axial force design value, and steel configuration;

[0011] S6: Process and standardize data on axial force size, material type, and structure size to enable better understanding by machine learning models;

[0012] S7: Then perform RNN feature extraction, input layer processing, and output layer processing to extract various data features for steel reinforcement component design and reinforcement calculation;

[0013] S8: Perform feature fusion to concatenate the RNN-extracted steel reinforcement component data features to form a new steel reinforcement component data feature vector F, resulting in a more comprehensive data representation;

[0014] S9: Train the steel reinforcement component data calculation model using the data to predict the optimal shear reinforcement area Asv under a given shear force;

[0015] S10: Then perform structural design to determine the steel reinforcement cover thickness; set adequate steel reinforcement spacing to ensure construction quality; consider construction convenience and reasonably set steel reinforcement hooks, welding construction measures;

[0016] S11: Perform review and optimization. Review the calculation results against the specifications and check if the reinforcement calculation results meet the requirements. On the premise of safety, economy, and reasonableness, optimize the design scheme;

[0017] S12: Submit the design drawings to the relevant departments for review. Based on the review comments, modify and improve the drawings. Deliver the final confirmed design drawings to the construction party and perform technical briefing.

[0018] In a preferred embodiment, in step S6, the reinforcement calculation caused by the bending moment uses the formula: where M represents the bending moment design value; fy represents the yield strength of the steel; As represents the tensile steel area; and d represents the effective height of the steel, which is the distance from the center of the steel to the edge of the compression zone.

[0019] The formula for calculating the tensile steel area is:

[0020] In a preferred embodiment, in step S6, the Min-Max normalization method is used to normalize the various data features of the reinforcement calculation. The specific steps are as follows:

[0021] S1. Calculate maximum and minimum: For each feature, find the maximum (Max) and minimum (Min) in the dataset;

[0022] S2. Apply transformation formula: For each feature value x in the dataset, transform using the following formula:

[0023] x-Min

[0024] x norm = Max -Min;

[0025] where x norm is the transformed value;

[0026] S3. Transform the dataset: Apply the above transformation to each feature of the entire dataset;

[0027] After min-max normalization, all features will be scaled to the range of 0 to 1.

[0028] In a preferred embodiment, the input layer processing of the RNN feature extraction in step S7 comprises:

[0029] Converting the original data text into word vectors using Word Embedding technology;

[0030] Let xi be the input vector of the i-th time step, V be the embedding matrix, and t be the number of time steps, then xi = V·ei, where ei is a one-hot encoding vector representing a word;

[0031] The hidden layer processing of the RNN feature extraction comprises: using LSTM (Long Short-Term Memory Network) RNN unit to process time series data;

[0032] Let hi be the hidden state of the i-th time step, and ci be the cell state, then the update formula is:

[0033] fi = σ(Wf·[hi-1, xi] + bf);

[0034] II = σ(Wi·[hi-1, xi] + bi);

[0035] c~i = tanh(Wc·[hi-1, xi] + bc)

[0036] ci = fi·ci-1 + i i ·c~i;

[0037] oi = σ(Wo·[hi-1, xi] + bo);

[0038] hi = oi·tanh(ci);

[0039] where Wf, Wi, Wc, Wo are weight matrices, bf, bi, bc, bo are bias terms, and σ is a sigmoid function.

[0040] The output layer processing of the RNN feature extraction includes using the hidden state ht of the last time step as the output feature.

[0041] In a preferred embodiment, the output layer processing of the RNN feature extraction in step S7 includes using the hidden state ht of the last time step as the output feature.

[0042] In a preferred embodiment, the model training process in step S9 includes:

[0043] Data division: divide the steel bar component data features Fpca into a training set Xtrain, a validation set Xval, and a test set Xtest

[0044] Model selection: select a logistic regression model as the probability discrimination model

[0045] Model training: train using the training set Xtrain and the label Ytrain;

[0046] Let θ be the model parameters and σ be the logistic function, then the model output is:

[0047] Calculate the loss function, the calculation formula is:

[0048]

[0049] where L(θ) is the loss function, representing the difference between the model predicted value and the true value; N is the total number of training samples;

[0050] yi is the true label of the i-th sample, taking values 0 or 1;

[0051] y^i is the predicted probability of the i-th sample by the model, calculated as y^i = σ(θT·Fpca,i), where σ is the logistic function, Fpca,i is the feature vector after dimensionality reduction of the i-th sample, and θ is the model parameters.

[0052] In a preferred embodiment, in step S10, in the process of reasonably setting the steel bar hook and welding structure measures

[0053] The hook length calculation formula is: lhook = diameter multiple · d

[0054] Welding length calculation: lweld = specification required welding length.

[0055] In a preferred embodiment, in step S10, the size, shape and arrangement of the reinforcement of the component are accurately expressed according to the requirements of structural design; during the drawing process, the plan view, elevation view and sectional view of the component are respectively made to ensure that the details of the component are fully displayed from different perspectives; the plan view mainly displays the plan position and reinforcement distribution of the component, the elevation view displays the elevation shape and elevation arrangement of the reinforcement of the component, and the sectional view is used to display the detailed position and shape of the internal reinforcement of the component.

[0056] In a preferred embodiment, in step S10, the reinforcement number, specification, quantity and spacing information are marked in detail in the drawing, and the concrete strength grade, protective layer thickness and embedded part position important information are also marked; after the drawing is completed, the drawing is audited to ensure the accuracy, integrity and constructability of the drawing; finally, construction instructions are written according to the drawing, including material requirements, construction methods, quality standards and acceptance requirements, so that construction personnel can construct according to the design intention to ensure the engineering quality and safety; during the whole process of drawing construction drawing, close communication with designers and construction personnel is needed to ensure that the drawing can accurately reflect the design intention and meet the actual needs of construction.

[0057] In a preferred embodiment, in step S11, the checking and optimization need to check the relevant design specifications, and the reinforcement calculation results are checked in detail to ensure that all calculation results meet the specification requirements; specific checking points include: first, check whether the bending moment, shear force and axial force internal force calculation is accurate, and whether the reinforcement cross-sectional area determined thereby is sufficient; second, verify whether the arrangement of the reinforcement is reasonable, including whether the reinforcement diameter, quantity and spacing meet the minimum diameter and maximum spacing specification requirements; third, confirm whether the anchorage length, lap length and construction details of the reinforcement meet the specification requirements; in addition, it is also necessary to check whether the concrete protective layer thickness meets the requirements of corrosion resistance and durability.

[0058] In summary, due to the adoption of the above technical scheme, the beneficial effects of the present application are:

[0059] 1、In the present application, a large amount of historical design data can be quickly processed and analyzed by a machine learning model, and the key factors affecting the calculation of axial force reinforcement are automatically identified, thereby improving the calculation efficiency. The machine learning model automatically completes the calculation through an algorithm, reducing errors in the calculation process. The well-trained machine learning model can accurately predict the reinforcement area based on the input parameters, improving the accuracy of the design. The machine learning model can discover hidden relationships between data by analyzing historical data, which is particularly important for reinforcement calculation caused by axial force, as axial force is related to various factors such as structural size, material properties, etc. Adaptability and generalization ability: the trained machine learning model not only applies to specific cases, but also has a certain generalization ability, which can handle similar but different design problems. At any stage of design, the machine learning model can provide instant calculation results to help make quick decisions. With the accumulation of more data, the machine learning model can continuously learn and optimize, constantly improving the accuracy and efficiency of prediction. The application of machine learning to the reinforcement calculation process caused by axial force can significantly improve the intelligent level of design, improve design efficiency and quality, and provide strong technical support for cost optimization.

[0060] 2、In the present application, through accurate calculation of bending moment and reinforcement design, the performance of the structure in actual use can be ensured to be more reliable, and the durability and service life of the structure can be improved. Optimized reinforcement design can reduce unnecessary material use, save natural resources, and reduce environmental pollution. Reasonable reinforcement design can simplify the construction process, improve construction efficiency, and reduce on-site operation difficulty. Optimization of structural design can reduce the cost of later maintenance and repair, and improve overall economic benefits. Optimized shear reinforcement design can make the structure better adapt to different environmental and geological conditions, improve the adaptability and reliability of the structure. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical scheme and advantages of the present application clearer, the following embodiments are used to further illustrate the present application. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application.

[0062] Embodiment:

[0063] A design method for a steel reinforcement member, the method comprising the following steps:

[0064] S1: preliminary preparation, determine the design task, understand the engineering demand, and clearly understand the purpose, size, shape, material requirement information of the steel reinforcement member;

[0065] S2: collect information, collect relevant design specifications, standards, atlas, material performance, etc. At the same time, analyze the engineering geological conditions to understand the geological, hydrological, climatic conditions of the project site, and provide basis for subsequent design;

[0066] S3: Preliminary design, determine the type of component According to the engineering requirements, select the appropriate component type, including beams, plates, columns, walls; Determine the size of the component According to the stress analysis, construction requirements and construction convenience, determine the cross-sectional size of the component;

[0067] S4: Stress analysis of components, determine internal forces, including bending moment, shear force, axial force, select appropriate steel and concrete grade according to the results of stress analysis;

[0068] S5: Reinforcement calculation, according to the internal force calculation results, determine the steel area, collect historical structure design case data, including structure type, material characteristics, axial force design value, steel configuration;

[0069] S6: Process and standardize the data of axial force size, material type and structure size so that the machine learning model can better understand;

[0070] S7: Then perform RNN feature extraction, input layer processing and output layer processing to extract various data features for steel reinforcement component design and reinforcement calculation;

[0071] S8: Feature fusion, splice the steel reinforcement component data features extracted by RNN to form a new steel reinforcement component data feature vector F, and form a more comprehensive data representation;

[0072] S9: Train the steel reinforcement component data calculation model, use the data to train the model, and predict the optimal shear steel area Asv under a given shear force;

[0073] S10: Then perform construction design, determine the steel reinforcement cover thickness; Set enough steel spacing to ensure construction quality; Consider construction convenience and reasonably set steel hooks, welding construction measures;

[0074] S11: Check and optimize, check the calculation results against the specifications, and check if the reinforcement calculation results meet the requirements. On the premise of safety, economy and reasonableness, optimize the design scheme;

[0075] S12: Review the drawings Submit the design drawings to the relevant departments for review, modify and improve the drawings according to the review opinions, and deliver the final confirmed design drawings to the construction party and perform technical briefing.

[0076] In step S6, the reinforcement calculation caused by the bending moment uses the formula: Where M represents the bending moment design value; fy represents the yield strength of the steel; As represents the area of the tensile steel, and d represents the effective height of the steel, which is the distance from the center of the steel to the edge of the compression zone;

[0077] The formula for calculating the area of the tensile reinforcement is:

[0078] In step S6, the Min-Max normalization method is used to normalize the data features calculated by the reinforcement, and the specific steps are as follows:

[0079] 1. Calculate the maximum and minimum values: for each feature, find the maximum (Max) and minimum (Min) values in the data set;

[0080] 2. Apply the transformation formula: for each feature value x in the data set, use the following formula for transformation:

[0081] x-Min

[0082] x norm = Max -Min;

[0083] Where x norm is the transformed value;

[0084] 3. Transform the data set: apply the above transformation to each feature of the entire data set.

[0085] After Min-Max normalization, all features will be scaled to the range of 0 to 1.

[0086] In step S7, the input layer processing of RNN feature extraction includes:

[0087] Using Word Embedding technology to convert original data text into word vectors;

[0088] Let xi be the input vector of the i-th time step, V be the embedding matrix, and t be the number of time steps, then xi = V·ei, where ei is a one-hot encoding vector representing a word;

[0089] The hidden layer processing of RNN feature extraction includes: using LSTM (Long Short-Term Memory Network) RNN unit to process time series data;

[0090] Let hi be the hidden state of the i-th time step, and ci be the cell state, then the update formula is:

[0091] fi = σ(Wf·[hi-1, xi] + bf);

[0092] II = σ(Wi·[hi-1, xi] + bi);

[0093] c~i = tanh(Wc·[hi-1, xi] + bc)

[0094] ci = fi·ci-1 + i i ·c~i;

[0095] oi = σ(Wo · [hi-1, xi] + bo);

[0096] hi = oi · tanh(ci);

[0097] where Wf, Wi, Wc, Wo are weight matrices, bf, bi, bc, bo are bias terms, and σ is a sigmoid function.

[0098] The output layer processing of the RNN feature extraction includes using the hidden state ht of the last time step as the output feature.

[0099] In step S7, the output layer processing of the RNN feature extraction includes using the hidden state ht of the last time step as the output feature.

[0100] In step S9, the model training process includes:

[0101] Data division: divide the steel bar component data features Fpca into a training set Xtrain, a validation set Xval, and a test set Xtest;

[0102] Model selection: select a logistic regression model as the probability discrimination model;

[0103] Model training: train using the training set Xtrain and the label Ytrain;

[0104] Let θ be the model parameters and σ be the logistic function, then the model output is:

[0105] Calculate the loss function, the calculation formula is:

[0106]

[0107] where L(θ) is the loss function, representing the difference between the model prediction and the true value; N is the total number of training samples;

[0108] yi is the true label of the i-th sample, taking values 0 or 1;

[0109] y^i is the predicted probability of the i-th sample by the model, calculated as y^i = σ(θT·Fpca,i), where σ is the logistic function, Fpca,i is the feature vector after dimensionality reduction of the i-th sample, and θ is the model parameters.

[0110] In step S10, in the process of reasonably setting the steel bar hook and welding structural measures:

[0111] The hook length calculation formula is: lhook = diameter multiple · d

[0112] Weld length calculation: lweld = specified weld length.

[0113] In step S10, computer-aided design (CAD) software or manual drawing methods are used to accurately represent the dimensions, shapes, and reinforcement placement of the components according to the requirements of the structural design. During the drawing process, plan views, elevation views, and section views of the components are created to ensure a comprehensive display of the details from different perspectives. Plan views mainly show the planar position and reinforcement distribution of the components, elevation views show the elevation shape and reinforcement placement of the components, and section views show the detailed position and shape of the internal reinforcement of the components.

[0114] In step S10, the drawing is annotated with detailed information such as reinforcement number, specification, quantity, and spacing, for example, "#8@200" represents a steel bar with a diameter of 8mm and a spacing of 200mm. In addition, important information such as the strength grade of the concrete, the thickness of the protective layer, and the location of the embedded parts is also annotated. After the drawing is completed, it is reviewed to ensure its accuracy, completeness, and constructability. Finally, construction instructions are written based on the drawing, including material requirements, construction methods, quality standards, and acceptance requirements, so that construction personnel can follow the design intent and ensure project quality and safety. Throughout the process of drawing construction drawings, close communication with designers and construction personnel is required to ensure that the drawings accurately reflect the design intent and meet the actual needs of construction.

[0115] In step S11, the checking and optimization is a key step to ensure that the structural design scheme is both safe and economical. In this process, the calculation results of reinforcement need to be checked in detail against relevant design specifications to ensure that all calculation results meet the specification requirements. Specific checking points include: first, checking whether the internal force calculation such as bending moment, shear force and axial force is accurate, and whether the steel reinforcement cross-sectional area determined thereby is sufficient; second, verifying whether the steel reinforcement arrangement is reasonable, including whether the steel reinforcement diameter, quantity and spacing meet the minimum diameter and maximum spacing requirements of the specification; third, confirming whether the steel reinforcement anchorage length and lap length meet the specification requirements; in addition, it is also necessary to check whether the concrete cover thickness meets the requirements of corrosion prevention and durability. In terms of optimization, under the premise of ensuring structural safety, economic and reasonable considerations are made for the design scheme, and the optimization points include: adjusting the steel reinforcement configuration, reducing the amount of steel reinforcement and reducing costs by using higher strength steel reinforcement or optimizing steel reinforcement layout; optimizing concrete design, such as adjusting the concrete strength grade to reduce material costs; considering the use of new materials or structural systems to improve structural performance or reduce construction difficulty; at the same time, multiple schemes are compared, and through iterative calculation and evaluation, the optimal design scheme that meets the requirements of structural function, safety and economy is found. Through this series of checking and optimization, the design scheme can meet the specification requirements while achieving the design goals of safety, economy and reasonableness

[0116] In the present application, the machine learning model can quickly process and analyze a large amount of historical design data, automatically identify the key factors affecting the calculation of axial force reinforcement, and improve the calculation efficiency. The machine learning model automatically completes the calculation through algorithms, reducing errors in the calculation process. The well-trained machine learning model can accurately predict the steel reinforcement area based on the input parameters, improving the accuracy of the design. The machine learning model can discover hidden relationships between data by analyzing historical data, which is particularly important for reinforcement calculation caused by axial force, as axial force is related to multiple factors such as structural size and material properties. Adaptability and generalization ability: the trained machine learning model not only applies to specific cases, but also has a certain generalization ability to handle similar but different design problems. The machine learning model can provide instant calculation results at any stage of the design process, helping to make quick decisions. With the accumulation of more data, the machine learning model can continuously learn and optimize, constantly improving the accuracy and efficiency of prediction. The application of machine learning to the process of reinforcement calculation caused by axial force can significantly improve the intelligent level of design, improve the efficiency and quality of design, and provide strong technical support for cost optimization.

[0117] In the present application, through the accurate calculation of bending moment and reinforcement design, the performance of the structure in actual use can be ensured to be more reliable, and the durability and service life of the structure are improved. The optimized reinforcement design can reduce unnecessary material use, save natural resources, and reduce environmental pollution. Reasonable reinforcement design can simplify the construction process, improve construction efficiency, and reduce on-site operation difficulty. The optimization of structure design can reduce the cost of later maintenance and repair, and improve the overall economic benefit. The optimized shear reinforcement design can make the structure better adapt to different environmental and geological conditions, and improve the adaptability and reliability of the structure.

[0118] It should be noted that, in this text, relational terms such as first and second are used only to distinguish one entity or operation from another, and do not necessarily require or imply any actual such relationship or order between the entities or operations. Moreover, the term "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or other elements inherent in such a process, method, article or device. Without more limitations, the element defined by the statement "comprising a" does not exclude the presence of other identical elements in the process, method, article or device including the element.

[0119] The above examples are only used to illustrate the technical solutions of the present application, and not to limit it; although the present application has been described in detail with reference to the foregoing examples, those skilled in the art should understand that they can still modify the technical solutions recorded in the foregoing examples, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solution deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for designing a reinforcing member, characterized by: The method comprises the following steps: S1: making preparations, determining design tasks, understanding engineering requirements, and clearly understanding the purpose, size, shape, and material requirements of the steel reinforcement component; S2: collecting information, collecting relevant design specifications, standards, atlases, and material performance data, and analyzing the geological, hydrological, and climatic conditions of the project site to provide a basis for subsequent design; S3: making preliminary design, determining component type, selecting appropriate component type according to engineering requirements, including beams, plates, columns, and walls; determining component size according to stress analysis, construction requirements, and construction convenience to determine the cross-sectional size of the component; S4: performing stress analysis on the component to determine internal forces, including bending moment, shear force, and axial force, and selecting appropriate steel reinforcement and concrete grade according to the stress analysis results; S5: performing reinforcement calculation, determining steel reinforcement cross-sectional area according to internal force calculation results, collecting historical structural design case data, including structure type, material characteristics, axial force design value, and steel reinforcement configuration; S6: processing and standardizing data on axial force size, material type, and structure size to enable the machine learning model to better understand; S7: then performing RNN feature extraction, input layer processing, and output layer processing to extract various data features for steel reinforcement component design and reinforcement calculation; S8: performing feature fusion to splice the RNN-extracted steel reinforcement component data features together to form a new steel reinforcement component data feature vector F, forming a more comprehensive data representation; S9: training the steel reinforcement component data calculation model using the data to train the model and predict the optimal shear reinforcement area Asv under a given shear force; S10: then performing construction design to determine the steel reinforcement cover thickness; setting sufficient steel reinforcement spacing to ensure construction quality; and considering construction convenience to reasonably set steel reinforcement hooks, welding construction measures; S11: performing checking and optimization to check the calculation results against the specifications and check whether the reinforcement calculation results meet the requirements; and optimizing the design scheme under the premise of safety, economy, and reasonableness; S12: review the drawings will design drawings submitted to the relevant departments for review, according to the comments, the drawings are modified and improved, the final confirmation of the design drawings to the construction side, and technical disclosure, the step S6, the moment caused by the reinforcement calculation using formula: Where M represents the design value of the bending moment; fy represents the yield strength of the reinforcement; As represents the area of the tensile reinforcement, d represents the effective height of the reinforcement, the distance from the center of the reinforcement to the edge of the compression zone; The formula for calculating the area of the tensile reinforcement is: In step S6, the Min-Max normalization method is used to normalize the data features calculated by the reinforcement, and the specific steps are as follows: S1. Calculate the maximum and minimum values: for each feature, find the maximum (Max) and minimum (Min) values in the dataset; S2. Apply the transformation formula: for each feature value x in the dataset, use the following formula for transformation: ； wherein is the transformed value; S3. Transform the dataset: apply the above transformation to each feature of the entire dataset; After min-max normalization, all features will be scaled to the range of 0 to 1, and in step S7, the input layer processing of RNN feature extraction includes: Using Word Embedding technology to convert the original data text into word vectors; Let xi be the input vector at the i-th time step, V be the embedding matrix, and t be the number of time steps, then xi=V·ei, where ei is a one-hot encoding vector representing a word; The hidden layer processing of RNN feature extraction includes: using LSTM (Long Short-Term Memory Network) RNN unit to process time series data; Let hi be the hidden state at the i-th time step, and ci be the cell state, then the update formula is: fi=σ(Wf·[hi−1,xi]+bf); II = σ(Wi·[hi-1, xi] + bi); ci = tanh(Wc·[hi-1, xi] + bc) ci = fi · ci−1+ i i · ci−1; oi = σ(Wo·[hi-1, xi] + bo); hi = oi·tanh(ci); where Wf, Wi, Wc, Wo are weight matrices, bf, bi, bc, bo are bias terms, and σ is a sigmoid function. The output layer processing of the RNN feature extraction includes using the hidden state ht of the last time step as the output feature.

2. A method of designing a reinforcing member as claimed in claim 1, wherein: The model training process in step S9 includes: Data division: divide the steel member data features Fpca into a training set Xtrain, a validation set Xval, and a test set Xtest Model selection: select a logistic regression model as the probability discrimination model Model training: train using the training set Xtrain and the label Ytrain; Let θ be the model parameters and σ be the logistic function, then the model output is: ; Calculate the loss function, and the calculation formula is: ； where: L(θ) is the loss function, representing the difference between the model predicted value and the true value; N is the total number of training samples; yi is the true label of the ith sample, taking values 0 or 1; y^i is the predicted probability of the model for the ith sample, calculated as y^i = σ(θT·Fpca,i), where σ is the logistic function, Fpca,i is the feature vector after dimensionality reduction of the ith sample, and θ is the model parameter.

3. A method of designing a reinforcing member as claimed in claim 1, wherein: In step S10, in the process of reasonably setting the steel hook and welding structural measures The hook length calculation formula is: lhook = diameter multiple · d Welding length calculation: lweld = specification required welding length.

4. A method of designing a reinforcing member as claimed in claim 1, wherein: In step S10, according to the requirements of structural design, accurately express the size, shape and arrangement of the member and the reinforcement; During drawing, the plan, elevation and section views of the member will be drawn respectively to ensure that the details of the member are fully displayed from different perspectives; the plan view mainly shows the plan position and reinforcement distribution of the member, the elevation view shows the elevation shape and reinforcement elevation arrangement of the member, and the section view is used to show the detailed position and shape of the internal reinforcement of the member.

5. A method of designing a reinforcing member as claimed in claim 1, wherein: In step S10, in the drawing, the reinforcement number, specification, quantity, spacing information are marked in detail, and the concrete strength grade, protective layer thickness, embedded part position important information are also marked; after drawing, the drawing is audited to ensure the accuracy, integrity and constructability of the drawing; finally, the construction instructions are written according to the drawing, including material requirements, construction methods, quality standards, acceptance requirements, so that construction personnel can construct according to the design intention, ensure the engineering quality and safety; during the whole process of drawing construction drawing, close communication with designers and construction personnel is needed to ensure that the drawing can accurately reflect the design intention and meet the actual needs of construction.

6. A method of designing a reinforcing member as claimed in claim 1, wherein: In the step S11, the checking and optimization need to be checked against the relevant design specification, and the reinforcement calculation results are checked in detail to ensure that all calculation results meet the specification requirements; the specific checking points include: first, check whether the bending moment, shear force and axial force internal force calculation is accurate, and whether the steel reinforcement cross-sectional area determined thereby is sufficient; secondly, verify whether the arrangement of steel reinforcement is reasonable, including whether the steel reinforcement diameter, quantity and spacing meet the minimum diameter and maximum spacing specification requirements; thirdly, confirm whether the anchorage length, lap length and construction details of the steel reinforcement meet the specification requirements; in addition, it is also necessary to check whether the concrete cover thickness meets the requirements of corrosion prevention and durability.

Images