Steel structure construction safety intelligent analysis method based on machine learning

Abstract

The invention discloses an intelligent steel structure safety construction analysis method based on machine learning, and the method achieves the real-time simulation and monitoring of the construction process of a steel structure through the introduction of a machine learning method and the combination of a digital twinning technology, thereby achieving the precise analysis and scientific pre-judgment of the safety performance of the structure. And the safety and the stability of steel structure construction are obviously improved. Through real-time simulation, problems in the construction process can be found and solved in time, and safety accidents are avoided. Meanwhile, on the basis of prediction and optimization suggestions of a machine learning model, constructors can be guided to carry out scientific construction, and the construction efficiency and quality are improved. In addition, the construction cost can be reduced, and the economic benefits of building enterprises can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of building construction, and in particular to an intelligent steel structure safety construction analysis method based on machine learning. Background Art

[0002] In modern construction projects, steel structures have been widely used for their advantages of light weight, high strength and fast construction speed. Steel structures face many challenges during construction, such as time-varying mechanical parameters, structural stability issues and construction quality control. There are many factors that affect the safe construction of steel structures, and there are complex nonlinear relationships between the various influencing factors. In addition, there are many difficulties in on-site monitoring of the forces at each important node. In engineering design, in order to ensure the safety of steel structures, sometimes conservativeness is used, resulting in high construction costs. In addition, when performing structural verification, a large amount of modeling calculations are required, and parameter optimization cannot be effectively guaranteed; most of them rely on a certain calculation software for structural force analysis, but for complex steel structures such as "high, large, special and specialized", only relying on design drawings for manual modeling, the operation is large and error-prone.

[0003] In order to ensure that both safety and economy can be taken into account when designing steel structures, it is a problem worth studying to establish an intelligent steel structure safety construction analysis method based on machine learning that takes into account multiple influencing factors, is efficient and accurate. Summary of the invention

[0004] The purpose of this invention is to provide an intelligent analysis method for steel structure construction safety based on machine learning. By introducing machine learning technology and combining it with digital twin technology, real-time simulation and monitoring of the steel structure construction process can be achieved, thereby achieving accurate analysis and scientific prediction of the structural safety performance.

[0005] The object of the present invention is achieved in that: An intelligent steel structure safety construction analysis method based on machine learning includes the following steps: Step 1: Use Python software to identify the original structural design drawings and layout drawings, obtain the data of each floor, and establish a model database; Step 2: Clean, integrate and standardize the data in step 1 to obtain a digital twin model, provide high-quality data support for subsequent analysis, and complete the construction of intelligent analysis algorithms; Step 3: Use machine learning methods to train the digital twin model. Through historical data and simulation data, a machine learning model that predicts the safety performance of steel structures is trained. The machine learning model analyzes various data during the construction process in real time and predicts the mechanical response and deformation of the structure, thereby achieving accurate analysis of the safety performance of the structure. Step 4: Establish a conversion interface file for finite element calculation, convert the machine learning model into a finite element analysis model for safe construction stability analysis, so as to reflect the various states of the steel structure during the construction process in real time; Step 5: Combine the workers' construction process, on-site construction equipment performance, material types, construction methods, and environmental factors to analyze and process the data, find the key factors affecting stability, and establish an optimization objective function to optimize the parameters; Step 6: After finding the optimal parameters, use Python to perform optimization analysis, form a post-processing database, obtain the optimized component dimensions, and re-import the optimized component dimensions into the model to perform the second step of stability calculation to analyze whether it meets the requirements of the specification. When the results meet the design requirements, export the calculation report.

[0006] In step 1, the data includes parts, components and bolts.

[0007] The specific operation of step 2 is as follows: step 2.1 imports data and verifies the integrity of the data; step 2.2. When the data is complete, the normalization formula is used to normalize the data, and when the data is incomplete, the missing value of the data is processed; step 2.3. After the data is normalized in step 2.2, the data set is divided.

[0008] The formula for the normalization process is: .

[0009] In step 4, various states in the construction process include mechanical response and deformation. According to different requirements, structural calculation and structural stability verification under construction conditions are performed.

[0010] In step 4, the finite element analysis model includes a pre-processing module, a calculation module and a post-processing module.

[0011] The beneficial effects of the present invention are as follows: the present invention significantly improves the safety and stability of steel structure construction. Through real-time simulation, problems in the construction process can be discovered and solved in a timely manner to avoid the occurrence of safety accidents. At the same time, the prediction and optimization suggestions based on the machine learning model can guide construction personnel to carry out scientific construction and improve construction efficiency and quality. In addition, the implementation of the present invention can also reduce construction costs and improve the economic benefits of construction companies. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 is a flow chart of the present invention, Figure 2 Flowchart of data processing. DETAILED DESCRIPTION

[0013] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0014] like Figure 1 As shown, an intelligent steel structure safety construction analysis method based on machine learning includes the following steps: Step 1: Use Python software to identify the original structural design drawings and layout drawings, obtain data for each floor, and establish a model database; the data includes parts, components, and bolts.

[0015] Step 2: Clean, integrate and standardize the data in step 1 to obtain a digital twin model, provide high-quality data support for subsequent analysis, and complete the construction of an intelligent analysis algorithm. The specific operations of step 2 are as follows: Figure 2 As shown, step 2.1 imports data and verifies the integrity of the data; step 2.2. When the data is complete, the normalization formula is used to normalize the data, and when the data is incomplete, the missing value of the data is processed; step 2.3. After the data is normalized in step 2.2, the data set is divided. The normalization of data is to normalize the input variables and output values ​​in the data set to [0, 1] before subsequent predictions.

[0016] The formula for the normalization process is: .

[0017] Method for handling missing values: Generally, the method of filling in data is adopted; 1) Manual filling: Manually fill in the missing data values ​​according to the characteristics of other data values ​​in the data sample. This method is the best one for filling in data, but it is not applicable when the number of missing values ​​is large.

[0018] 2) Mean value filling: The missing values ​​are filled with the average of all other values ​​in the feature corresponding to the missing value. In essence, it is to use the existing data to infer the missing data.

[0019] 3) Nearest filling: Find a group of numbers that are most similar to the missing values ​​in the sample with complete data, and use this group of data to fill in the missing values. The difficulty of this method lies in the difficulty in defining the similarity standard and the judgment standard is too subjective.

[0020] Step 3: Use machine learning methods to train the digital twin model. Through historical data and simulation data, a machine learning model that predicts the safety performance of steel structures is trained. The machine learning model analyzes various data during the construction process in real time and predicts the mechanical response and deformation of the structure, thereby achieving accurate analysis of the structural safety performance.

[0021] Step 4: Establish a conversion interface file for finite element calculation, and convert the machine learning model into a finite element analysis model for safe construction stability analysis, so as to reflect the various states of the steel structure during the construction process in real time; Step 5: Combine the workers' construction procedures, on-site construction machinery performance, material types, construction methods, and environmental factors to perform data analysis and processing to find the key factors affecting stability, and establish an optimization objective function for parameter optimization; the various states in the construction process include mechanical response and deformation, and according to different needs, structural calculations and structural stability verifications are performed under construction conditions.

[0022] The finite element analysis model includes a pre-processing module, a calculation module and a post-processing module.

[0023] Pre-processing module The pre-processing module is the starting point of the finite element analysis process. Its task is to establish a finite element model and provide all necessary raw data for subsequent finite element calculations. Specific functions include: Geometric Modeling: Ability to create, edit, modify, and validate geometric shapes to accurately simulate actual physical structures or systems.

[0024] Material property definition: Set the material properties of different parts or regions in the model, such as elastic modulus, Poisson's ratio, density, etc. These properties will directly affect the analysis results.

[0025] Meshing: Divide the geometric model into many small, interconnected units (such as triangles, quadrilaterals, tetrahedrons, etc.), and define nodes inside each unit for numerical calculation. The quality of meshing has an important impact on the accuracy of analysis results and computational efficiency.

[0026] Boundary conditions and load application: Set the model's boundary conditions (such as fixed constraints, sliding constraints, etc.) and external loads (such as force, pressure, temperature, etc.) to simulate the physical field under actual working conditions.

[0027] The calculation module is the core of finite element analysis. Its main function is to receive the finite element model information provided by the pre-processing module, solve the finite element equations, and calculate the response of the model under specific working conditions. Specific functions include: Algorithm selection: Select an appropriate solution algorithm, such as direct method, iterative method, etc., based on the type and complexity of the analysis problem.

[0028] Solution process: Solve the finite element equations and calculate the displacement, stress, strain and other physical quantities of each node in the model.

[0029] Result saving: Save the solution results to a database or file for subsequent analysis and processing.

[0030] The main task of the post-processing module is to read, process and display the results of the calculation module, providing analysts with an intuitive and easy-to-understand evaluation basis. Specific functions include: Result visualization: Using graphics and image technology, the calculation results are displayed in the form of contour maps, cloud maps, vector maps, animations, etc., to help analysts intuitively understand the stress distribution, deformation, etc. of the model.

[0031] Data extraction and analysis: allows users to extract physical values ​​of specific nodes or areas from the result data for further data analysis, such as comparing the results under different working conditions and evaluating the reliability of the model.

[0032] Report generation: Automatically generate reports based on analysis results, including model description, analysis process, result display, etc., for easy sharing with team members or customers.

[0033] Step 5: Combine the workers' construction process, on-site construction equipment performance, material types, construction methods, and environmental factors to conduct data analysis and processing to find the key factors affecting stability, and establish an optimization objective function to optimize parameters.

[0034] Step 6: After finding the optimal parameters, use Python to perform optimization analysis, form a post-processing database, obtain the optimized component dimensions, and re-import the optimized component dimensions into the model to perform the second step of stability calculation to analyze whether it meets the requirements of the specification. When the results meet the design requirements, export the calculation report.

Claims

1. An intelligent steel structure safety construction analysis method based on machine learning, characterized by: The following steps are involved: Step 1: Use Python software to identify the original structural design drawings and layout drawings, obtain the data of each floor, and establish a model database; Step 2: Clean, integrate and standardize the data in step 1 to obtain a digital twin model, provide high-quality data support for subsequent analysis, and complete the construction of intelligent analysis algorithms; Step 3: Use machine learning methods to train the digital twin model. Use historical data and simulation data to train a machine learning model that predicts the safety performance of steel structures. The machine learning model analyzes various data during the construction process in real time and predicts the mechanical response and deformation of the structure, thereby achieving accurate analysis of the structural safety performance; the machine learning method includes six steps: data preparation, feature engineering, model selection, model training, model evaluation and tuning, and model deployment and application. Each step has its specific content and purpose, and together constitutes the complete process of machine learning. In practical applications, it is necessary to select appropriate machine learning methods and steps based on the characteristics and needs of specific problems; Step 4: Establish a conversion interface file for finite element calculation, convert the machine learning model into a finite element analysis model for safe construction stability analysis, so as to reflect the various states of the steel structure during the construction process in real time; Step 5: Combine the workers' construction process, on-site construction equipment performance, material types, construction methods, and environmental factors to analyze and process the data, find the key factors affecting stability, and establish an optimization objective function to optimize the parameters; Step 6: After finding the optimal parameters, use Python to perform optimization analysis, form a post-processing database, obtain the optimized component dimensions, and re-import the optimized component dimensions into the model to perform the second step of stability calculation to analyze whether it meets the requirements of the specification. When the results meet the design requirements, export the calculation report.

2. The intelligent steel structure safety construction analysis method based on machine learning according to claim 1 is characterized by: In step 1, the data includes parts, components and bolts.

3. The intelligent steel structure safety construction analysis method based on machine learning according to claim 1 is characterized by: The specific operation of step 2 is as follows: step 2.1 imports data and verifies the integrity of the data; step 2.

2. when the data is complete, normalization is performed using a normalization formula, and when the data is incomplete, missing value processing is performed on the data; Step 2.

3. After normalizing the data in step 2.2, divide the data set.

4. The intelligent steel structure safety construction analysis method based on machine learning according to claim 3 is characterized by: The formula for the normalization process is: 。 5. The intelligent steel structure safety construction analysis method based on machine learning according to claim 1 is characterized by: In step 4, various states in the construction process include mechanical response and deformation. According to different requirements, structural calculation and structural stability verification under construction conditions are performed.

6. The intelligent steel structure safety construction analysis method based on machine learning according to claim 1 is characterized by: In step 4, the finite element analysis model includes a pre-processing module, a calculation module and a post-processing module.

Images