Large-span steel structure refined construction method based on simulation model

Abstract

According to the large-span steel structure refined construction method based on the simulation model, through whole-process refined simulation and optimization, the construction level of a large-span steel structure is remarkably improved, the construction scheme is determined based on a multi-objective optimization algorithm, accurate design of parameters such as hoisting, supporting and monitoring is achieved, a temperature field model is combined with actual measurement correction, and the construction efficiency is improved. The integration of optimization closure temperature, node semi-rigidity and initial defects enables simulation errors, support stability evaluation and optimization guarantee of safety coefficients, most construction risks can be predicted in advance on the whole, the field adjustment time is shortened, the formal construction period is shortened, the construction efficiency and safety are greatly improved, the obvious engineering value and economic benefits are achieved, and the method is suitable for popularization and application. The problem that an existing large-span steel structure construction method lacks a calculation model for predicting the risk in advance and shortening the construction period by combining multi-aspect data is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of fine construction, in particular to a large-span steel structure fine construction method based on a simulation model. BACKGROUND

[0002] Large-span steel structures are widely used in large buildings such as stadiums and exhibition centers due to their large span and complex spatial stress. However, existing construction techniques rely heavily on experience-based design, and there are significant shortcomings. The construction plan lacks systematic simulation support, and parameters such as lifting points and hoisting sequences are selected based on experience, which can easily lead to structural stress exceeding limits or excessive displacement. The time-varying influence of the solar temperature field on the structure is not fully considered, and improper closure temperature selection can easily produce residual stress. The semi-rigid properties of the node and the initial defects of the component are often simplified and ignored, and the simulation model deviates significantly from the actual working conditions. The stability evaluation of temporary supports often uses simplified formulas, making it difficult to predict instability risks in a timely manner.

[0003] These problems result in a rough construction plan, with frequent on-site rework and adjustment, which not only wastes a lot of time and cost, but also can cause safety hazards. There is an urgent need for a technical solution that can achieve fine simulation and optimization of the entire construction process to predict risks in advance, shorten the construction period, and improve construction efficiency. SUMMARY

[0004] The main purpose of the present application is to provide a large-span steel structure fine construction method based on a simulation model, which solves the problem of lack of calculation model for predicting risks in advance and shortening the construction period in combination with multiple data in the existing large-span steel structure construction method.

[0005] To solve the above technical problems, the technical solution adopted by the present application is: a large-span steel structure fine construction method based on a simulation model, the method comprising: S1, using numerical simulation analysis technology to construct a construction whole-process simulation model containing structure system conversion, node stress change and temperature influence parameters; S2, based on the construction whole-process simulation model, using a multi-objective optimization algorithm to simulate and analyze the lifting parameters, to determine the optimal scheme of lifting points, hoisting sequence, speed, temporary support arrangement and sensor distribution; S3, establishing a steel structure temperature field calculation model under the action of solar radiation by numerical simulation, simulating the temperature distribution law under solar radiation and shadow shielding, and determining the best layout position of temperature sensors; S4, arranging on-site temperature monitoring equipment to collect key part data, comparing and correcting the simulation model, obtaining the time-varying law of the temperature field during construction and optimizing the closure temperature; S5, using the calculation theory of node semi-rigid properties and component initial defects to integrate related parameters into the steel structure construction period calculation model; S6, based on the construction period calculation model, perfecting the temporary support stability limit bearing capacity calculation theory, evaluating the stability of the support system and optimizing the design; S7, integrating the optimization scheme of steps S1-S6 to form a construction implementation scheme, synchronously enabling the monitoring system and temperature monitoring equipment, real-time feedback of data and dynamic adjustment of construction parameters.

[0006] In the preferred scheme, in step S2, the steps of constructing a construction whole-process simulation model containing structure system transformation, node stress change and temperature influence parameters by using numerical simulation analysis technology are as follows: Define the optimization variable vector , including the three-dimensional coordinates of the lifting point , the component lifting sequence , the lifting speed , the temporary support parameters , and the sensor position ; Determine the objective functions of minimizing the maximum stress of the structure , minimizing the maximum displacement of the structure , minimizing the total construction period , and minimizing the construction cost , and assign the weight coefficients of each target; Establish lifting balance constraints to ensure that the resultant force and moment at the lifting point are zero, , ; Set the lifting sequence constraints based on the component topology dependency relationship to limit the post-component construction after the pre-component lifting is completed; Define the equipment and material constraints, including the lifting speed range, lifting point spacing, sensor spacing, and the limit value of the temporary support bearing capacity; Improve the NSGA-III algorithm iteration calculation, generate initial solutions according to uniform distribution, and select feasible solutions that meet the balance constraints; Call the finite element model in step S1 to calculate the objective functions of each group of solutions and evaluate them, and normalize them using the fuzzy membership function; Classify the population according to the dominance relationship, calculate the Euclidean distance between each solution and the reference point, and determine the niche; Generate offspring solutions using SBX crossover, introduce randomness using polynomial mutation, and adjust the crossover probability and the mutation probability according to the generation distance; Perform gradient descent search on non-dominated solutions to optimize key variables; Stop iteration when the hypervolume index change rate is less than 1% for 5 consecutive generations, and output the Pareto optimal solution set; TOPSIS method is used to select the optimal solution from the Pareto set, a standardized decision matrix is constructed, the distance between each solution and the positive and negative ideal solutions is calculated, the solution with the largest closeness degree is selected, and the hoisting scheme file is output.

[0007] In the preferred scheme, in step S2, the improved NSGA-Ⅲ algorithm iteration includes: Non-dominated sorting, the population is divided into different dominant levels according to the objective function value; Reference point correlation, calculate the Euclidean distance between each solution and the reference point; Cross mutation, SBX crossover is used to generate offspring solutions; Local search, gradient optimization is performed on the optimal solution: When the change rate of hypervolume of continuous 5 generations meets the following conditions, stop iteration, and output the Pareto optimal solution set: ; Wherein is the hypervolume index of the kth generation population; Optimal set screening, extract non-dominated solutions to form a candidate scheme set: ; TOPSIS decision includes: standardized decision matrix, closeness calculation, and selection The solution with the largest closeness degree is selected as the optimal solution, and the optimization result is output.

[0008] In the preferred scheme, in step S3, the temperature field calculation model of the steel structure under the action of solar radiation is established through numerical simulation, the temperature distribution law under solar radiation and shadow shielding is simulated, and the steps to determine the best layout position of the temperature sensor are as follows: Collect local solar radiation data to determine the maximum radiation intensity at noon , sunrise time and sunset time ; Calculate the solar zenith angle , calculate the solar elevation angle and azimuth angle through astronomical formula according to the project latitude, date and time, and convert it to zenith angle; Establish a shadow shielding model, based on the three-dimensional coordinates of the components, use the ray tracing method to determine the shielding relationship between components at any time, and generate a shielding state function ; Set the temperature field control equation and boundary conditions, and establish the heat conduction differential equation: based on Fourier's law, introduce the volume heat capacity , thermal conductivity , and define the heat flux source term ; Set the boundary conditions, consider solar radiation absorption and long-wave radiation heat dissipation on the sunlit surface, consider convective heat transfer on the non-sunlit surface, and determine the absorption coefficient , emissivity , heat transfer coefficient , etc. parameters; Initialize ambient temperature , set according to the daily average temperature of the construction season, reserve real-time update interface, convenient to correct in step S4 combined with measured data.

[0009] In the preferred scheme, in step S3, the temperature field is solved by using the thermal analysis unit to discretize the structure and determine the time step. Construct a thermal matrix to generate a heat capacity matrix , heat conduction matrix , heat load vector ; Transient solution, Newmark method is used to solve the heat conduction equation, and the temperature difference between adjacent time steps is less than or equal to 0.5℃, and the temperature time history of each node is output ; Extract the characteristic parameters of the temperature field: Calculate the temperature gradient , take the modulus value to get the gradient distribution; Analyze the extreme value of temperature difference, and count the maximum temperature difference in 24 hours , mark the high temperature variable area; Generate heat flux density vector , calculate the direction and size of heat flux according to Fourier's law, and provide basis for sensor layout; Optimize temperature sensor layout point: Select candidate monitoring points, select calculation nodes from the high temperature variable area as candidate points, the number is 2~3 times the planned sensor layout; Establish an information entropy optimization model to maximize the monitoring information , determine the constraint sensor spacing; Greedy algorithm optimization, sort candidate points by information entropy , select points that meet the constraints in turn until the upper limit of the number of sensors is reached.

[0010] In the preferred scheme, in step S4, arrange the field temperature monitoring equipment to collect key part data, compare and correct the simulation model, get the temperature field time-varying law during construction period and optimize the closure temperature; Install sensors according to the layout scheme output in step S3, install distributed optical fiber sensors at the corresponding coordinates , ensure that the sensors correspond one by one to the model nodes; Set data acquisition parameters, determine the sampling frequency of sunshine period, and store the data format as ; Debugging system, power on test sensor stability, collect initial ambient temperature, compare laboratory calibration value, error should be less than or equal to 0.5℃; Analyze temperature field model error, calculate single point deviation, for each monitoring point, calculate simulation temperature Absolute deviation of the measured temperature ;

[0011] Statistical average deviation, calculate global average temperature deviation When , start model correction; Weighted least squares model correction, determine correction weight, temperature gradient calculated in step S3 As weight basis, high temperature variable area monitoring point weight is higher, ; Correction calculation, for each model node, substitute into correction formula ; Verify effectiveness, repeat correction iteration until maximum single point deviation≤1℃, average deviation≤0.8℃, output corrected temperature field model; Extract the time-varying law of temperature field during construction period: Decompose time series, use wavelet transform to separate trend item of corrected temperature field And fluctuation item ; Identify fluctuation characteristics, perform Fourier transform on fluctuation item, extract amplitude , main period , phase angle ; Establish time-varying law model, integrate trend item and fluctuation item, generate global temperature time-varying function: .

[0012] In the preferred scheme, in step S4, the closure temperature optimization calculation steps are as follows: Define optimization goal, take minimum closure structure temperature difference stress as goal ), wherein ; Determine constraint interval, statistics 10% quantile And 90% quantile Of basic temperature , as closure temperature range; Use golden section method to solve, search for optimal closure temperature In Interval, ensure temperature difference stress≤design allowable residual stress ; Determine closure window, intensive monitoring 24h before construction, when the real-time temperature and deviation ≤0.5℃ and temperature change rate less than or equal to 0.1℃ / h, determine the closure window.

[0013] In the preferred embodiment, in step S5, the calculation theory of node semi-rigid characteristics and component initial defects is used to integrate relevant parameters into the steel structure construction period calculation model. Classify node types, divide nodes into bolted connections, , welded connections, , mixed connections, , and count the number and location of each type of node. Test mechanical parameters, select typical nodes for bending test, load to the limit state, record the bending moment-rotation curve, extract the initial stiffness , yield bending moment , and ultimate rotation ; Fit the constitutive model, based on the test curve, fit the exponential bending moment-rotation relationship: ; Determine the damage accumulation coefficient , shape parameter ; Parameterized modeling of component initial defects: Establish the initial bending model, for each component, take the maximum initial bending amplitude according to the specification , use a sine curve: to describe the bending distribution along the length. Establish the initial eccentricity model, take the maximum allowable eccentricity of the section , introduce the random distribution coefficient , calculate the initial eccentricity: ; Build a residual stress model, use a bilinear distribution model , take 0.3 times the yield strength as the stress peak , determine the distribution coefficient according to the section type ; Integrate the construction period calculation model: Modify the element stiffness matrix, integrate the node semi-rigid characteristics into the element stiffness, calculate the stiffness reduction matrix: ; Get the modified element stiffness: ; Build a defect equivalent load, convert initial bending, initial eccentricity, and residual stress into equivalent load vectors: ; Coupling thermal-mechanical-defect, integrating temperature field data in step S4, establishing equilibrium equation: ; Wherein is the temperature stiffness matrix, is the temperature equivalent load; Adjust the model verification and parameters: Field test comparison, select multiple typical nodes and components, paste strain gauges to monitor actual stress, calculate stress error rate: , require ; Analyze parameter sensitivity, use control variable method to adjust node parameters , defect parameters and , calculate stress change rate, keep parameters that have significant impact on the results; Confirm the final model, adjust sensitive parameters to meet the error requirements, and output the construction period calculation model.

[0014] In the preferred scheme, in step S6, based on the construction period calculation model, improve the calculation theory of the ultimate bearing capacity of temporary support stability, evaluate the stability of the support system and optimize the design steps as follows: Determine the basic parameters of the temporary support: Calculate the cross-sectional properties, calculate the cross-sectional moment of inertia , area , and turning radius ; Determine the material properties, specify the support steel grade, and obtain the yield strength from the specification , use the elastic modulus in step S1 ; Statistical geometric parameters, record the support length , height , plane coordinates , determine the length coefficient according to the constraint conditions , calculate the slenderness ratio ; Calculate the stability ultimate bearing capacity: Axial compression stability coefficient solution, according to the slenderness ratio , use the modified formula to calculate the stability coefficient , Use the polynomial formula when , use the Euler formula when ; Calculate the ultimate bearing capacity of a single support, substitute the formula , and correct it to: ; Wherein ; Check local stability: for steel pipe support, check the ratio of wall thickness to outer diameter.

[0015] In the preferred embodiment, in step S6, the stability of the support system is evaluated: Calculate the overall stability coefficient, and calculate the system stability coefficient using the rigid frame stability theory: ; wherein is the number of support groups, is the actual axial force borne by the kth support; Nonlinear stability analysis, respectively, calculate the geometric nonlinear and material nonlinear, using the arc length method to solve: wherein is the tangent stiffness matrix of the support system, is the geometric stiffness matrix, is the load coefficient, when instability , the stability safety margin is ; Step S6.3.3, local stability check, for steel pipe section, check the ratio of wall thickness to outer diameter: ; wherein is the wall thickness of the steel pipe; Optimization design of support system: Optimization variable definition: wherein is the number of supports, is the support section size, is the support height, is the support plane coordinate; Objective function: wherein is the total cost of the support system, is the material / installation rate; Constraint conditions: ; Optimization algorithm, using particle swarm optimization (PSO) algorithm to solve, and output the final scheme.

[0016] The present application provides a large-span steel structure fine construction method based on simulation model, which has the following beneficial effects: 1. The technical scheme improves the construction level of large-span steel structure through full-process fine simulation and optimization, and the construction scheme is determined based on a multi-objective optimization algorithm, realizing accurate design of hoisting, support, monitoring and other parameters, the temperature field model is combined with the measured correction, and the closure temperature optimization accuracy reaches ±0.5℃. 2. The integration of semi-rigid nodes and initial defects reduces the simulation error to less than or equal to 10%, and the support stability assessment and optimization ensure a safety factor of less than or equal to 1.5. Overall, most construction risks can be predicted in advance, reducing on-site adjustment time by more than 30%, saving 20%-25% of the formal construction cycle, greatly improving construction efficiency and safety, and possessing significant engineering value and economic benefits. Attached Figure Description

[0017] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the refined construction method for large-span steel structures according to the present invention; Figure 2 This is a simulation diagram of the stress on the support structure of the present invention; Figure 3 This is a simulation diagram of the stress on the top truss of the present invention; Figure 4 This is a schematic diagram of the sensor distribution and key deformation monitoring areas of this invention; Figure 5 This is an axonometric schematic diagram of the sensor distribution and semi-rigid connection points of the present invention. Detailed Implementation

[0018] Example 1 like Figures 1-5 As shown, a refined construction method for large-span steel structures based on simulation models is proposed. This method includes: S1. A simulation model of the entire construction process, including parameters such as structural system transformation, nodal stress changes, and temperature influence, is constructed using numerical simulation analysis technology. S2. Based on the simulation model of the entire construction process, a multi-objective optimization algorithm is used to simulate and analyze the hoisting parameters to determine the optimal scheme for the hoisting point, hoisting sequence, speed, temporary support layout and sensor distribution. S3. Establish a calculation model of the temperature field of steel structure under solar radiation through numerical simulation, simulate the temperature distribution law under solar radiation and shadow blocking, and determine the optimal placement of temperature sensors. S4. Deploy on-site temperature monitoring equipment to collect data from key locations, compare and correct the simulation model, obtain the time-varying law of the temperature field during the construction period, and optimize the closure temperature. S5. The calculation theory of semi-rigid characteristics of nodes and initial defects of components is adopted, and relevant parameters are integrated into the calculation model of steel structure construction period. S6. Based on the construction period calculation model, improve the calculation theory of the ultimate bearing capacity of temporary support stability, evaluate the stability of the support system and optimize the design. S7. Integrate the optimization schemes of steps S1-S6 to form a construction implementation plan, and simultaneously activate the monitoring system and temperature monitoring equipment to provide real-time data feedback and dynamically adjust construction parameters.

[0019] In the preferred embodiment, in step S2, the hoisting parameter multi-objective optimization construction step is as follows: Step S2.1, optimization parameter initialization, define variable dimension: Step S2.1.1, determine the optimization variable vector and boundary: (1); Wherein is the lifting point coordinate, hoisting sequence, is the hoisting speed, is the temporary support parameter, is the sensor position coordinate.

[0020] Step S2.1.2, set the target function weight, and allocate the weight coefficient according to the engineering priority: (2); Step S2.2, input finite element model parameters: Step S2.2.1, structure parameter import: Component section attribute: moment of inertia , cross-sectional area , elastic modulus ; Material parameters: yield strength , density ; Step S2.2.2, set the working condition parameters: Lifting force calculation model: (3); is the distance from the lifting point to the center of gravity, safety factor, is the total weight of the component; Temporary support stiffness: (4), wherein is the support height; Step S2.3, multi-objective optimization iteration calculation: Step S2.3.1, generate initial population: Generate 200 initial solutions according to uniform distribution, meet : (5); Wherein is the Z-axis component force of the i-th lifting point, is the moment generated by the i-th lifting point; Step S2.3.2, calculate the objective function: For each group of solutions, finite element simulation is called, and the output is: Maximum stress: where is the Mises stress of the jth point of the ith member, which needs to satisfy ; Maximum displacement: where is the deflection of the jth point of the ith member, which needs to satisfy ; Hoisting construction period: where is the hoisting stroke of the ith member, is the hoisting speed of the ith member; Step S2.3.3, check the constraint conditions: Verify the constraint equations one by one and eliminate the items that do not satisfy: (6); where is the allowable stress of the material, is the stress when hoisting statically, is the axial force borne by the ith temporary support, is the length coefficient, is the information entropy of sensor distribution, is the maximum allowable information entropy.

[0021] In the preferred scheme, step S2.3.4, improve the NSGA-III algorithm iteration: ① Non-dominated sorting: divide the population into different dominance levels according to the objective function value ② Reference point association: calculate the Euclidean distance between each solution and the reference point: , is the tth reference point; ③ Crossover and mutation: generate offspring solutions by SBX crossover; ④ Local search: implement gradient optimization on the optimal solution: ; Step S2.3.5, convergence judgment: Stop iteration when the following conditions are met: ; where is the hypervolume index of the kth generation population; Step S2.4, optimal scheme decision: Step S2.4.1, optimal set screening: Extract non-dominated solutions to form a candidate scheme set: ; Step S2.4.2, TOPSIS decision: ①Standardized decision matrix: (7); ② Calculate the closeness: (8); ③Choose The largest possible solution is considered the optimal solution. Step S2.5, output the optimization results: Construction plan including the following parameters: lifting point coordinates, as shown below. Figure 4 The yellow markings in the middle and Figure 5 The location indicated by the black square in the middle: and lifting force distribution ; Speed ​​parameters: Lifting speed at each stage and the corresponding dynamic load factor ; Support Layout Diagram: Temporary Support Coordinates ,high and bearing capacity verification value, Figure 5 The yellow five-pointed star in the center indicates the location of the temporary support point below. Figure 5 The two purple pentagrams on the left are vertical fixing points, and the purple pentagram in the middle of the truss is the diagonal brace fixing point. Sensor distribution map: Coordinates of monitoring points and sensitivity coefficient ,like Figure 4 and Figure 5 The red dot in the image indicates this. The abstract optimization objective is transformed into executable construction parameters through algorithmic formulas. Each calculation step corresponds to clear engineering constraints, ensuring the feasibility and optimality of the output solution.

[0022] In the preferred scheme, step S3 involves the calculation of the solar radiation temperature field and the optimization of sensor deployment: Step S3.1, Solar radiation parameter modeling: Step S3.1.1, Calculation of solar radiation intensity: Determine the solar radiation flux density: (9); in for Solar radiation intensity at any given time This represents the local maximum radiation intensity at noon. For sunrise and sunset times, The solar zenith angle; Step S3.1.1, Shadow Occlusion Model: Establish the occlusion judgment function between components: ; Step S3.2, establish temperature field control equation: Step S3.2.1, heat conduction differential equation: (10); Wherein is the volume heat capacity of steel, is the thermal conductivity of steel, is the three-dimensional transient temperature field, is the heat flow density source term; Step S3.2.2, boundary conditions: Sunlight surface: ; Non-sunlight surface: ; Wherein is the solar radiation absorption coefficient, is the long-wave radiation emissivity, is the Stefan-Boltzmann constant, is the convective heat transfer coefficient, is the ambient temperature.

[0023] In the preferred scheme, step S3.2, solve the temperature field value: Step S3.2.1, finite element discrete, eight-node solid element discrete structure, time step is 10 min: ; Wherein is the heat capacity matrix, is the heat conduction matrix, is the heat load vector; Step S3.2.2, solve the iteration, use Newmark method to solve the transient temperature field, convergence criterion: ; Step S3.3, extract temperature field characteristic parameters: Step S3.3.1, calculate the temperature gradient modulus: (11); Step S3.3.2, analyze the temperature difference extreme value, calculate the maximum temperature difference in 24 hours: ; Step S3.3.3, heat flow density vector: ; Step S3.4, optimize sensor layout point: Step S3.4.1, screen candidate points, select the monitoring candidate points that meet the following conditions: ; Step S3.4.2, optimize the objective function: Maximize the amount of temperature monitoring information: ; in Let the gradient weights be the gradient weights of the p-th candidate point. The number of candidate points; Step S3.4.3, the constraints are as follows: Spatial distribution: The distance between any two points; Quantity limit: Determined based on the budget; Step S3.4.4: Optimize the algorithm by using a greedy algorithm to select the optimal combination of monitoring points, and output the results: (1) Temperature field calculation report: including temperature cloud map and gradient distribution curve at typical times. (2) Sensor deployment scheme: Coordinate list: (p=1,2,...,k); Priority sorting: based on Value and The overall score.

[0024] In the preferred scheme, step S4 involves the temperature field model correction and closure temperature optimization construction steps: Step S4.1, Deployment of the on-site temperature monitoring system and matching of monitoring points: Step S4.1.1: Based on the sensor deployment plan output in step S3, at the corresponding coordinates... Install distributed fiber optic sensors to ensure that the monitoring points correspond one-to-one with the simulation model calculation nodes in step S3; Step S4.1.2, set the data acquisition parameters: Set the sampling frequency, and the data storage format is: ,in for The measured temperature at the p-th monitoring point at time p; Step S4.2, the steps of the simulation model correction algorithm are as follows: Step S4.2.1, calculate the error analysis index: Calculate the deviation between the simulated temperature and the measured temperature in step S3: (12); in For the p-th node output by the S3 simulation model, Temperature at any given time This represents the average temperature deviation. The model is corrected when starting; Step S4.2.2, the model is corrected by weighted least squares: (13); wherein is the corrected temperature of the pth node, and the weight coefficient is: , is the temperature gradient of the qth point calculated in step S3.3.1, and the high-temperature variable region has a higher weight; Step S4.2.3, verify the correction effectiveness, and repeat the iterative correction until the following condition is met:

[0025] Step S4.2, extract the time-varying law of the temperature field during construction, and decompose the corrected temperature field by wavelet transform to separate the trend item and the fluctuation item:

[0026] wherein is the temperature trend item, i.e., the basic temperature field, is the periodic fluctuation item; Step S4.2.3, identify the fluctuation characteristic parameters, and extract the characteristics of the fluctuation item by Fourier transform: (14); wherein is the fluctuation amplitude, is the main period, is the phase angle; Step S4.2.4, output the time-varying law model, and generate the global temperature time-varying function: (15).

[0027] In the preferred scheme, step S4.2, the optimization calculation of the closure temperature is as follows: Step S4.2.1, optimize the objective function, with the minimum thermal stress of the structure after closure as the target: ; wherein is the closure temperature, is the thermal stress of the structure under the closure temperature, which can be calculated by thermal-mechanical coupling: ; is the design allowable residual stress, and the temperature difference distribution during closure is: ; Step S4.2.2, constraint condition, determine the closure temperature interval based on the time-varying law: ; wherein is the 10% quantile of the last 30 days, is the 90% quantile of the last 30 days; is the 10% quantile of the last 30 days, is the 90% quantile of the last 30 days; Step S4.2.3, optimal closure temperature solving, golden section method is used to search the optimal solution in the constraint interval.

[0028] In the preferred scheme, in step S5, the node semi-rigid and initial defect calculation model construction steps are as follows: Step S5.1, node semi-rigid characteristic parameter identification: Step S5.1.1, node type classification, according to the node construction form, such as bolt connection or welding, a classification matrix is established: ; Step S5.1.2, semi-rigid mechanical parameter test, the characteristic parameters can be obtained through the node bending resistance test: Initial stiffness: , yield bending moment: , ultimate rotation angle: , subscript t is the node type corresponding to ; In step S5.1.3, the bending moment-rotation angle relationship model, the exponential constitutive equation is used to describe the nonlinear behavior of the node: (16); Wherein is the node bending moment, is the node rotation angle, is the ultimate bending moment, is the characteristic rotation angle, is the damage accumulation coefficient, is the maximum historical rotation angle, is the shape parameter; Step S5.2, component initial defect parameterized modeling: Step S5.2.1, initial bending model, a sinusoidal curve is used to describe the initial bending of the component along the length direction: ; Wherein is the initial bending value at x from the left end of the component, is the maximum initial bending amplitude, is the component length; Step S5.2.2, initial eccentricity model, considering the deviation of the section centroid and the force center: , Wherein is the initial eccentricity, is the maximum allowable eccentricity, is the random distribution coefficient; Step S5.2.3, residual stress distribution, adopts a bilinear distribution model: (17); wherein is the residual stress at the section neutral axis y, is the residual stress peak value, is the section height, is the stress distribution coefficient; Step S5.3, construction period calculation model integration: Step S5.3.1, unit stiffness matrix correction, introduces the node semi-rigid characteristics into the unit stiffness matrix: ; ; wherein is the corrected unit stiffness matrix, is the ideal rigid node unit stiffness matrix, is the stiffness reduction matrix, is the node secant stiffness under the current rotation , is the node stiffness contribution matrix; Step S5.3.2, defect load vector construction, converts the initial defect into an equivalent load: (18); wherein is the defect equivalent load vector, is the shape function matrix, is the member section flexural rigidity, is the section area; Step S5.3.3, temperature-defect coupling equation, establishes a balance equation considering temperature effect: ; wherein is the temperature stiffness matrix, which can be calculated in step S4 temperature field, is the displacement vector, is the external load vector, is the temperature equivalent load vector; Step S5.4, model verification and parameter adjustment: Step S5.4.1, field test comparison, selects 3-5 typical nodes and members, monitors the actual stress through strain gauges, and calculates the error index: (19); wherein Calculate stress for model, Calculate stress for field test, require , otherwise go back to adjust in step S5.1.3 , ; Step S5.4.2, parameter sensitivity analysis, use control variable method to identify key parameters: For node parameters : Calculate , keep sensitive parameters with S>5MPa; For defect parameters : When , improve modeling accuracy; Step S5.4.3, final model output, generate construction period calculation model containing: Component semi-rigid connection point diagram, as shown in Figure 4 and Figure 5 green dot, Figure 4 two green dots in the upper part of the figure represent tensioning supports, which are used to tension the truss upwards by setting cables at the top, and the two green dots at the bottom represent fixed supports; Modified overall stiffness matrix and load vector.

[0029] In the preferred scheme, in step S6, temporary support stability ultimate bearing capacity calculation and system optimization construction steps: Step S6.1, determine the mechanical parameters of temporary support components: Step S6.1.1, section property calculation, for commonly used steel pipe sections of temporary supports, circular or square, calculate key parameters: ; ; Where is the support section moment of inertia, is the outer diameter / inner diameter of the circular section, the outer side length of the square section, is the inner side length of the square section; Step S6.1.2, material performance parameters: The yield strength is , the elastic modulus is in step S1, The slenderness ratio is , where is the section rotation radius, is the length coefficient Step S6.2, stability ultimate bearing capacity calculation theory: Step S6.2.1, axial compression stability coefficient, using the correction coefficient considering residual stress and initial bending: (20); (21); Step S6.2.2, basic formula of ultimate bearing capacity: wherein is the ultimate bearing capacity of single support under axial compression, is the support section area; Step S6.2.2, two-way load correction, considering the composite stress when the horizontal force acts: (22); wherein is the horizontal shear force borne by the support, is the shear ultimate bearing capacity.

[0030] In the preferred solution, step S6.3, evaluating the stability of the support system: Step S6.3.1, overall stability coefficient calculation, using the rigid frame stability theory to calculate the system stability coefficient: (23); wherein is the number of support groups, is the actual axial force borne by the kth support group; Step S6.3.2, nonlinear stability analysis, respectively calculating geometric nonlinear and material nonlinear, using the arc length method to solve: wherein is the tangent stiffness matrix of the support system, is the geometric stiffness matrix, is the load coefficient, when instability , the stability safety margin is

[0031] Step S6.3.3, local stability checking, for steel pipe section, checking the ratio of wall thickness to outer diameter: ; wherein is the steel pipe wall thickness; Step S6.4, optimization design of support system: Step S6.4.1, optimization variable definition: wherein is the number of supports, is the support section size, is the support height, is the support plane coordinate; Step S6.4.2, Objective function: ,in To support the total system cost, For material / installation rates; Step S6.4.3, Constraints: (twenty four); Step S6.4.3, optimize the algorithm, using the Particle Swarm Optimization (PSO) algorithm to solve: 1. Initialize the particle swarm; the range of variable values ​​should be set based on engineering experience. 2. Calculate fitness, including cost and constraint penalties; 3. Update particle positions: ; 4. The cost change rate is less than 3% for five consecutive generations. Step S6.4.4, final solution output: (1) Supporting design parameters, including cross-sectional dimensions, quantity, and layout coordinates; (2) Stability verification report, including single-strand bearing capacity, overall stability coefficient, and safety margin; (3) Construction details, including the connection nodes between the support and the main structure, and the preloading scheme. The preloading value is taken as 1.1 times the design load. The actual implementation scheme is shown in Table 1. Table 1. Detailed Implementation of Steps S1-S6

[0032] Example 2 Further explanation in conjunction with Example 1, such as Figures 1-5 The structure shown illustrates a refined construction method for large-span steel structures based on a simulation model. This method includes: S1. A simulation model of the entire construction process, including parameters such as structural system transformation, nodal stress changes, and temperature influence, is constructed using numerical simulation analysis technology. Step S1.1, Structural parameter collection and organization: Step S1.1.1: Collect the geometric parameters of the steel structure components, such as cross-sectional dimensions, length, and quantity, and material property parameters, such as elastic modulus. Yield strength ,density thermal conductivity and the construction methods of nodes, such as bolts, welding, and mixed connections; Step S1.1.2: Compile construction load parameters, including component self-weight, hoisting equipment load, temporary support reaction force, wind load, and the basic wind pressure and construction live load that can be taken once every 50 years according to local meteorological data. Step S1.2, construct the finite element model: Step S1.2.1, adopt solid element discrete members, shell element simulate thin plate members, beam element simplify secondary support members; Step S1.2.2, define node connection properties, initially model as rigid nodes, later supplement semi-rigid properties in step S5, set member-to-member contact relations, such as weld surface binding, bolt connection pre-tightening force; Step S1.3, set construction conditions and boundary conditions: Step S1.3.1, divide construction stages, such as member hoisting, temporary support installation, system transformation, closure, etc., and clearly define the load application sequence and unloading path for each stage; Step S1.3.2, set boundary conditions: fixed support constraints, determine the constraint direction according to the design drawings, temporary support constraints, initially model as elastic support, and the stiffness is determined according to the support material.

[0033] Step S1.4, incorporate temperature influence parameters: Step S1.4.1, import local weather data, annual average temperature, daily temperature fluctuation range, sunshine duration, and reserve temperature field calculation interfaces in the model for convenience in supplementing temperature field data in subsequent steps S3-S4; Step S1.4.2, define material thermal expansion coefficient , set up thermal-mechanical coupling calculation module to ensure that temperature changes can be converted into structural internal forces and deformations; Step S1.5, initial model verification Step S1.5.1, select simple members for static calculation, compare theoretical solutions with model calculation results, and the error should be ≤5%; Step S1.5.2, adjust the unit grid density, the grid size can be taken as 1 / 5~1 / 3 of the smallest cross-sectional dimension of the member, to ensure grid independence verification.

[0034] S2, based on the construction whole-process simulation model, use multi-objective optimization algorithm to simulate and analyze the hoisting parameters, determine the optimal scheme of lifting points, hoisting sequence, speed, temporary support arrangement and sensor distribution; In step S2, the steps for constructing a construction whole-process simulation model including structural system transformation, node stress change and temperature influence parameters are as follows: Define optimization variable vector , including three-dimensional coordinates of lifting points , member hoisting sequence , hoisting speed , temporary support parameters , sensor position ; Determine the minimum structural maximum stress , minimize the maximum displacement of the structure minimize the total construction period minimize the construction cost objective functions, and assign each objective weight coefficient; establish hoisting balance constraints to ensure that the resultant force and moment at the hoisting point are zero, , ; set hoisting sequence constraints based on the topological dependence relationship of components to limit the construction of post-component after the completion of pre-component hoisting; clearly define equipment and material constraints, including hoisting speed range, hoisting point spacing, sensor spacing, and temporary support bearing capacity limit; improve the NSGA-III algorithm iteration calculation, generate initial solutions according to uniform distribution, and screen feasible solutions that meet the balance constraints; call the finite element model in step S1 to calculate the objective function of each solution and evaluate it, and normalize it using the fuzzy membership function; classify the population according to the dominance relationship, calculate the Euclidean distance between each solution and the reference point, and determine the niche; generate offspring solutions using SBX crossover, introduce randomness using polynomial mutation, and adjust the crossover probability and mutation probability according to the generation distance; perform gradient descent search on non-dominated solutions to optimize key variables; stop iteration when the change rate of hypervolume index is less than 1% for 5 consecutive generations, and output the Pareto optimal solution set; select the optimal solution from the Pareto set using the TOPSIS method, construct a standardized decision matrix, calculate the distance between each solution and the positive and negative ideal solutions, select the solution with the closest degree, and output the hoisting scheme file.

[0035] In the preferred scheme, the improved NSGA-III algorithm iteration in step S2 includes: non-dominated sorting, dividing the population into different dominance levels according to the objective function value; reference point association, calculating the Euclidean distance between each solution and the reference point; crossover and mutation, generating offspring solutions using SBX crossover; local search, performing gradient optimization on the optimal solution: stop iteration when the change rate of hypervolume satisfies the following conditions for 5 consecutive generations, and output the Pareto optimal solution set: ; where is the hypervolume index of the kth generation population; optimal set screening, extracting non-dominated solutions to form a candidate scheme set: ; TOPSIS decision-making includes: standardization of decision matrix, calculation of closeness degree, selection The largest scheme as the optimal solution, output optimization results.

[0036] In the preferred scheme, in step S3, the temperature field calculation model of steel structure under the action of solar radiation is established by numerical simulation, the temperature distribution law under solar radiation and shadow shielding is simulated, and the steps of determining the best layout position of temperature sensor are as follows: Collect local solar radiation data to determine the maximum radiation intensity at noon , sunrise time , and sunset time ; Calculate the solar zenith angle , calculate the solar elevation angle and azimuth angle through astronomical formula according to the latitude, date and time of the project, and convert it into zenith angle; Establish a shadow shielding model, judge the shielding relationship between components at any time based on the three-dimensional coordinates of the components, and generate a shielding state function ; Set the temperature field control equation and boundary conditions, and establish the heat conduction differential equation: based on Fourier's law, introduce the volume heat capacity , thermal conductivity , and define the heat flux source term ; Set the boundary conditions, consider solar radiation absorption and long-wave radiation heat dissipation on the sunlit surface, consider convective heat transfer on the non-sunlit surface, and determine the absorption coefficient , emissivity , convective heat transfer coefficient , etc. Initialize the ambient temperature , set it according to the daily temperature in the construction season, and reserve a real-time update interface for easy correction in step S4 combined with measured data.

[0037] In the preferred scheme, in step S3, the temperature field is solved by numerical method, and the thermal analysis unit is used to discretize the structure to determine the time step Build a thermal matrix to generate a heat capacity matrix , heat conduction matrix , and heat load vector ; Transient solution, solve the heat conduction equation by Newmark method, iterate until the temperature difference between adjacent time steps is less than or equal to 0.5℃, and output the temperature time history of each node ; Extract the characteristic parameters of the temperature field: Calculate the temperature gradient , take the modulus of the spatial derivative of each node temperature to get the gradient distribution; Analysis of temperature difference extreme value, statistics of 24h global maximum temperature difference , mark high temperature variable area; Generate heat flux density vector , according to Fourier law to calculate the heat flow direction and size, provide the basis for sensor layout; Optimization of temperature sensor layout point: Screening candidate monitoring points, selecting calculation nodes from high temperature variable area as candidate points, the number is 2~3 times the planned sensor layout; Establish information entropy optimization model, with maximum monitoring information as the goal , determine the constraint sensor spacing; Greedy algorithm optimization, according to information entropy Sort candidate points, select points that meet the constraints in turn, until the upper limit of the number of sensors is reached.

[0038] In the preferred scheme, in step S4, arrange on-site temperature monitoring equipment to collect key part data, compare and correct the simulation model, get the construction period temperature field time-varying law and optimize the joint temperature; Install sensors according to the layout scheme output in step S3, install distributed optical fiber sensors at the corresponding coordinates , ensure that the sensors correspond to the model nodes one by one; Set data acquisition parameters, determine the sampling frequency of sunshine period, and store the data format as ; Debug the system, test the stability of the sensor, collect the initial environmental temperature, and compare the laboratory calibration value, the error should be less than or equal to 0.5℃; Analyze the temperature field model error, calculate the single point deviation, for each monitoring point, calculate the absolute deviation between the simulation temperature and the measured temperature ;

[0039] Calculate the average deviation, calculate the global average temperature deviation , when , start model correction; Weighted least squares model correction, determine the correction weight, take the temperature gradient calculated in step S3 as the weight basis, the weight of high temperature variable area monitoring point is higher, ; Correction calculation, for each model node, substitute into the correction formula ; Verify the effectiveness, repeat the correction iteration until the maximum single point deviation is less than or equal to 1℃, and the average deviation is less than or equal to 0.8℃, output the corrected temperature field model; Extract the construction period temperature field time-varying law: Decompose the time series and use wavelet transform to separate the trend term of the corrected temperature field. With fluctuation term ; Identify wave characteristics, perform Fourier transform on the wave term, and extract the amplitude. Main cycle Phase angle ; Establish a time-varying pattern model, integrate the trend term and fluctuation term, and generate a global temperature time-varying function: .

[0040] In the preferred scheme, the calculation steps for the closure temperature optimization in step S4 are as follows: Define the optimization objective as minimizing the thermal stress of the structure after closure. ),in ; Determine the constraint interval and statistically analyze the baseline temperature. 10th percentile ( ) and 90th percentile ( ), as the closure temperature range; The solution is obtained using the golden section method. Search for the optimal closure temperature within the interval Ensure that the thermal stress is less than or equal to the design allowable residual stress. ; The closure window was determined, and intensive monitoring was conducted for 24 hours prior to construction. When the real-time temperature was... When the deviation is ≤0.5℃ and the temperature change rate is ≤0.1℃ / h, it is determined as the closure window.

[0041] In the preferred scheme, in step S5, the calculation theory of semi-rigid characteristics of nodes and initial defects of components is adopted to integrate relevant parameters into the calculation model of steel structure construction period. Classify node types, and classify nodes according to their construction form into bolted connections (… ), welding connection ( ), hybrid connection ( ), to count the number and location of various types of nodes; Test mechanical parameters, select typical nodes for bending tests, load to the ultimate state, record the moment-rotation curve, and extract the initial stiffness. Yield bending moment Extreme turning angle ; By fitting a constitutive model and based on experimental curves, an exponential moment-rotation relationship is obtained. ; Determine the damage accumulation coefficient Shape parameters ; Component initial defect parameterized modeling: Establish initial curvature model, take the maximum initial curvature amplitude according to the specification for each component , adopt sinusoidal curve: , describe the curvature distribution along the length; Establish initial eccentricity model, take the maximum allowable eccentricity , introduce random distribution coefficient , calculate the initial eccentricity: ; Establish residual stress model, adopt bilinear distribution model , take 0.3 times the yield strength as the stress peak value , determine the distribution coefficient according to the section type ; Integrate the construction period calculation model: Modify the element stiffness matrix, integrate the node semi-rigid characteristics into the element stiffness, and calculate the stiffness reduction matrix: ; Get the modified element stiffness: ; Build defect equivalent load, convert initial curvature, initial eccentricity, and residual stress into equivalent load vector: ; Couple thermal-force-defects, integrate the temperature field data in step S4, and establish the equilibrium equation: ; Wherein is the temperature stiffness matrix, is the temperature equivalent load; Adjust model verification and parameters: Compare with field test, select multiple typical nodes and components, paste strain gauges to monitor actual stress, and calculate stress error rate: , require ; Analyze parameter sensitivity, adjust node parameters , defect parameters and respectively by using the control variable method, calculate the stress change rate, and retain the parameters that have significant impact on the results; Confirm the final model, adjust the sensitive parameters to meet the error requirements, and output the construction period calculation model.

[0042] In the preferred scheme, based on the construction period calculation model in step S6, the temporary support stability limit bearing capacity calculation theory is improved, the support system stability is evaluated, and the design is optimized as follows: Determine temporary support base parameters: Calculate cross-section properties, calculate cross-section moment of inertia according to support type , area , radius of gyration ; Determine material properties, specify support steel grade, obtain yield strength from specification , use the elastic modulus in step S1 ; Statistical geometric parameters, record support length , height , plane coordinates , determine the length coefficient according to the constraint condition , calculate the slenderness ratio ; Calculate the ultimate bearing capacity of stability: Axial compression stability coefficient solution, according to the slenderness ratio , use the modified formula to calculate the stability coefficient , , use the polynomial formula when , use the Euler formula; Calculate the ultimate bearing capacity of a single support, substitute the formula , and correct it to: ; Where ; - Local stability check: for steel pipe support, check the ratio of wall thickness to outer diameter.

[0043] In the preferred scheme, in step S6, the stability of the support system is evaluated: Calculate the overall stability coefficient, calculate the system stability coefficient using the rigid frame stability theory: ; Where is the number of support groups, is the actual axial force borne by the kth support group; Nonlinear stability analysis, respectively calculate geometric nonlinear and material nonlinear, solve using the arc length method: , where is the tangent stiffness matrix of the support system, is the geometric stiffness matrix, is the load coefficient, when , the stability safety margin is ; Step S6.3.3, local stability check, for steel pipe section, check the ratio of wall thickness to outer diameter: ; wherein is the steel pipe wall thickness; Optimization design support system: Optimization variable definition: wherein is the support number, is the support cross-sectional size, is the support height, is the support plane coordinates; Objective function: wherein is the total cost of the support system, is the material / installation rate; Constraint condition: ; Optimization algorithm, using particle swarm optimization (PSO) algorithm to solve, and output the final scheme.

[0044] The above embodiments are only preferred technical solutions of the present application, and should not be regarded as a limitation of the present application, the protection scope of the present application should be the technical solutions recited in the claims, including the equivalent replacement schemes of the technical features recited in the claims as the protection scope. That is, within this range, equivalent replacement improvements are also within the protection scope of the present application.

Claims

1. A large-span steel structure fine construction method based on a simulation model, characterized by: The method comprises: S1, constructing a construction whole-process simulation model containing structure system conversion, node stress change and temperature influence parameters by using numerical simulation analysis technology; S2, based on the construction whole-process simulation model, simulating and analyzing hoisting parameters by using a multi-objective optimization algorithm to determine an optimal scheme of hoisting points, hoisting sequence, speed, temporary support arrangement and sensor distribution; S3, establishing a steel structure temperature field calculation model under the action of solar radiation by numerical simulation to simulate temperature distribution rules under solar radiation and shadow shielding and determine the best arrangement position of temperature sensors; S4, arranging on-site temperature monitoring equipment to collect key position data, comparing and correcting the simulation model to obtain temperature field time-varying rules during construction and optimize the joining temperature; S5, using the calculation theory of node semi-rigid characteristics and component initial defects to integrate relevant parameters into the steel structure construction period calculation model; S6, based on the construction period calculation model, improving the calculation theory of the stability limit bearing capacity of temporary supports to evaluate the stability of the support system and optimize the design; S7, integrating the optimization scheme of steps S1-S6 to form a construction implementation scheme, simultaneously starting the monitoring system and temperature monitoring equipment, feeding back data in real time and dynamically adjusting construction parameters.

2. The fine construction method of a long-span steel structure based on a simulation model according to claim 1, characterized in that: In step S2, the steps of constructing a construction whole-process simulation model containing structure system conversion, node stress change and temperature influence parameters by using numerical simulation analysis technology are as follows: Definition of optimization variable vector , including the three-dimensional coordinates of the lifting point , the component lifting sequence , the lifting speed , the temporary support parameters , the sensor position ; minimizing the maximum stress of the structure minimizing the maximum displacement of the structure minimizing the total construction period minimizing the construction cost the objective functions, and assigning each objective weight coefficients; Establish the lifting balance constraint to ensure that the resultant force and moment of the lifting point are zero, , ; Setting hoisting sequence constraints, limiting the construction of post-positioned components after the completion of pre-positioned components based on the topological dependence relationship of components; Defining equipment and material constraints, including the speed range of hoisting, the distance between hoisting points, the distance between sensors, and the limit value of the bearing capacity of temporary supports; Improving the iterative calculation of the NSGA-III algorithm, generating initial solutions according to uniform distribution, and selecting feasible solutions that satisfy the balance constraints; Calling the finite element model in step S1, calculating the objective function of each solution and evaluating it, and normalizing it using the fuzzy membership function; Classifying the population according to the dominance relationship, calculating the Euclidean distance between each solution and the reference point, and determining the niche; The offspring solution is generated by SBX crossover, and the randomness is introduced by polynomial mutation, and the crossover probability and mutation probability is dynamically adjusted according to the generation distance; Implementing gradient descent search on non-dominated solutions to optimize key variables; Stopping iteration when the change rate of hypervolume index is less than 1% for 5 consecutive generations, and outputting the Pareto optimal solution set; Selecting the optimal solution from the Pareto set using the TOPSIS method, constructing a standardized decision matrix, calculating the distance between each scheme and the positive and negative ideal solutions, selecting the scheme with the largest closeness degree, and outputting the hoisting scheme file.

3. The fine construction method of a long-span steel structure based on a simulation model according to claim 2, characterized in that: In step S2, the improvement of the NSGA-III algorithm iteration includes: Non-dominated sorting, dividing the population into different dominance levels according to the objective function value; Reference point association, calculating the Euclidean distance between each solution and the reference point; Crossing and mutation, generating offspring solutions by SBX crossover; Local search, gradient optimization of the optimal solution: Stopping iteration when the change rate of hypervolume satisfies the following conditions for 5 consecutive generations, and outputting the Pareto optimal solution set: ； wherein is the hypersurface index of the kth generation population; Optimal set screening, extracting non-dominated solutions to form candidate set: ; TOPSIS decision-making includes: standardization decision matrix, calculate closeness, select The largest scheme as the optimal solution, output optimization results.

4. The fine construction method of a long-span steel structure based on a simulation model according to claim 1, characterized in that: In step S3, the steps of establishing a steel structure temperature field calculation model under the action of solar radiation by numerical simulation to simulate temperature distribution rules under solar radiation and shadow shielding and determine the best arrangement position of temperature sensors are as follows: Collecting local solar radiation data to determine the maximum radiation intensity at noon , sunrise time and sunset time ; Calculate the solar zenith angle According to the project latitude, date, time, calculate the solar altitude angle and azimuth angle through astronomical formula, and convert into the zenith angle; A shadow blocking model is established, and a ray tracing method is used to determine the blocking relationship between components at any time based on the three-dimensional coordinates of the components to generate a blocking state function ; Setting temperature field control equation and boundary conditions, establishing heat conduction differential equation: based on Fourier's law, introducing volume heat capacity , thermal conductivity , define heat flux density source term ; The boundary conditions are set, the solar radiation absorption and long-wave radiation heat dissipation are considered for the sunlit surface, the convection heat transfer is considered for the non-sunlit surface, and parameters such as the absorption coefficient, the emissivity, and the convection heat transfer coefficient are determined .​​ Initialization ambient temperature , according to the average daily temperature of construction season, reserve real-time update interface, convenient steps S4 combined with measured data correction.

5. The fine construction method of a long-span steel structure based on a simulation model according to claim 4, characterized in that: In step S3, the temperature field is solved by using a thermal analysis unit to discretize the structure and determine the time step; constructing a thermal matrix, generating a heat capacity matrix , a heat conduction matrix , a heat load vector ; Transient solution, Newmark method is used to solve heat conduction equation, iteration to adjacent time step temperature difference ≤0.5℃, output each node temperature time history ; Extract the characteristic parameters of the temperature field: Computing temperature gradient The spatial derivative of the temperature at each node is taken to obtain the gradient distribution. Analyze the temperature difference extreme value, count the global maximum temperature difference within 24 hours , mark the high temperature variable area; Generating heat flow density vectors According to Fourier law, the heat flow direction and size are calculated to provide a basis for sensor layout. Optimize the temperature sensor layout points: Select candidate monitoring points from the high-temperature variation area and select calculation nodes as candidate points, with a quantity of 2-3 times the planned sensor layout; An information entropy optimization model is established to monitor information maximization as the target , determine the constraint sensor spacing; Greedy algorithm selects the optimal one according to information entropy Sort the candidate points, and select the points that meet the constraints in turn until the upper limit of the number of sensors is reached.

6. The fine construction method of a long-span steel structure based on a simulation model according to claim 1, characterized in that: In step S4, arrange the on-site temperature monitoring equipment to collect key part data, compare and correct the simulation model, obtain the temperature field time-varying law during construction, and optimize the closure temperature; Install the sensor according to the layout scheme output in step S3, and ensure that the sensor corresponds to the model node one by one Install the distributed optical fiber sensor, and ensure that the sensor corresponds to the model node one by one; Set data acquisition parameters, determine the frequency of the sunshine period, data format storage as ; Debug the system, test the sensor stability, collect the initial environmental temperature, compare with the laboratory calibration value, and the error should be less than or equal to 0.5℃; Analyzing the temperature field model error, calculating single point deviation, for each monitoring point, calculating the simulation temperature The absolute deviation of the measured temperature ;​ Statistical average bias, calculate global average temperature bias Model correction is initiated when Model correction is initiated when Weighted least squares model revision, determine the correction weight, to the temperature gradient calculated in step S3 For weight basis, high temperature variable area monitoring point weight is higher, ; Amend the calculation, for each model node, substitute the amendment formula ; Verify the effectiveness, repeat the correction iteration until the maximum single-point deviation is less than or equal to 1℃, and the average deviation is less than or equal to 0.8℃, and output the corrected temperature field model; Extract the temperature field time-varying law during construction: Decompose the time series, using wavelet transform to separate the trend of the corrected temperature field and the wave term ; Identify the fluctuation characteristics, Fourier transform the fluctuation term, and extract the amplitude , main period , phase angle ; Establish the time-varying law model, integrate the trend term and the fluctuation term, and generate the global temperature time-varying function: 。 7. The method according to claim 6, wherein the method is characterized by: In step S4, the closure temperature optimization calculation steps are as follows: The optimization objective is defined to minimize the temperature stress of the closed structure , wherein ; Determine the constraint interval, the statistical basis temperature The 10% quantile ( ) and 90% quantile ( ) of the statistical basis temperature are determined as the closing temperature range. Solve by golden section method, in Search for the optimal temperature in the interval , to ensure that the temperature difference stress is less than or equal to the design allowable residual stress ; Determine closure window, intensive monitoring 24h before construction, when real-time temperature and deviation ≤0.5℃ and temperature change rate less than or equal to 0.1℃ / h, determine closure window.

8. The fine construction method of a long-span steel structure based on a simulation model according to claim 1, characterized in that: In step S5, the calculation theory of node semi-rigid characteristics and component initial defects is used to integrate relevant parameters into the steel structure construction period calculation model; Classify node types, divide nodes into bolted connections, , welded connections, , hybrid connections, , and count and locate each type of node. Test the mechanical parameters, select the typical node to carry out the bending test, load to the limit state, record the bending moment-rotation angle curve, extract the initial stiffness , yield bending moment , limit rotation angle ; Fit the constitutive model based on the test curve to fit the exponential moment-rotation relationship: ； Determining an injury accumulation coefficient , shape parameters ; Parameterized modeling of component initial defects: Establish the initial bending model, according to the maximum initial bending amplitude of each component, take the standard , using sinusoidal curve: , describing a distribution of bending along the length; Build the eccentricity model, take the maximum allowable eccentricity , introduce a random distribution coefficient , calculate the initial eccentricity: ; Build the residual stress model, adopt the bilinear distribution model , take 0.3 times yield strength as stress peak value , determine distribution coefficient according to section type ; Integrate the construction period calculation model: Modify the element stiffness matrix to integrate the node semi-rigid characteristics into the element stiffness, and calculate the stiffness reduction matrix: ； Get the corrected element stiffness: ； Construct the defect equivalent load to convert the initial bending, initial eccentricity, and residual stress into equivalent load vectors: ； Couple heat-force-defects, integrate the temperature field data in step S4, and establish the balance equation: ； wherein is the temperature stiffness matrix, is the temperature equivalent load; Adjust the model verification and parameters: Field test comparison, select multiple typical nodes and components, paste strain gauge to monitor actual stress, calculate stress error rate: , request ; The parameter sensitivity is analyzed, the control variable method is adopted, and the node parameters are adjusted respectively , defect parameters and , the stress change rate is calculated, and the parameters which have significant influence on the results are reserved; Confirm the final model, adjust the sensitive parameters to meet the error requirements, and output the construction period calculation model.

9. The simulation model-based refined construction method for long-span steel structures according to claim 1, characterized in that: In step S6, based on the construction period calculation model, improve the temporary support stability limit bearing capacity calculation theory, evaluate the support system stability and optimize the design steps as follows: Determine the temporary support foundation parameters: Calculate section properties, depending on support type, calculate section moment of inertia , area , radius of gyration ; Determine material properties, identify supporting steel grade, find yield strength in specification , use the elastic modulus from step S1 ; statistical geometry parameters, record support length , height , planar coordinates , determine length coefficient according to constraint condition , calculate slenderness ratio ; Calculate the stability limit bearing capacity: The solution of the stability coefficient of the shaft center under pressure is based on the slenderness ratio , the stability coefficient is calculated by a modified formula , , a polynomial formula is used when , an Euler formula is used when The single support ultimate bearing capacity is calculated, and the formula is substituted , and the horizontal shear is corrected as ； wherein ; Check the local stability, and check the wall thickness to diameter ratio of the steel pipe support structure.

10. The method according to claim 9, wherein the method is characterized by: In step S6, evaluate the stability of the support system: Calculate the overall stability coefficient, and calculate the system stability coefficient by using the rigid frame stability theory: ； wherein is the number of support groups, is the actual axial force supported by the kth support group; Nonlinear stability analysis, respectively, calculates geometric nonlinear and material nonlinear, using the arc length method to solve: where is the tangent stiffness matrix of the support system, is the geometric stiffness matrix, is the load coefficient, when instability , the stability safety margin is ; Check the local stability, and check the wall thickness to diameter ratio of the steel pipe section: ； wherein t is the steel pipe wall thickness; Optimize the design of the support system: Optimization variable definitions: where is the number of supports, is the support cross-sectional dimension, is the support height, is the support planar coordinate; Objective function: wherein Ctotai is the total cost of the support system, Cmateri is the material / installation rate; Constraints: ; Optimize the algorithm by using the particle swarm optimization (PSO) algorithm to solve and output the final scheme.

Images