AI-based steel structure lifting point intelligent arrangement method

Abstract

The invention relates to an AI-based steel structure lifting point intelligent arrangement method, which belongs to the technical field of steel structure building construction, and comprises the following operation steps: 1, establishing a multi-data-source steel structure lifting point AI model; and 2, carrying out AI steel structure lifting point intelligent design. And 3, designing and constructing integrated intelligent arrangement by adopting steel structure lifting points of AI. Fourthly, the maximum lifting deformation and the maximum stress ratio are judged, and local precision control is conducted on the deviation between the actual deformation and stress of the butt joint position of the (i + 1) th lifted steel structure and the theoretical value of the AI model; and detecting the overall actual maximum deformation and stress of the first i + 1 lifted steel structures after hoisting, and carrying out overall precision control. The AI model, numerical simulation and an intelligent algorithm are combined, and the butt joint precision of steel structure lifting installation is improved. The problem of insufficient consideration of butt joint precision and field construction factors is solved. The safety of steel structure lifting point arrangement is improved, and meanwhile real-time feedback adjustment of steel structure lifting point arrangement is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of steel structure building construction, and in particular to an AI-based intelligent arrangement method for lifting points of steel structures. Background Art

[0002] The setting of lifting points plays a vital role in lifting operations. It refers to the reasonable determination of the connection points of the lifting ropes or slings on the lifting objects according to the characteristics of the lifting objects and the lifting requirements. The selection of these lifting points is directly related to the safety, stability and efficiency of the lifting operation. The correct setting of lifting points can ensure that the lifting objects remain balanced during the lifting process, avoid tilting or tipping, and thus protect the safety of the operators and the surrounding environment. At the same time, reasonable setting of lifting points can also reduce friction and loss during the lifting process, improve lifting efficiency and reduce operating costs.

[0003] The number and distribution of lifting points are important aspects of lifting point setting. Generally speaking, the number of lifting points should be determined according to the weight, shape and size of the lifting object. For objects with heavy weight or complex shapes, the number of lifting points should be appropriately increased to improve the stability during the lifting process. The distribution of lifting points should follow the principle of uniformity and symmetry to ensure that the lifting object is evenly stressed during the lifting process and avoid damage or tilting of the lifting object due to excessive local stress.

[0004] The existing technical solutions that are most similar to the present invention generally take the minimum bending moment or maximum natural frequency and minimum strain energy of the hoisted structure as the objective function, the allowable load of the crane as the constraint condition, and use genetic algorithms, particle swarm algorithms or enumeration methods to select the lifting points.

[0005] The existing lifting point arrangement mainly considers the overall deformation, stress and strain energy of the steel structure, ignoring the control of the docking accuracy between steel structures. At the same time, the existing lifting point arrangement is mainly based on theory, and various algorithms are applied in the design stage to obtain the optimal lifting point arrangement, ignoring the influence of the lifting site construction factors. After the existing steel structure lifting point arrangement is determined in the design stage, it will not be adjusted at the construction site. If an emergency occurs at the construction site and the actual situation does not match the design, it is difficult to adjust the lifting point arrangement in real time. Summary of the invention

[0006] The present invention mainly solves the deficiencies in the prior art and provides an AI-based intelligent arrangement method for lifting and slinging points of steel structures. It combines AI models, numerical simulations and intelligent algorithms to improve the docking accuracy of steel structure lifting and installation. It solves the problem that the existing lifting and slinging point arrangement methods do not take into account the steel structure docking accuracy and on-site construction factors. It improves the safety of the steel structure lifting and slinging point arrangement, and realizes the real-time feedback adjustment of the steel structure lifting point arrangement.

[0007] The above technical problems of the present invention are mainly solved by the following technical solutions: An AI-based intelligent arrangement method for lifting points of steel structures includes the following steps: Step 1: Establish an AI model for steel structure lifting points based on multiple data sources.

[0008] Step 2: Conduct intelligent design of steel structure lifting points using AI technology.

[0009] Determine the layout of the lifting points of the i-th lifted steel structure; collect the characteristic parameters of the i-th lifted steel structure, and use the multi-data source steel structure lifting point AI model to preliminarily plan the number of lifting points and the area Si where the lifting points are located; the AI ​​model combines finite element analysis and multi-objective genetic algorithm to determine the specific lifting nodes in area Si.

[0010] The characteristic parameters of the i+1 lifting steel structure are collected, with the goal of minimizing the overall strain energy after the i-th and i+1 lifting steel structures are connected, and minimizing the deformation of the rods at the connection between the i+1 steel structure and the i+2 steel structure. The AI ​​model combines finite element analysis and multi-objective genetic algorithm to determine the layout of the lifting points of the i+1 lifting steel structure.

[0011] The AI ​​model of steel structure lifting points from multiple data sources records the number and arrangement of lifting points of each lifting steel structure determined by the simulation of the entire lifting process, the maximum lifting deformation, and the maximum stress ratio, and determines whether the maximum lifting deformation and the maximum stress ratio exceed the specified value during the entire lifting process.

[0012] Step 3: Adopt AI-based integrated intelligent layout of steel structure lifting point design and construction. According to the determined number and layout of lifting points, lift the i-th lifted steel structure on site, and monitor the determined maximum lifting deformation, maximum stress ratio, and deformation and stress at the joint with the i+1th lifted steel structure.

[0013] Step 4: Collect monitoring data and the temperature and wind speed of the construction site on the day for the AI ​​model to determine the maximum lifting deformation, maximum stress ratio, actual deformation at the joint with the i+1 lifted steel structure, and stress deviation from the AI ​​model theoretical value for local precision control; detect the overall actual maximum deformation and stress of the first i+1 lifted steel structures after hoisting for overall precision control.

[0014] As a preferred option, a multi-data source steel structure hoisting database is established based on the existing steel structure lifting point data of the construction party, supervision party, and inspection party. The database mainly includes data on the size and weight of the lifted steel structure, the number and arrangement of lifting points, lifting deformation and stress ratio, lifting technology, and environmental parameters. The characteristic parameters of the database are taken to train the multi-data source steel structure lifting point AI model.

[0015] As a preferred method, the weight and size, structural form, node number, weather forecast temperature and wind speed of the i-th lifting steel structure rod are collected. The AI ​​model of lifting points of steel structure with multiple data sources aims to optimize the economy and minimize the lifting deformation and stress ratio. The number of lifting points and the area where the lifting points are located are preliminarily planned using database data and collected data. i , where each S i Contains ten adjacent nodes.

[0016] As a preferred method, the AI ​​model of lifting points of steel structures with multiple data sources outputs the information model of the lifted steel structure for finite element analysis. The strain energy and the minimum deformation of the rod at the joint with the i+1 steel structure are taken as the objective functions. The multi-objective genetic algorithm is used to determine the area S i The specific lifting node inside; the fitness function is as follows: Formula 1.

[0017] Where: BIAS is the deformation of the docking point of the lifted steel structure, Di,j is the deformation of the jth docking point of the i-th lifted steel structure; energy is the strain energy of the lifted structure.

[0018] As a preferred method, the number of lifting points of the i+1 lifting steel structure and the area S where the lifting points are located are preliminarily planned with the goal of optimizing economic efficiency and minimizing lifting deformation and stress ratio. i+1 ; Then, the objective function is to minimize the overall strain energy after the i-th and i+1-th lifting steel structures are connected, and the deformation of the rod at the connection between the i+1-th and i+2-th steel structures is minimized. The multi-objective genetic algorithm is used for numerical simulation to determine the arrangement of the lifting points of the i+1-th lifting steel structure until the arrangement of the lifting points of the last lifting steel structure is confirmed; the fitness function is as shown in Formula 2: Formula 2.

[0019] Where: BIAS is the deformation of the docking point of the lifted steel structure, Di,j is the deformation of the jth docking point of the i-th lifted steel structure; energy is the strain energy of the lifted structure.

[0020] Preferably, if the maximum lifting deformation and the maximum stress ratio exceed the specified values, return to reconfirm the number and arrangement of lifting points until there is no situation that exceeds the specified values; when it is confirmed that there is no situation that exceeds the specified values, the AI ​​model of lifting points of steel structures with multiple data sources is reconstructed based on the determined number and arrangement of lifting points, the maximum lifting deformation, the maximum stress ratio, and the deformation of the joints of the lifted steel structures, so that the AI ​​model reaches the most accurate state.

[0021] As a preference, when performing local precision control, when the actual value is greater than the theoretical value by more than 10%, the characteristic value is extracted based on the current actual value, and the lifting point arrangement of the i+1th lifted steel structure is adjusted; when the actual value is less than 110% of the theoretical value, the i+1th lifted steel structure is lifted according to the lifting points arranged by the AI ​​model.

[0022] As a preferred option, when performing overall precision control, if the overall actual maximum deformation and stress exceed the specification value, the AI ​​model predicts whether the overall actual maximum deformation and stress can be improved after lifting the i+2th lifted steel structure. If so, continue to arrange the lifting points for the remaining lifted steel structures; if not, stop arranging the lifting points for the i+2th lifted steel structure, and use reinforcement measures to strengthen the installed steel structure so that its deformation and stress meet the specification requirements. The AI ​​model re-collects the characteristic parameters after reinforcement, and plans the lifting point arrangement of the i+2th lifted steel structure until all steel structures are lifted into place.

[0023] The present invention can achieve the following effects: The present invention provides an AI-based intelligent arrangement method for lifting and slinging points of steel structures. Compared with the prior art, the method combines AI models, numerical simulations and intelligent algorithms to improve the docking accuracy of steel structure lifting and installation. The method solves the problem that the existing lifting and slinging point arrangement methods do not adequately consider the steel structure docking accuracy and on-site construction factors. The method improves the safety of steel structure lifting and slinging point arrangement, and realizes real-time feedback adjustment of steel structure slinging point arrangement. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0025] The technical solution of the invention is further specifically described below through embodiments and in conjunction with the accompanying drawings.

[0026] Example 1: Figure 1 As shown, an AI-based intelligent arrangement method for lifting points of steel structures includes the following steps: Step 1: Establish an AI model for lifting points of steel structures from multiple data sources. Use the existing lifting point data of steel structures from previous construction parties, supervisors, and inspectors to establish a multi-data source steel structure lifting database, which mainly includes data on the size and weight of the steel structure being lifted, the number and arrangement of lifting points, lifting deformation and stress ratio, lifting process, and environmental parameters. Extract database feature parameters and train the AI ​​model for lifting points of steel structures from multiple data sources.

[0027] Step 2: Conduct intelligent design of steel structure lifting points using AI technology.

[0028] Determine the layout of the lifting points of the i-th steel structure to be lifted; collect the characteristic parameters of the i-th lifting steel structure, and use the multi-data source steel structure lifting point AI model to preliminarily plan the number of lifting points and the area Si where the lifting points are located; the AI ​​model uses finite element analysis and multi-objective genetic algorithm to determine the specific lifting nodes in area Si. Collect the characteristic parameters such as the weight and size of the i-th lifting steel structure rod, structural form, node number, weather forecast temperature and wind speed at the construction site. The multi-data source steel structure lifting point AI model aims to optimize economic efficiency and minimize lifting deformation and stress ratio, and preliminarily plans the number of lifting points and the area S where the lifting points are located using database data and collected data i , where each S i Contains ten adjacent nodes.

[0029] The AI ​​model of lifting points of steel structures with multiple data sources outputs the information model of the lifted steel structure for finite element analysis. The strain energy and the minimum deformation of the rod at the joint with the i+1th steel structure are taken as the objective function. The multi-objective genetic algorithm is used to determine the specific lifting nodes in the area Si. The fitness function is as shown in Formula 1: Formula 1.

[0030] Where: BIAS is the deformation of the docking point of the lifted steel structure, Di,j is the deformation of the jth docking point of the i-th lifted steel structure; energy is the strain energy of the lifted structure.

[0031] The characteristic parameters of the i+1 lifting steel structure are collected. With the goal of minimizing the overall strain energy after the i+1 lifting steel structure is connected to the i+1 lifting steel structure and minimizing the deformation of the rods at the connection between the i+1 steel structure and the i+2 steel structure, the AI ​​model uses finite element analysis and multi-objective genetic algorithm to determine the layout of the lifting points of the i+1 lifting steel structure. With the goal of optimizing economic efficiency and minimizing lifting deformation and stress ratio, the number of lifting points of the i+1 lifting steel structure and the area S where the lifting points are located are preliminarily planned. i+1 ; Then, the objective function is to minimize the overall strain energy after the i-th and i+1-th lifting steel structures are connected, and the deformation of the rod at the connection between the i+1-th and i+2-th steel structures is minimized. The multi-objective genetic algorithm is used for numerical simulation to determine the arrangement of the lifting points of the i+1-th lifting steel structure until the arrangement of the lifting points of the last lifting steel structure is confirmed; the fitness function is as shown in Formula 2: Formula 2.

[0032] Where: BIAS is the deformation of the docking point of the lifted steel structure, Di,j is the deformation of the jth docking point of the i-th lifted steel structure; energy is the strain energy of the lifted structure.

[0033] The AI ​​model of lifting points for steel structures from multiple data sources records the number and arrangement of lifting points for each steel structure, the maximum lifting deformation, and the maximum stress ratio determined by the simulation of the entire lifting process, and determines whether the maximum lifting deformation and the maximum stress ratio exceed the specified value during the entire lifting process. If the maximum lifting deformation and the maximum stress ratio exceed the specified value, the AI ​​model of lifting points from multiple data sources is reconstructed based on the determined number and arrangement of lifting points, the maximum lifting deformation, the maximum stress ratio, and the deformation of the joints of the lifted steel structure, so that the AI ​​model reaches the most accurate state.

[0034] Step 3: Adopt AI-based integrated intelligent layout of steel structure lifting point design and construction. According to the determined number and layout of lifting points, lift the i-th lifted steel structure on site, and monitor the determined maximum lifting deformation, maximum stress ratio, and deformation and stress at the joint with the i+1th lifted steel structure.

[0035] Step 4: Collect monitoring data and the temperature and wind speed of the construction site on the day for the AI ​​model to determine the maximum lifting deformation, maximum stress ratio, actual deformation at the joint with the i+1th lifted steel structure, and stress deviation from the AI ​​model theoretical value for local precision control. When the actual value is greater than the theoretical value by more than 10%, extract the characteristic value based on the current actual value, and return to adjust the lifting point arrangement of the i+1th lifted steel structure; when the actual value is less than 110% of the theoretical value, lift the i+1th lifted steel structure according to the lifting point arranged by the AI ​​model.

[0036] The overall actual maximum deformation and stress of the first i+1 lifted steel structure after hoisting are detected and the overall accuracy is controlled. If the overall actual maximum deformation and stress exceed the specification value, the AI ​​model is used to predict whether the overall actual maximum deformation and stress can be improved after lifting the i+2 lifted steel structure. If so, the lifting points of the remaining lifted steel structures will continue to be arranged; if not, the lifting points of the i+2 lifted steel structure will be stopped, and reinforcement measures will be taken to strengthen the installed steel structure so that its deformation and stress meet the specification requirements. The AI ​​model will re-collect the characteristic parameters after reinforcement, and plan the lifting point arrangement of the i+2 lifted steel structure until all steel structures are lifted in place.

[0037] Embodiment 2: Example: Figure 1 As shown, an AI-based intelligent arrangement method for lifting points of steel structures includes the following steps: Step 1: Establish an AI model for lifting points of steel structures from multiple data sources. Use the existing lifting point data of steel structures from previous construction parties, supervisors, and inspectors to establish a multi-data source steel structure lifting database, which mainly includes data on the size and weight of the steel structure being lifted, the number and arrangement of lifting points, lifting deformation and stress ratio, lifting process, and environmental parameters. Extract database feature parameters and train the AI ​​model for lifting points of steel structures from multiple data sources.

[0038] The dataset of the database used for the training of the AI ​​model of lifting points of steel structures from multiple data sources contains the following: Economic efficiency refers to the cost of lifting a piece of steel structure, represented by Q. The maximum deflection during the lifting process of the steel structure, represented by L. The maximum stress ratio during the lifting process of the steel structure, represented by B. The weight of a piece of steel structure, represented by M. The temperature difference W and wind speed difference F between the construction site and the design state. The number and layout of lifting points, represented by S, N L is the number of lifting points on the left side of the symmetry axis, N R Increase the number of suspension points to the right of the axis of symmetry.

[0039]

[0040] The data set is divided into a training set and a test set in an 8:2 ratio. The machine learning algorithm is used for model training to obtain an AI model between the parameters M, W, F, S and Q, L, B.

[0041] Step 2: Intelligent design of steel structure hanging points based on AI.

[0042] (1) The characteristic parameters of the weight M of the i-th lifting steel structure member, the temperature difference W at the construction site, and the wind speed difference F are collected, and the steel structure is divided into two parts, the left and right parts. The lifting points are randomly placed in the left and right parts to obtain different indicators S. The AI ​​model for lifting points of steel structures with multiple data sources aims to optimize the economic efficiency Q and minimize the deflection L and stress ratio B, and determine the number of lifting points and their left and right distribution.

[0043] (2) The AI ​​model outputs the lifting point of steel structure from multiple data sources. The information model of the lifted steel structure is subjected to finite element analysis. The objective function is to minimize the strain energy and the deformation of the rod at the joint with the i+1th steel structure. The multi-objective genetic algorithm is used to determine the specific distribution position of the lifting point. The fitness function is as shown in Formula 1, where i=1.

[0044] Formula 1 Where: BIAS is the deformation of the lifting steel structure joint point, D i,j is the deformation size of the jth docking point of the i-th lifted steel structure; energy is the strain energy of the lifted structure, k is the component number, m ik is the total number of components of the i-th lifted steel structure.

[0045] When i>1, the specific distribution position of the hanging points is determined according to the fitness function formula 2.

[0046] Formula 2

[0047] Where: BIAS is the deformation of the docking point of the lifted steel structure, Di,j is the deformation of the jth docking point of the i-th lifted steel structure; energy is the total strain energy of the first i lifted structures, and d represents all lifted steel structures.

[0048] The AI ​​model of lifting points for steel structures from multiple data sources records the number and arrangement of lifting points for each lifting steel structure, the maximum lifting deflection L, and the maximum stress ratio B determined by the simulation of the entire lifting process, and determines whether the maximum lifting deflection L and the maximum stress ratio B exceed the specified values ​​during the entire lifting process. If the maximum lifting deflection L and the maximum stress ratio B exceed the specified values ​​in the specification, the number and arrangement of lifting points are reconfirmed until there is no situation where the specified values ​​are exceeded; when it is confirmed that there is no situation where the specified values ​​are exceeded, the lifting point arrangement model for the entire lifting process is constructed with the determined number and arrangement of lifting points.

[0049] Step 3: According to the determined number and arrangement of lifting points, lift the first lifted steel structure on site, and monitor the determined maximum lifting deflection L, maximum stress ratio B, and deformation at the joint with the second lifted steel structure.

[0050] Step 4: The AI ​​model collects the parameters of the first steel structure to be lifted on site, such as M, W, F, and S, to determine the actual maximum lifting deflection L, the maximum stress ratio B, and the actual deformation BIAS at the joint with the second steel structure to be lifted, and compares them with the theoretical values ​​of the AI ​​model. When any of the actual values ​​L, B, and BIAS is greater than the theoretical value by more than 10%, the lifting point arrangement of the second steel structure to be lifted is adjusted based on the current actual value, and the adjustment method is as shown in Formula 2; when the actual value is less than 110% of the theoretical value, the second steel structure to be lifted is lifted according to the lifting point arranged by the AI ​​model.

[0051]

[0052] After the first two lifted steel structures are hoisted, the overall actual maximum deflection L and stress ratio B are tested for overall accuracy control. If the overall actual maximum deflection L and stress ratio B exceed the specified values ​​in the specification, the AI ​​model is used to predict whether the overall actual maximum deflection L and stress ratio B can be improved after the third lifted steel structure is lifted. If so, the lifting points for the remaining lifted steel structures will continue to be arranged; if not, the lifting points for the third lifted steel structure will be stopped, and reinforcement measures will be taken to strengthen the installed steel structure so that its deflection L and stress ratio B meet the requirements of the specification. The AI ​​model will re-collect the characteristic parameters after on-site reinforcement, and re-plan the lifting point arrangement for the third lifted steel structure until all steel structures are lifted in place.

[0053] In summary, this AI-based steel structure lifting point intelligent layout method uses a combination of AI models, numerical simulation and intelligent algorithms to improve the docking accuracy of steel structure lifting installation. It solves the problem of insufficient consideration of steel structure docking accuracy and on-site construction factors in the existing lifting point layout method. It improves the safety of steel structure lifting point layout, and realizes real-time feedback adjustment of steel structure lifting point layout.

[0054] Combining multiple data sources, AI models are used to predict the number of lifting points of the steel structure being lifted and the area Si where the lifting points are located, and then intelligent algorithms are used to determine the method of arranging the lifting points. Through intelligent design of the lifting points of the steel structure being lifted, combined with on-site feedback, the AI ​​model is used to adjust the lifting point arrangement in real time to control local and overall accuracy.

[0055] The above description is only a specific embodiment of the present invention, but the structural features of the present invention are not limited thereto. Any changes or modifications made by any technician in the field of the present invention are included in the patent scope of the present invention.

Claims

1. An AI-based intelligent arrangement method for lifting points of steel structures, characterized in that The steps are as follows: Step 1: Establish an AI model for lifting points of steel structures using multiple data sources; Step 2: Conduct intelligent design of steel structure lifting points using AI technology; Determine the layout of the lifting points of the i-th steel structure to be lifted; collect the characteristic parameters of the i-th steel structure to be lifted, and use the AI ​​model of lifting points of steel structures from multiple data sources to preliminarily plan the number of lifting points and the area Si where the lifting points are located; the AI ​​model combines finite element analysis with a multi-objective genetic algorithm to determine the specific lifting nodes in area Si; The characteristic parameters of the i+1 lifting steel structure are collected, and the overall strain energy after the i+1 lifting steel structure is connected to the i+1 lifting steel structure is minimized, and the deformation of the rods at the connection between the i+1 steel structure and the i+2 steel structure is minimized. The AI ​​model uses finite element analysis and multi-objective genetic algorithm to determine the arrangement of the lifting points of the i+1 lifting steel structure; The AI ​​model of lifting points of steel structures with multiple data sources records the number and arrangement of lifting points of each steel structure, the maximum lifting deformation, and the maximum stress ratio determined by the simulation of the entire lifting process, and determines whether the maximum lifting deformation and the maximum stress ratio exceed the specified value during the entire lifting process; Step 3: Adopt AI-based steel structure lifting point design and construction integrated intelligent layout, lift the i-th lifted steel structure on site according to the determined number and layout of lifting points, and monitor the determined maximum lifting deformation, maximum stress ratio, and deformation and stress at the joint with the i+1th lifted steel structure; Step 4: Collect monitoring data and the temperature and wind speed of the construction site on the day for the AI ​​model to determine the maximum lifting deformation, maximum stress ratio, actual deformation at the joint with the i+1 lifted steel structure, and stress deviation from the AI ​​model theoretical value for local precision control; detect the overall actual maximum deformation and stress of the first i+1 lifted steel structures after hoisting for overall precision control.

2. The AI-based intelligent arrangement method for lifting points of steel structures according to claim 1 is characterized in that: A multi-data source steel structure hoisting database is established based on the existing steel structure lifting point data of the construction party, supervision party, and inspection party. The database mainly includes data on the size and weight of the lifted steel structure, the number and layout of lifting points, lifting deformation and stress ratio, lifting technology and environmental parameters. The database feature parameters are extracted to train the multi-data source steel structure lifting point AI model.

3. The AI-based intelligent arrangement method for lifting points of steel structures according to claim 1 is characterized in that: The weight and size, structural form, node number, construction site weather forecast temperature and wind speed of the i-th lifting steel structure rod are collected. The multi-data source steel structure lifting point AI model aims to optimize the economy and minimize the lifting deformation and stress ratio. It uses the database data and the collected data to preliminarily plan the number of lifting points and the area S where the lifting points are located. i , where each S i Contains ten adjacent nodes.

4. The AI-based intelligent arrangement method for lifting points of steel structures according to claim 3 is characterized in that: The AI ​​model of lifting points of steel structures with multiple data sources outputs the information model of the lifted steel structure for finite element analysis. The strain energy and the minimum deformation of the rod at the joint with the i+1 steel structure are taken as the objective function. The multi-objective genetic algorithm is used to determine the area S i The specific lifting node inside; the fitness function is as follows: Formula 1; Where: BIAS is the deformation of the docking point of the lifted steel structure, Di,j is the deformation of the jth docking point of the i-th lifted steel structure; energy is the strain energy of the lifted structure.

5. The AI-based intelligent arrangement method for lifting points of steel structures according to claim 4 is characterized in that: With the goal of optimizing economic efficiency and minimizing lifting deformation and stress ratio, the number of lifting points of the i+1 lifting steel structure and the area S where the lifting points are located are preliminarily planned. i+1 ; Then, the objective function is to minimize the overall strain energy after the i-th and i+1-th lifting steel structures are connected, and the deformation of the rod at the connection between the i+1-th and i+2-th steel structures is minimized. The multi-objective genetic algorithm is used for numerical simulation to determine the arrangement of the lifting points of the i+1-th lifting steel structure until the arrangement of the lifting points of the last lifting steel structure is confirmed; the fitness function is as shown in Formula 2: Formula 2; Where: BIAS is the deformation of the docking point of the lifted steel structure, Di,j is the deformation of the jth docking point of the i-th lifted steel structure; energy is the strain energy of the lifted structure.

6. The AI-based intelligent arrangement method for lifting points of steel structures according to claim 5 is characterized in that: If the maximum lifting deformation and the maximum stress ratio exceed the specified values, the system will return to reconfirm the number and arrangement of lifting points until they no longer exceed the specified values. After confirming that they do not exceed the specified values, the AI ​​model of lifting points of steel structures based on the determined number and arrangement of lifting points, the maximum lifting deformation, the maximum stress ratio, and the deformation at the joints of the lifted steel structures is reconstructed to make the AI ​​model as accurate as possible.

7. The AI-based intelligent arrangement method for lifting points of steel structures according to claim 1 is characterized in that: When performing local precision control, when the actual value is greater than the theoretical value by more than 10%, the characteristic value is extracted based on the current actual value, and the lifting point layout of the i+1th lifted steel structure is adjusted; when the actual value is less than 110% of the theoretical value, the i+1th lifted steel structure is lifted according to the lifting points arranged by the AI ​​model.

8. The AI-based intelligent arrangement method for lifting points of steel structures according to claim 1 is characterized in that: When performing overall precision control, if the overall actual maximum deformation and stress exceed the specification value, the AI ​​model predicts whether the overall actual maximum deformation and stress can be improved after lifting the i+2th lifted steel structure. If so, continue to arrange the lifting points for the remaining lifted steel structures; if not, stop arranging the lifting points for the i+2th lifted steel structure, and use reinforcement measures to strengthen the installed steel structure so that its deformation and stress meet the specification requirements. The AI ​​model re-collects the characteristic parameters after reinforcement, and plans the lifting point arrangement of the i+2th lifted steel structure until all steel structures are lifted into place.

Images