Methods and systems for dynamic carbon accounting of coal-based solid waste road base layers throughout their entire life cycle
By refining the entire life cycle stages, establishing a classification and identification system, integrating historical and real-time data into a database, constructing a phased carbon emission quantification model and scenario correction matrix, and combining LSTM neural networks for prediction, the problems of incomplete full life cycle coverage and poor data adaptability in existing technologies have been solved, achieving accurate and dynamic carbon accounting and providing a basis for long-term planning.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SHANXI TRAFFIC PLANNING PROSPECTING & DESIGN INST
- Filing Date
- 2025-10-12
- Publication Date
- 2026-05-26
AI Technical Summary
Existing carbon accounting methods for coal-based solid waste roadbeds fail to fully cover the entire life cycle, neglect the calculation of regenerated carbon balance in the waste recycling stage, and lack real-time monitoring equipment and classification and identification systems. This results in poor data adaptability, weak predictive ability, and an inability to cope with biases caused by variables such as the type of modifier and climate differences.
By refining the entire life cycle stages, establishing a classification and identification system, integrating historical and real-time data into a database, constructing a phased carbon emission quantification model and scenario correction matrix, combining LSTM neural network for prediction, performing renewable carbon balance calculations, and analyzing uncertainties through Monte Carlo simulation to output carbon emission placement confidence intervals.
It enables accurate and dynamic full life-cycle carbon accounting, provides a basis for long-term planning, improves the scientific nature and accuracy of accounting, and takes into account both environmental protection and resource recycling.
Smart Images
Figure CN121684351B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of carbon accounting, and in particular to a method and system for dynamic carbon accounting of coal-based solid waste road base throughout its entire life cycle. Background Technology
[0002] Solid wastes such as coal gangue and fly ash generated during coal mining, washing, and utilization accumulate over time, occupying land and easily causing environmental pollution. Using these wastes as road base materials can both dispose of solid waste and replace traditional energy-intensive building materials, demonstrating significant ecological and carbon reduction potential. However, current carbon accounting for solid waste resource utilization often focuses on a single stage, lacking a dynamic lifecycle assessment from raw material acquisition, processing, construction, operation to waste recycling, making it difficult to quantify its true carbon reduction benefits.
[0003] Current carbon accounting methods for coal-based solid waste roadbed construction lack comprehensive lifecycle coverage, often neglecting the regeneration carbon balance calculation during the waste recycling stage. They focus only on a single stage or partial links, failing to reflect the actual carbon emissions across the entire chain from generation to disposal to recycling. Data collection relies heavily on static historical coefficients, lacking dynamic parameter support from real-time monitoring equipment. Furthermore, they fail to establish classification and labeling systems for different solid wastes such as fly ash and coal gangue, resulting in poor data adaptability. In terms of models, few design independent calculation units for each stage, and none construct scenario-carbon emission correction matrices, making it unable to handle biases introduced by variables such as modifier type and climate differences. Predictive capabilities are weak, failing to incorporate LSTM neural networks for medium- to long-term forecasts, and generally lacking uncertainty analysis using Monte Carlo simulations, making it difficult to output reliable carbon emission placement confidence intervals. Summary of the Invention
[0004] To improve existing methods and systems, this paper presents a method and system for dynamic carbon accounting of coal-based solid waste road base courses throughout their entire life cycle. This method refines the entire life cycle stages, integrates historical and real-time data into a database, and combines phased quantitative models, scenario correction, LSTM prediction, regenerated carbon balance calculation, and Monte Carlo uncertainty analysis. It can accurately, dynamically, and comprehensively complete the carbon accounting of coal-based solid waste road base courses and provide a basis for long-term planning.
[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:
[0006] A dynamic carbon accounting method for the entire life cycle of coal-based solid waste road base courses, including:
[0007] The entire life cycle of coal-based solid waste road base is divided into stages, and coal-based solid waste is classified into types according to its composition and characteristics, and a classification and identification system is established.
[0008] Based on historical data, the emission coefficients of various coal-based solid wastes at the generation stage are collected. Monitoring equipment is deployed in the road base construction area and operating road section to collect dynamic parameters in each life cycle stage in real time, forming a basic database.
[0009] A phased carbon emission quantification model is constructed based on data from the basic database, and carbon emission quantification data for each life cycle stage is calculated and obtained respectively.
[0010] Based on key scenario variables affecting carbon emissions, a scenario-carbon emission correction matrix is established to correct the quantitative carbon emission data for each life cycle stage.
[0011] Based on real-time collected dynamic parameters during the operation phase, a long short-term memory neural network model is used to predict carbon emission quantification data for the next 5-10 years and obtain dynamic carbon emission prediction curves.
[0012] Based on the first and second recycling of recycled aggregates after the abandonment of coal-based solid waste road base, the carbon emission coefficients of crushing energy consumption and screening energy consumption corresponding to different recycling, and the carbon emission reduction of recycled aggregates replacing natural aggregates, a carbon balance equation is established to calculate the carbon balance in the waste recycling stage.
[0013] Based on updated data from the basic database and dynamically monitored parameters, the life-cycle carbon emissions are recalculated. The Monte Carlo simulation method is used to analyze the impact of the uncertainty of parameters at each stage on the accounting results and output the confidence interval of the carbon emissions.
[0014] Preferably, the step of dividing the entire life cycle of coal-based solid waste road base into stages, classifying coal-based solid waste into types according to its composition and characteristics, and establishing a classification and identification system specifically includes:
[0015] The entire life cycle of the coal-based solid waste road base course covers the coal-based solid waste generation stage, pretreatment stage, transportation stage, road base course construction stage, operation and maintenance stage, and waste recycling stage.
[0016] Based on the composition and physical and mechanical properties of coal-based solid waste as the main classification criteria, coal-based solid waste is divided into fly ash, coal gangue, coal slime and coal-related kaolin.
[0017] Design unique labeling information for each type of coal-based solid waste, including the solid waste category name, core characteristic indicators, source of generation, and application scenarios, and establish a classification labeling system.
[0018] Preferably, the step of collecting emission coefficients of various coal-based solid wastes at their generation stages based on historical data, deploying monitoring equipment in road base construction areas and operational road sections, and collecting dynamic parameters in real time at each life cycle stage to form a basic database specifically includes:
[0019] For each stage of the entire lifecycle, corresponding coefficients and dynamic parameters are collected, including:
[0020] The primary carbon emission coefficient of associated solid waste generated during coal mining and the primary carbon emission coefficient of solid waste separated from coal washing plants during the coal-based solid waste generation stage.
[0021] Energy consumption and carbon emission coefficients of each process in the pretreatment stage of coal-based solid waste;
[0022] Carbon emission coefficient per unit mileage of transportation vehicles during the transportation of coal-based solid waste is determined by collecting data on actual vehicle load, transportation distance, and actual driving speed.
[0023] The carbon emission coefficient per man-hour of construction machinery during the road base construction stage is determined by collecting data on the actual compaction degree of the road base, the material temperature during mixing, and the actual operating time of each construction machine.
[0024] Carbon emission coefficients of various materials and energy consumption carbon emission coefficients of maintenance operations during the road base operation and maintenance phase are collected, along with daily traffic volume, real-time road surface temperature, daily rainfall, and strain data of the base structure layer.
[0025] The carbon emission coefficients of the recycling process of road base layer crushing and screening and the carbon emission coefficients of solid waste landfill disposal were collected. Data on the utilization rate of recycled aggregate after crushing of coal-based solid waste base layer and the amount of unrecycled solid waste landfilled were also collected.
[0026] The above data is integrated, categorized, and stored to form a basic database.
[0027] Preferably, the step of constructing a phased carbon emission quantification model based on data from the basic database, and calculating and obtaining carbon emission quantification data for each life cycle stage, specifically includes:
[0028] Obtain the temporal and spatial boundaries of each life cycle stage, identify the core carbon emission sources for each stage, and form a stage-carbon emission source inventory.
[0029] Based on data in the basic database, a corresponding basic coefficient is matched for each carbon emission source, and the basic coefficients are classified and calibrated.
[0030] A phased carbon emission quantification model is constructed, which adopts a phased calculation and total output structure. An independent calculation unit is set up for each phase, and the unit includes basic coefficient retrieval, actual data input, correction coefficient calculation, and phase result output.
[0031] The carbon emission quantification data for each stage of the life cycle is obtained based on the phased carbon emission quantification model.
[0032] Preferably, the step of establishing a scenario-carbon emission correction matrix based on key scenario variables affecting carbon emissions, and correcting the quantitative carbon emission data for each life cycle stage, specifically includes:
[0033] Based on the full life cycle characteristics of coal-based solid waste road base, key scenario variables were obtained, including the type of coal-based solid waste modifier, the slope of the transportation route, and the climate type of the operating area.
[0034] Construct a three-dimensional matrix of scenario variable combination - life cycle stage - correction coefficient. The row dimension represents all possible combinations of scenario variables, the column dimension represents the life cycle stage affected by the scenario, and the numerical dimension represents the correction coefficient of each variable combination-stage.
[0035] Carbon emission quantification data for each stage is extracted from the phased carbon emission quantification model. Single-stage correction calculations are performed using correction coefficients, and multivariate superposition corrections are applied to stages affected by multiple scenario variables to obtain corrected carbon emission quantification data.
[0036] Preferably, the process of predicting carbon emission quantification data for the next 5-10 years based on real-time collected dynamic parameters during the operational phase, using a long short-term memory neural network model, and obtaining a dynamic carbon emission prediction curve specifically includes:
[0037] A prediction model is constructed using a long short-term memory neural network. The neural network is trained based on the carbon emission quantification data in the historical dataset. The mean squared error is used as the loss function to measure the difference between the predicted value and the historical data.
[0038] The model consists of an input layer, a hidden layer, and an output layer. There are a total of 3 hidden layers, each containing 64 neurons, and all of them use the ReLU activation function to process the data.
[0039] Based on the corrected carbon emission quantification data during the operation phase, the data is input into the model to obtain the predicted carbon emission quantification data for the next 5-10 years.
[0040] The quantitative carbon emission forecasts for the next 5-10 years are organized into time series data based on time order, and a dynamic carbon emission forecast curve is plotted.
[0041] Preferably, the carbon balance equation established based on the first and second recycling of recycled aggregates from coal-based solid waste road base courses, the carbon emission coefficients of crushing energy consumption and screening energy consumption corresponding to different recycling processes, and the carbon emission reduction of recycled aggregates replacing natural aggregates, and the carbon balance calculation for the waste recycling stage specifically includes:
[0042] The actual operating energy consumption of the crushing and screening equipment for the first and second recycling processes is obtained, and multiplied by the corresponding carbon emission coefficients for the crushing and screening energy consumption of the first and second recycling processes to obtain the carbon emissions of the crushing process of the aggregates for the first and second recycling processes.
[0043] The total carbon emissions of the crushing stage in the first and second regeneration processes are added together with the total carbon emissions of the screening stage to obtain the total carbon emissions of the first and second regeneration processes.
[0044] By determining the proportion of recycled aggregate replacement for each recycling cycle, the carbon emission reduction after replacing the primary and secondary recycled aggregates is obtained by multiplying the total weight of the primary and secondary recycled aggregates by the carbon emission coefficient of the natural aggregates.
[0045] The total carbon emissions from the recycling process and the total carbon reductions from substitution are summed up. The carbon balance value of the waste recycling stage is obtained by subtracting the total carbon reductions from the total carbon emissions from the recycling process.
[0046] Preferably, the step of recalculating the life-cycle carbon emission value based on updated data from the basic database and dynamic monitoring parameters, and analyzing the impact of uncertainties in parameters at each stage on the calculation results using Monte Carlo simulation, specifically includes the following:
[0047] Based on database update data and dynamic monitoring parameters, carbon emission quantification data for each life cycle stage is calculated using a phased carbon emission quantification model to obtain carbon emission values.
[0048] Parameters that cause fluctuations in carbon emission results are selected from each stage of the entire life cycle. The life cycle carbon emission value is calculated based on a set number of simulations using the Monte Carlo simulation method, forming a simulation result dataset.
[0049] Sort the simulation results from smallest to largest, remove the top 2.5% and bottom 2.5% extreme values, and the remaining minimum and maximum values are the carbon emission placement confidence intervals at the 95% confidence level.
[0050] Furthermore, a dynamic carbon accounting system for the entire life cycle of coal-based solid waste road base is proposed, including:
[0051] Life cycle stage segmentation module: The module segments the entire life cycle of coal-based solid waste road base, and constructs a classification and identification system for fly ash / coal gangue / coal slime / coal-based kaolin according to its composition characteristics;
[0052] Phased carbon emission quantification module: The module establishes independent calculation units according to the life cycle stage, matches the carbon emission source inventory with the basic coefficient, and outputs quantification data for each stage;
[0053] Scene correction module: The module constructs a three-dimensional correction matrix of key scene variables and life cycle stages, and performs single-stage or multi-variable superposition correction on the quantified data;
[0054] LSTM Dynamic Prediction Module: Based on corrected operational data, the module uses a three-layer LSTM neural network to predict carbon emissions for the next 5-10 years and generate a dynamic prediction curve.
[0055] Recycled carbon balance module: The module calculates the energy consumption and carbon emissions of recycled aggregate crushing and screening, compares the carbon emission reduction of replacing natural aggregate, and outputs the carbon balance value of the waste recycling stage;
[0056] Processor: The processor is used to handle the calculation process of each formula and the construction calculation process of each model.
[0057] Compared with the prior art, the advantages of the present invention are:
[0058] The entire life cycle is meticulously divided, and a classification and identification system is established based on the composition and characteristics of coal-based solid waste, laying the foundation for accurate accounting. A basic database is built by integrating historical data and real-time monitoring to ensure comprehensive and dynamically updated data, improving accounting accuracy. A quantitative model is built in stages, coupled with a scenario-carbon emission correction matrix, which can specifically correct carbon emission data under different scenarios. LSTM neural networks can be used to predict carbon emissions in the next 5-10 years, providing a basis for long-term planning. A carbon balance equation is established considering the regeneration situation in the waste recycling stage, taking into account both environmental protection and resource recycling. Finally, Monte Carlo simulation is used to analyze parameter uncertainties and output confidence intervals, making the accounting results more valuable. Overall, this provides a scientific, dynamic, and comprehensive accounting solution for carbon management of coal-based solid waste in roadbed areas. Attached Figure Description
[0059] Figure 1 This is a schematic diagram of the method proposed in this invention;
[0060] Figure 2 This is a schematic diagram illustrating the lifecycle stage division proposed in this invention;
[0061] Figure 3 This is a schematic diagram illustrating the basic database structure proposed in this invention;
[0062] Figure 4 This is a schematic diagram illustrating the data for calculating carbon emission quantification proposed in this invention;
[0063] Figure 5 This is a schematic diagram illustrating the carbon emission quantification data correction proposed in this invention;
[0064] Figure 6 This is a schematic diagram of obtaining the dynamic carbon emission prediction curve proposed in this invention;
[0065] Figure 7 This is a schematic diagram of carbon balance calculation in the waste recycling stage proposed in this invention;
[0066] Figure 8 This is a schematic diagram illustrating the confidence interval for the carbon emission values proposed in this invention. Detailed Implementation
[0067] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.
[0068] A dynamic carbon accounting system for the entire life cycle of coal-based solid waste road base courses, including:
[0069] Life cycle stage segmentation module: The module segments the entire life cycle of coal-based solid waste road base, and constructs a classification and identification system for fly ash / coal gangue / coal slime / coal-based kaolin according to its composition characteristics;
[0070] Phased carbon emission quantification module: The module establishes independent calculation units according to the life cycle stage, matches the carbon emission source inventory with the basic coefficient, and outputs quantification data for each stage;
[0071] Scene correction module: The module constructs a three-dimensional correction matrix of key scene variables and life cycle stages, and performs single-stage or multi-variable superposition correction on the quantified data;
[0072] LSTM Dynamic Prediction Module: Based on corrected operational data, the module uses a three-layer LSTM neural network to predict carbon emissions for the next 5-10 years and generate a dynamic prediction curve.
[0073] Recycled carbon balance module: The module calculates the energy consumption and carbon emissions of recycled aggregate crushing and screening, compares the carbon emission reduction of replacing natural aggregate, and outputs the carbon balance value of the waste recycling stage;
[0074] Processor: The processor is used to handle the calculation process of each formula and the construction calculation process of each model.
[0075] See Figure 1 As shown, the dynamic carbon accounting method for the entire life cycle of coal-based solid waste road base courses includes:
[0076] Step 1: Divide the entire life cycle of coal-based solid waste road base into stages, classify coal-based solid waste into types according to its composition and characteristics, and establish a classification and identification system;
[0077] Step 2: Based on historical data, collect emission coefficients of various coal-based solid wastes at the generation stage, deploy monitoring equipment in the road base construction area and operating road section, collect dynamic parameters in each life cycle stage in real time, and form a basic database;
[0078] Step 3: Construct a phased carbon emission quantification model based on data in the basic database, and calculate and obtain carbon emission quantification data for each life cycle stage;
[0079] Step 4: Based on key scenario variables affecting carbon emissions, establish a scenario-carbon emission correction matrix to correct the quantitative carbon emission data for each life cycle stage;
[0080] Step 5: Based on the dynamic parameters collected in real time during the operation phase, predict the quantitative data of carbon emissions for the next 5-10 years through a long short-term memory neural network model, and obtain the dynamic carbon emission prediction curve.
[0081] Step Six: Based on the first and second recycling of recycled aggregates after the abandonment of coal-based solid waste road base, the carbon emission coefficients of crushing energy consumption and screening energy consumption corresponding to different recycling, and the carbon emission reduction of recycled aggregates replacing natural aggregates, establish a carbon balance equation and perform carbon balance calculations in the waste recycling stage.
[0082] Step 7: Based on the updated data in the basic database and the dynamic monitoring parameters, recalculate the carbon emission value for the entire life cycle. Using the Monte Carlo simulation method, analyze the impact of the uncertainty of parameters at each stage on the accounting results and output the confidence interval of the carbon emission value.
[0083] See Figure 2 As shown, the entire life cycle of coal-based solid waste road base is divided into stages, and coal-based solid waste is classified into types according to its composition and characteristics. A classification and identification system is established, specifically including:
[0084] The entire life cycle of the coal-based solid waste road base course covers the coal-based solid waste generation stage, pretreatment stage, transportation stage, road base course construction stage, operation and maintenance stage, and waste recycling stage.
[0085] Based on the composition and physical and mechanical properties of coal-based solid waste as the main classification criteria, coal-based solid waste is divided into fly ash, coal gangue, coal slime and coal-related kaolin.
[0086] Design unique labeling information for each type of coal-based solid waste, including the solid waste category name, core characteristic indicators, source of generation, and application scenarios, and establish a classification labeling system.
[0087] See Figure 3 As shown, based on historical data, emission coefficients of various coal-based solid wastes at their generation stages are collected. Monitoring equipment is deployed in road base construction areas and operational road sections to collect dynamic parameters in real time at each life cycle stage, forming a basic database that specifically includes:
[0088] For each stage of the entire lifecycle, corresponding coefficients and dynamic parameters are collected, including:
[0089] The primary carbon emission coefficient of associated solid waste generated during coal mining and the primary carbon emission coefficient of solid waste separated from coal washing plants during the coal-based solid waste generation stage.
[0090] Energy consumption and carbon emission coefficients of each process in the pretreatment stage of coal-based solid waste;
[0091] Carbon emission coefficient per unit mileage of transportation vehicles during the transportation of coal-based solid waste is determined by collecting data on actual vehicle load, transportation distance, and actual driving speed.
[0092] The carbon emission coefficient per man-hour of construction machinery during the road base construction stage is determined by collecting data on the actual compaction degree of the road base, the material temperature during mixing, and the actual operating time of each construction machine.
[0093] Carbon emission coefficients of various materials and energy consumption carbon emission coefficients of maintenance operations during the road base operation and maintenance phase are collected, along with daily traffic volume, real-time road surface temperature, daily rainfall, and strain data of the base structure layer.
[0094] The carbon emission coefficients of the recycling process of road base layer crushing and screening and the carbon emission coefficients of solid waste landfill disposal were collected. Data on the utilization rate of recycled aggregate after crushing of coal-based solid waste base layer and the amount of unrecycled solid waste landfilled were also collected.
[0095] The above data is integrated, categorized, and stored to form a basic database.
[0096] Specifically, during the coal-based solid waste generation stage, monthly coal mining volume and corresponding associated solid waste carbon emissions are statistically analyzed. The carbon emissions per unit of coal mining volume are calculated, and the average value over the past 12 months is used as the primary carbon emission coefficient for associated solid waste from coal mining at that coal mine. During the coal-based solid waste pretreatment stage, daily equipment operating hours and energy consumption are recorded, and the daily solid waste processed by the equipment is also statistically analyzed. During the coal-based solid waste transportation stage, vehicles are categorized by type, and fuel consumption per 100 kilometers for each type of vehicle is statistically analyzed. Combined with diesel carbon emission factors, carbon emissions per unit mileage are calculated as the carbon emission coefficient per unit mileage for the transportation vehicle. During the road base construction stage, machinery is categorized by type, and daily operating hours and corresponding energy consumption for each machine are recorded. Combined with diesel carbon emission factors, carbon emissions per unit work hour are calculated as the carbon emission coefficient per work hour for the construction machinery. During the road base operation and maintenance stage, the unit operating energy consumption of maintenance equipment is statistically analyzed, and the carbon emission coefficient for maintenance operation energy consumption is calculated by combining electricity carbon emission factors. During the road base waste recycling stage, the carbon emission coefficient for landfilling per unit mass of solid waste is obtained as the carbon emission coefficient for solid waste landfill disposal.
[0097] The system adopts a hierarchical classification structure. The top layer is divided into 6 main modules according to the life cycle stage; each main module is divided into 2 sub-modules according to data type; and each sub-module is further subdivided into data tables according to solid waste type.
[0098] See Figure 4 As shown, a phased carbon emission quantification model is constructed based on data from the basic database, and the carbon emission quantification data for each life cycle stage is calculated and obtained, specifically including:
[0099] Obtain the temporal and spatial boundaries of each life cycle stage, identify the core carbon emission sources for each stage, and form a stage-carbon emission source inventory.
[0100] Based on data in the basic database, a corresponding basic coefficient is matched for each carbon emission source, and the basic coefficients are classified and calibrated.
[0101] A phased carbon emission quantification model is constructed, which adopts a phased calculation and total output structure. An independent calculation unit is set up for each phase, and the unit includes basic coefficient retrieval, actual data input, correction coefficient calculation, and phase result output.
[0102] The carbon emission quantification data for each stage of the life cycle is obtained based on the phased carbon emission quantification model.
[0103] Specifically, the carbon emission source list is matched one by one according to the stage. The generation stage of coal mining machine fuel combustion corresponds to the generation stage carbon emission coefficient submodule - the primary carbon emission coefficient of coal mining associated solid waste in the database; the transportation stage of diesel truck emissions corresponds to the transportation stage carbon emission coefficient submodule - the carbon emission coefficient per unit mileage of transportation vehicle; if the same emission source corresponds to multiple coefficients, the coefficient is retrieved according to the actual equipment model used.
[0104] The top-level aggregated output unit of the phased carbon emission quantification model summarizes carbon emission data from six phases, outputting total life-cycle carbon emissions and the percentage of each phase. The middle layer comprises six independent phase calculation units. The bottom layer includes a basic data interface connecting to a basic database, supporting coefficient retrieval and dynamic parameter input. The formula for calculating total life-cycle carbon emissions is as follows:
[0105]
[0106] in, The total carbon emissions at a certain stage of the life cycle. Emissions from a single carbon emission source. The number of carbon emission sources within a certain life cycle stage. For the first The basic coefficient of each carbon emission source For the first Actual parameters of each carbon emission source For the first Correction factor for each carbon emission source.
[0107] See Figure 5 As shown, based on key scenario variables affecting carbon emissions, a scenario-carbon emission correction matrix is established to correct the quantitative carbon emission data for each life cycle stage. Specifically, this includes:
[0108] Based on the full life cycle characteristics of coal-based solid waste road base, key scenario variables were obtained, including the type of coal-based solid waste modifier, the slope of the transportation route, and the climate type of the operating area.
[0109] Construct a three-dimensional matrix of scenario variable combination - life cycle stage - correction coefficient. The row dimension represents all possible combinations of scenario variables, the column dimension represents the life cycle stage affected by the scenario, and the numerical dimension represents the correction coefficient of each variable combination-stage.
[0110] Carbon emission quantification data for each stage is extracted from the phased carbon emission quantification model. Single-stage correction calculations are performed using correction coefficients, and multivariate superposition corrections are applied to stages affected by multiple scenario variables to obtain corrected carbon emission quantification data.
[0111] Specifically, the multivariate superposition correction calculation is applied to stages affected by two or more scenario variables, and is applicable to the construction and operation and maintenance stages. Taking the construction stage as an example:
[0112] The construction phase is affected by both the type of modifier and the climate. When the variables have independent and cumulative effects, such as the modifier affecting mixing energy consumption and the climate affecting curing temperature energy consumption, they do not interfere with each other, and the correction coefficient is the product of the corresponding coefficients of each variable. When the variables have correlated and cumulative effects, such as in high-temperature climates where cement modifiers set faster and mixing time needs to be shortened to offset some energy consumption, the correlation coefficient correction value needs to be marked in the matrix. The corrected carbon emission data for the construction phase = the original carbon emission data for the construction phase × the cumulative correction coefficient for the construction phase. The contribution of the cumulative coefficient is broken down, and the climate influence accounts for 30%, which is consistent with the actual situation that temperate humid climates require increased curing frequency. If the contribution of a certain variable exceeds 50%, it is necessary to confirm whether the variable is the main influencing factor of the phase. For example, in cold temperate climates, the climate contribution during the construction phase exceeds 60%, and the energy consumption for antifreeze curing needs to be controlled first.
[0113] The data from the four stages of correction were compiled to form a carbon emission inventory for the target engineering scenario. The total carbon emissions before and after correction were compared, and the main variables affecting the scenario were analyzed. For example, in this case, climate type and type of modifier were the main sources of increase. Based on the correction results, an optimization scheme was proposed.
[0114] See Figure 6 As shown, based on real-time collected dynamic parameters during the operational phase, a long short-term memory neural network model is used to predict carbon emission quantification data for the next 5-10 years, obtaining a dynamic carbon emission prediction curve, specifically including:
[0115] A prediction model is constructed using a long short-term memory neural network. The neural network is trained based on the carbon emission quantification data in the historical dataset. The mean squared error is used as the loss function to measure the difference between the predicted value and the historical data.
[0116] The model consists of an input layer, a hidden layer, and an output layer. There are a total of 3 hidden layers, each containing 64 neurons, and all of them use the ReLU activation function to process the data.
[0117] Based on the corrected carbon emission quantification data during the operation phase, the data is input into the model to obtain the predicted carbon emission quantification data for the next 5-10 years.
[0118] The quantitative carbon emission forecasts for the next 5-10 years are organized into time series data based on time order, and a dynamic carbon emission forecast curve is plotted.
[0119] Specifically, since it is necessary to predict carbon emissions for the next 5-10 years, the annual data is selected as the basic time unit, and historical data and real-time parameters are all summarized into annual data. A time series dataset is constructed, and the data is organized in the format of year-corrected carbon emissions-annual traffic volume-annual average road surface temperature-annual rainfall-annual base layer strain, forming a structured data table with each row corresponding to one year.
[0120] The model uses core parameters affecting carbon emissions during the operational phase as input features, including annual traffic volume, annual average pavement temperature, annual rainfall, annual subgrade strain, and carbon emissions corrected for the previous year. The number of features in the input layer corresponds to the number of input features, and the time step is set to 3, meaning the model predicts carbon emissions for the fourth year using input features from three consecutive years. There are three hidden layers, each with 64 neurons, and the number of neurons in each layer is fixed at 64. ReLU activation is used for all layers to enhance the model's ability to fit nonlinear data and avoid traditional activation functions prone to gradient vanishing. Interlayer connections are used: the output of the first hidden layer is directly used as the input of the second hidden layer, and the output of the second layer is used as the input of the third layer. Dropout layers are added to each layer to prevent the model from over-relying on certain data. The output dimension is set to 5 or 10 to predict annual carbon emissions for the next 5-10 years. Linear activation is used because carbon emission prediction is a continuous numerical prediction, and linear activation can directly output prediction results within the actual numerical range.
[0121] The mean squared error (MSE) is used as the loss function. Its core logic is to calculate the average of the squares of the differences between the carbon emission values predicted by the model and the historical actual carbon emission values. The smaller the value, the more accurate the prediction. During the training process, the MSE of the training set and the MSE of the validation set are displayed in real time. If the MSE of the validation set no longer decreases after 5 consecutive training rounds, it means that the model has converged and training needs to be stopped.
[0122] See Figure 7As shown, based on the first and second recycling of recycled aggregates from coal-based solid waste road base courses, the carbon emission coefficients of crushing energy consumption and screening energy consumption corresponding to different recycling processes, and the carbon emission reduction of recycled aggregates replacing natural aggregates, a carbon balance equation is established to perform carbon balance calculations in the waste recycling stage. Specifically, this includes:
[0123] The actual operating energy consumption of the crushing and screening equipment for the first and second recycling processes is obtained, and multiplied by the corresponding carbon emission coefficients for the crushing and screening energy consumption of the first and second recycling processes to obtain the carbon emissions of the crushing process of the aggregates for the first and second recycling processes.
[0124] The total carbon emissions of the crushing stage in the first and second regeneration processes are added together with the total carbon emissions of the screening stage to obtain the total carbon emissions of the first and second regeneration processes.
[0125] By determining the proportion of recycled aggregate replacement for each recycling cycle, the carbon emission reduction after replacing the primary and secondary recycled aggregates is obtained by multiplying the total weight of the primary and secondary recycled aggregates by the carbon emission coefficient of the natural aggregates.
[0126] The total carbon emissions from the recycling process and the total carbon reductions from substitution are summed up. The carbon balance value of the waste recycling stage is obtained by subtracting the total carbon reductions from the total carbon emissions from the recycling process.
[0127] See Figure 8 As shown, based on updated data from the basic database and dynamically monitored parameters, the life-cycle carbon emissions are recalculated. The Monte Carlo simulation method is used to analyze the impact of parameter uncertainties at each stage on the calculation results. The confidence intervals for the output carbon emissions values specifically include:
[0128] Based on database update data and dynamic monitoring parameters, carbon emission quantification data for each life cycle stage is calculated using a phased carbon emission quantification model to obtain carbon emission values.
[0129] Parameters that cause fluctuations in carbon emission results are selected from each stage of the entire life cycle. The life cycle carbon emission value is calculated based on a set number of simulations using the Monte Carlo simulation method, forming a simulation result dataset.
[0130] Sort the simulation results from smallest to largest, remove the top 2.5% and bottom 2.5% extreme values, and the remaining minimum and maximum values are the carbon emission placement confidence intervals at the 95% confidence level.
[0131] Specifically, the phased recalculation process follows the independent calculation unit structure of the phased carbon emission quantification model. Each unit calls the updated basic data and completes the recalculation of the remaining stages in sequence, ensuring that each stage uses the updated coefficients and the latest dynamic parameters, and avoiding the mixing of old and new data.
[0132] The carbon emission values for each stage are summarized after recalculation to obtain the updated total carbon emissions for the entire life cycle; the differences between the total carbon emissions before and after the update are compared and the reasons for the differences are analyzed, such as the reduction in transportation distance and the reduction in equipment energy consumption, to form the updated comparison results of carbon emissions for the entire life cycle;
[0133] When the key fluctuation parameters for the entire life cycle are selected, parameters whose carbon emission fluctuation exceeds 5% in the corresponding stage when the parameter fluctuation is 10% are selected; parameters that are fixed or have very small fluctuations are excluded, and parameters that are greatly affected by the environment, process, and operation are given priority; parameters that are supported by historical fluctuation data are selected.
[0134] For each key fluctuation parameter, a random value is generated according to its probability distribution in a single Monte Carlo simulation operation. All random values of parameters in this simulation are substituted into the phased carbon emission quantification model to recalculate the carbon emissions for each phase. The carbon emissions for each phase are summed to obtain the total life-cycle carbon emission value for one simulation. The single simulation process is repeated 1000 times, and the simulation number and the corresponding total life-cycle carbon emission value are recorded each time to form a simulation result dataset containing 1000 data points.
[0135] Arrange the 1000 simulation results in ascending order, calculate the number of extreme values to be removed, remove the top 25 minimum values and the bottom 25 maximum values after sorting, and select the minimum and maximum values from the remaining 950 data points. This range is the full life cycle carbon emission placement confidence interval at a 95% confidence level.
[0136] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.
[0137] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.
[0138] 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 principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A dynamic carbon accounting method for the entire life cycle of coal-based solid waste road base courses, characterized in that, include: The entire life cycle of coal-based solid waste road base is divided into stages, and coal-based solid waste is classified into types according to its composition and characteristics, and a classification and identification system is established. Based on historical data, the emission coefficients of various coal-based solid wastes at the generation stage are collected. Monitoring equipment is deployed in the road base construction area and operating road section to collect dynamic parameters in each life cycle stage in real time, forming a basic database. A phased carbon emission quantification model is constructed based on data from the basic database, and carbon emission quantification data for each life cycle stage is calculated and obtained respectively. Based on key scenario variables affecting carbon emissions, a scenario-carbon emission correction matrix is established to correct the carbon emission quantification data at each life cycle stage. Specifically, based on the full life cycle characteristics of coal-based solid waste road base, key scenario variables are obtained, including the type of coal-based solid waste modifier, the slope of the transportation route, and the climate type of the operating area. A three-dimensional matrix of scenario variable combination-life cycle stage-correction coefficient is constructed, with the row dimension representing all possible combinations of scenario variables, the column dimension representing the life cycle stage affected by the scenario, and the numerical dimension representing the correction coefficient of each variable combination-stage. Carbon emission quantification data for each stage is extracted from the phased carbon emission quantification model, and single-stage correction calculation is performed in combination with the correction coefficients. Multi-variable superposition correction is also performed on stages affected by multiple scenario variables to obtain the corrected carbon emission quantification data. Based on real-time collected dynamic parameters during the operation phase, a long short-term memory neural network model is used to predict carbon emission quantification data for the next 5-10 years and obtain dynamic carbon emission prediction curves. Based on the primary and secondary recycling of recycled aggregates from coal-based solid waste road base courses, the carbon emission coefficients of crushing and screening energy consumption corresponding to different recycling processes, and the carbon emission reduction of recycled aggregates replacing natural aggregates, a carbon balance equation is established to calculate the carbon balance of the waste recycling stage. Specifically, the actual operating energy consumption of crushing and screening equipment in primary and secondary recycling is obtained and multiplied by the corresponding carbon emission coefficients of crushing and screening energy consumption in primary and secondary recycling to obtain the carbon emissions of the crushing process of primary and secondary recycled aggregates. The total carbon emissions of the crushing process in primary and secondary recycling are added to the total carbon emissions of the screening process to obtain the total carbon emissions of the primary and secondary recycling processes. By determining the recycled aggregate replacement ratio for each recycling cycle, the total weight of primary and secondary recycled aggregates is multiplied by the carbon emission coefficient of natural aggregates to obtain the carbon emission reduction after replacement by primary and secondary recycled aggregates. The total carbon emissions of the recycling process and the total carbon emission reduction of replacement are summarized, and the carbon balance value of the waste recycling stage is obtained by subtracting the total carbon emission reduction of replacement from the total carbon emissions of the recycling process. Based on updated data from the basic database and dynamically monitored parameters, the life-cycle carbon emissions are recalculated. The Monte Carlo simulation method is used to analyze the impact of the uncertainty of parameters at each stage on the accounting results and output the confidence interval of the carbon emissions.
2. The method for dynamic carbon accounting of coal-based solid waste road base courses throughout their entire life cycle, as described in claim 1, is characterized in that... The process of dividing the entire life cycle of coal-based solid waste road base into stages, classifying coal-based solid waste according to its composition and characteristics, and establishing a classification and identification system specifically includes: The entire life cycle of the coal-based solid waste road base course covers the coal-based solid waste generation stage, pretreatment stage, transportation stage, road base course construction stage, operation and maintenance stage, and waste recycling stage. Based on the composition and physical and mechanical properties of coal-based solid waste as the main classification criteria, coal-based solid waste is divided into fly ash, coal gangue, coal slime and coal-related kaolin. Design unique labeling information for each type of coal-based solid waste, including the solid waste category name, core characteristic indicators, source of generation, and application scenarios, and establish a classification labeling system.
3. The method for dynamic carbon accounting of coal-based solid waste road base courses throughout their entire life cycle, as described in claim 1, is characterized in that... The aforementioned method involves collecting emission coefficients of various coal-based solid wastes at their generation stages based on historical data, deploying monitoring equipment in road base construction areas and operational road sections, and collecting dynamic parameters in real time at each life cycle stage to form a basic database. This database specifically includes: For each stage of the entire lifecycle, corresponding coefficients and dynamic parameters are collected, including: The primary carbon emission coefficient of associated solid waste generated during coal mining and the primary carbon emission coefficient of solid waste separated from coal washing plants during the coal-based solid waste generation stage. Energy consumption and carbon emission coefficients of each process in the pretreatment stage of coal-based solid waste; Carbon emission coefficient per unit mileage of transportation vehicles during the transportation of coal-based solid waste is determined by collecting data on actual vehicle load, transportation distance, and actual driving speed. The carbon emission coefficient per man-hour of construction machinery during the road base construction stage is determined by collecting data on the actual compaction degree of the road base, the material temperature during mixing, and the actual operating time of each construction machine. Carbon emission coefficients of various materials and energy consumption carbon emission coefficients of maintenance operations during the road base operation and maintenance phase are collected, along with daily traffic volume, real-time road surface temperature, daily rainfall, and strain data of the base structure layer. The carbon emission coefficients of the recycling process of road base layer crushing and screening and the carbon emission coefficients of solid waste landfill disposal were collected. Data on the utilization rate of recycled aggregate after crushing of coal-based solid waste base layer and the amount of unrecycled solid waste landfilled were also collected. The above data is integrated, categorized, and stored to form a basic database.
4. The method for dynamic carbon accounting of coal-based solid waste road base courses throughout their entire life cycle, as described in claim 1, is characterized in that... The step of constructing a phased carbon emission quantification model based on data from the basic database, and calculating and obtaining carbon emission quantification data for each life cycle stage, specifically includes: Obtain the temporal and spatial boundaries of each life cycle stage, identify the core carbon emission sources for each stage, and form a stage-carbon emission source inventory. Based on data in the basic database, a corresponding basic coefficient is matched for each carbon emission source, and the basic coefficients are classified and calibrated. A phased carbon emission quantification model is constructed, which adopts a phased calculation and total output structure. An independent calculation unit is set up for each phase, and the unit includes basic coefficient retrieval, actual data input, correction coefficient calculation, and phase result output. The carbon emission quantification data for each stage of the life cycle is obtained based on the phased carbon emission quantification model.
5. The method for dynamic carbon accounting of coal-based solid waste road base courses throughout their entire life cycle, as described in claim 1, is characterized in that... The process of using real-time collected dynamic parameters from the operational phase, and employing a long short-term memory neural network model to predict carbon emission quantification data for the next 5-10 years to obtain a dynamic carbon emission prediction curve specifically includes: A prediction model is constructed using a long short-term memory neural network. The neural network is trained based on the carbon emission quantification data in the historical dataset. The mean squared error is used as the loss function to measure the difference between the predicted value and the historical data. The model consists of an input layer, a hidden layer, and an output layer. There are a total of 3 hidden layers, each containing 64 neurons, and all of them use the ReLU activation function to process the data. Based on the corrected carbon emission quantification data during the operation phase, the data is input into the model to obtain the predicted carbon emission quantification data for the next 5-10 years. The quantitative carbon emission forecasts for the next 5-10 years are organized into time series data based on time order, and a dynamic carbon emission forecast curve is plotted.
6. The method for dynamic carbon accounting of coal-based solid waste road base courses throughout their entire life cycle, as described in claim 1, is characterized in that... The process involves recalculating the life-cycle carbon emissions based on updated data from the basic database and dynamic monitoring parameters. Using Monte Carlo simulation, the impact of uncertainties in parameters at each stage on the calculation results is analyzed, and the confidence intervals for the output carbon emissions values are specifically included: Based on database update data and dynamic monitoring parameters, carbon emission quantification data for each life cycle stage is calculated using a phased carbon emission quantification model to obtain carbon emission values. Parameters that cause fluctuations in carbon emission results are selected from each stage of the entire life cycle. The life cycle carbon emission value is calculated based on a set number of simulations using the Monte Carlo simulation method, forming a simulation result dataset. Sort the simulation results from smallest to largest, remove the top 2.5% and bottom 2.5% extreme values, and the remaining minimum and maximum values are the carbon emission placement confidence intervals at the 95% confidence level.
7. A dynamic carbon accounting system for the entire life cycle of coal-based solid waste road base courses, used to implement the dynamic carbon accounting method for the entire life cycle of coal-based solid waste road base courses as described in any one of claims 1-6, characterized in that, include: Life cycle stage segmentation module: Divide the entire life cycle of coal-based solid waste road base into stages, and construct a classification and identification system for fly ash / coal gangue / coal slime / coal-based kaolin according to composition characteristics; Phased carbon emission quantification module: Establishes independent calculation units according to life cycle stages, matches carbon emission source inventory with basic coefficients, and outputs quantified data for each stage; Scene correction module: Constructs a three-dimensional correction matrix of key scene variables and life cycle stages, and performs single-stage or multi-variable superposition correction on quantitative data; LSTM dynamic prediction module: Based on corrected operational data, it predicts carbon emissions for the next 5-10 years using a three-layer LSTM neural network and generates dynamic prediction curves; Recycled Carbon Balance Module: Calculates the energy consumption and carbon emissions of recycled aggregate crushing and screening, compares the carbon emission reduction of replacing natural aggregate, and outputs the carbon balance value of the waste recycling stage; Processor: The processor is used to handle the calculation process of each formula and the construction calculation process of each model.