Green certificate circulation relation analysis method, device and system and medium

Abstract

The invention relates to the technical field of green certificate transaction, and provides a green certificate circulation relation analysis method, device and system and a medium. According to the implementation scheme, variables in a causal loop diagram are extracted, and all the variables and the types of the variables are obtained; executing a Granger causal test on each variable to obtain a Granger causal relationship among the variables so as to construct a causal relationship matrix; based on the type of each variable and the causal relationship matrix, constructing a stock flow graph; performing quantitative estimation on the influence coefficient of each causal link in the stock flow diagram to obtain a target coefficient matrix; based on the stock flow diagram and the target coefficient matrix, constructing a system dynamics model; and performing simulation calculation on the system dynamics model to obtain a circulation relationship analysis result of the green evidence among the multiple markets. According to the embodiment of the invention, quantitative analysis can be carried out on the green evidence production and elimination balance and circulation relationship in a multi-market environment, and sustainable development of the green evidence market is promoted.

Description

Technical Field

[0001] This invention relates to the field of green certificate trading technology, and in particular to a method, apparatus, system and medium for analyzing the circulation relationship of green certificates. Background Technology

[0002] With the continuous expansion of renewable energy installed capacity, green electricity consumption certificates (hereinafter referred to as green certificates), as an important market-based tool for measuring the relationship between renewable energy power production and consumption, are playing an increasingly important role in promoting energy structure transformation and achieving "dual carbon" goals. Currently, green certificates have gradually formed a complex pattern of multiple markets operating in parallel and influencing each other, including the electricity market and carbon trading market. The price formation mechanism, supply and demand balance, and cross-market circulation path of green certificates are all affected by a variety of market factors.

[0003] However, in the context of multi-market coupling, the circulation of green certificates among the production, consumption, and trading markets exhibits highly dynamic and non-linear characteristics. Cross-influences and feedback effects exist between different market variables, making it difficult to accurately depict the balance between green certificate production and consumption and its evolution from a single market perspective. Existing technologies mostly focus on static analysis of single markets or local factors, lacking systematic identification and quantitative modeling of causal relationships between green certificate-related variables in a multi-market environment, thus failing to reveal the true circulation paths and evolutionary patterns of green certificates across different markets. Summary of the Invention

[0004] This invention provides a method, apparatus, system, and medium for analyzing the circulation relationship of green certificates, which can solve at least one of the above-mentioned technical problems.

[0005] In a first aspect, embodiments of the present invention provide a method for analyzing the circulation relationship of green certificates, including: Extract the variables from the preset causal loop diagram to obtain each variable and the type of each variable, wherein the causal loop diagram represents the qualitative structure diagram of the circulation of green certificates among multiple markets; Granger causality tests were performed on each of the variables to obtain the Granger causal relationships between the variables, and a causal relationship matrix was constructed. Based on the types of each variable and the causal relationship matrix, a stock flow diagram is constructed; The impact coefficients of each causal link in the existing traffic graph are quantitatively estimated to obtain the target coefficient matrix; Based on the existing flow diagram and the target coefficient matrix, a system dynamics model is constructed; The system dynamics model is simulated to obtain the analysis results of the circulation relationship of the green certificate in multiple markets. The circulation analysis results include the evolution path of the green certificate price and the circulation path of the green certificate in multiple markets.

[0006] Secondly, embodiments of the present invention provide an analysis device for the circulation relationship of green certificates, comprising: The variable extraction module is used to extract variables from a preset causal loop diagram to obtain each variable and the type of each variable. The causal loop diagram represents the qualitative structure of the green certificate circulating among multiple markets. The Granger causality test module is used to perform Granger causality tests on each of the variables to obtain the Granger causal relationships between the variables and construct a causal relationship matrix. The stock flow graph construction module is used to construct a stock flow graph based on the types of each variable and the causal relationship matrix. The quantification estimation module is used to quantify and estimate the influence coefficients of each causal link in the existing traffic diagram to obtain the target coefficient matrix; The model building module is used to construct a system dynamics model based on the stock flow diagram and the target coefficient matrix; The simulation calculation module is used to perform simulation calculations on the system dynamics model to obtain the analysis results of the circulation relationship of the green certificate in multiple markets. The circulation analysis results include the evolution path of the green certificate price and the circulation path of the green certificate in multiple markets.

[0007] Thirdly, embodiments of the present invention also provide an analysis system for the circulation relationship of green certificates, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method described in any one of the embodiments of the present invention.

[0008] Fourthly, embodiments of the present invention also provide a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to perform the method described in any one of the embodiments of the present invention.

[0009] This invention employs a technical solution that extracts variables from a causal loop diagram representing the circulation of green certificates across multiple markets. Granger causality tests are then performed on each variable to identify and screen statistically significant causal relationships, thereby constructing a causal relationship matrix reflecting the interactive influence of multiple markets. This objectively characterizes the direction and path of action of green certificate-related variables. Based on this, a stock flow diagram is constructed by combining the variable types and the causal relationship matrix, transforming the qualitative causal structure into a computable system dynamics topology. This clearly expresses the accumulation, change, and adjustment relationships of green certificates across production, consumption, and trading markets. Furthermore, by quantitatively estimating the influence coefficients of each causal link in the stock flow diagram, a target coefficient matrix is ​​obtained to quantitatively characterize the influence intensity of various market factors. Finally, the system dynamics model constructed based on the stock flow diagram and the target coefficient matrix can dynamically simulate the green certificate circulation process under multi-variable and multi-feedback loop conditions, thereby obtaining the analysis results of the circulation relationship of green certificates across multiple markets. Thus, this embodiment of the invention enables a quantitative analysis of the balance and circulation of green certificate production and consumption in a multi-market environment, which is conducive to promoting the sustainable development of the green certificate market.

[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0011] The accompanying drawings are provided for a better understanding of this solution and do not constitute a limitation of the invention. Wherein: Figure 1 This is a flowchart of a method for analyzing the circulation relationship of green certificates according to an embodiment of the present invention; Figure 2 This is a causal loop diagram in the analysis method of green certificate circulation relationship according to an embodiment of the present invention; Figure 3 This is a stock flow diagram in the green certificate circulation relationship analysis method of an embodiment of the present invention; Figure 4 This is a structural block diagram of an analysis device for the circulation relationship of green certificates according to an embodiment of the present invention; Figure 5 This is a schematic block diagram of an electronic device used to implement the methods of embodiments of the present invention. Detailed Implementation

[0012] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of the present invention, including various details to aid understanding. These details should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of the invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0013] Figure 1 This is a flowchart of an embodiment of the method for analyzing the circulation relationship of green certificates according to the present invention.

[0014] like Figure 1 As shown, the analysis method for the transfer relationship of the green certificate may include: S110, extract the variables from the preset causal loop diagram to obtain each variable and the type of each variable, wherein the causal loop diagram represents the qualitative structure diagram of the green certificate circulating among multiple markets. S120, Perform Granger causality tests on each of the variables to obtain the Granger causal relationships between the variables, and construct a causal relationship matrix; S130, construct a stock flow diagram based on the types of each variable and the causal relationship matrix; S140, Quantitatively estimate the influence coefficients of each causal link in the existing traffic diagram to obtain the target coefficient matrix; S150, Based on the stock flow diagram and the target coefficient matrix, construct a system dynamics model; S160, Simulation calculation is performed on the system dynamics model to obtain the analysis results of the circulation relationship of the green certificate in multiple markets, wherein the circulation analysis results include the evolution path of the green certificate price and the circulation path of the green certificate in multiple markets.

[0015] like Figure 2 As shown, for example, a causal loop diagram refers to a qualitative structural diagram pre-constructed based on domain knowledge (such as economic theory, current policies, and industrial structure) to describe the interaction between green certificates in multiple markets such as the production market, the consumption market, and the transaction market. It represents the positive or negative influence relationship between variables through directed connections and allows the formation of feedback loops.

[0016] For example, in a multi-market coupled scenario for green certificates, variables such as "green certificate supply", "green certificate demand", "green certificate price", "new energy power generation" and "compliance cost" can be set in the causal loop diagram, and directed arrows can be used to represent feedback relationships such as "increased demand for green certificates → increased price of green certificates" and "increased price of green certificates → increased investment in new energy power generation".

[0017] For example, a variable refers to a quantitative element used to characterize the multi-market operation status, evolution process, or adjustment mechanism of green certificates. Variables can be classified into state variables, rate variables, auxiliary variables, or constant variables according to system dynamics modeling rules.

[0018] In this example, a total of 38 variables are selected as shown below.

[0019] State variables: These are stock quantities, referring to the cumulative quantities of system states, such as electricity prices, green certificate holdings by buyers, green certificate holdings by green energy producers, changes in green certificate prices, Gross Domestic Product (GDP), carbon allowance holdings by buyers, and carbon allowance inventory held by buyers.

[0020] Rate variables: These are flows, the rate at which changes in stock. Specifically, they include: the amount of green certificates required to meet the Renewable Portfolio Standard (RPS), excess demand for green certificates, circulation of green certificates, issuance of green certificates, annual GDP growth, carbon quota supply, carbon quota sales, and carbon quota demand.

[0021] Auxiliary variables: GDP annual growth rate, green electricity installed capacity, green power unit operating time, thermal power unit operating time, thermal power installed capacity, electricity demand, expected carbon quota sales, expected carbon quota purchase, carbon quota supply-demand ratio, thermal power plant profit, thermal power unit power generation, green power unit power generation, green electricity plant profit, green certificate price, expected green certificate sales, expected green certificate purchase, carbon price, electricity supply, electricity supply-demand ratio.

[0022] Constant variables: network loss coefficient, RPS quota growth rate, carbon dioxide emission coefficient of thermal power units, carbon emission intensity, and the proportion of carbon emissions in the power industry.

[0023] For example, a causal relationship matrix refers to a two-dimensional matrix constructed based on the results of the Granger causality test, used to represent in matrix form whether there is a statistically significant causal relationship between any variable and another variable and its direction.

[0024] For example, if the test results show that the historical changes in "green certificate demand" can significantly improve the predictive ability of the current value of "green certificate price", then a causal relationship from "green certificate demand" to "green certificate price" is recorded in the corresponding row and column position in the causal relationship matrix.

[0025] like Figure 3 As shown, exemplarily, a stock-flow graph refers to a quantitative structural graph within a system dynamics framework, which maps state variables to stock nodes, rate variables to flow nodes, and constructs the system topology through directed connections, based on variable types and causal relationship matrices. In the figure, <time>→Electricity demand indicates how electricity demand changes over time.

[0026] For example, the "cumulative inventory of green certificates" can be mapped to stock nodes, the "green certificate generation rate" and "green certificate consumption rate" can be mapped to the corresponding inflow and outflow flows, and the flows can be adjusted by the auxiliary variable "market demand intensity".

[0027] For example, Figure 3 " in <time>→V” indicates that variable V changes over time; "V x →V y " represents variable V y With variable V x change.

[0028] For example, the target coefficient matrix refers to the parameter matrix obtained after quantitatively estimating the influence intensity of each causal link in the stock flow diagram, used to characterize the quantitative interaction relationship between different variables. For instance, in the target coefficient matrix, the "influence coefficient of green certificate price change on green certificate trading rate" can be represented in numerical form to reflect the moderating effect of price change on market trading activity.

[0029] For example, the flow relationship analysis results refer to the analysis output results obtained by simulating the system dynamics model and used to characterize the evolution characteristics of green certificates in a multi-market environment.

[0030] For example, the results of the circulation relationship analysis may include curves showing the price of green certificates over time under different scenario parameters, as well as the circulation paths and trends of green certificates between the production end, the consumption end, and the trading market.

[0031] For example, in step S110, firstly, based on the domain knowledge of multi-market coupled trading of green certificates, a causal loop diagram describing the interaction between green certificates in the production market, the consumption market, and the trading market is pre-constructed; then, all nodes appearing in the causal loop diagram are traversed, and the corresponding variable names are identified one by one; next, according to the functional attributes of the variables in the system, the variables are divided into state variables, rate variables, auxiliary variables, and constant variables. Among them, variables reflecting the cumulative state of the system are marked as state variables, variables representing the rate of change are marked as rate variables, and variables used to regulate relationships are marked as auxiliary variables or constant variables, thereby obtaining each variable and its corresponding variable type.

[0032] For example, in step S160, the completed system dynamics model is loaded into the simulation environment, and corresponding time series data or initial parameters are configured for each input variable in the model; then, the system dynamics equations are iteratively calculated according to the preset simulation time step to obtain the evolution results of each state variable and auxiliary variable within the simulation period; finally, the simulation output results are sorted and analyzed to extract the analysis results reflecting the trend of green certificate price changes and the circulation path of green certificates between different markets.

[0033] Based on the above implementation method, by introducing causal loop diagrams, Granger causality tests, and system dynamics modeling, a systematic characterization of the production-consumption balance and circulation process of green certificates in a multi-market environment is provided, moving from qualitative to quantitative analysis and from structure to behavior. First, a causal loop diagram clarifies the basic interaction structure of green certificates between the production end, the consumption end, and the trading market. Then, Granger causality tests are used to screen and quantify the causal relationships between variables. Based on this, a stock-flow diagram and a system dynamics model are constructed, and the model is simulated to obtain the dynamic evolution results of green certificate prices and circulation paths. This effectively avoids the biases caused by relying solely on empirical assumptions or single-market analysis, providing verifiable causal evidence and calculable dynamic evolution results for multi-market coupling relationships. This provides more objective and comprehensive analytical support for green certificate market operation evaluation, policy parameter adjustment, and production-consumption balance optimization.

[0034] In one implementation, Granger causality tests are performed on each variable to obtain Granger causal relationships between the variables, thereby constructing a causal matrix. This includes: pairwise combining each variable to obtain variable pairs, where each variable pair includes an explained variable and an explanatory variable; constructing first autoregressive models based on the time series data and historical lags corresponding to each explained variable, where the first autoregressive models are used to predict the current value of the explained variable based on the historical lags of the explained variable, resulting in first residual sums of squares; constructing second autoregressive models based on the time series data and historical lags corresponding to each explained variable, and the time series data and historical lags corresponding to each explanatory variable, where the second autoregressive models are used to predict the current value of the explained variable based on the combined historical lags of the explained variable and the historical lags of the explanatory variables, resulting in second residual sums of squares; and performing causal hypothesis tests on each variable pair based on the first and second residual sums of squares to obtain a causal matrix.

[0035] For example, first, all variables obtained in step S110 are acquired. Then, using a nested loop or combination algorithm, each variable is systematically paired with all other variables to form a series of ordered variable pairs. For any two variables (e.g., variable A and variable B), the system generates two independent variable pairs (A, B) and (B, A) to ensure that bidirectional causal relationships can be tested. In each generated variable pair, the first variable is designated as the "explanatory variable" (i.e., the potential cause), and the second variable is designated as the "explained variable" (i.e., the potential outcome).

[0036] For example, suppose there are three variables: green certificate price (GC_Price), carbon market price (Carbon_Price), and renewable energy generation (RE_Power). The system will generate all possible variable pairs as follows: (GC_Price, Carbon_Price), (Carbon_Price, GC_Price), (GC_Price, RE_Power), (RE_Power, GC_Price), (Carbon_Price, RE_Power), (RE_Power, Carbon_Price). When dealing with the variable pair (GC_Price, Carbon_Price), GC_Price is the explanatory variable, and Carbon_Price is the dependent variable.

[0037] For example, the system iterates through all the variable pairs generated in the previous step. For each variable pair, the system first selects the time series data corresponding to its "explained variable". Next, the system needs to determine an optimal lag order (p), i.e., how many past time points of data to use for prediction. This is usually determined by information criteria such as the Akaike Information Criterion (AIC) or the Bayesian Information Criterion (BIC). Then, an autoregressive model containing only the historical lags of the explained variable itself is constructed, and the model is fitted using the least squares method. After the model is fitted, the system uses it to predict historical data, subtracting the predicted value from the actual value at each time point to obtain a series of residuals. Finally, the system squares all these residuals and sums them to obtain a total value, i.e., the first residual sum of squares (p). ).

[0038] For example, taking the variable pair (GC_Price, Carbon_Price) as an example, the explained variable is Carbon_Price. Assume the optimal lag order is determined to be 2 (p=2). The model constructed by the system (the first autoregressive model) is: Carbon_Price(t) = c + Carbon_Price(t-1)+ Carbon_Price(t-2)+ In the formula, Carbon_Price(t) is the observed value of the explained variable at the current time point t; c is a constant term; is the coefficient of the first-order lag term; Carbon_Price(t-1) is the observed value of the explained variable at the previous time point (t minus 1 period); is the coefficient of the second-order lag term; Carbon_Price(t-2) is the observed value of the explained variable at the first two time points (t minus 2 periods); This is the residual term.

[0039] The system trains the model using historical carbon price data and then uses it to predict the carbon price for each historical day. For example, if the actual carbon price on a certain day is 55 yuan and the model predicts 54 yuan, the residual is 1. The residuals for all time periods are squared and summed to obtain the first residual sum of squares. = 150.8.

[0040] For example, the system iterates through all variable pairs again. For each variable pair, the system selects the time series data for its "explained variable" and "explained variable," and uses the same lag order (p) determined in the previous step to ensure fairness in the comparison. Next, the system constructs a new, more complex autoregressive model that includes not only the historical lags of the explained variable itself but also the historical lags of the explanatory variables as additional predictors. The system also uses the least squares method to fit this new model and uses it to generate predicted values. By calculating the difference between the predicted and actual values, a new residual sequence is obtained, and its squares are summed to obtain the second residual sum of squares (…). ).

[0041] For example, continuing with the variable pair (GC_Price, Carbon_Price), the lag order remains 2 (p=2). The model now constructed by the system (the second autoregressive model) is: Carbon_Price(t) = c + Carbon_Price(t-1)+ Carbon_Price(t-2)+ GC_Price(t-1)+ GC_Price(t-2)+μ(t). Where Carbon_Price(t) is the observed value of the explained variable at the current time point t; c is a constant term; is the coefficient of the first-order lag term; Carbon_Price(t-1) is the observed value of the explained variable at the previous time point (t minus 1 period); GC_Price(t-2) represents the coefficient of the second-order lag term; Carbon_Price(t-2) represents the observed value of the explained variable at the first two time points (t minus 2 periods); GC_Price(t-1) represents the observed value of the explanatory variable (green certificate price) at the first time point; GC_Price(t-2) represents the observed value of the explanatory variable (green certificate price) at the first two time points. The coefficients of the first-order lag terms of the explanatory variables; The coefficients of the second-order lag term are the explanatory variables; μ(t) is the residual term.

[0042] When predicting today's carbon price, this model considers not only yesterday's and the day before's carbon prices, but also yesterday's and the day before's green certificate prices. After training and prediction, the system calculates a new sum of squared residuals. = 95.2. Because information about green certificate prices was incorporated, the model's prediction error was reduced. < .

[0043] For example, the system uses the first sum of squared residuals obtained in the previous two steps for each pair of variables ( ) and the second residual sum of squares ( The F-statistic is calculated using an F-test formula. The null hypothesis of this test is that "historical lags of the explanatory variables have no predictive power for the explained variable." The system evaluates the calculated F-statistic and its corresponding p-value, comparing it to a pre-set significance level (e.g., α=0.05). If the p-value is less than this significance level, the null hypothesis is rejected, and a significant Granger causal relationship is established between the variable pairs. The system fills the test results of all variable pairs (e.g., 1 indicates the existence of a causal relationship, and 0 indicates the absence of one) into an N×N matrix, where the rows represent explanatory variables and the columns represent the explained variable. This final matrix is ​​the causal relationship matrix.

[0044] For example, for the variable pair (GC_Price, Carbon_Price), the system uses =150.8 and =95.2 An F-test was performed. The calculated p-value is assumed to be 0.02. Because 0.02 < 0.05, the system rejects the null hypothesis and concludes that "green certificate price" is a Granger cause of "carbon market price". Therefore, in the causal relationship matrix, the cell with row name GC_Price and column name Carbon_Price is assigned a value of 1. The system will repeat this process for all other variable pairs until the entire causal relationship matrix is ​​filled.

[0045] According to the above implementation method, by combining each variable pairwise, a baseline autoregressive model considering only the historical information of the explained variable itself and an extended autoregressive model incorporating the historical information of the explanatory variables are constructed. The sum of squared residuals of the two models are compared to perform causal hypothesis testing, thereby systematically identifying whether Granger causal relationships exist between variables and forming a causal relationship matrix. In this way, based on fully utilizing the dynamic characteristics of time series, it is possible to quantitatively distinguish between changes driven solely by their own inertia and changes caused by other variables, reducing the interference of spurious correlations and providing a scientific and reliable causal basis for subsequent causal structure construction and model analysis.

[0046] In one implementation, a causal hypothesis test is performed on each pair of variables based on the first and second residual sums of squares to obtain a causal matrix. This includes: calculating a statistic based on the first and second residual sums of squares; performing an F-test based on the statistic to obtain the p-value of the F-test; if the p-value is less than a preset threshold, it is determined that the explanatory variable and the explained variable constitute a Granger causal relationship, and the explanatory variable and the explained variable are connected in a directed manner to obtain the causal link of the variable pair; and the causal matrix is ​​determined based on the causal link of each pair of variables.

[0047] For example, the system obtains the first residual sum of squares for each pair of variables ( (from a model containing only the history of the explained variable) and the second residual sum of squares ( (From a model that includes the history of both the explained and explanatory variables). Then, the system substitutes these two values, along with the total number of observations in the time series (T) and the model's lag order (p), into a standard F-statistic calculation formula. The specific form of this formula is F=[( ) / p] / [ [ / (T-2p-1)]. Performing this calculation yields a specific floating-point value, which is the F-statistic used for subsequent hypothesis testing.

[0048] For example, consider the variable pair (GC_Price, Carbon_Price). Suppose we have the following data: the first residual sum of squares. =150.8; Second residual sum of squares =95.2; Time series length T=250 (days); Lag order p=2; Substitute these values ​​into the formula to calculate: F=[(150.8-95.2) / 2] / [95.2 / (250-2 2-1)];F=[55.6 / 2] / [95.2 / 245];F=27.8 / 0.38857≈71.54;Therefore, the calculated F statistic is 71.54.

[0049] For example, the F-statistic calculated in the previous step, along with its corresponding two degrees of freedom (molecular degrees of freedom) =p and degrees of freedom in the denominator =T-2p-1) is used as input to an F-distribution. The probability of obtaining the current F-statistic or a more extreme value is calculated by querying the cumulative distribution function (CDF) of the F-distribution or using a specialized function in a statistics library. This calculated probability value is the P-value of this F-test. A smaller P-value indicates an improvement in the observed predictive power of the model (i.e., ...). Much larger The less likely it is to be caused by random factors.

[0050] For example, the system calculates the F-statistic as 71.54, along with its degrees of freedom. =2 and =245, input into the statistical function. The statistical library will calculate the probability of P(F(2,245)>71.54), assuming the calculation result is 1.88e-24, that is, the P value of this F test is 1.88e-24.

[0051] For example, the p-value obtained in the previous step is compared with a pre-set significance level threshold (α, typically 0.05, 0.01, or 0.1). If the p-value is strictly less than this threshold, the system rejects the null hypothesis that "the explanatory variable has no predictive power over the explained variable" and determines that a significant Granger causal relationship exists between the two. Once a causal relationship is determined, the system creates a directed link representing this relationship. This link starts from the node representing the "explanatory variable" and points to the node representing the "explanated variable," forming a unidirectional causal chain that visually represents the direction of the influence. If the p-value is not less than the threshold, no link is created.

[0052] For example, the p-value 1.88e-24 is compared with the significance level threshold α = 0.05. Since 1.88e-24 < 0.05, the condition is met. Therefore, the system determines that "GC_Price" and "Carbon_Price" constitute a Granger causal relationship. Subsequently, the system creates a directed edge in its internal graph structure from node GC_Price to node Carbon_Price, i.e., GC_Price → Carbon_Price. This directed edge is the causal link for this variable pair.

[0053] For example, initialize an N×N zero matrix, where N is the total number of variables to be analyzed, and the rows and columns of the matrix are indexed in the same order by the names of all variables. Then, iterate through all variable pairs and their corresponding causal test results. For each variable pair that is determined to have a Granger causal relationship (i.e., the variable pair for which a causal link was successfully created in the previous step), find the corresponding cell in the matrix and change the value of that cell from 0 to 1. The position of the cell (row_i, col_j) in the matrix, with row index i corresponding to the explanatory variable (cause) and column index j corresponding to the explained variable (effect). After completing the traversal and assignment of values ​​for all variable pairs, this matrix containing all causal relationship information is the final causal relationship matrix.

[0054] For example, assuming there are three variables (GC_Price, Carbon_Price, RE_Power), the system will create a 3×3 zero matrix. Based on the results of the previous step, for the causal link GC_Price→Carbon_Price, the system will find the cell in the matrix at row GC_Price and column Carbon_Price and set its value to 1. If, after checking all six variable pairs, a causal relationship RE_Power→GC_Price is also found, then the cell at row RE_Power and column GC_Price will also be set to 1.

[0055] According to the above implementation method, by comparing the changes in the sum of squared residuals before and after the introduction of explanatory variables, the corresponding statistics are calculated and F-tests are performed to determine whether the historical information of the explanatory variables can statistically improve the predictive ability of the explained variable. Based on this, directed causal links between variables are identified, and finally, a causal relationship matrix is ​​formed. In this way, significant causal relationships can be screened based on objective statistical tests, avoiding causal misjudgments caused by subjective assumptions, improving the reliability and interpretability of the causal relationship matrix, and providing a robust data foundation for subsequent structural modeling and system analysis.

[0056] In one implementation, a stock flow graph is constructed based on the types of each variable and the causal relationship matrix, including: mapping the variables corresponding to each element in the causal relationship matrix to nodes in a graphical modeling environment according to the types of each variable, to obtain each node; establishing directed connections between each node based on the directed connection relationships between the variables corresponding to each element, to obtain a topological connection structure; and verifying the topological connection structure based on a preset domain knowledge rule base to eliminate or correct connection links that do not conform to domain logic, to obtain the stock flow graph.

[0057] For example, the system first reads the input causal matrix and obtains a list of all unique variable names represented by their row and column indices. Next, it queries a predefined variable type mapping table to determine which type of node each variable should be represented as in the graphical modeling environment (e.g., a "stock" node or a "flow" node). Then, the system creates a corresponding visual node object on the canvas of the graphical modeling environment for each variable in the list.

[0058] For example, each cell (i, j) in the causal relationship matrix is ​​traversed using nested loops. For each cell (element), its value is checked to see if it is 1. If it is 1, it indicates that a causal relationship exists from row variable i to column variable j. Once the relationship is confirmed, node objects representing these two variables can be found in the graphical modeling environment. Then, a directed connection line with an arrow is drawn between the node representing the cause variable i and the node representing the effect variable j. The arrow starts from the cause node and points to the effect node, visually representing the direction of influence. This process is repeated until all connections represented by elements with a value of 1 in the matrix have been drawn on the canvas, thus forming a preliminary topological connection structure.

[0059] For example, a domain knowledge rule base containing the basic logic and axioms of a specific field (such as economics or ecology) is loaded. Then, each directed connection in the topology generated in the previous step is examined one by one. For each connection, it is determined whether it conforms to a rule in the rule base. If a connection is found to violate domain knowledge (e.g., one "flow" points to another "flow," but in this domain, this usually requires a "stock" as an intermediary), a pre-defined correction operation is performed. The correction operation may include directly removing the illogical connection or correcting it according to the rules, such as automatically inserting a stock node representing a cumulative effect between the two flow nodes. After verifying all connections and making possible corrections, the final network graph is a stock-flow graph that reflects both data-driven causal relationships and conforms to domain-specific logic.

[0060] According to the above implementation method, by mapping the variables in the causal relationship matrix to different nodes in the graphical modeling environment according to their types, and constructing a topological connection structure between nodes based on the directed causal relationships between variables, a domain knowledge rule base is introduced to verify and correct the connection relationships, thereby forming a stock flow graph that conforms to the business mechanism. In this way, the abstract causal relationship matrix can be transformed into a system dynamics model structure with a clear structure and explicit semantics. On the one hand, this effectively avoids unreasonable causal links that violate domain rules, improving the rationality and credibility of the model structure. On the other hand, it provides a standardized and interpretable structural foundation for subsequent parameter estimation, system dynamics modeling, and simulation analysis, which is conducive to improving the reliability of the overall modeling and analysis results.

[0061] In one implementation, the influence coefficients of each causal link in the existing traffic graph are quantitatively estimated to obtain a target coefficient matrix. This includes: obtaining each first variable corresponding to each causal link in the existing traffic graph, wherein the first variable includes independent variables and dependent variables; constructing an independent variable matrix based on each independent variable, and constructing a dependent variable matrix based on each dependent variable; constructing a coefficient matrix based on the dimensions of the independent variable matrix and the dependent variable matrix, wherein each element in the coefficient matrix represents the influence coefficient of the independent variable on the dependent variable; if there is no causal relationship between the independent variable and the dependent variable, the corresponding elements of the independent variable and the dependent variable are set to zero; if there is a causal relationship between the independent variable and the dependent variable, the corresponding elements of the independent variable and the dependent variable are determined as parameters to be estimated to initialize the coefficient matrix, obtaining an initialized coefficient matrix; and estimating the coefficient matrix by minimizing the sum of squared residuals based on the dependent variable matrix, the initialized coefficient matrix, and the independent variable matrix to obtain the target coefficient matrix.

[0062] For example, the first variable refers to the set of variables located at both ends of any causal link in the stock flow diagram and included in the process of quantifying the impact coefficient. It includes at least independent variables as the source of influence and dependent variables as the affected objects. The independent variables are used to characterize the factors that affect other variables, and the dependent variables are used to characterize the objects whose values ​​change under the interaction relationship.

[0063] In this example, the first variable corresponds to the variable entity that occupies a row or column position in the independent variable matrix and the dependent variable matrix, respectively. Its value is obtained through time series observation and serves as the basic variable for constructing the coefficient matrix and estimating the influence coefficients of each causal link.

[0064] For example, first, all nodes in the stock flow graph are traversed, and the names of these nodes (representing variables) are collected into a list. Then, the role of each variable is distinguished according to the direction of all directed connections (causal links) in the graph. For any link x→y, x is an independent variable (cause), and y is a dependent variable (effect). The time series data of all identified unique independent variables are arranged in columns to construct the independent variable matrix X; at the same time, the time series data of all identified unique dependent variables are also arranged in columns to construct the dependent variable matrix Y.

[0065] For example, the independent variables include: electricity demand, grid loss coefficient, thermal power installed capacity, thermal power unit operating time, annual GDP growth rate, the proportion of carbon emissions from the power industry, carbon emission intensity, carbon dioxide emission coefficient of thermal power units, RPS quota growth rate, green power unit operating time, green power installed capacity, carbon price, electricity price, electricity supply, thermal power unit power generation, GDP, annual GDP growth, carbon quota supply, seller's carbon quota holdings, carbon quota sales, buyer's carbon quota holdings, carbon quota demand, expected purchases, expected sales of carbon quotas, thermal power plant profits, green power unit power generation, green certificate issuance, green power plant green certificate holdings, green certificate circulation, buyer's green certificate holdings, RPS quota, green certificate quantity required to meet RPS, expected purchase of green certificates, expected sale of green certificates, green certificate price, excess demand for green certificates, and changes in green certificate price.

[0066] The dependent variables include: electricity supply-demand ratio, carbon allowance supply-demand ratio, green energy producer profits, electricity price, electricity supply, thermal power generation, GDP, annual GDP growth, carbon allowance supply, sellers' carbon allowance holdings, carbon allowance sales, buyers' carbon allowance holdings, carbon allowance demand, expected purchases, expected sales, thermal power producer profits, green power generation, green certificate issuance, green energy producer green certificate holdings, green certificate circulation, buyers' green certificate holdings, RPS allowance, green certificate quantity required to meet RPS, expected green certificate purchases, expected green certificate sales, green certificate price, excess green certificate demand, and green certificate price changes.

[0067] For example, firstly, a k×m zero matrix is ​​created based on the number of columns in the independent variable matrix X (number of independent variables k) and the number of columns in the dependent variable matrix Y (number of dependent variables m), serving as the initial framework for the coefficient matrix. Next, the system iterates through each causal link in the stock flow graph. For each link x → y, the system finds the cell in the coefficient matrix whose row index corresponds to the independent variable x and whose column index corresponds to the dependent variable y. Since this causal link explicitly exists, the system changes the value of this cell from 0 to a special placeholder or marker indicating that it is a "parameter to be estimated". All cells without a direct causal link remain at 0. This matrix containing 0 and the "parameter to be estimated" marker is the initialized coefficient matrix.

[0068] According to the above implementation method, by extracting the variable set corresponding to the causal links from the stock flow graph, constructing the independent variable matrix, dependent variable matrix, and their corresponding coefficient matrix, and using the causal pointing relationship to structurally constrain and initialize the coefficient matrix, and then estimating the parameters of each influence coefficient by minimizing the sum of squared residuals, the target coefficient matrix is ​​obtained. In this way, qualitative causal structure and quantitative parameter estimation can be organically combined. On the one hand, this avoids invalid estimation of variable pairs without causal relationships, reducing model dimensionality and computational complexity; on the other hand, it ensures that the obtained coefficient matrix strictly conforms to the topological structure of the stock flow graph, which is beneficial to improving the stability, interpretability, and accuracy of subsequent system dynamics model simulation analysis of the parameter estimation results.

[0069] In one implementation, a target coefficient matrix is ​​obtained by minimizing the sum of squared residuals to estimate the coefficient matrix based on the dependent variable matrix, the initialized coefficient matrix, and the independent variable matrix. This includes: determining the residual matrix based on the difference between the product of the dependent variable matrix and the products of the initialized coefficient matrix and the independent variable matrix; constructing an objective function with the parameters to be estimated as variables by minimizing the sum of squared residuals based on the residual matrix; solving the objective function to obtain the coefficient values; and updating the values ​​to be estimated in the coefficient matrix based on the coefficient values ​​to obtain the target coefficient matrix.

[0070] For example, the dependent variable matrix Y, the initialized coefficient matrix A, and the independent variable matrix X are substituted into the linear relationship assumption Y = XA + ε. To solve for A that best reflects the true relationship between the variables, an objective function is constructed based on the dependent variable matrix Y, the initialized coefficient matrix A, and the independent variable matrix X. The solution method employs ordinary least squares (OLS), by applying the objective function... Find information about the parameters to be estimated. Take the partial derivatives of the objective function and set them to zero, then find the coefficients that minimize the objective function (i.e., the sum of squared residuals).

[0071] For example, the estimated values ​​of each band in the coefficient matrix are updated with the obtained coefficient values, and thus the target coefficient matrix can be obtained.

[0072] According to the above implementation method, by constructing the residual relationship between the observed values ​​of the dependent variable and the predicted values ​​of the model, the problem of solving the coefficient matrix is ​​transformed into a parameter optimization problem with the objective of minimizing the sum of squared residuals. Furthermore, by solving the objective function, quantitative estimation of each causal influence coefficient is achieved, ultimately yielding a target coefficient matrix that truly reflects the strength of the interaction between variables. In this way, while maintaining the model structure constraints, objective parameter estimation and automatic calibration can be achieved, which helps reduce biases caused by subjective settings, improves the accuracy of characterizing the influence strength of causal links and the overall model fitting effect, thereby enhancing the credibility and stability of subsequent system dynamics simulation analysis results.

[0073] In one implementation, a system dynamics model is constructed based on the stock flow diagram and the target coefficient matrix, including: defining a difference equation for each state variable and each rate variable in the stock flow diagram to obtain various difference equations; using the values ​​of each influence coefficient in the target coefficient matrix as weight coefficients of the corresponding independent variables in the difference equations to parameterize each difference equation; integrating the parameterized difference equations to obtain a system of simultaneous difference equations; and encapsulating the system of simultaneous difference equations to obtain the system dynamics model, wherein the system dynamics model is used to calculate the values ​​of each state variable, each rate variable, and each auxiliary variable at the current time step.

[0074] For example, first, all nodes in the stock-flow graph are traversed and categorized into state variables (stock), rate variables (flow), and auxiliary variables. Then, based on the fundamental principles of system dynamics, rules governing the time-varying behavior of each state and rate variable are defined. For each state variable, the system defines its value at the next time step as its current value plus the product of all inflow rates and the time step, minus the product of all outflow rates and the time step. For each rate variable, the system defines its value at the next time step as a function relating all its input variables (which can be state variables, other rate variables, or auxiliary variables). These equations, based on graph connectivity and describing how the variables evolve over time, are difference equations.

[0075] Understandably, the structure of the difference equation of a state variable is defined as the sum or subtraction of the previous time value of the state variable and the rate variables of all inflows or outflows of the state variable; the structure of the difference equation of a state variable is defined as the sum or subtraction of the previous time value of the state variable and the rate variables of all inflows or outflows of the state variable.

[0076] For example, the system obtains the target coefficient matrix estimated using methods such as least squares in the previous example. Then, the system iterates through all the difference equations defined in the previous step. For each equation, the system finds the value of the influence coefficient corresponding to its independent variable in the target coefficient matrix within its function. The system uses these specific values ​​as weight coefficients for the corresponding independent variables in the equation, thereby transforming a structured, qualitative equation into a computable parameterized difference equation containing precise numerical values.

[0077] Subsequently, the system of simultaneous difference equations obtained in the previous step is encapsulated into an executable function or computational module, which is the system dynamics model. The system dynamics model can be abstracted into the following functional expression: In the formula, The output of the model is the simulation result of how all variables in the system evolve over time. The inputs to the model are the initial values ​​of each variable in the system (state variables, rate variables, and auxiliary variables) and the external driving variables. These are the model parameters, also known as the target coefficient matrix; The graph structure of the model, i.e., the topology of the stock flow graph; This refers to the system dynamics calculation process.

[0078] The model receives the values ​​of each state variable, rate variable, and auxiliary variable at the current time (i.e., t=0), along with external input variables, as model input X. Internally, the model initiates a time loop, starting from t=0. At each time step, it strictly follows the calculation order defined by the simultaneous difference equations, using the current value to calculate the value at the next time step, and continuously iterates and updates. The model records all variable values ​​calculated at each time step, ultimately outputting a result set (i.e., output Y) containing time-series data of all variables throughout the entire simulation period. For example, it could provide complete data on the cumulative issuance of green certificates, the monthly increase in green certificates, and the price changes of green certificates over time, to reflect the evolution path of green certificate prices and their circulation path across multiple markets.

[0079] According to the above implementation method, by transforming the structural relationships in the stock-flow diagram into difference equations and parameterizing each equation in conjunction with the target coefficient matrix, a system of simultaneous difference equations capable of characterizing the interactions of multiple variables is constructed. Based on this, a system dynamics model is formed, enabling unified modeling and iterative calculation of the system state evolution over time. This not only rigorously inherits the causal logic and feedback structure expressed by the stock-flow diagram at the mathematical level but also transforms qualitative analysis into a computable and simulable quantitative model. This improves the accuracy and interpretability of system evolution analysis, providing a reliable basis for dynamic prediction and decision support of complex systems under different scenarios.

[0080] Figure 4 This is a structural block diagram of an analysis device for the circulation relationship of green certificates according to an embodiment of the present invention.

[0081] like Figure 4 As shown, the analysis device for the circulation relationship of the green certificate may include: The variable extraction module 510 is used to extract variables from a preset causal loop diagram to obtain each variable and the type of each variable, wherein the causal loop diagram represents the qualitative structure diagram of the circulation of green certificates among multiple markets. Granger causality test module 520 is used to perform Granger causality tests on each of the variables to obtain the Granger causal relationships between the variables and to construct a causal relationship matrix; The stock flow graph construction module 530 is used to construct a stock flow graph based on the types of each variable and the causal relationship matrix. The quantification estimation module 540 is used to quantify and estimate the influence coefficients of each causal link in the existing traffic diagram to obtain the target coefficient matrix. The model building module 550 is used to build a system dynamics model based on the stock flow diagram and the target coefficient matrix; The simulation calculation module 560 is used to perform simulation calculations on the system dynamics model to obtain the analysis results of the circulation relationship of the green certificate in multiple markets. The circulation analysis results include the evolution path of the green certificate price and the circulation path of the green certificate in multiple markets.

[0082] In one implementation, the Granger causality test module includes: A variable combination unit is used to combine each of the variables in pairs to obtain variable pairs, wherein each variable pair includes the explained variable and the explanatory variable; The first autoregressive model construction unit is used to construct each first autoregressive model based on the time series data and historical lag terms corresponding to each of the explained variables. The first autoregressive model is used to predict the current value of the explained variable based on the historical lag terms of the explained variable, and to obtain each first residual sum of squares. The second autoregressive model construction unit is used to construct each second autoregressive model based on the time series data and historical lag terms corresponding to each of the explained variables, and the time series data and historical lag terms corresponding to each of the explanatory variables. The second autoregressive model is used to predict the current value of the explained variable based on the historical lag terms of the explained variable and the historical lag terms of the explanatory variables, and to obtain each second residual sum of squares. The causal relationship hypothesis testing unit is used to perform causal relationship hypothesis testing on each of the variable pairs based on each of the first residual sum of squares and each of the second residual sum of squares, to obtain the causal relationship matrix.

[0083] In one implementation, the causal hypothesis testing includes: The statistics calculation subunit is used to calculate statistics based on the first residual sum of squares and the second residual sum of squares; The F-test subunit is used to perform an F-test based on the statistic to obtain the P-value of the F-test. The first directed connection subunit is used to determine that the explanatory variable constitutes a Granger causal relationship with the explained variable if the P value is less than a preset threshold, and to connect the explanatory variable and the explained variable in a directed manner to obtain the causal link of the variable pair. The causal relationship matrix determination subunit is used to determine the causal relationship matrix based on the causal links of each of the variable pairs.

[0084] In one implementation, the stock flow map construction module includes: The mapping unit is used to map the variables corresponding to each element in the causal relationship matrix to nodes in the graphical modeling environment according to the type of each variable, so as to obtain each node; The second directed connection unit is used to establish directed connections between the nodes based on the directed connection relationships between the variables corresponding to each element, thereby obtaining a topological connection structure. The verification unit is used to verify the topology connection structure based on a preset domain knowledge rule base, so as to eliminate or correct the connection links that do not conform to the domain logic, and obtain the existing traffic map.

[0085] In one implementation, the quantization estimation module includes: The first variable acquisition unit is used to acquire each first variable corresponding to each causal link in the stock flow diagram, wherein the first variable includes independent variables and dependent variables; A variable matrix construction unit is used to construct an independent variable matrix based on each of the independent variables, and to construct a dependent variable matrix based on each of the dependent variables; A coefficient matrix construction unit is used to construct a coefficient matrix based on the dimensions of the independent variable matrix and the dependent variable matrix, wherein each element in the coefficient matrix represents the influence coefficient of the independent variable on the dependent variable; The coefficient matrix initialization unit is used to initialize the coefficient matrix by setting the corresponding elements of the independent variable and the dependent variable to zero if there is no causal relationship between the independent variable and the dependent variable, and by determining the corresponding elements of the independent variable and the dependent variable as parameters to be estimated if there is a causal relationship between the independent variable and the dependent variable, so as to obtain the initialized coefficient matrix. An estimation unit is used to estimate the coefficient matrix by minimizing the sum of squared residuals based on the dependent variable matrix, the initialized coefficient matrix, and the independent variable matrix, thereby obtaining the target coefficient matrix.

[0086] In one embodiment, the estimation unit includes: The residual matrix determination subunit is used to determine the residual matrix based on the difference between the product of the dependent variable matrix and the product of the initialized coefficient matrix and the independent variable matrix. The objective function construction sub-unit is used to construct an objective function with the parameter to be estimated as a variable by minimizing the sum of squared residuals based on the residual matrix. The solution sub-unit is used to solve the objective function and obtain the values ​​of each coefficient. An update subunit is used to update each of the estimated values ​​in the coefficient matrix based on each of the coefficient values, thereby obtaining the target coefficient matrix.

[0087] In one implementation, the model building module includes: The difference equation definition unit is used to define a difference equation for each state variable and each rate variable in the stock flow diagram to obtain the difference equations. The parameterization unit is used to take the values ​​of each influence coefficient in the target coefficient matrix as the weight coefficients of the corresponding independent variables in the difference equation, so as to parameterize each of the difference equations. The integration unit is used to integrate the various parameterized difference equations to obtain a system of simultaneous difference equations. An encapsulation unit is used to encapsulate the simultaneous difference equations to obtain the system dynamics model, wherein the system dynamics model is used to calculate the values ​​at the next time step based on the values ​​of each of the state variables, each of the rate variables, and each of the auxiliary variables at the current time step.

[0088] The specific functions and examples of each module and submodule of the system in this embodiment of the invention can be found in the relevant descriptions of the corresponding steps in the above method embodiments, and will not be repeated here.

[0089] The acquisition, storage, and application of user personal information involved in the technical solution of this invention all comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0090] This invention also provides an analysis system for the transfer relationship of green certificates, comprising: At least one processor; and a memory communicatively connected to said at least one processor; The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the method described in any one of the embodiments of the present invention.

[0091] The beneficial effects of the green certificate circulation relationship analysis system in this embodiment of the invention are equivalent to the beneficial effects of the green certificate circulation relationship analysis method described above, and will not be repeated here.

[0092] This invention also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to perform the method described in any one of the embodiments of this invention.

[0093] The beneficial effects of the storage medium of the present invention are equivalent to the beneficial effects of the above-mentioned method for analyzing the circulation relationship of green certificates, and will not be repeated here.

[0094] Figure 5 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present invention is shown. Electronic device 800 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 800 may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0095] like Figure 5 As shown, the electronic device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. The RAM 803 may also store various programs and data required for the operation of the electronic device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0096] Multiple components in electronic device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of displays, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows electronic device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0097] The computing unit 801 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as the method for analyzing the flow of green certificates. For example, in some embodiments, the method for analyzing the flow of green certificates can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the method for analyzing the flow of green certificates described above can be performed. Alternatively, in other embodiments, the computing unit 801 may be configured by any other suitable means (e.g., by means of firmware) to perform an analysis method for green certificate transfer relationships.

[0098] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0099] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0100] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0101] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0102] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0103] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0104] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this invention can be achieved, and this is not limited herein.

[0105] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of this invention should be included within the scope of protection of this invention.< / time> < / time>

Claims

1. A method for analyzing the transfer relationship of green certificates, characterized in that, include: Extract the variables from the preset causal loop diagram to obtain each variable and the type of each variable, wherein the causal loop diagram represents the qualitative structure diagram of the circulation of green certificates among multiple markets; Granger causality tests were performed on each of the variables to obtain the Granger causal relationships between the variables, and a causal relationship matrix was constructed. Based on the types of each variable and the causal relationship matrix, a stock flow diagram is constructed; The impact coefficients of each causal link in the existing traffic graph are quantitatively estimated to obtain the target coefficient matrix; Based on the existing flow diagram and the target coefficient matrix, a system dynamics model is constructed; The system dynamics model is simulated to obtain the analysis results of the circulation relationship of the green certificate in multiple markets. The circulation analysis results include the evolution path of the green certificate price and the circulation path of the green certificate in multiple markets.

2. The method according to claim 1, characterized in that, The step of performing Granger causality tests on each of the variables to obtain the Granger causal relationships between the variables and constructing a causal relationship matrix includes: Each of the variables is combined in pairs to obtain variable pairs, wherein each variable pair includes the explained variable and the explanatory variable; Based on the time series data and historical lag terms corresponding to each of the explained variables, each first autoregressive model is constructed, wherein the first autoregressive model is used to predict the current value of the explained variable based on the historical lag terms of the explained variable, and to obtain each first residual sum of squares; Based on the time series data and historical lag terms corresponding to each of the explained variables, and the time series data and historical lag terms corresponding to each of the explanatory variables, each second autoregressive model is constructed. The second autoregressive model is used to predict the current value of the explained variable based on the historical lag terms of the explained variable and the historical lag terms of the explanatory variables, and to obtain each second residual sum of squares. Based on each of the first residual sums of squares and each of the second residual sums of squares, a causal relationship hypothesis test is performed on each of the variable pairs to obtain the causal relationship matrix.

3. The method according to claim 2, characterized in that, The step of performing causal hypothesis testing on each pair of variables based on the first residual sum of squares and the second residual sum of squares to obtain the causal relationship matrix includes: Calculate the statistic based on the first residual sum of squares and the second residual sum of squares; Perform an F-test based on the statistic to obtain the P-value of the F-test; If the P value is less than a preset threshold, then it is determined that the explanatory variable and the explained variable constitute a Granger causal relationship, and the explanatory variable and the explained variable are connected in a directed manner to obtain the causal link of the variable pair; The causal relationship matrix is ​​determined based on the causal links of each variable pair.

4. The method according to claim 1, characterized in that, The construction of the stock flow graph based on the types of each variable and the causal relationship matrix includes: Based on the type of each variable, the variables corresponding to each element in the causal relationship matrix are mapped to nodes in the graphical modeling environment to obtain each node; Based on the directed connection relationship between the variables corresponding to each element, directed connections are established between each node to obtain the topological connection structure; Based on a preset domain knowledge rule base, the topology connection structure is verified to eliminate or correct connection links that do not conform to domain logic, thereby obtaining the existing traffic graph.

5. The method according to claim 1, characterized in that, The quantitative estimation of the impact coefficients of each causal link in the existing traffic graph, to obtain the target coefficient matrix, includes: Obtain each first variable corresponding to each causal link in the existing traffic graph, wherein the first variable includes an independent variable and a dependent variable; Based on each of the independent variables, construct an independent variable matrix, and based on each of the dependent variables, construct a dependent variable matrix; Based on the dimensions of the independent variable matrix and the dependent variable matrix, a coefficient matrix is ​​constructed, wherein each element in the coefficient matrix represents the influence coefficient of the independent variable on the dependent variable; If there is no causal relationship between the independent variable and the dependent variable, the corresponding elements of the independent variable and the dependent variable are set to zero. If there is a causal relationship between the independent variable and the dependent variable, the corresponding elements of the independent variable and the dependent variable are determined as parameters to be estimated, so as to initialize the coefficient matrix and obtain the initialized coefficient matrix. Based on the dependent variable matrix, the initialized coefficient matrix, and the independent variable matrix, the target coefficient matrix is ​​obtained by minimizing the sum of squared residuals to estimate the coefficient matrix.

6. The method according to claim 5, characterized in that, The step of estimating the target coefficient matrix by minimizing the sum of squared residuals based on the dependent variable matrix, the initialized coefficient matrix, and the independent variable matrix includes: The residual matrix is ​​determined based on the difference between the dependent variable matrix and the product of the initialized coefficient matrix and the independent variable matrix. Based on the residual matrix, an objective function with the parameter to be estimated as a variable is constructed by minimizing the sum of squared residuals. Solve the objective function to obtain the values ​​of each coefficient; Based on each of the coefficient values, each of the values ​​to be estimated in the coefficient matrix is ​​updated to obtain the target coefficient matrix.

7. The method according to claim 1, characterized in that, The construction of a system dynamics model based on the stock flow diagram and the target coefficient matrix includes: Define a difference equation for each state variable and each rate variable in the stock flow diagram to obtain the various difference equations; The values ​​of each influence coefficient in the target coefficient matrix are used as weight coefficients of the corresponding independent variables in the difference equations to parameterize each difference equation. The parameterized difference equations are integrated to obtain a system of simultaneous difference equations. The system of simultaneous difference equations is encapsulated to obtain the system dynamics model, wherein the system dynamics model is used to calculate the values ​​of each state variable, each rate variable and each auxiliary variable at the current time.

8. An analysis device for the circulation relationship of green certificates, characterized in that, include: The variable extraction module is used to extract variables from a preset causal loop diagram to obtain each variable and the type of each variable. The causal loop diagram represents the qualitative structure of the green certificate circulating among multiple markets. The Granger causality test module is used to perform Granger causality tests on each of the variables to obtain the Granger causal relationships between the variables and construct a causal relationship matrix. The stock flow graph construction module is used to construct a stock flow graph based on the types of each variable and the causal relationship matrix. The quantification estimation module is used to quantify and estimate the influence coefficients of each causal link in the existing traffic diagram to obtain the target coefficient matrix; The model building module is used to construct a system dynamics model based on the stock flow diagram and the target coefficient matrix; The simulation calculation module is used to perform simulation calculations on the system dynamics model to obtain the analysis results of the circulation relationship of the green certificate in multiple markets. The circulation analysis results include the evolution path of the green certificate price and the circulation path of the green certificate in multiple markets.

9. A system for analyzing the transfer relationship of green certificates, characterized in that, include: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-7.

Images