A method and system for identifying and apportioning the responsibility of a subject carbon emission boundary based on copula-kernel density estimation

Abstract

The application discloses a source-load-storage integrated main body carbon emission boundary identification and responsibility allocation method and system based on a Copula-nuclear density estimation, relates to the technical field of low-carbon operation of a power system and carbon emission management of a comprehensive energy system, and realizes accurate calculation of full-link carbon emission of a source-load-storage integrated main body, non-overlapping and non-vacuum division of a responsibility boundary, fair allocation of multi-main-body carbon emission responsibility and effective incentive of low-carbon behavior, and provides accurate, reliable and engineering-landable core technical support for user-side carbon asset management, landing of a low-carbon incentive policy and source-network-load-storage collaborative low-carbon operation of a new-type power system.

Description

Technical Field

[0001] This invention relates to the fields of low-carbon operation of power systems and carbon emission accounting and management of integrated energy systems, specifically to a method for identifying the carbon emission boundaries and allocating responsibilities of integrated source-load-storage entities based on Copula-kernel density estimation. Background Technology

[0002] With the deepening of restrictions and controls on greenhouse gas emissions, the integrated source-load-storage system has become the core carrier for the low-carbon transformation of the user side of the new power system. This system is formed by the deep coupling of distributed renewable energy (wind / solar), electrochemical energy storage, and flexible loads. It has the core characteristics of multivariate nonlinear correlation, bidirectional flow of electricity, and cross-time transfer of carbon responsibility. Through the coordinated operation of source-storage-load, it achieves efficient energy utilization and low-carbon emission reduction.

[0003] Dynamic carbon emission factors are the core signals for characterizing the carbon intensity of electricity consumption behavior and allocating carbon emission responsibilities. Accurate boundary identification and fair responsibility allocation are crucial foundations for achieving user-side carbon management. However, existing technologies still have several core shortcomings when dealing with the complex operational characteristics of integrated energy source-load-storage systems, mainly in the following aspects: 1) Insufficient accuracy in carbon emission boundary identification. Existing methods mostly adopt a one-way traceability mode, which cannot adapt to the bidirectional flow characteristics of electricity in integrated energy source-load-storage systems, leading to overlapping, vacuum, or mismatched carbon emission responsibilities, making it difficult to meet the needs of refined carbon accounting; 2) Severely insufficient adaptability of carbon factors. Traditional static carbon emission factors cannot adapt to the spatiotemporal dynamics of grid carbon factors, fluctuations in the proportion of green electricity, and the nonlinear coupling of multiple factors in energy storage charging and discharging behavior. In particular, existing methods cannot achieve charging and discharging due to the cross-time period transfer characteristics of energy storage carbon responsibility. 3) The accurate source matching of carbon factors in different time periods leads to a serious disconnect between carbon accounting results and actual carbon emissions; 4) The responsibility sharing mechanism is unbalanced. Existing sharing methods either ignore the incentive effect of low-carbon behavior based solely on physical power paths, or rely solely on behavior evaluation without the fairness support of physical power flow constraints, resulting in a lack of credibility in the sharing results and making it difficult to implement; 5) The integration of data-driven and physical constraints is low. Existing methods mostly rely on preset physical models or single statistical methods, which cannot simultaneously characterize the nonlinear correlation of multiple factors and the physical conservation constraints of the power system. They either rely too much on empirical parameters and have poor generalization ability, or deviate from physical rules and do not conform to engineering practice, making it difficult to achieve large-scale application.

[0004] Therefore, this invention proposes a method for identifying the carbon emission boundary and allocating responsibility for integrated source-load-storage entities based on Copula-kernel density estimation. The aim is to achieve accurate identification of the carbon emission boundary and fair allocation of responsibility for integrated source-load-storage entities by integrating physical power flow constraints of the power system with multi-factor data-driven modeling, thereby solving the core pain points of existing technologies and improving the accuracy and engineering practicality of user-side carbon accounting. Summary of the Invention

[0005] The purpose of this invention is to address the core shortcomings of existing integrated source-load-storage carbon emission accounting and responsibility allocation technologies, such as insufficient accuracy in identifying carbon emission boundaries, poor adaptability of dynamic carbon factors, imbalance between fairness and incentives in responsibility allocation mechanisms, and low integration of data-driven modeling with power system physical constraints. This invention provides a Copula-kernel density estimation-based method for identifying carbon emission boundaries and allocating responsibility for integrated source-load-storage entities. This method achieves accurate carbon emission accounting across the entire chain of integrated source-load-storage entities, non-overlapping and vacuum-free boundary delineation of responsibility, fair allocation of carbon emission responsibilities among multiple entities, and effective incentives for low-carbon behavior. It provides accurate, reliable, and engineering-feasible core technical support for user-side carbon asset management, implementation of low-carbon incentive policies, and collaborative low-carbon operation of new power systems.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A method for identifying the carbon emission boundary and allocating responsibility for integrated source-load-storage entities based on Copula-kernel density estimation includes the following steps:

[0008] Step 1: Construct a standardized data acquisition system and physical constraint framework for the integrated source-load-storage system;

[0009] First, the core operational characteristics of the main body are extracted; then, a standardized metering point system covering the entire link is configured; finally, node power conservation constraints are established, and the collection and verification of time-series operational data within the accounting cycle are completed.

[0010] Step 2: Achieve precise delineation of the carbon emission responsibility boundaries for integrated source-load-storage entities;

[0011] First, a directed weighted topology model integrating physical characteristics and data correlation characteristics is constructed. Then, the power flow tracing method of proportional allocation is used to complete the source tracing of the entire link power path. Finally, the source tracing results are corrected through Copula correlation analysis, and carbon emission responsibility boundary units with no overlap and no vacuum are divided.

[0012] Step 3: Generate a dynamic carbon emission factor that adapts to the source-load-storage coupling characteristics;

[0013] First, the time series data of the core influencing factors of the carbon factor are extracted and preprocessed. Then, the joint distribution model of multiple factors is completed by kernel density estimation and Copula function. Finally, the dynamic carbon factor is generated, and the time series source matching of the carbon factor of energy storage charging and discharging is completed.

[0014] Step 4: Complete the comprehensive allocation of carbon emission responsibilities among multiple stakeholders;

[0015] First, the basic allocation coefficient of electricity path is calculated based on the power flow tracking results. Then, a comprehensive index of low-carbon behavior of the main body is constructed and the behavior correction coefficient is calculated. Finally, the carbon emission responsibility of each main body is calculated by combining the dynamic carbon factor, and the total carbon emission conservation constraint is met throughout the process.

[0016] Step 5: Perform full-process verification of the calculation results and model performance;

[0017] First, the consistency of the total amount of carbon is verified through carbon conservation verification. Then, the accuracy of data-driven modeling is verified through correlation restoration verification. Finally, the effectiveness of the low-carbon incentive mechanism is verified through incentive effectiveness verification.

[0018] Preferably, in step 1, the processing is as follows:

[0019] Step 11: Extract the three core operational characteristics of the integrated source-load-storage system: physical operation, electricity-carbon coupling, and data correlation, and clarify the core requirements for data acquisition and modeling;

[0020] Step 12: Configure a standardized metering point system covering the source side, storage side, load side, grid connection interface, and common branch, and complete the standardized collection of full-link time-series operation data within the accounting cycle;

[0021] Step 13: Establish node power conservation constraints, verify the rationality of the collected data, and provide a physical constraint basis for subsequent calculations.

[0022] Preferably, in step 2, the processing is as follows:

[0023] Step 21: Construct a directed weighted topology model that integrates physical features and data correlation features to provide a topological foundation for power path tracing;

[0024] Step 22: Use the proportional allocation method for power flow tracing, following the core logic of "carbon follows electricity and is split according to power ratio", to complete the full-link power path tracing and clarify the power source composition of each branch;

[0025] Step 23: Calculate the correlation coefficient between branch power and core influencing factors using the Copula function, correct the power flow tracking results, and divide the carbon emission responsibility boundaries into three categories: direct responsibility units, indirect responsibility units, and common responsibility units.

[0026] Preferably, in step 3, the processing is as follows:

[0027] Step 31: Extract time-series data of the three core influencing factors of carbon factors: spatiotemporal dynamics, power attributes, and energy storage behavior, and complete data cleaning and standardization preprocessing;

[0028] Step 32: Fit the marginal distribution of each factor by kernel density estimation, and use the Copula function to characterize the nonlinear correlation between multiple factors to complete the joint distribution modeling of multiple factors;

[0029] Step 33: Calculate the correlation coordination correction coefficient, generate a dynamic carbon factor that adapts to the source-load-storage coupling characteristics, and complete the time-series source matching of energy storage charging and discharging carbon factors to solve the problem of cross-time period transfer of energy storage carbon responsibility.

[0030] Preferably, in step 4, the processing is as follows:

[0031] Step 41: Based on the power path results of power flow tracking, calculate the effective power consumption of each entity, calculate the basic power path allocation coefficient, and ensure the fairness of allocation;

[0032] Step 42: Construct a comprehensive index of main low-carbon behaviors that includes demand response participation and energy storage low-carbon charging and discharging matching degree, calculate behavior correction coefficients, and achieve effective incentives for low-carbon behaviors;

[0033] Step 43: Combining dynamic carbon factors and comprehensive allocation weights, complete the accounting of direct carbon emission responsibilities and public carbon emission responsibilities of each entity, and fully meet the total carbon emission conservation constraints.

[0034] Preferably, in step 5, the processing is as follows:

[0035] Step 51: Verify the consistency between the total carbon emission responsibility and the sum of the responsibilities shared by each entity through carbon conservation verification to ensure that the accounting results are unbiased;

[0036] Step 52: Verify the model's accuracy in restoring the nonlinear correlations of multiple factors through correlation restoration verification, ensuring the accuracy of data-driven modeling;

[0037] Step 53: Verify the effectiveness of incentives to check the responsibility reduction effect of low-carbon behavior subjects and ensure that the low-carbon incentive mechanism is effectively implemented.

[0038] A source-load-storage integrated entity carbon emission boundary identification and responsibility allocation system based on Copula-kernel density estimation includes: a source-load-storage integrated entity operation characteristic analysis and metering point configuration module; a carbon emission boundary identification module based on power flow tracking + Copula correlation; a dynamic carbon factor adaptation module based on Copula-kernel density estimation; a carbon emission responsibility allocation module based on power path + entity behavior; and a closed-loop verification module, wherein:

[0039] The module for analyzing the operational characteristics of the integrated source-load-storage system and configuring metering points is used to construct a standardized data acquisition system and physical constraint framework for the integrated source-load-storage system.

[0040] A carbon emission boundary identification module based on current tracing and Copula correlation is used to accurately delineate the carbon emission responsibility boundaries of integrated source-load-storage entities.

[0041] A dynamic carbon factor adaptation module based on Copula-kernel density estimation is used to generate dynamic carbon emission factors that adapt to source-load-storage coupling characteristics.

[0042] The carbon emission responsibility allocation module based on electricity path and subject behavior is used to complete the comprehensive allocation of carbon emission responsibilities among multiple subjects.

[0043] The closed-loop verification module is used to verify the calculation results and model performance throughout the entire process.

[0044] Preferably, the integrated source-load-storage main operation characteristic analysis and metering point configuration module is specifically used for:

[0045] The three core operational characteristics of the integrated source-load-storage system—physical operation, electro-carbon coupling, and data correlation—are extracted, and the core requirements for data acquisition and modeling are clarified.

[0046] Configure a standardized metering point system covering the source side, storage side, load side, grid connection interface, and common branch, and complete the standardized collection of time-series operation data of the entire link within the accounting cycle;

[0047] Establish node power conservation constraints, verify the rationality of the collected data, and provide a physical constraint basis for subsequent calculations.

[0048] Preferred: The carbon emission boundary identification module based on tidal current tracking and Copula correlation is specifically used for:

[0049] Construct a directed weighted topology model that integrates physical characteristics and data correlation characteristics to provide a topological foundation for power path tracing;

[0050] The proportional allocation method is used for power flow tracing, following the logic of carbon following electricity and splitting according to power ratio, to complete the traceability of the entire power path and clarify the power source composition of each branch.

[0051] The correlation coefficient between branch power and core influencing factors is calculated using the Copula function, and the power flow tracking results are corrected to divide the carbon emission responsibility boundaries into three categories: direct responsibility units, indirect responsibility units, and public responsibility units.

[0052] Preferred: The dynamic carbon factor adaptation module based on Copula-kernel density estimation is specifically used for:

[0053] Extract time-series data of the three core influencing factors of carbon factors: spatiotemporal dynamics, power attributes, and energy storage behavior, and complete data cleaning and standardization preprocessing.

[0054] By fitting the marginal distributions of each factor through kernel density estimation, and using the Copula function to characterize the nonlinear correlations among multiple factors, the joint distribution model of multiple factors is completed.

[0055] Calculate the correlation-coordinated correction coefficient, generate a dynamic carbon factor that adapts to the source-load-storage coupling characteristics, and complete the time-series source matching of the energy storage charging and discharging carbon factor.

[0056] Preferred: The carbon emission responsibility sharing module based on electricity path + subject behavior is specifically used for:

[0057] Based on the power path results from power flow tracking, the effective power consumption of each entity is calculated, and the basic power path allocation coefficient is calculated.

[0058] Construct a comprehensive index of main low-carbon behaviors covering demand response participation and energy storage low-carbon charging and discharging matching degree, and calculate behavior correction coefficients;

[0059] By combining dynamic carbon factors and comprehensive allocation weights, the direct carbon emission responsibilities and public carbon emission responsibilities of each entity are calculated, and the total carbon emission conservation constraint is met throughout the process.

[0060] The closed-loop verification module is specifically used for:

[0061] The carbon conservation check verifies the consistency between the total carbon emission responsibility and the sum of the responsibilities shared by each entity.

[0062] The correlation restoration test verifies the model's accuracy in restoring the nonlinear correlations of multiple factors.

[0063] By validating the effectiveness of incentives, we can verify the impact of reducing the responsibility of low-carbon actors.

[0064] Compared with existing technologies, this invention has the following significant advantages: Addressing the problems of insufficient boundary identification accuracy, poor carbon factor adaptability, imbalanced responsibility allocation mechanisms, and low integration of data-driven and physical constraints in existing dynamic carbon factor accounting and responsibility allocation methods, this invention proposes a source-load-storage integrated entity carbon emission boundary identification and responsibility allocation method based on Copula-kernel density estimation. By constructing a standardized metrology system and physical constraint framework, the reliability and compliance of carbon accounting data are ensured at the fundamental level. By introducing a boundary identification method based on power flow tracking and Copula correlation correction, the limitations of traditional unidirectional traceability are overcome, accurately solving the responsibility mismatch problem under bidirectional power flow. By designing a dynamic carbon factor adaptation mechanism based on Copula-kernel density estimation, accurate carbon factor adaptation and time-series traceability of energy storage carbon responsibility under multi-factor coupling are achieved. By establishing a two-dimensional responsibility allocation mechanism of power path + entity behavior, the fairness of allocation and low-carbon incentives are balanced, effectively guiding the coordinated low-carbon operation of source-storage-load. Through a three-layer rigid closed-loop verification system, the rigor of the accounting results and engineering feasibility are guaranteed. This series of technological innovations enables the present invention to achieve accurate carbon emission accounting, clear boundary delineation, and fair allocation of responsibility under the complex operating conditions of the integrated power generation, load, and storage system, providing strong technical support for the low-carbon transformation of the user side and the construction of new power systems. Attached Figure Description

[0065] Figure 1 This is a system architecture diagram of the present invention;

[0066] Figure 2 This is a flowchart of a method for identifying the carbon emission boundary and allocating responsibility for an integrated source-load-storage entity based on Copula-kernel density estimation. Detailed Implementation

[0067] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Specific implementation method one:

[0069] Referring to Figure 1, the present invention provides an integrated source-load-storage carbon emission boundary identification and responsibility allocation system based on Copula-kernel density estimation, which includes five core modules connected in sequence. The modules communicate with each other to form a complete carbon accounting and responsibility allocation closed loop.

[0070] Furthermore, Figure 2The flowchart below shows the steps of the source-load-storage integrated entity carbon emission boundary identification and responsibility allocation method based on Copula-kernel density estimation. The specific implementation process is as follows:

[0071] Step 1: Analysis of the Main Operational Characteristics of the Integrated Source-Load-Storage System and Configuration of Metering Points

[0072] This module is responsible for building a standardized data acquisition system and physical constraint framework for the integrated source-load-storage system, extracting the core operating characteristics of the system, configuring a standardized metering point system, establishing node power conservation constraints, and completing the acquisition and verification of time-series operating data within the accounting cycle.

[0073] The specific implementation process is as follows:

[0074] Step 11: Extraction of Core Operational Features

[0075] For integrated energy source, load, and storage systems, three core operational characteristics are identified:

[0076] Physical operation characteristics: The core manifestation is the bidirectional flow of electricity, that is, the power of the grid connection port, source-storage, and storage-load branch frequently switch in both directions; the source-storage-load time sequence is strongly coupled, that is, the time sequence mismatch of the photovoltaic clock output curve, the wind power night peak output, and the load double peak characteristics, which requires dynamic tracking of the electricity path and adaptation to time sequence data modeling.

[0077] Electricity-carbon coupling characteristics: The core manifestation is the mixed power attributes, that is, the coexistence of self-generated and self-consumed green electricity and supplementary power supply from the grid; the carbon responsibility of energy storage is transferred across time periods, that is, carbon factors are locked during charging and carbon responsibility is released during discharging, requiring multiple factors to coordinate and adapt to carbon factors and trace cross-time period responsibility.

[0078] Data correlation characteristics: The core manifestation is that the spatiotemporal dynamics, power source attributes, and energy storage behavior are nonlinearly correlated, and the synergistic effect of multiple factors needs to be characterized by the Copula function.

[0079] Step 12: Configuration of Standardized Metering Point System and Data Acquisition

[0080] A standardized metering point system is configured for the five core nodes: source, storage, load, grid connection interface, and common branch, to collect time-series operational data within the accounting period.

[0081] Source-side metering points: configured at the output of distributed photovoltaic and wind power grid-connected inverters to collect 15-minute-level power generation and curtailment rate data for power attribute quantification and edge distribution fitting;

[0082] Storage-side metering point: Configured at the AC / DC side interface of the energy storage PCS, it collects 15-minute-level charge and discharge capacity and SOC data for energy storage behavior modeling and carbon responsibility transfer tracing;

[0083] Load-side metering point: Located at the inlet of flexible load, it collects 15-minute-level electricity consumption and demand response reduction data for the purpose of quantifying the main behavior and determining responsibility.

[0084] Grid connection point metering: Configured at the grid connection point between the main body and the public power grid, it collects bidirectional electricity consumption data at the 15-minute level and real-time marginal carbon factor data of the power grid for spatiotemporal dynamic characterization and total carbon responsibility accounting;

[0085] Public branch line metering points: These are located at the load points of public lines and public facilities to collect hourly network loss and public electricity consumption data, which are used for the division of public responsibility units.

[0086] Step 13: Constructing Node Power Conservation Constraints

[0087] For the five core nodes, a node power conservation constraint is established to verify the rationality of the collected time-series data, remove outlier data, and ensure the convergence and accuracy of subsequent calculations. The constraint formula is as follows:

[0088]

[0089] In the formula, / Let n be the set of inflow / outflow branches. Let n be the self-generated power of node n (only the source node is non-zero). For the power loss of node n, Let n be the load power of node n.

[0090] Step 2: Carbon emission boundary identification module based on tidal current tracking and Copula correlation

[0091] This module is responsible for accurately delineating the carbon emission responsibility boundaries of the integrated source-load-storage system, constructing a directed weighted topology model that integrates physical characteristics and data correlation characteristics, using the proportional allocation method for power flow tracing to complete the source tracing of the entire power path, correcting the source tracing results through Copula correlation analysis, and dividing three types of carbon emission responsibility boundary units that are non-overlapping and non-vacuum.

[0092] The specific implementation process is as follows:

[0093] Step 21: Construction of Directed Weighted Topology Model

[0094] Construct a directed weighted topology model G=(N,L,W) for the integrated source-load-storage entity, where:

[0095] Node set N: covers five core nodes: source node, storage node, load node, parallel gateway interface node, and common branch node;

[0096] Branch set L: The electrical connection between corresponding nodes. The branch direction is consistent with the real-time power flow direction and is dynamically updated as power flows bidirectionally.

[0097] The weight set W is the product of the branch power ratio and the Copula correlation coefficient, which integrates the physical characteristics of the power system and the correlation characteristics of multi-factor data.

[0098] Step 22: Power flow tracing and energy path tracing using the proportional allocation method

[0099] Power flow tracing is achieved using a proportional allocation method, following the core logic of "carbon follows electricity and is split according to power ratio," to realize full-link power path tracing. The specific process is as follows:

[0100] Determine the starting point for tracing the source: take the grid-connected wind and solar power nodes (green electricity inlet) and grid connection outlet nodes (mains power inlet) as the core power source starting points;

[0101] Node-by-node power allocation: For each node, the three types of electricity flowing into the node—green electricity, mains electricity, and energy storage discharge—are simultaneously allocated to each outflow branch according to the power ratio of the outflow branch.

[0102] Power balance constraint: During the splitting process, the node power conservation constraint is strictly followed to ensure that the sum of the electrical components of each branch is equal to the total electrical quantity of the branch;

[0103] Source tracing results output: Finally, the green electricity ratio, grid electricity ratio, and energy storage discharge ratio of each branch and each time period within the accounting period are obtained, clarifying the physical flow and source composition of electricity.

[0104] The formula for splitting the electricity percentage of any node n, any inflow branch lin, and any outflow branch lout is as follows:

[0105]

[0106] In the formula, branch road The proportion of the k-th type of electricity, where k=1 is green electricity, k=2 is mains electricity, and k=3 is energy storage discharge. branch road Real-time power.

[0107] Step 23: Copula Correlation Correction and Boundary Delineation

[0108] The Kendall rank correlation coefficient τ between the power consumption of each branch and the three core influencing factors is calculated using the Copula function. The power consumption ratio obtained from power flow tracing is adjusted based on the magnitude of the coefficient to avoid responsibility mismatch caused by relying solely on physical paths. A correlation threshold is set. , The carbon emission responsibility boundary units are divided into three categories with no overlap and no vacuum:

[0109] Directly responsible unit: The criteria for determination are unique path and correlation coefficient. The responsibility attribution rule is that it is directly attributed to the corresponding load-bearing entity;

[0110] Indirect responsibility unit: The judgment criterion is that it contains an energy storage charging and discharging path and The responsibility for disposal is transferred to the final consumer as the charging and discharging sequence progresses;

[0111] Public responsibility unit: The determination criteria are the corresponding public branch / network loss and the correlation coefficient. The responsibility allocation rule is based on the proportion of electricity consumption by the beneficiary.

[0112] Step 3: Dynamic carbon factor adaptation module based on Copula-kernel density estimation

[0113] This module is responsible for generating dynamic carbon emission factors that adapt to the source-load-storage coupling characteristics. It extracts time-series data of three core influencing factors: spatiotemporal dynamics, power source attributes, and energy storage behavior. It fits the marginal distribution of each factor through kernel density estimation, constructs a multi-factor joint distribution model using the Copula function, generates dynamic carbon factors, and completes the time-series source matching of energy storage charging and discharging carbon factors.

[0114] The specific implementation process is as follows:

[0115] Step 31: Preprocessing of core factor data

[0116] Extract 15-minute time-series data within the accounting period T to construct sequences of three core influencing factors:

[0117] Spatiotemporal dynamic factor X1: Real-time marginal carbon factor sequence of the power grid, reflecting the spatiotemporal dynamic characteristics of the carbon factor;

[0118] Power Attribute Factor X2: Real-time green electricity ratio sequence of the integrated source-load-storage entity, reflecting the dynamic changes in the power structure;

[0119] Energy Storage Behavior Factor X3: Real-time net charge and discharge value sequence of energy storage, with charging being positive and discharging being negative, reflecting the charging and discharging behavior characteristics of energy storage.

[0120] Data cleaning, outlier removal, and standardization were completed to provide a high-quality data foundation for subsequent distribution fitting.

[0121] Step 32: Multi-factor joint distribution modeling

[0122] Kernel density estimation and marginal distribution fitting: A Gaussian kernel function is used, and the optimal bandwidth is automatically calculated using the Silverman rule to fit the marginal distributions of the three core factors. No pre-defined distribution type is required, accurately capturing the inherent distribution characteristics of the data. The kernel density estimation formula is as follows:

[0123]

[0124] In the formula, The bandwidth is automatically calculated using Silverman's rule, with k=1, 2, and 3 corresponding to the three core factors. Let N be the Gaussian kernel function, and N be the number of samples.

[0125] Copula Joint Distribution Modeling: A joint distribution model of the three core factors is constructed using the AMH-Copula function. The relevant parameters θ of the Copula function are solved using the maximum likelihood estimation method to accurately characterize the nonlinear correlation among the three factors. The AMH-Copula function formula is as follows:

[0126]

[0127] In the formula, For the AMH-Copula connection function, F Xk For factor X k The cumulative distribution function is given by θ, which is a parameter of the Copula function and takes values ​​in the range [-1, 1].

[0128] Step 33: Dynamic carbon factor generation and energy storage carbon factor source tracing and matching

[0129] Calculation of the correlation co-correction coefficient: The correlation co-correction coefficient is defined as reflecting the degree of deviation of the joint distribution density of multiple factors at the current moment from the independent case. After normalization to the [0,1] interval, the formula is as follows:

[0130]

[0131] In the formula, Let f be the probability density function of the Copula function. Xk (x k (t) represents the factor X at time t. k The marginal probability density.

[0132] Dynamic carbon factor generation: By combining correlation-based synergistic correction coefficients, a dynamic carbon factor adapted to the source-load-storage coupling characteristics is generated. The dynamic constraint coefficient α is used to smooth out fluctuations caused by multi-factor coupling, ensuring that the dynamic carbon factor remains within the range of [0.8]. 1.2 The measurement compliance range is defined by the formula:

[0133]

[0134] In the formula, α is the dynamic constraint coefficient of the carbon factor, typically ranging from 0.2 to 0.5. Let t be the marginal carbon factor of the power grid at time t.

[0135] Carbon factor time-series traceability matching for energy storage charging and discharging: Considering the cross-time-period transfer characteristic of carbon responsibility in energy storage, the carbon factor during charging is directly taken from the current dynamic carbon factor; during discharging, based on the charging amount and corresponding dynamic carbon factor of the historical charging period corresponding to this discharge, the equivalent carbon emission factor for discharge is calculated, achieving accurate traceability matching of carbon responsibility between charging and discharging periods. The formula is as follows:

[0136]

[0137] In the formula, This is the set of historical charging periods corresponding to this discharge. The energy stored at time τ is the amount of charge. Let τ be the dynamic carbon factor at time τ.

[0138] Step 4: Carbon emission responsibility sharing module based on electricity path and subject behavior

[0139] This module is responsible for the comprehensive allocation of carbon emission responsibilities among multiple entities. It calculates the basic allocation coefficient of electricity path based on power flow tracking results, constructs a comprehensive index of low-carbon behavior of entities and calculates behavior correction coefficients, and combines dynamic carbon factors to complete the final accounting of carbon emission responsibilities of each entity, thus fully meeting the carbon emission conservation constraints.

[0140] The specific implementation process is as follows:

[0141] Step 41: Calculation of Basic Power Path Allocation Coefficient

[0142] Based on the electricity path results from power flow tracking, the effective electricity consumption of the four responsible entities—source-side operators, storage-side operators, load-side users, and public management entities—is calculated. The basic allocation coefficient for each entity is then calculated, strongly linking carbon emission responsibility to physical electricity consumption to ensure fair allocation. The formula is as follows:

[0143]

[0144] In the formula, The effective power consumption of subject i is the direct power consumption + energy storage transfer power - transferred power; I is the set of four major responsible subjects.

[0145] Step 42: Calculation of Comprehensive Indicators and Behavior Correction Coefficients for Main Subject Low-Carbon Behaviors

[0146] Construction of a comprehensive indicator for low-carbon behavior of various entities: Extracting the demand response participation of each entity within the accounting period. Energy storage low-carbon charging and discharging matching degree ,through After standardization, a comprehensive index for the main body's low-carbon behavior is constructed, with the following formula:

[0147]

[0148] In the formula, The participation rate in the demand response of subject i is calculated by weighting the number of responses and the response capacity. The matching degree of low-carbon energy storage charging and discharging for subject i is the matching degree between the proportion of green electricity during charging and the demand for green electricity from the load during discharging; weight , Typical reference values ​​are determined using the analytic hierarchy process or entropy weight method. , .

[0149] Calculation of Behavioral Correction Coefficient: Based on the comprehensive index of low-carbon behavior of the main body, the behavioral correction coefficient is calculated. The correction range is controlled by the incentive intensity coefficient λ, so that the value of the behavioral correction coefficient is controlled between [0.8, 1.2]. The better the behavior, the smaller the correction coefficient, and the lower the carbon emission responsibility. The formula is as follows:

[0150]

[0151] In the formula, λ is the excitation intensity coefficient, with a typical value of λ = 0.5. This represents the average value of the comprehensive indicators of low-carbon behavior across all stakeholders.

[0152] Step 43: Comprehensive apportionment and final liability accounting

[0153] Calculate the comprehensive allocation weight for each entity, combine it with the dynamic carbon factor to account for the direct carbon emission responsibility of each entity, allocate public carbon emission responsibility according to the comprehensive allocation weight, and finally obtain the total carbon emission responsibility of each entity. The formula is as follows:

[0154]

[0155]

[0156] In the formula, The overall allocation weight for subject i, The direct carbon emission responsibility of subject i is calculated based on dynamic carbon factors and direct electricity consumption. The total public carbon emission responsibility includes carbon emissions from public branch roads and carbon emissions from network losses.

[0157] The allocation process fully complies with the total carbon emission conservation constraint: ,in To ensure that the total carbon emissions of the integrated source-load-storage system are measured within the accounting period, there is no duplication or omission in the accounting.

[0158] Step 5: Closed-loop verification module

[0159] This module is responsible for verifying the calculation results and model performance throughout the entire process. It uses three layers of rigid verification—carbon conservation verification, correlation restoration verification, and excitation effectiveness verification—to ensure the rigor and rationality of the solution.

[0160] The specific implementation process is as follows:

[0161] Step 51 Carbon conservation check

[0162] To verify the consistency between the total carbon emission responsibility and the sum of the responsibilities shared by each entity, and to ensure there are no accounting biases, the formula is as follows:

[0163]

[0164] In the formula, The allowable deviation threshold is typically 0.5%. If the deviation meets the threshold requirement, the carbon conservation check is passed; otherwise, the data acquisition and allocation calculation stages are traced back to investigate any anomalies.

[0165] Step 52 Correlation Restoration Verification

[0166] To verify the model's accuracy in reproducing the nonlinear correlations of multiple factors and ensure the accuracy of data-driven modeling, the formula is as follows:

[0167]

[0168] In the formula, The actual Kendall rank correlation coefficient of the original data. The correlation coefficient is the model fit coefficient. The allowable deviation threshold is typically 0.1. If the deviation meets the threshold requirement, the correlation restoration check is performed; otherwise, the Copula function selection and parameter calibration are optimized.

[0169] Step 53: Incentive Validity Verification

[0170] To verify the effectiveness of the low-carbon incentive mechanism, two key aspects need to be verified: first, entities with better low-carbon behavior have lower carbon emission responsibility per unit of effective electricity than the industry benchmark; second, the behavior correction coefficient and carbon emission responsibility show a significant negative correlation. If both aspects are met, the incentive effectiveness verification is passed; otherwise, the incentive intensity coefficient and weighting parameters are optimized to ensure the effective implementation of the incentive mechanism. Specific Implementation Method Two:

[0172] Reference Figure 1A source-load-storage integrated entity carbon emission boundary identification and responsibility allocation system based on Copula-kernel density estimation includes: a source-load-storage integrated entity operation characteristic analysis and metering point configuration module; a carbon emission boundary identification module based on power flow tracking + Copula correlation; a dynamic carbon factor adaptation module based on Copula-kernel density estimation; a carbon emission responsibility allocation module based on power path + entity behavior; and a closed-loop verification module, wherein:

[0173] The integrated source-load-storage main body operation characteristic analysis and metering point configuration module is used to construct a standardized data acquisition system and physical constraint framework for the integrated source-load-storage main body. Specifically, it extracts three core operation characteristics of the integrated source-load-storage main body: physical operation, electric-carbon coupling, and data correlation, and clarifies the core requirements for data acquisition and modeling; configures a standardized metering point system covering the source side, storage side, load side, grid connection interface, and common branch, and completes the standardized acquisition of time-series operation data of the entire link within the accounting cycle; establishes node power conservation constraints, verifies the rationality of the collected data, and provides a physical constraint basis for subsequent calculations.

[0174] This module covers all nodes from the source side, storage side, load side, grid connection interface, and common branch, establishing a unified measurement standard to solve the data gap problem caused by the scattered collection points and inconsistent standards in traditional methods. It provides a complete time-series data foundation for full-link carbon traceability. It extracts three core characteristics in advance: physical operation, electro-carbon coupling, and data correlation, and clarifies collection requirements and data dimensions. This avoids invalid data redundancy and ensures that no core modeling variables are omitted, balancing collection efficiency and data validity. In addition, the module uses node power conservation as the core constraint to perform rationality verification on the collected data, eliminating abnormal data and correcting power imbalance deviations from the source. This provides a reliable physical basis for subsequent power flow tracking and responsibility allocation calculations, ensuring algorithm convergence and result credibility.

[0175] A carbon emission boundary identification module based on power flow tracing and Copula correlation is used to accurately delineate the carbon emission responsibility boundaries of integrated source-load-storage entities. Specifically, it constructs a directed weighted topology model that integrates physical characteristics and data correlation characteristics to provide a topological foundation for power path tracing; it adopts a proportional allocation method for power flow tracing, following the logic of carbon following electricity and splitting by power proportion, to complete the full-link power path tracing and clarify the power source composition of each branch; it calculates the correlation coefficient between branch power and core influencing factors through the Copula function, corrects the power flow tracing results, and delineates three types of carbon emission responsibility boundaries: direct responsibility units, indirect responsibility units, and common responsibility units.

[0176] This module integrates physical topology features and data correlation features to construct a directed weighted topology, overcoming the limitation that pure physical topology cannot reflect the coupling relationship between source, load, and storage, and is more suitable for the power flow pattern in strongly coupled scenarios. It uses the proportional allocation method to complete the source tracing of the entire power path, clarifying the power source composition of each branch, and deeply binding carbon emission responsibility with physical power flow, ensuring the physical rationality of responsibility attribution from the underlying logic. It quantifies the nonlinear correlation between branch power and influencing factors through the Copula function, corrects the calculation error of pure power flow tracing, and divides the responsibility units into three categories: direct, indirect, and common, to achieve no overlap or vacuum in the carbon emission responsibility boundary, solving the industry pain point of overlapping or blank responsibility attribution in traditional methods.

[0177] A dynamic carbon factor adaptation module based on Copula-kernel density estimation is used to generate dynamic carbon emission factors that adapt to the source-load-storage coupling characteristics. Specifically, it extracts time-series data of the three core influencing factors of carbon factors: spatiotemporal dynamics, power source attributes, and energy storage behavior, and completes data cleaning and standardization preprocessing; it fits the marginal distribution of each factor through kernel density estimation, uses the Copula function to characterize the nonlinear correlation between multiple factors, and completes multi-factor joint distribution modeling; it calculates the correlation synergistic correction coefficient, generates dynamic carbon factors that adapt to the source-load-storage coupling characteristics, and completes time-series source matching of energy storage charging and discharging carbon factors.

[0178] This module fits the marginal distribution of each influencing factor through kernel density estimation and combines it with the Copula function to characterize the nonlinear relationship between spatiotemporal dynamics, power source attributes, and energy storage behavior, thus constructing a multi-factor joint distribution model. Compared with traditional static and single-factor carbon factors, it better reflects the real carbon emission characteristics under the scenario of source-load-storage fluctuations. It generates dynamic carbon factors corresponding to time-series operating data, rather than fixed averages, which can respond to changes in carbon emission intensity caused by power output fluctuations, load changes, and energy storage actions, significantly improving the time dimension accuracy of carbon accounting. In view of the spatiotemporal offset characteristics of energy storage charging and discharging, it performs time-series source matching of energy storage charging and discharging carbon factors, accurately mapping the carbon emission responsibility of the charging period to the absorption side of the discharging period, solving the difficulties of cross-time period transfer and accounting misalignment of energy storage carbon responsibility.

[0179] The carbon emission responsibility allocation module based on power path and entity behavior is used to complete the comprehensive allocation of carbon emission responsibilities among multiple entities. Specifically, it calculates the effective power consumption of each entity based on power path results from power flow tracking and calculates the basic allocation coefficient of power path; it constructs a comprehensive indicator of entity low-carbon behavior covering demand response participation and energy storage low-carbon charging and discharging matching degree, and calculates behavior correction coefficients; and it combines dynamic carbon factors and comprehensive allocation weights to complete the calculation of direct carbon emission responsibilities and public carbon emission responsibilities of each entity, fully meeting the total carbon emission conservation constraint.

[0180] This module calculates the basic allocation coefficient based on the power path obtained from power flow tracking, strongly linking carbon emission responsibility with the actual power consumption of the entities, adhering to the fair principle of "whoever uses electricity, bears the responsibility," and avoiding unfair allocation caused by subjective weighting. This module constructs a comprehensive low-carbon behavior index covering demand response participation and the matching degree of low-carbon charging and discharging of energy storage. By adjusting the allocation results through behavior correction coefficients, it reduces the responsibility of entities that take proactive low-carbon actions, achieving a positive incentive of "lower responsibility for low carbon" and guiding entities to actively reduce carbon emissions. This module also covers the accounting of direct carbon emission responsibility and public carbon emission responsibility, and strictly follows the carbon emission total conservation constraint throughout the process, ensuring both clear responsibility attribution and no deviation or additional increase or decrease in the total amount of accounting.

[0181] The closed-loop verification module is used to verify the calculation results and model performance throughout the entire process. Specifically, it verifies the consistency between the total carbon emission responsibility and the sum of the responsibilities shared by each subject through carbon conservation verification; it verifies the model's degree of restoration of the nonlinear correlation of multiple factors through correlation restoration verification; and it verifies the responsibility reduction effect of low-carbon behavior subjects through incentive effectiveness verification.

[0182] This module establishes a verification mechanism from three dimensions: total consistency, model accuracy, and mechanism effectiveness, covering the entire process from data input, model calculation, to result implementation. Compared to a single verification method, it is more comprehensive, systematically identifying issues such as carbon conservation bias, model distortion, and incentive failure. By verifying the consistency between total carbon emission responsibility and the sum of the responsibilities shared by each entity, it ensures that the accounting results are free from systematic bias at the total level, meeting the basic compliance requirements of carbon accounting. This module adds an incentive effectiveness verification to validate the correspondence between low-carbon behaviors and responsibility reductions, ensuring that the solution is not only theoretically sound but also plays a guiding role in low-carbon practices in real-world scenarios, enhancing the overall engineering practicality and implementation value of the solution.

[0183] The accompanying drawings of the embodiments disclosed in this invention only involve structures relevant to the embodiments disclosed in this invention. Other structures can be referred to with common designs. Unless otherwise specified, the same embodiment and different embodiments of this invention can be combined with each other.

Claims

1. A method for identifying the carbon emission boundary and allocating responsibility for integrated source-load-storage entities based on Copula-kernel density estimation, characterized in that, Includes the following steps: Step 1: Construct a standardized data acquisition system and physical constraint framework for the integrated source-load-storage system; First, the core operational characteristics of the main body are extracted; then, a standardized metering point system covering the entire link is configured; finally, node power conservation constraints are established, and the collection and verification of time-series operational data within the accounting cycle are completed. Step 2: Achieve precise delineation of the carbon emission responsibility boundaries for integrated source-load-storage entities; First, a directed weighted topology model integrating physical characteristics and data correlation characteristics is constructed. Then, the power flow tracing method of proportional allocation is used to complete the source tracing of the entire link power path. Finally, the source tracing results are corrected through Copula correlation analysis, and carbon emission responsibility boundary units with no overlap and no vacuum are divided. Step 3: Generate a dynamic carbon emission factor that adapts to the source-load-storage coupling characteristics; First, the time series data of the core influencing factors of the carbon factor are extracted and preprocessed. Then, the joint distribution model of multiple factors is completed by kernel density estimation and Copula function. Finally, the dynamic carbon factor is generated, and the time series source matching of the carbon factor of energy storage charging and discharging is completed. Step 4: Complete the comprehensive allocation of carbon emission responsibilities among multiple stakeholders; First, the basic allocation coefficient of electricity path is calculated based on the power flow tracking results. Then, a comprehensive index of low-carbon behavior of the main body is constructed and the behavior correction coefficient is calculated. Finally, the carbon emission responsibility of each main body is calculated by combining the dynamic carbon factor, and the total carbon emission conservation constraint is met throughout the process. Step 5: Perform full-process verification of the calculation results and model performance; First, the consistency of the total amount of carbon is verified through carbon conservation verification. Then, the accuracy of data-driven modeling is verified through correlation restoration verification. Finally, the effectiveness of the low-carbon incentive mechanism is verified through incentive effectiveness verification.

2. The method for identifying the carbon emission boundary and allocating responsibility for an integrated source-load-storage entity based on Copula-kernel density estimation as described in claim 1, characterized in that, In step 1, the processing is as follows: Step 11: Extract the three core operational characteristics of the integrated source-load-storage system: physical operation, electricity-carbon coupling, and data correlation, and clarify the core requirements for data acquisition and modeling; Step 12: Configure a standardized metering point system covering the source side, storage side, load side, grid connection interface, and common branch, and complete the standardized collection of full-link time-series operation data within the accounting cycle; Step 13: Establish node power conservation constraints, verify the rationality of the collected data, and provide a physical constraint basis for subsequent calculations.

3. The method for identifying the carbon emission boundary and allocating responsibility for an integrated source-load-storage entity based on Copula-kernel density estimation as described in claim 1, characterized in that, In step 2, the processing is as follows: Step 21: Construct a directed weighted topology model that integrates physical features and data correlation features to provide a topological foundation for power path tracing; Step 22: Use the proportional allocation method for power flow tracing, following the core logic of "carbon follows electricity and is split according to power ratio", to complete the full-link power path tracing and clarify the power source composition of each branch; Step 23: Calculate the correlation coefficient between branch power and core influencing factors using the Copula function, correct the power flow tracking results, and divide the carbon emission responsibility boundaries into three categories: direct responsibility units, indirect responsibility units, and public responsibility units.

4. The method for identifying the carbon emission boundary and allocating responsibility for an integrated source-load-storage entity based on Copula-kernel density estimation as described in claim 1, characterized in that, In step 3, the processing is as follows: Step 31: Extract time-series data of the three core influencing factors of carbon factors: spatiotemporal dynamics, power attributes, and energy storage behavior, and complete data cleaning and standardization preprocessing; Step 32: Fit the marginal distribution of each factor by kernel density estimation, and use the Copula function to characterize the nonlinear correlation between multiple factors to complete the joint distribution modeling of multiple factors; Step 33: Calculate the correlation coordination correction coefficient, generate a dynamic carbon factor that adapts to the source-load-storage coupling characteristics, and complete the time-series source matching of energy storage charging and discharging carbon factors to solve the problem of cross-time period transfer of energy storage carbon responsibility.

5. The method for identifying the carbon emission boundary and allocating responsibility for an integrated source-load-storage entity based on Copula-kernel density estimation as described in claim 1, characterized in that, In step 4, the processing is as follows: Step 41: Based on the power path results of power flow tracking, calculate the effective power consumption of each entity, calculate the basic power path allocation coefficient, and ensure the fairness of allocation; Step 42: Construct a comprehensive index of main low-carbon behaviors that includes demand response participation and energy storage low-carbon charging and discharging matching degree, calculate behavior correction coefficients, and achieve effective incentives for low-carbon behaviors; Step 43: Combining dynamic carbon factors and comprehensive allocation weights, complete the accounting of direct carbon emission responsibilities and public carbon emission responsibilities of each entity, and fully meet the total carbon emission conservation constraints.

6. The method for identifying the carbon emission boundary and allocating responsibility for an integrated source-load-storage entity based on Copula-kernel density estimation as described in claim 1, characterized in that, In step 5, the processing is as follows: Step 51: Verify the consistency between the total carbon emission responsibility and the sum of the responsibilities shared by each entity through carbon conservation verification to ensure that the accounting results are unbiased; Step 52: Verify the model's accuracy in restoring the nonlinear correlations of multiple factors through correlation restoration verification, ensuring the accuracy of data-driven modeling; Step 53: Verify the effectiveness of incentives to check the responsibility reduction effect of low-carbon behavior subjects and ensure that the low-carbon incentive mechanism is effectively implemented.

7. A source-load-storage integrated entity carbon emission boundary identification and responsibility allocation system based on Copula-kernel density estimation, characterized in that, include: The system comprises the following modules: integrated source-load-storage main operation characteristic analysis and metering point configuration module; carbon emission boundary identification module based on power flow tracking and Copula correlation; dynamic carbon factor adaptation module based on Copula-kernel density estimation; carbon emission responsibility allocation module based on power path and main behavior module; and closed-loop verification module. The module for analyzing the operational characteristics of the integrated source-load-storage system and configuring metering points is used to construct a standardized data acquisition system and physical constraint framework for the integrated source-load-storage system. A carbon emission boundary identification module based on current tracing and Copula correlation is used to accurately delineate the carbon emission responsibility boundaries of integrated source-load-storage entities. A dynamic carbon factor adaptation module based on Copula-kernel density estimation is used to generate dynamic carbon emission factors that adapt to source-load-storage coupling characteristics. The carbon emission responsibility allocation module based on electricity path and subject behavior is used to complete the comprehensive allocation of carbon emission responsibilities among multiple subjects. The closed-loop verification module is used to verify the calculation results and model performance throughout the entire process.

8. The source-load-storage integrated entity carbon emission boundary identification and responsibility allocation system based on Copula-kernel density estimation according to claim 7, characterized in that, The integrated source-load-storage main operation characteristic analysis and metering point configuration module is specifically used for: The three core operational characteristics of the integrated source-load-storage system—physical operation, electro-carbon coupling, and data correlation—are extracted, and the core requirements for data acquisition and modeling are clarified. Configure a standardized metering point system covering the source side, storage side, load side, grid connection interface, and common branch, and complete the standardized collection of time-series operation data of the entire link within the accounting cycle; Establish node power conservation constraints, verify the rationality of the collected data, and provide a physical constraint basis for subsequent calculations.

9. A source-load-storage integrated entity carbon emission boundary identification and responsibility allocation system based on Copula-kernel density estimation as described in claim 7, characterized in that, The carbon emission boundary identification module based on tidal current tracking and Copula correlation is specifically used for: Construct a directed weighted topology model that integrates physical characteristics and data correlation characteristics to provide a topological foundation for power path tracing; The proportional allocation method is used for power flow tracing, following the logic of carbon following electricity and splitting according to power ratio, to complete the traceability of the entire power path and clarify the power source composition of each branch. The correlation coefficient between branch power and core influencing factors is calculated using the Copula function, and the power flow tracking results are corrected to divide the carbon emission responsibility boundaries into three categories: direct responsibility units, indirect responsibility units, and public responsibility units.

10. A source-load-storage integrated entity carbon emission boundary identification and responsibility allocation system based on Copula-kernel density estimation according to claim 7, characterized in that, The dynamic carbon factor adaptation module based on Copula-kernel density estimation is specifically used for: Extract time-series data of the three core influencing factors of carbon factors: spatiotemporal dynamics, power attributes, and energy storage behavior, and complete data cleaning and standardization preprocessing. By fitting the marginal distributions of each factor through kernel density estimation, and using the Copula function to characterize the nonlinear correlations among multiple factors, the joint distribution model of multiple factors is completed. Calculate the correlation-coordinated correction coefficient, generate a dynamic carbon factor that adapts to the source-load-storage coupling characteristics, and complete the time-series source matching of the energy storage charging and discharging carbon factor. Based on the power path results from power flow tracking, the effective power consumption of each entity is calculated, and the basic power path allocation coefficient is calculated. Construct a comprehensive index of main low-carbon behaviors covering demand response participation and energy storage low-carbon charging and discharging matching degree, and calculate behavior correction coefficients; By combining dynamic carbon factors and comprehensive allocation weights, the direct carbon emission responsibilities and public carbon emission responsibilities of each entity are calculated, and the total carbon emission conservation constraint is met throughout the process. The closed-loop verification module is specifically used for: The carbon conservation check verifies the consistency between the total carbon emission responsibility and the sum of the responsibilities shared by each entity. The correlation restoration test verifies the model's accuracy in restoring the nonlinear correlations of multiple factors. By validating the effectiveness of incentives, we can verify the impact of reducing the responsibility of low-carbon actors.

Images