A method for predicting the life cycle carbon emissions of a solid waste-based concrete structure
By constructing a dynamic prediction model covering the entire life cycle, the problem of incomplete carbon emission assessment of solid waste-based concrete structures was solved, achieving high-precision carbon emission prediction and supporting green design and low-carbon decision-making.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGXI UNIVERSITY OF TECHNOLOGY
- Filing Date
- 2025-12-04
- Publication Date
- 2026-08-04
AI Technical Summary
Existing technologies lack a systematic and dynamic quantitative method for assessing the carbon emissions of solid waste-based concrete structures throughout their entire life cycle, resulting in incomplete assessment results, low prediction accuracy, and difficulty in supporting low-carbon optimization design and operation and maintenance decisions.
A dynamic prediction model for the production, service and decommissioning stages is constructed. Through principal component analysis and neural networks, combined with accelerated aging tests and reliability theory, a service performance evolution function and a full life cycle material reliability model are established. Physical constraint loss and sparsity regularization are introduced to achieve high-precision prediction of carbon emissions.
It achieves dynamic integrated prediction of carbon emissions throughout the entire life cycle, improves the scientific nature of carbon emissions during the service phase and the accuracy of predictions during the dismantling phase, provides more complete and realistic carbon footprint assessment results, and supports green design and low-carbon decision-making.
Smart Images
Figure CN121581899B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of carbon emission prediction technology for building materials, and in particular to a method for predicting carbon emissions throughout the entire life cycle of solid waste-based concrete structures. Background Technology
[0002] As a key area of global energy consumption and carbon emissions, the construction industry's green and low-carbon transformation has become a crucial path to achieving the "dual carbon" goals. Against this backdrop, utilizing industrial solid waste to prepare cementitious materials and applying them to concrete structures can not only dispose of a large amount of solid waste and reduce the consumption of natural resources, but also significantly reduce the high carbon emissions generated by cement production, resulting in outstanding environmental and economic benefits.
[0003] However, to scientifically assess and compare the comprehensive low-carbon benefits of different solid waste-based concrete formulations, simply measuring their carbon footprint during the production stage is far from sufficient. A life-cycle carbon emission assessment system covering raw material acquisition, production, service maintenance, and even demolition is essential. Traditional concrete carbon emission assessments often focus on the production stage, lacking systematic and dynamic quantitative methods for assessing carbon emissions generated during long-term service, such as performance degradation, maintenance and reinforcement activities, and demolition. This results in incomplete assessments and low prediction accuracy, making it difficult to support low-carbon optimization design and operation and maintenance decisions for solid waste-based concrete structures. Therefore, this invention proposes a life-cycle carbon emission prediction method for solid waste-based concrete structures. It constructs dynamic prediction models for the production and demolition stages using principal component analysis and neural networks, establishes service performance evolution functions and repair decision models based on accelerated aging tests and reliability theory, and introduces physical constraint loss, sparsity regularization, and adaptive weighting mechanisms to improve model robustness. This achieves high-precision, integrated prediction of life-cycle carbon emissions, providing scientific support for green design, life extension assessment, and carbon management decisions for solid waste-based concrete structures. Summary of the Invention
[0004] The purpose of this invention is to provide a method for predicting carbon emissions throughout the entire life cycle of solid waste-based concrete structures.
[0005] To achieve the above objectives, the present invention is implemented according to the following technical solution: This invention includes the following steps: Historical production process data of different solid waste-based concrete were obtained and the characteristics of the historical production process and the corresponding carbon emission factors were extracted. The carbon emission of the historical production process was calculated, and a carbon emission prediction model of the production process was constructed to obtain the predicted carbon emission of the solid waste-based concrete production process. Construct a service performance evolution function and a full life cycle material reliability model for solid waste cementitious materials, identify aging repair moments, and predict the number of aging repairs within the design life by combining the service environment of the solid waste-based concrete structure to be predicted. Extract standard aging repair operation data and corresponding carbon emission factors to calculate the carbon emission of a single standard aging repair operation, and calculate the predicted carbon emission during service life based on the number of aging repairs and the carbon emission of a single standard aging repair operation. Historical demolition process data of different solid waste-based concrete were obtained to extract the characteristics of the historical demolition process and the corresponding carbon emission factors. The carbon emission of the historical demolition process was calculated, and a carbon emission prediction model for the demolition process was constructed to obtain the predicted carbon emission of the solid waste-based concrete demolition process. The predicted carbon emissions of solid waste-based concrete structures throughout their entire life cycle are determined by the predicted carbon emissions during the production process, service life, and demolition process. The production process data includes raw material data, transportation data, and preparation data for solid waste-based concrete; the historical production process characteristics include material characteristics, transportation characteristics, and preparation characteristics. The carbon emission factors are obtained by querying a publicly available carbon emission factor database. The production process includes raw material acquisition, raw material transportation, and concrete preparation.
[0006] Furthermore, the method for constructing a carbon emission prediction model for the production process includes: Historical production process data of different solid waste-based concrete were obtained. Carbon emission factors corresponding to raw material production, transportation, power generation, and mechanical operation were extracted from public databases. Principal component analysis was used to extract historical production process data to obtain historical production process characteristics. Based on historical production process data and corresponding carbon emission factors, the carbon emissions of the production process were calculated, including raw material carbon emissions, transportation carbon emissions, and preparation carbon emissions. A carbon emission prediction model for the production process is trained using the carbon emissions from the production process and historical production process characteristics. The carbon emission prediction model for the production process includes an input layer, a feature enhancement layer, a prediction regression layer, and a policy output layer. The feature enhancement layer uses a fully connected neural network to perform high-dimensional mapping of historical production process features; The prediction regression layer uses the raw material prediction head, transportation prediction head, and preparation prediction head to predict the carbon emissions of raw materials, carbon emissions of carbon transportation, and carbon emissions of preparation, respectively. The strategy output layer sums the outputs of the three prediction heads through a fully connected network to obtain the predicted carbon emissions of the production process. The carbon emission prediction model for the production process adjusts the prediction accuracy through prediction loss, physical constraint loss, and sparsity loss, as expressed in the following expression: ; ; ; ; in For the total loss function, To predict losses, For physical constraint loss, For sparsity loss, , , For weight hyperparameters, The number of training samples. , , , To predict weights, , For the first The actual and predicted values of raw material carbon emissions for the sample group. , For the actual and predicted values of carbon emissions from transportation, , To prepare true and predicted values of carbon emissions, , This represents the actual and predicted values of carbon emissions during the production process. The output of the feature enhancement layer represents the production process feature dimension. For the first The weight parameters of each production process feature in the prediction head. This is the task weight index.
[0007] Furthermore, the method for constructing the service performance evolution function and the full life cycle material reliability model of solid waste cementitious materials includes: Accelerated aging tests were conducted on the solid waste-based concrete to be predicted. The inherent aging coefficient was determined by fitting the accelerated aging test results. The environmental impact coefficient was calculated based on the service environment parameters of the solid waste-based concrete structure to be predicted. An initial service performance evolution function for the solid waste-based concrete was constructed from the inherent aging coefficient and the environmental impact coefficient, with the following expression: ; ; in for The service performance of solid waste-based concrete. According to monthly statistics For the initial service performance of solid waste-based concrete, The inherent aging coefficient, for Environmental impact coefficient at any time for Cumulative number of months at any time For cumulative month index, For the first Monthly time interval , The orthogonal influence coefficient is... The chloride ion concentration in the service environment. The service environment temperature; Evolution function of service performance of solid waste-based concrete Construct a life-cycle material reliability model, the expression of which is:
[0008] in for Reliability of material properties at all times This is the performance failure threshold. , for The mean and standard deviation of the difference between the service performance and the performance failure threshold of solid waste-based concrete within the sliding window at any given time; The aging repair of solid waste-based concrete is assessed based on the reliability of material properties. After aging repair, the repair effect coefficient is determined based on the interfacial bond strength of the repair materials. A service performance evolution function for solid waste-based concrete is then constructed, expressed as: ; ; in For the first The repair efficiency coefficient of secondary aging repair For the first Performance recovery coefficient after secondary aging repair This represents the total number of aging repair cycles. For correction factor, For the first Secondary aging repair interface adhesion strength This represents the initial tensile strength of solid waste-based concrete.
[0009] Furthermore, the method for predicting the number of aging repairs within the design life includes: Starting from the service life of the solid waste-based concrete structure, the reliability of the solid waste-based concrete material performance is calculated monthly using the initial solid waste-based concrete service performance evolution function and the full life cycle material reliability model. When the reliability of the solid waste-based concrete material performance is less than the reliability warning threshold, the first aging repair is carried out and the corresponding aging repair time is extracted. The reliability of solid waste-based concrete material performance is calculated monthly using the service performance evolution function of solid waste-based concrete and the full life cycle material reliability model. The next aging repair time is determined, and the calculation is iterated until the aging repair time is greater than the structural design service life. The iteration stops when the time is greater than the structural design service life. The number of aging repairs corresponding to the previous aging repair time is taken as the number of aging repairs within the design service life of solid waste-based concrete.
[0010] Furthermore, the method for constructing a carbon emission prediction model for the demolition process includes: Historical demolition process data of different solid waste-based concrete were obtained. Carbon emission factors corresponding to blasting raw material production, blasting process, mechanical operation, excavator operation and transportation were extracted from public databases. Principal component analysis was used to extract historical demolition process data to obtain historical demolition process characteristics. Carbon emissions of historical demolition process were calculated based on historical demolition process data and corresponding carbon emission factors. A carbon emission prediction model for the demolition process is trained using historical demolition process characteristics and historical demolition process carbon emissions; the carbon emission prediction model for the demolition process includes an input layer, a feature processing layer, a prediction layer, and an output layer; The feature processing layer calculates a demolition difficulty influence factor based on the demolition process features, and enhances the demolition process features based on the demolition difficulty influence factor to obtain enhanced features; the expression for the demolition difficulty influence factor is:
[0011] in Factors affecting the difficulty of demolition, , , To integrate the difficulty of dismantling, For the influence function of building type, For building type, , This provides the usage and reference dosage of concrete for construction. , The compressive strength and reference compressive strength of solid waste-based concrete. , The elastic modulus and reference elastic modulus of solid waste-based concrete are given. This is the sensitivity coefficient for concrete performance. The prediction layer captures the nonlinear relationship between enhancement features, dismantling process features, and carbon emissions during the dismantling process using a three-layer BP neural network to obtain the predicted carbon emissions during the dismantling process. The three-layer BP neural network includes an input layer, a hidden layer, and an output layer. The hidden layer uses an activation function. The carbon emission prediction model for the demolition process employs an adaptive weight loss function to improve prediction accuracy. This adaptive weight loss function is determined based on the demolition difficulty influencing factors, and its expression is: ; in For adaptive weight loss function, The number of training samples. For adaptive weights, For the first The difficulty of dismantling the sample group is an influencing factor. The average value of the factors affecting the difficulty of demolition. The standard deviation of the factor affecting the difficulty of demolition. This represents the actual carbon emissions during the demolition process. This is a predicted amount of carbon emissions from the demolition process. The regularization coefficient is . For the mixed parameters of the elastic network, This is the model weight matrix; Input the characteristics of the solid waste-based concrete demolition process to be predicted into the carbon emission prediction model of the demolition process to obtain the predicted carbon emission of the demolition process.
[0012] Furthermore, the solid waste-based concrete formula includes: solid waste cementitious materials, coarse aggregate, fine aggregate, and alkali activator; the cementitious materials include metakaolin and granulated blast furnace slag; the coarse aggregate is commercial crushed stone; the fine aggregate is manufactured sand; and the alkali activator is prepared from water, solid sodium hydroxide, and sodium silicate solution.
[0013] The beneficial effects of this invention are: This invention provides a method for predicting the carbon emissions of solid waste-based concrete structures throughout their entire life cycle. Compared with existing technologies, this invention has the following technical advantages: It achieves dynamic integrated prediction of carbon emissions throughout the entire life cycle: covering all stages of raw material production, transportation, preparation, service maintenance and dismantling, and dynamically links maintenance activities during the service stage with carbon emissions through a performance degradation model, providing a more complete and realistic carbon footprint assessment result; An innovative service carbon emission prediction mechanism based on performance reliability has been developed: by constructing a service performance evolution function of solid waste cementitious materials and a full life cycle material reliability model, the performance degradation process of materials under environmental influence can be scientifically simulated, and the timing of repair and reinforcement can be accurately identified based on reliability indicators. This allows for dynamic prediction of the number of aging repairs within the design life and the corresponding carbon emissions, transforming service-stage carbon emission prediction from empirical estimation to scientific calculation based on performance degradation. This invention improves the accuracy and scientific rigor of carbon emission prediction during the demolition phase: In response to the complexity of the demolition process, this invention proposes a demolition difficulty influencing factor, quantifies the comprehensive impact of building type, concrete usage, and material mechanical properties on demolition energy consumption, and uses an intelligent prediction model that integrates this factor for regression prediction, effectively capturing the nonlinear relationship between various factors and carbon emissions, and significantly improving the accuracy and adaptability of demolition carbon emission prediction. With significant engineering application value and low-carbon guidance significance: The prediction method provided by this invention can quickly predict and compare the carbon footprint of solid waste-based concrete with different formulations, structural forms and environments throughout its entire life cycle. This provides key data support for the research and development of green building materials, the design of low-carbon structures and the optimization of operation and maintenance solutions, and powerfully promotes energy conservation and emission reduction and solid waste resource utilization in the construction industry. Attached Figure Description
[0014] Figure 1 This is a flowchart illustrating the steps of a method for predicting the carbon emissions of a solid waste-based concrete structure throughout its entire life cycle, as described in this invention. Detailed Implementation
[0015] The present invention will be further described below through specific embodiments. The illustrative embodiments and descriptions herein are used to explain the present invention, but are not intended to limit the present invention.
[0016] The present invention provides a method for predicting the carbon emissions of solid waste-based concrete structures throughout their entire life cycle, comprising the following steps: like Figure 1 As shown, this embodiment includes the following steps: Historical production process data of different solid waste-based concrete were obtained and the characteristics of the historical production process and the corresponding carbon emission factors were extracted. The carbon emission of the historical production process was calculated, and a carbon emission prediction model of the production process was constructed to obtain the predicted carbon emission of the solid waste-based concrete production process. Construct a service performance evolution function and a full life cycle material reliability model for solid waste cementitious materials, identify aging repair moments, and predict the number of aging repairs within the design life by combining the service environment of the solid waste-based concrete structure to be predicted. Extract standard aging repair operation data and corresponding carbon emission factors to calculate the carbon emission of a single standard aging repair operation, and calculate the predicted carbon emission during service life based on the number of aging repairs and the carbon emission of a single standard aging repair operation. Historical demolition process data of different solid waste-based concrete were obtained to extract the characteristics of the historical demolition process and the corresponding carbon emission factors. The carbon emission of the historical demolition process was calculated, and a carbon emission prediction model for the demolition process was constructed to obtain the predicted carbon emission of the solid waste-based concrete demolition process. The predicted carbon emissions of solid waste-based concrete structures throughout their entire life cycle are determined by the predicted carbon emissions during the production process, service life, and demolition process. The production process data includes raw material data, transportation data, and preparation data for solid waste-based concrete; the historical production process characteristics include material characteristics, transportation characteristics, and preparation characteristics. The carbon emission factors are obtained by querying a publicly available carbon emission factor database. The production process includes raw material acquisition, raw material transportation, and concrete preparation.
[0017] In this embodiment, the method for constructing a carbon emission prediction model for the production process includes: Historical production process data of different solid waste-based concrete were obtained. Carbon emission factors corresponding to raw material production, transportation, power generation, and mechanical operation were extracted from public databases. Principal component analysis was used to extract historical production process data to obtain historical production process characteristics. Based on historical production process data and corresponding carbon emission factors, the carbon emissions of the production process were calculated, including raw material carbon emissions, transportation carbon emissions, and preparation carbon emissions. A carbon emission prediction model for the production process is trained using the carbon emissions from the production process and historical production process characteristics. The carbon emission prediction model for the production process includes an input layer, a feature enhancement layer, a prediction regression layer, and a policy output layer. The feature enhancement layer uses a fully connected neural network to perform high-dimensional mapping of historical production process features; The prediction regression layer uses the raw material prediction head, transportation prediction head, and preparation prediction head to predict the carbon emissions of raw materials, carbon emissions of carbon transportation, and carbon emissions of preparation, respectively. The strategy output layer sums the outputs of the three prediction heads through a fully connected network to obtain the predicted carbon emissions of the production process. The carbon emission prediction model for the production process adjusts the prediction accuracy through prediction loss, physical constraint loss, and sparsity loss, as expressed in the following expression: ; ; ; ; in For the total loss function, To predict losses, For physical constraint loss, For sparsity loss, , , For weight hyperparameters, The number of training samples. , , , To predict weights, , For the first The actual and predicted values of raw material carbon emissions for the sample group. , For the actual and predicted values of carbon emissions from transportation, , To prepare true and predicted values of carbon emissions, , This represents the actual and predicted values of carbon emissions during the production process. The output of the feature enhancement layer represents the production process feature dimension. For the first The weight parameters of each production process feature in the prediction head. For task weight index; In the actual assessment, historical production process data of different solid waste-based concretes were obtained; the production process data includes raw material data, transportation data, and preparation data of solid waste-based concrete; the raw material data includes the material proportions of solid waste-based concrete; the transportation data includes the solid waste origin, construction site, vehicle type, and transportation materials; the preparation data includes preparation steps, equipment used, working time of each step, and power consumption of each step. Carbon emission factors corresponding to raw material production, transportation, power generation, and mechanical operation are extracted from public databases, and historical production process characteristics are obtained by extracting historical production process data using principal component analysis. The historical production process characteristics include material characteristics, transportation characteristics, and preparation characteristics; the material characteristics include solid waste type, solid waste dosage, and concrete material properties; the transportation characteristics include transportation distance of each raw material, solid waste dosage, and concrete material properties; the preparation characteristics include preparation steps, solid waste dosage, and concrete material properties. The concrete material properties are obtained through material property tests, including concrete strength grade and elastic modulus. The concrete material properties are affected by the material mix ratio (coarse / fine aggregate, cement content), and therefore have an indirect relationship with raw material carbon emissions, transportation carbon emissions (the amount of work done by transportation vehicles), and preparation carbon emissions (the amount of work done by equipment screening / mixing). The carbon emissions from the production process are calculated based on historical production process data and corresponding carbon emission factors, specifically as follows: Carbon emissions from raw materials: The mass dosage of each raw material in a unit mass of solid waste-based concrete is determined based on the material mix ratio. The carbon emissions of each raw material in a unit mass of solid waste-based concrete are obtained by multiplying the mass dosage of each raw material with the corresponding carbon emission factor. The carbon emissions are then summed and multiplied by the density of solid waste-based concrete to obtain the carbon emissions of raw materials per unit volume of solid waste-based concrete. Carbon emissions from transportation: (1) Determine the transportation distance of solid waste materials based on the geographic codes of the solid waste production site and the construction site (based on the means of transportation, the distance is divided into land transportation distance and water transportation distance. The same material may be transferred by multiple means of transportation. The carbon emissions from transportation are calculated in segments and then accumulated). Based on the construction site, directly search the nearest sand and gravel mining area to determine the transportation distance of sand and gravel raw materials (which may also include both land transportation distance and water transportation distance). The transportation distance of other raw materials is taken as 50km, and the transportation method is land transportation. (2) Calculate the carbon emission factor of the means of transportation and the product of the transportation distance, and then divide it by the single load of raw materials of the means of transportation to obtain the carbon emissions from transportation per unit mass of raw materials. (3) Calculate the carbon emissions from transportation per unit mass of raw materials and the product of mass admixture, accumulate the carbon emissions from transportation per unit mass of solid waste-based concrete, and multiply it by the density of solid waste-based concrete to obtain the carbon emissions from transportation per unit volume of solid waste-based concrete. Carbon emissions from preparation: The process power (equipment power), equipment working time, and grid power generation carbon emission factor of each process are calculated sequentially according to the preparation process. The carbon emissions from electricity consumption are then accumulated. The mechanical carbon emissions are calculated sequentially according to the preparation process and the equipment working time and equipment mechanical carbon emission factor of each process are then accumulated. The sum of the carbon emissions from electricity consumption and mechanical carbon emissions per unit volume of solid waste-based concrete structure is taken as the carbon emissions from the preparation of solid waste-based concrete per unit volume. In the carbon emission prediction model for the production process: The input layer needs to preprocess the production process features (standardization, unique encoding) and convert them into vectors of fixed dimensions; The feature enhancement layer uses a two-layer fully connected neural network to map the production process features (the feature dimension was originally 1 / 4d) to a higher dimension. The first layer expands the dimension to 1 / 2d through the ReLU activation function, and the second layer expands the dimension to d through the ReLU activation function. In the prediction regression layer, the raw material prediction head captures the linear relationship between material characteristics and raw material carbon emissions through a linear layer to predict raw material carbon emissions; the transportation prediction head captures the nonlinear relationship between transportation characteristics and transportation carbon emissions through a BP neural network to predict carbon transportation carbon emissions; and the preparation prediction head uses a recurrent neural network to capture the time nonlinear relationship between preparation characteristics and preparation carbon emissions to predict preparation carbon emissions. The strategy output layer integrates the outputs of the three prediction heads through the Sigmoid activation function to obtain the predicted carbon emissions of the production process (in 1m). 3 (Unit: solid waste-based concrete) based on Using a set of sample data, a carbon emission prediction model for the production process is trained based on the total loss function, and weighted hyperparameters are selected. , , The prediction weights are 0.5 / 0.3 / 0.2. , , , It is 0.25 / 0.25 / 0.25 / 0.25.
[0018] In this embodiment, the method for constructing the service performance evolution function and the full life cycle material reliability model of solid waste cementitious materials includes: Accelerated aging tests were conducted on the solid waste-based concrete to be predicted. The inherent aging coefficient was determined by fitting the accelerated aging test results. The environmental impact coefficient was calculated based on the service environment parameters of the solid waste-based concrete structure to be predicted. An initial service performance evolution function for the solid waste-based concrete was constructed from the inherent aging coefficient and the environmental impact coefficient, with the following expression: ; ; in for The service performance of solid waste-based concrete. According to monthly statistics For the initial service performance of solid waste-based concrete, The inherent aging coefficient, for Environmental impact coefficient at any time for Cumulative number of months at any time For cumulative month index, For the first Monthly time interval , The orthogonal influence coefficient is... The chloride ion concentration in the service environment. The service environment temperature; Evolution function of service performance of solid waste-based concrete Construct a life-cycle material reliability model, the expression of which is:
[0019] in for Reliability of material properties at all times This is the performance failure threshold. , for The mean and standard deviation of the difference between the service performance and the performance failure threshold of solid waste-based concrete within the sliding window at any given time; The aging repair of solid waste-based concrete is assessed based on the reliability of material properties. After aging repair, the repair effect coefficient is determined based on the interfacial bond strength of the repair materials. A service performance evolution function for solid waste-based concrete is then constructed, expressed as: ; ; in For the first The repair efficiency coefficient of secondary aging repair For the first Performance recovery coefficient after secondary aging repair This represents the total number of aging repair cycles. For correction factor, For the first Secondary aging repair interface adhesion strength This represents the initial tensile strength of solid waste-based concrete.
[0020] 4. The method for predicting the carbon emissions of a solid waste-based concrete structure throughout its entire life cycle according to claim 1, characterized in that the method for predicting the number of aging repairs within the design life includes: Starting from the service life of the solid waste-based concrete structure, the reliability of the solid waste-based concrete material performance is calculated monthly using the initial solid waste-based concrete service performance evolution function and the full life cycle material reliability model. When the reliability of the solid waste-based concrete material performance is less than the reliability warning threshold, the first aging repair is carried out and the corresponding aging repair time is extracted. The service performance evolution function of solid waste-based concrete and the material reliability model of the whole life cycle are used to calculate the material performance reliability of solid waste-based concrete on a monthly basis to determine the next aging repair time. The calculation is iterated until the aging repair time is greater than the structural design service life, at which point the iteration stops and the number of aging repairs corresponding to the previous aging repair time is taken as the number of aging repairs within the design service life of solid waste-based concrete. In the actual evaluation, accelerated aging tests (300 freeze-thaw cycles) were conducted, and the test data were analyzed using an exponential decay model. The inherent aging coefficient was obtained by fitting. The orthogonal influence coefficients were calibrated through orthogonal experiments. , The environmental impact factor is updated monthly based on the historical monthly average service environment temperature and annual average service environment chloride ion concentration. ; The service performance (specifically, the compressive strength of solid waste-based concrete) of the concrete was calculated monthly using an initial solid waste-based concrete service performance evolution function. The material reliability of the solid waste-based concrete was also calculated monthly using a life-cycle material reliability model. When the concrete material reliability fell below a reliability warning threshold, the first aging repair was performed, and the corresponding aging repair time was extracted. In the life-cycle material reliability model, the performance failure threshold... Taking 70% of the initial service performance, a reliability warning threshold is set based on a structural failure probability of 5% to 0.1%. The range is 2.33 to 3.09; After the first aging repair, a repair efficiency factor is introduced. and performance recovery coefficient The initial service performance evolution function of solid waste-based concrete is updated to obtain the service performance evolution function of solid waste-based concrete. Combined with the full life-cycle material reliability model, the material performance reliability of solid waste-based concrete is calculated monthly. The next aging repair time is extracted, and the calculation is iteratively performed until the aging repair time exceeds the structural design service life of 40 years. Then, the number of aging repairs corresponding to the previously calculated aging repair time is selected as the number of aging repairs within the design service life of the solid waste-based concrete. Among these, a correction coefficient is used. Take a performance recovery coefficient of 0.8 to 1.2. The value ranges from 0.3 to 0.8, and decreases with increasing repair frequency. A paste with a higher strength than the current solid waste-based concrete paste is selected for aging repair, and the bonding strength at the aging repair interface is [not specified]. A fixed value is taken; the mean and standard deviation of the difference between the service performance and the performance failure threshold of solid waste-based concrete reflect... Fluctuations in the service performance of solid waste-based concrete caused by temperature within the sliding window prior to the time point; During aging repair, the structural surface is thinned by grinding, and new reinforcement material is poured at the grinding interface. Taking the aging repair of a standard beam structure as an example, the reinforcement replacement rate for a single aging repair is determined to be 5%. The carbon emission factor of the reinforcement material is extracted, and the replacement volume of the reinforcement material per unit volume of solid waste-based concrete beam structure in a single aging repair is calculated (unit volume * reinforcement replacement rate). The carbon emission of the material in a single aging repair is obtained by calculating (reinforcement material carbon emission factor * reinforcement material replacement volume). The grinding time is obtained from (reinforcement material replacement volume / grinding machine efficiency). The mechanical carbon emission factor of the grinding machine and the carbon emission factor of power grid generation are extracted, and (grinding machine mechanical carbon emission factor * grinding time) is calculated. The carbon emissions from grinding during a single aging repair are obtained by summing (grinding time * grinding machine power * grid power generation carbon emission factor); the welding time is taken as a fixed value, the carbon emission factor of the welding machinery is extracted, and the carbon emissions from welding during a single aging repair are obtained by summing (welding time * welding machinery carbon emission factor) and (welding time * welding power * grid power generation carbon emission factor); the carbon emissions from temporary electricity consumption during a single aging repair are obtained by calculating (other temporary power consumption * temporary power consumption time * grid power generation carbon emission factor); the carbon emissions from materials, grinding, welding, and temporary electricity consumption during a single aging repair are summed as the carbon emissions for a single standard aging repair operation. The predicted carbon emissions during service life are calculated based on the number of aging repairs and the carbon emissions from a single standard aging repair operation.
[0021] In this embodiment, the method for constructing a carbon emission prediction model for the demolition process includes: Historical demolition process data of different solid waste-based concrete were obtained. Carbon emission factors corresponding to blasting raw material production, blasting process, mechanical operation, excavator operation and transportation were extracted from public databases. Principal component analysis was used to extract historical demolition process data to obtain historical demolition process characteristics. Carbon emissions of historical demolition process were calculated based on historical demolition process data and corresponding carbon emission factors. A carbon emission prediction model for the demolition process is trained using historical demolition process characteristics and historical demolition process carbon emissions; the carbon emission prediction model for the demolition process includes an input layer, a feature processing layer, a prediction layer, and an output layer; The feature processing layer calculates a demolition difficulty influence factor based on the demolition process features, and enhances the demolition process features based on the demolition difficulty influence factor to obtain enhanced features; the expression for the demolition difficulty influence factor is:
[0022] in Factors affecting the difficulty of demolition, , , To integrate the difficulty of dismantling, For the influence function of building type, For building type, , This provides the usage and reference dosage of concrete for construction. , The compressive strength and reference compressive strength of solid waste-based concrete. , The elastic modulus and reference elastic modulus of solid waste-based concrete are given. This is the sensitivity coefficient for concrete performance. The prediction layer captures the nonlinear relationship between enhancement features, dismantling process features, and carbon emissions during the dismantling process using a three-layer BP neural network to obtain the predicted carbon emissions during the dismantling process. The three-layer BP neural network includes an input layer, a hidden layer, and an output layer. The hidden layer uses an activation function. The carbon emission prediction model for the demolition process employs an adaptive weight loss function to improve prediction accuracy. This adaptive weight loss function is determined based on the demolition difficulty influencing factors, and its expression is: ; in For adaptive weight loss function, The number of training samples. For adaptive weights, For the first The difficulty of dismantling the sample group is an influencing factor. The average value of the factors affecting the difficulty of demolition. The standard deviation of the factor affecting the difficulty of demolition. This represents the actual carbon emissions during the demolition process. This is a predicted amount of carbon emissions from the demolition process. The regularization coefficient is . For the mixed parameters of the elastic network, This is the model weight matrix; Input the characteristics of the solid waste-based concrete demolition process to be predicted into the carbon emission prediction model of the demolition process to obtain the predicted carbon emission of the demolition process. In actual assessments, historical demolition process data for different solid waste-based concrete is obtained, specifically including building information (building type, concrete usage), blasting material usage, demolition equipment power and cumulative duration, excavator cumulative working time, construction site, and transportation trips. Blasting carbon emissions are calculated by multiplying blasting material usage by (carbon emission factor from blasting raw material production + carbon emission factor from blasting process). Mechanical demolition carbon emissions are calculated by summing (cumulative demolition equipment time * mechanical operation carbon emission factor) and (demolition equipment power * cumulative time * grid power generation carbon emission factor). The carbon emissions from excavator demolition are calculated as (cumulative working hours of excavator * carbon emission factor of excavator operation); the carbon emissions from construction waste transportation are calculated as (number of transportation trips * construction waste transportation distance * carbon emission factor of transportation); the carbon emissions from blasting, mechanical demolition, excavator demolition, and construction waste transportation are summed as the carbon emissions from building demolition, and the carbon emissions from building demolition are calculated as (building demolition carbon emissions / amount of concrete used) to obtain the carbon emissions from the demolition process (per 1m³). 3 (Unit: solid waste-based concrete) The characteristics of the demolition process include the transportation distance of construction waste, transportation method, building type, amount of concrete used in construction, and the material properties of solid waste-based concrete; the amount of concrete used in construction is taken as (building volume * concrete mix ratio * loss coefficient). In the carbon emission prediction model for the demolition process, the enhanced feature vector is specifically: The input layer of a three-layer backpropagation (BP) neural network consists of m neurons, the hidden layer consists of 2m neurons, and the output layer consists of m neurons; the regularization coefficient of the adaptive weight loss function. Take 0.01, elastic mesh hybrid parameters Take 0.5; Input the characteristics of the solid waste-based concrete demolition process to be predicted into the carbon emission prediction model of the demolition process to obtain the predicted carbon emission of the demolition process. The predicted carbon emissions per unit volume of solid waste-based concrete structure are obtained by summing the predicted carbon emissions during the production process, service life, and demolition process.
[0023] In this embodiment, the solid waste-based concrete formula includes: solid waste cementitious material, coarse aggregate, fine aggregate, and alkali activator; the cementitious material includes metakaolin and granulated blast furnace slag; the coarse aggregate is commercial crushed stone; the fine aggregate is manufactured sand; and the alkali activator is prepared from water, solid sodium hydroxide, and sodium silicate solution.
[0024] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A method for predicting the carbon emissions of solid waste-based concrete structures throughout their entire life cycle, characterized in that, Includes the following steps: S1. Obtain historical production process data of different solid waste-based concrete and extract the characteristics of historical production process and corresponding carbon emission factors. Calculate the carbon emission of historical production process and construct a carbon emission prediction model of production process to obtain the predicted carbon emission of solid waste-based concrete production process. S2. Construct the service performance evolution function of solid waste cementitious materials and the material reliability model for the whole life cycle, identify the aging repair time, and predict the number of aging repairs within the design life by combining the service environment of the solid waste-based concrete structure to be predicted. S3. Extract standard aging repair operation data and corresponding carbon emission factors to calculate the carbon emission of a single standard aging repair operation. Calculate the predicted carbon emission during service life based on the number of aging repairs and the carbon emission of a single standard aging repair operation. S4. Obtain historical demolition process data of different solid waste-based concrete, extract the characteristics of the historical demolition process and the corresponding carbon emission factors, calculate the carbon emission of the historical demolition process, and construct a carbon emission prediction model of the demolition process to obtain the predicted carbon emission of the solid waste-based concrete demolition process. S5. Determine the total life-cycle carbon emission forecast of solid waste-based concrete structures based on the predicted carbon emissions during the production process, service life, and demolition process. The production process data includes raw material data, transportation data, and preparation data for solid waste-based concrete; the historical production process characteristics include material characteristics, transportation characteristics, and preparation characteristics. The carbon emission factors are obtained by querying a publicly available carbon emission factor database. The production process includes raw material acquisition, raw material transportation, and concrete preparation.
2. The method for predicting the carbon emissions of solid waste-based concrete structures throughout their entire life cycle according to claim 1, characterized in that, The method for constructing a carbon emission prediction model for the production process includes: Historical production process data of different solid waste-based concrete were obtained. Carbon emission factors corresponding to raw material production, transportation, power generation, and mechanical operation were extracted from public databases. Principal component analysis was used to extract historical production process data to obtain historical production process characteristics. Based on historical production process data and corresponding carbon emission factors, the carbon emissions of the production process were calculated, including raw material carbon emissions, transportation carbon emissions, and preparation carbon emissions. A carbon emission prediction model for the production process is trained using the carbon emissions from the production process and historical production process characteristics. The carbon emission prediction model for the production process includes an input layer, a feature enhancement layer, a prediction regression layer, and a policy output layer. The feature enhancement layer uses a fully connected neural network to perform high-dimensional mapping of historical production process features; The prediction regression layer uses the raw material prediction head, transportation prediction head, and preparation prediction head to predict the carbon emissions of raw materials, carbon emissions of carbon transportation, and carbon emissions of preparation, respectively. The strategy output layer sums the outputs of the three prediction heads through a fully connected network to obtain the predicted carbon emissions of the production process. The carbon emission prediction model for the production process adjusts the prediction accuracy through prediction loss, physical constraint loss, and sparsity loss, as expressed in the following expression: ； ； ； ； in For the total loss function, To predict losses, For physical constraint loss, For sparsity loss, , , For weight hyperparameters, The number of training samples. , , , To predict weights, , For the first The actual and predicted values of raw material carbon emissions for the sample group. , For the actual and predicted values of carbon emissions from transportation, , To prepare true and predicted values of carbon emissions, , This represents the actual and predicted values of carbon emissions during the production process. The output of the feature enhancement layer represents the production process feature dimension. For the first The weight parameters of each production process feature in the prediction head. This is the task weight index.
3. The method for predicting the carbon emissions of solid waste-based concrete structures throughout their entire life cycle according to claim 1, characterized in that, The method for constructing the service performance evolution function and the full life cycle material reliability model of solid waste cementitious materials includes: Accelerated aging tests were conducted on the solid waste-based concrete to be predicted. The inherent aging coefficient was determined by fitting the accelerated aging test results. The environmental impact coefficient was calculated based on the service environment parameters of the solid waste-based concrete structure to be predicted. An initial service performance evolution function for the solid waste-based concrete was constructed from the inherent aging coefficient and the environmental impact coefficient, with the following expression: ； ； in for The service performance of solid waste-based concrete. According to monthly statistics, For the initial service performance of solid waste-based concrete, The inherent aging coefficient, for Environmental impact coefficient at any time for Cumulative number of months at any time For cumulative month index, For the first Monthly time interval, , The orthogonal influence coefficient is... The chloride ion concentration in the service environment. The service environment temperature; Evolution function of service performance of solid waste-based concrete Construct a life-cycle material reliability model, the expression of which is: ； in for Reliability of material properties at all times This is the performance failure threshold. , for The mean and standard deviation of the difference between the service performance and the performance failure threshold of solid waste-based concrete within the sliding window at any given time; The aging repair of solid waste-based concrete is assessed based on the reliability of material properties. After aging repair, the repair effect coefficient is determined based on the interfacial bond strength of the repair materials. A service performance evolution function for solid waste-based concrete is then constructed, expressed as: ; ； in For the first The repair efficiency coefficient of secondary aging repair For the first Performance recovery coefficient after secondary aging repair This represents the total number of aging repair cycles. For correction factor, For the first Secondary aging repair interface adhesion strength This represents the initial tensile strength of solid waste-based concrete.
4. The method for predicting the carbon emissions of solid waste-based concrete structures throughout their entire life cycle according to claim 1, characterized in that, The method for predicting the number of aging repairs within the design life includes: Starting from the service life of the solid waste-based concrete structure, the reliability of the solid waste-based concrete material performance is calculated monthly using the initial solid waste-based concrete service performance evolution function and the full life cycle material reliability model. When the reliability of the solid waste-based concrete material performance is less than the reliability warning threshold, the first aging repair is carried out and the corresponding aging repair time is extracted. The reliability of solid waste-based concrete material performance is calculated monthly using the service performance evolution function of solid waste-based concrete and the full life cycle material reliability model. The next aging repair time is determined, and the calculation is iterated until the aging repair time is greater than the structural design service life. The iteration stops when the time is greater than the structural design service life. The number of aging repairs corresponding to the previous aging repair time is taken as the number of aging repairs within the design service life of solid waste-based concrete.
5. The method for predicting the carbon emissions of a solid waste-based concrete structure throughout its entire life cycle according to claim 1, characterized in that, The method for constructing a carbon emission prediction model for the demolition process includes: Historical demolition process data of different solid waste-based concrete were obtained. Carbon emission factors corresponding to blasting raw material production, blasting process, mechanical operation, excavator operation and transportation were extracted from public databases. Principal component analysis was used to extract historical demolition process data to obtain historical demolition process characteristics. Carbon emissions of historical demolition process were calculated based on historical demolition process data and corresponding carbon emission factors. A carbon emission prediction model for the demolition process is trained using historical demolition process characteristics and historical demolition process carbon emissions; the carbon emission prediction model for the demolition process includes an input layer, a feature processing layer, a prediction layer, and an output layer; The feature processing layer calculates a demolition difficulty influence factor based on the demolition process features, and enhances the demolition process features based on the demolition difficulty influence factor to obtain enhanced features; the expression for the demolition difficulty influence factor is: ； in Factors affecting the difficulty of demolition, , , To integrate the difficulty of dismantling, For the influence function of building type, For building type, , This provides the usage and reference dosage of concrete for construction. , The compressive strength and reference compressive strength of solid waste-based concrete. , The elastic modulus and reference elastic modulus of solid waste-based concrete are given. This is the sensitivity coefficient for concrete performance. The prediction layer captures the nonlinear relationship between enhancement features, dismantling process features, and carbon emissions during the dismantling process using a three-layer BP neural network to obtain the predicted carbon emissions during the dismantling process. The three-layer BP neural network includes an input layer, a hidden layer, and an output layer. The hidden layer uses an activation function. The carbon emission prediction model for the demolition process employs an adaptive weight loss function to improve prediction accuracy. This adaptive weight loss function is determined based on the demolition difficulty influencing factors, and its expression is: ； in For adaptive weight loss function, The number of training samples. For adaptive weights, For the first The difficulty of dismantling the sample group is an influencing factor. The average value of the factors affecting the difficulty of demolition. The standard deviation of the factor affecting the difficulty of demolition. This represents the actual carbon emissions during the demolition process. This is a predicted amount of carbon emissions from the demolition process. The regularization coefficient is . For the mixed parameters of the elastic network, This is the model weight matrix; Input the characteristics of the solid waste-based concrete demolition process to be predicted into the carbon emission prediction model of the demolition process to obtain the predicted carbon emission of the demolition process.
6. The method for predicting the carbon emissions of a solid waste-based concrete structure throughout its entire life cycle according to claim 1, characterized in that, The solid waste-based concrete formula includes: solid waste cementitious materials, coarse aggregate, fine aggregate, and alkali activator; the cementitious materials include metakaolin and granulated blast furnace slag; the coarse aggregate is commercial crushed stone; the fine aggregate is manufactured sand; and the alkali activator is prepared from water, solid sodium hydroxide, and sodium silicate solution.