Rapid prediction and low-carbon optimization method for earthquake-resistant demand of liquefied stratum
By combining finite element analysis and machine learning, and utilizing ANN models and multi-objective optimization methods, we have achieved rapid and accurate prediction of the seismic resistance requirements of liquefied strata and low-carbon optimization, solving the problems of long calculation time and insufficient environmental impact in traditional methods, and improving prediction accuracy and design efficiency.
Patent Information
- Application Number
- CN202510759187.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-09
- Publication Date
- 2025-09-12
AI Technical Summary
Traditional methods for predicting seismic demand for liquefied strata have long calculation times, large data requirements, difficulty in providing real-time feedback, and lack consideration of environmental impact and cost optimization.
Combining finite element analysis and machine learning techniques, using ANN model training data sets, and combining multi-objective optimization methods, we can achieve rapid prediction of seismic resistance requirements and low-carbon optimization of liquefied strata. We simulate earthquake responses through a finite element platform, construct a sample data set, use artificial neural networks for training, and select the optimal solution through grid search and multi-objective optimization algorithms.
It achieves high-precision and rapid prediction of the seismic resistance requirements of liquefied strata, reduces carbon emissions and construction costs, improves the efficiency and sustainability of seismic resistance design for liquefied strata, and solves the problem of finding a balance between efficiency and sustainability in traditional methods.
Smart Images
Figure CN120633320A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of earthquake resistance demand prediction, and in particular to a method for rapid prediction of earthquake resistance demand in liquefied strata and a low-carbon optimization method. Background Art
[0002] Liquefied strata can cause significant ground deformation under strong earthquakes, causing serious damage to infrastructure, especially critical facilities such as bridges and roads. Traditional seismic demand prediction methods typically rely on finite element analysis and physical testing, but in practical applications they suffer from long calculation times, large data requirements, and difficulty in providing real-time feedback. Therefore, a rapid prediction method that integrates seismic demand prediction with low-carbon optimization is needed. This method can not only efficiently and accurately predict seismic demand, but also minimize the cost and environmental impact of gravel pile foundation treatment through low-carbon optimization design. Summary of the Invention
[0003] In response to the above problems, the present invention proposes a rapid prediction and low-carbon optimization method for the seismic demand of liquefied strata. The ANN model is trained using a data set constructed on a finite element platform. By combining finite element analysis and machine learning techniques, it can achieve higher accuracy and efficiency in the prediction of the seismic demand of liquefied strata. Combined with a low-carbon optimization method based on multi-objective optimization, it can not only accurately predict the seismic demand of liquefied strata, but also minimize the cost and environmental impact of the gravel pile treatment process, solving the problems of low prediction accuracy, long calculation time, and lack of consideration of environmental impact and cost optimization in traditional methods.
[0004] The specific steps of the rapid prediction and low-carbon optimization method for liquefied strata seismic demand are as follows:
[0005] S1, construct liquefied strata and gravel pile parameters;
[0006] S2: Based on the parameters of the liquefied stratum and gravel piles, the seismic response of the liquefied stratum is simulated in a finite element analysis platform to obtain the lateral displacement of the site caused by the liquefaction of saturated sand. The seismic demand index is obtained by data fitting of the lateral displacement of the site; a sample data set is constructed based on the parameters of the liquefied stratum, gravel piles, and the seismic intensity in the seismic response;
[0007] S3, using the seismic demand index as the true label, the sample data set as input, and the predicted seismic demand index as output, trains the artificial neural network (ANN) model to obtain a trained ANN model;
[0008] S4, through the grid search method to exhaustively enumerate all possible groups of gravel pile parameters, input the data of each group of gravel pile parameters into the trained ANN model to obtain the corresponding predicted seismic demand index, and calculate the corresponding carbon emissions, construction costs and site lateral displacement of each group of gravel pile parameters based on the gravel pile parameters and the predicted seismic demand index; use the multi-objective optimization algorithm to select the comprehensive optimal solution from the carbon emissions, construction costs and site lateral deformation of all groups of gravel pile parameters to complete low-carbon optimization.
[0009] Preferably, in S2, when simulating the seismic response of liquefied strata, the triggering conditions for sand liquefaction are: when the moment magnitude M is 7.5, the single amplitude shear strain reaches 3%, and the vertical effective stress σ′ v The load is 1 atm, and liquefaction occurs after 15 uniform loading cycles. The cycle resistance ratio at this time is expressed as:
[0010]
[0011] in, represents the effective stress σ′ v The circulation resistance ratio when the moment magnitude M is 7.5 and the moment magnitude M is 1 atm; N 1,60 It represents the number of blows in the standard penetration test, corrected by the overburden pressure and hammer energy.
[0012] Preferably, the liquefied stratum parameters include the site inclination, the thickness of the overlying non-liquefied soil layer, the hardness of the upper non-liquefied soil layer, the thickness of the liquefiable soil layer, N 1,60 and groundwater level depth; the gravel pile parameters include gravel pile diameter and gravel pile replacement rate.
[0013] Preferably, the gravel pile replacement rate is expressed as:
[0014]
[0015] Where D is the diameter of the gravel pile, and S is the spacing between the gravel piles.
[0016] Preferably, the seismic demand index is the relationship between the median lateral displacement of the site and the earthquake intensity, and its expression is:
[0017] S D =aIM b
[0018] Among them, S D represents the median lateral displacement of the site; IM represents the seismic intensity; a and b represent the seismic demand indicators.
[0019] Preferably, the carbon emissions include carbon emissions calculated by a process-based life cycle assessment method, carbon emissions calculated by an economic input-output life cycle assessment method, and carbon emissions calculated by a hybrid life cycle assessment method;
[0020] The life cycle assessment method is expressed as:
[0021]
[0022] Where E[Carbon(P-LCA)] represents the carbon emissions calculated based on the process-based life cycle assessment method; N represents the total number of projects; EF i represents the P-LCA carbon emission factor of the i-th project; Q i represents the construction work volume of the i-th project;
[0023] The economic input-output life cycle assessment method is expressed as:
[0024]
[0025] Among them, E[Carbon] represents the carbon emissions calculated by the economic input-output life cycle assessment method; EIO i represents the EIO-LCA carbon emission factor of the i-th project; EC i represents the project cost of the i-th project;
[0026] The hybrid life cycle assessment method is specifically as follows: if the carbon emissions of a subproject belong to construction carbon emissions or transportation carbon emissions, the carbon emissions of the subproject are calculated using the life cycle assessment method; if the carbon emissions of a subproject belong to raw material carbon emissions, the carbon emissions of the subproject are calculated using the economic input-output life cycle method. The carbon emissions of all subprojects are summed up to the carbon emissions calculated using the hybrid life cycle assessment method.
[0027] Preferably, the construction cost includes the total cost of gravel piles, which is expressed as:
[0028]
[0029] Where EC represents the total cost of gravel piles; N represents the total number of projects; UC i represents the unit cost of the engineering project of the i-th project; Q i represents the construction workload of the i-th project.
[0030] Preferably, the finite element analysis platform is the Opensees platform.
[0031] Preferably, the method further comprises: after training the ANN model, performing hyperparameter optimization using a cross-validation method.
[0032] Compared with the prior art, the present invention has the following beneficial effects:
[0033] (1) The present invention combines the seismic demand prediction of liquefied strata with low-carbon optimization design to provide a fast and accurate prediction method, which can effectively reduce carbon emissions and construction costs, and improve the efficiency and sustainability of seismic design for liquefied strata. It solves the problem in traditional design methods that it is difficult to ensure high efficiency while taking into account sustainability, significantly improves the accuracy and efficiency of seismic demand prediction and low-carbon optimization design, and ensures the high efficiency and sustainability of disaster prevention design for liquefied strata.
[0034] (2) The present invention combines the sustainability indicators of cost and carbon emissions and uses a multi-objective optimization method to perform low-carbon optimization design, thereby minimizing carbon emissions and construction costs and improving the seismic toughness of liquefied strata. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] The present invention will be described in further detail below with reference to the accompanying drawings;
[0036] Figure 1 This is a flow chart of a method for rapid prediction of seismic demand for liquefied formations and low-carbon optimization according to an embodiment of the present invention;
[0037] Figure 2 A schematic diagram of a method for estimating total construction costs and carbon dioxide emissions of a rapid prediction method for seismic demand of liquefied strata and a low-carbon optimization method according to an embodiment of the present invention;
[0038] Figure 3 A map of representative locations of stone pillars for improving gentle slope ground with different properties in a method for rapid prediction of seismic demand for liquefied strata and low-carbon optimization according to an embodiment of the present invention, including points S1-S9;
[0039] Figure 4 Probabilistic earthquake prediction model diagrams and damage exceedance probability diagrams of the maximum allowable displacement (0.3m) for 9 locations taken for the rapid prediction of liquefied formation seismic demand and low-carbon optimization method according to an embodiment of the present invention; wherein, the upper left and lower left are respectively the probabilistic earthquake prediction model diagrams and damage exceedance probability diagrams of the maximum allowable displacement (0.3m) for points S1-S3; the upper middle and lower middle are respectively the probabilistic earthquake prediction model diagrams and damage exceedance probability diagrams of the maximum allowable displacement (0.3m) for points S7-S9; the upper right and lower right are respectively the probabilistic earthquake prediction model diagrams and damage exceedance probability diagrams of the maximum allowable displacement (0.3m) for points S4-S6;
[0040] Figure 5 Schematic diagram of multi-objective optimization results of the rapid prediction of liquefied strata seismic demand and low-carbon optimization method according to an embodiment of the present invention. DETAILED DESCRIPTION
[0041] The present invention is further described below through specific embodiments.
[0042] See also Figure 1 As shown in the figure, the rapid prediction and low-carbon optimization method of liquefied strata seismic demand are as follows:
[0043] S1, construct liquefied strata and gravel pile parameters.
[0044] The key parameters of liquefiable soil layer and gravel pile include site slope, thickness of overlying non-liquefiable soil layer, soil properties (softness and hardness of overlying non-liquefiable soil layer), thickness of liquefiable soil layer, N 1,60 , groundwater level depth, gravel pile diameter, and replacement rate. A parametric database was constructed. The softness and hardness of the upper non-liquefiable soil layer were categorized into three levels: soft, medium, and hard, based on the soil's compaction modulus (CM).
[0045] Gravel column replacement rate A rr The calculation expression is:
[0046]
[0047] Where D is the diameter of the gravel pile, and S is the spacing between the gravel piles.
[0048] S2: Based on the parameters of the liquefied stratum and gravel piles, the seismic response of the liquefied stratum was simulated in a finite element analysis platform to obtain the lateral displacement of the site caused by the liquefaction of saturated sand. The seismic demand index was obtained by data fitting of the lateral displacement of the site; a sample data set was constructed based on the liquefied stratum parameters, gravel pile parameters, and the seismic intensity in the seismic response.
[0049] The finite element analysis platform of this embodiment is OpenSees. OpenSees is used to establish a nonlinear finite element model, simulate the seismic response of the liquefied stratum, and calculate the site lateral deformation caused by saturated sand liquefaction and the seismic demand indicators a and b.
[0050] Specifically, the parameters of the parameterized database are input into OpenSees as parameters for the PDMY and PIMY material models in OpenSees. The PDMY material model simulates the basic response characteristics of pressure-sensitive soil materials under general loading conditions, while the PIMY material model simulates non-liquefiable overburden.
[0051] In the earthquake response simulation of this embodiment, the intensity of the ground motion is measured by moment magnitude, with the moment magnitude ranging from 5.8 to 7.2 and the epicentral distance ranging from 0 km to 60 km. Based on the magnitude of the moment magnitude and epicentral distance, the earthquake is divided into five earthquake combinations: small magnitude large distance (SMLR), small magnitude small distance (SMSR), large magnitude large distance (LMLR), large magnitude small distance (LMSR), and near-fault earthquakes (fault distance R < 15 km). These effectively represent the impact of different ground motion characteristics on the seismic resistance of liquefied formations.
[0052] The parameters of the finite element analysis are calibrated based on the sand liquefaction triggering standard, which is defined as the vertical effective stress σ′ when the single shear strain of the sand reaches 3% under an earthquake with a magnitude of M7.5. v When the load is 1 atm, liquefaction is triggered after 15 uniform loading cycles. This standard is converted into an equivalent replacement of the cyclic resistance ratio CRR, and its expression is:
[0053]
[0054] in, represents the vertical effective stress σ′ of sand v Cyclic Resistance Ratio when the earthquake moment magnitude M is 7.5 and the earthquake moment magnitude M is 1 atm. 1,60 It is the number of hammer blows in the Standard Penetration Test (SPT), corrected by the overburden pressure and hammer energy.
[0055] The sample data set of this embodiment includes liquefied stratum parameters, gravel pile parameters and earthquake intensity; liquefied stratum parameters include site inclination, thickness of overlying non-liquefied soil layer, hardness of overlying non-liquefied soil layer, thickness of liquefiable soil layer, N 1,60 , groundwater level depth; gravel pile parameters include gravel pile diameter and gravel pile replacement rate.
[0056] The seismic demand of liquefied sites is expressed by the power law model as the median lateral deformation S of the site. D The relationship between IM and seismic intensity is expressed as follows:
[0057] S D =aIM b
[0058] IM uses Cumulative Absolute Velocity (CAV), and parameters a and b represent seismic demand indicators, which are obtained by fitting a power-law model to the finite element analysis results. To conveniently and accurately derive a and b, the above equation is converted into a linear equation in logarithmic space, which is expressed as follows:
[0059] ln(S D )=ln(a)+b·ln(IM)
[0060] S3, using the seismic demand index as the true label, the sample data set as the input, and the predicted seismic demand index as the output, trains the artificial neural network ANN model to obtain a trained ANN model.
[0061] The artificial neural network (ANN) model used in this embodiment is a general model; after the artificial neural network model is trained, the fast prediction model constructed based on the artificial neural network uses a cross-validation method to perform hyperparameter optimization, which significantly improves the prediction accuracy.
[0062] S4, through the grid search method to exhaustively enumerate all possible groups of gravel pile parameters, input the data of each group of gravel pile parameters into the trained ANN model to obtain the corresponding predicted seismic demand index, and calculate the corresponding carbon emissions, construction costs and site lateral displacement of each group of gravel pile parameters based on the gravel pile parameters and the predicted seismic demand index; use the multi-objective optimization algorithm to select the comprehensive optimal solution from the carbon emissions, construction costs and site lateral deformation of all groups of gravel pile parameters to complete low-carbon optimization.
[0063] The multi-objective optimization method aims to minimize carbon emissions and construction costs while simultaneously improving the seismic resilience of liquefied strata. The detailed process is as follows: First, a grid search method is used to exhaustively enumerate all possible parameter combinations (including gravel pile replacement ratio and gravel diameter) within the hyperparameter space. Each possible parameter combination is then used to predict seismic demand indicators using an artificial neural network (ANN) model. Finally, the carbon emissions and cost impacts of each parameter combination that meets the seismic requirements are evaluated to identify a solution that maximizes seismic performance (or seismic resistance) while minimizing carbon emissions and construction costs.
[0064] The seismic performance assessment specifically involves setting a permissible displacement threshold of 0.3m, calculating the site's lateral deformation (or displacement), and comparing it with the threshold. Based on the predicted seismic demand, if the calculated lateral deformation is less than the threshold, the site's seismic performance meets the requirements with the current configuration of stone pile diameter and replacement ratio. If it is greater than the threshold, the corresponding stone pile configuration is not adopted.
[0065] See also Figure 2 As shown in Figure 2, carbon emission methods include process-based life cycle assessment (P-LCA), economic input-output life cycle assessment (EIO-LCA), and hybrid life cycle assessment methods (hybrid LCA, or Hybrid LCD), as follows:
[0066] P-LCA is a traditional life cycle assessment method that calculates environmental impacts (such as carbon emissions) based on the production and transportation processes of materials. The P-LCA method usually requires detailed process data, such as energy consumption and emissions data for each stage of raw material production, transportation, manufacturing, use and disposal. Its calculation expression is:
[0067]
[0068] Where N represents the total number of projects, EF i represents the P-LCA carbon emission factor, Q i Indicates the construction work volume.
[0069] EIO-LCA is a life cycle assessment method based on an economic input-output model. It estimates the environmental impact of a product or service using data from the entire economic system, rather than relying on specific production process data. EIO-LCA uses the economic relationships between different industries to estimate indirect emissions and other environmental impacts, such as emissions from raw material extraction, manufacturing, and transportation involved in the supply chain. Its calculation expression is:
[0070]
[0071] Where N represents the total number of projects, EIO i represents the EIO-LCA carbon emission factor, EC i Represents the project cost.
[0072] The hybrid life cycle assessment approach specifically includes: if a subproject's carbon emissions are construction carbon emissions (such as building activities) or transportation carbon emissions (such as transportation), the carbon emissions of that subproject are calculated using the life cycle assessment approach; if a subproject's carbon emissions are raw material carbon emissions (such as embodied automobile carbon emissions), the carbon emissions of that subproject are calculated using the economic input-output life cycle approach; the carbon emissions of all subprojects are combined with the carbon emissions calculated using the hybrid life cycle assessment approach. This comprehensive approach provides greater flexibility and higher accuracy in carbon emissions across all project stages.
[0073] The construction cost includes the total cost of gravel piles, which is calculated as:
[0074]
[0075] Among them, UC i Indicates the unit cost of the project.
[0076] In a specific embodiment, the representative locations of the stone pillars for improving the gentle slope of the area with different attributes are shown in FIG. Figure 3 As shown, including points S1-S9, PGA represents the earthquake intensity index (i.e. IM value). The probability earthquake prediction model diagram of the 9 locations and the damage exceedance probability diagram of the maximum allowable displacement (0.3m) are shown in Figure 4 As shown; the upper left and lower left are respectively the probabilistic earthquake prediction model diagrams of points S1-S3 and the damage exceedance probability diagrams of the maximum allowable displacement (0.3m); the upper middle and lower middle are respectively the probabilistic earthquake prediction model diagrams of points S7-S9 and the damage exceedance probability diagrams of the maximum allowable displacement (0.3m); the upper right and lower right are respectively the probabilistic earthquake prediction model diagrams of points S4-S6 and the damage exceedance probability diagrams of the maximum allowable displacement (0.3m). Figure 5 Schematic diagram of the multi-objective optimization results of this embodiment.
[0077] In this embodiment, the slope of the site, the thickness of the overlying non-liquefied soil layer, the soil properties (the hardness of the upper non-liquefied soil layer), the thickness of the liquefiable soil layer, N 1,60 , groundwater level depth, gravel pile diameter, and replacement rate as input variables, while the seismic demand indicator of the liquefied stratum is used as the output variable. By training and optimizing this data, a rapid prediction and low-carbon optimization method for the seismic demand of liquefied strata was established, enabling high-precision and rapid prediction of the seismic demand of liquefied strata. Furthermore, a multi-objective optimization method was used to implement a low-carbon optimization design, combining cost and sustainability indicators, to minimize carbon emissions and construction costs while simultaneously improving the seismic resilience of the liquefied stratum.
[0078] This embodiment constructs a parameterized model library for liquefied strata and gravel pile foundation treatment, covering the key design properties of liquefiable soil strata and gravel piles; based on the OpenSees finite element analysis platform, the seismic response of the liquefied stratum is simulated, and the lateral deformation of the site and the seismic demand index caused by the liquefaction of saturated sand are calculated; a sample data set based on an artificial neural network prediction agent model is generated, with the liquefied stratum parameters, gravel pile diameter, replacement rate and seismic intensity as input, and the seismic demand index of the liquefied stratum as output; through training on the data set, a rapid prediction and low-carbon optimization method for the seismic demand of the liquefied stratum is established, which is used for high-precision and rapid prediction of the seismic demand of the liquefied stratum; combined with cost and sustainability indicators, a multi-objective optimization method is used for low-carbon optimization design to minimize carbon emissions and construction costs, while simultaneously improving the seismic toughness of the liquefied stratum.
[0079] This invention combines seismic demand forecasting for liquefied strata with low-carbon optimization design, providing a rapid and accurate prediction method that effectively reduces carbon emissions and construction costs, while improving the efficiency and sustainability of seismic design for liquefied strata. This approach addresses the difficulty of balancing high efficiency with sustainability in traditional design methods, significantly improving the accuracy and efficiency of seismic demand forecasting and low-carbon optimization design, ensuring the high efficiency and sustainability of disaster prevention design for liquefied strata.
[0080] The above is only a specific implementation of the present invention, but the design concept of the present invention is not limited to this. Any non-substantial changes to the present invention using this concept shall be deemed as an infringement of the protection scope of the present invention.
Claims
1. A method for rapid prediction of seismic demand for liquefied strata and low-carbon optimization, characterized in that: The steps include: S1, construct liquefied strata and gravel pile parameters; S2: Based on the parameters of the liquefied stratum and gravel piles, the seismic response of the liquefied stratum is simulated in a finite element analysis platform to obtain the lateral displacement of the site caused by the liquefaction of saturated sand. The seismic demand index is obtained by data fitting of the lateral displacement of the site; a sample data set is constructed based on the parameters of the liquefied stratum, gravel piles, and the seismic intensity in the seismic response; S3, using the seismic demand index as the true label, the sample data set as input, and the predicted seismic demand index as output, trains the artificial neural network (ANN) model to obtain a trained ANN model; S4, through the grid search method to exhaustively enumerate all possible groups of gravel pile parameters, input the data of each group of gravel pile parameters into the trained ANN model to obtain the corresponding predicted seismic demand index, and calculate the corresponding carbon emissions, construction costs and site lateral displacement of each group of gravel pile parameters based on the gravel pile parameters and the predicted seismic demand index; use the multi-objective optimization algorithm to select the comprehensive optimal solution from the carbon emissions, construction costs and site lateral deformation of all groups of gravel pile parameters to complete low-carbon optimization.
2. The method for rapid prediction of seismic demand for liquefied strata and low-carbon optimization according to claim 1 is characterized in that: In S2, when simulating the seismic response of liquefied strata, the triggering conditions for sand liquefaction are: when the moment magnitude M is 7.5, the single amplitude shear strain reaches 3%, and the vertical effective stress σ′ v The load is 1 atm, and liquefaction occurs after 15 uniform loading cycles. The cycle resistance ratio at this time is expressed as: in, represents the effective stress σ′ v The circulation resistance ratio when the moment magnitude M is 7.5 and the moment magnitude M is 1 atm; N 1,60 It represents the number of blows in the standard penetration test, corrected by the overburden pressure and hammer energy.
3. The method for rapid prediction of seismic demand for liquefied strata and low-carbon optimization according to claim 1 is characterized in that: The liquefied stratum parameters include the site inclination, the thickness of the overlying non-liquefied soil layer, the hardness of the upper non-liquefied soil layer, the thickness of the liquefied soil layer, N 1,60 and groundwater level depth; the gravel pile parameters include gravel pile diameter and gravel pile replacement rate.
4. The method for rapid prediction of seismic demand for liquefied strata and low-carbon optimization according to claim 3 is characterized in that: The gravel pile replacement rate is expressed as: Where D is the diameter of the gravel pile, and S is the spacing between the gravel piles.
5. The method for rapid prediction of seismic demand for liquefied strata and low-carbon optimization according to claim 1 is characterized in that: The seismic demand index is the relationship between the median lateral displacement of the site and the earthquake intensity, and its expression is: S D =aIM b Among them, S D represents the median lateral displacement of the site; IM represents the seismic intensity; a and b represent the seismic demand indicators.
6. The method for rapid prediction of seismic demand for liquefied strata and low-carbon optimization according to claim 1, characterized in that: The carbon emissions include carbon emissions calculated by process-based life cycle assessment method, carbon emissions calculated by economic input-output life cycle assessment method and carbon emissions calculated by hybrid life cycle assessment method; The life cycle assessment method is expressed as: Where E[Carbon(P-LCA)] represents the carbon emissions calculated based on the process-based life cycle assessment method; N represents the total number of projects; EF i represents the P-LCA carbon emission factor of the i-th project; Q i represents the construction work volume of the i-th project; The economic input-output life cycle assessment method is expressed as: Among them, E[Carbon] represents the carbon emissions calculated by the economic input-output life cycle assessment method; EIO i represents the EIO-LCA carbon emission factor of the i-th project; EC i represents the project cost of the i-th project; The hybrid life cycle assessment method is specifically as follows: if the carbon emissions of a subproject belong to construction carbon emissions or transportation carbon emissions, the carbon emissions of the subproject are calculated using the life cycle assessment method; if the carbon emissions of a subproject belong to raw material carbon emissions, the carbon emissions of the subproject are calculated using the economic input-output life cycle method. The carbon emissions of all subprojects are summed up to the carbon emissions calculated using the hybrid life cycle assessment method.
7. The method for rapid prediction of seismic demand for liquefied strata and low-carbon optimization according to claim 1, characterized in that: The construction cost includes the total cost of gravel piles, which is expressed as: Where EC represents the total cost of gravel piles; N represents the total number of projects; UC i represents the unit cost of the engineering project of the i-th project; Q i represents the construction workload of the i-th project.
8. The method for rapid prediction of seismic demand for liquefied strata and low-carbon optimization according to claim 1 is characterized in that: The finite element analysis platform is the Opensees platform.
9. The method for rapid prediction of seismic demand for liquefied strata and low-carbon optimization according to claim 1, characterized in that: Also includes: After training the ANN model, cross-validation method was used to optimize hyperparameters.