Method, device and system for generating future scenario of low-carbon energy social economy

Abstract

The invention provides a low-carbon energy social economy future scene generation method, device and system. The method comprises the following steps: acquiring interaction data of social economy and an energy system; on the basis of the interaction data, key driving factor identification processing and state definition processing are carried out, and a plurality of core driving factors influencing low-carbon transformation and state definition information corresponding to the core driving factors are obtained; generating a cross influence matrix corresponding to the plurality of core driving factors; generating a plurality of candidate scenes based on the state definition information of the plurality of core driving factors, performing stability scoring on the plurality of candidate scenes based on the cross influence matrix, and determining a plurality of target scenes from the plurality of candidate scenes according to a stability scoring result; and performing quantitative modeling based on the target scene to generate a low-carbon energy social economy future scene model.

Description

Technical Field

[0001] This application relates to the technical field of energy economic system modeling and climate policy scenario analysis, specifically to a method, device, and system for generating future socio-economic scenarios for low-carbon energy. Background Technology

[0002] Energy economic scenario analysis is a forward-looking systems analysis method. Its core is to construct multiple differentiated scenario plans by setting different future development assumptions, simulating the interaction between the energy system and the macroeconomic system, and then assessing the impact of different energy development paths on economic growth, industrial structure, energy security, carbon emissions, and other dimensions. This provides a scientific basis for energy policy formulation, industrial planning, and investment decisions. Low-carbon energy socio-economic future scenario generation refers to constructing a series of logical, quantifiable, and differentiated future development blueprints based on predictions of future energy transition, policy guidance, technological progress, and socio-economic evolution. These blueprints are used to analyze the interaction between low-carbon energy and the socio-economic system under different paths, providing a basis for policy formulation, planning, and risk assessment.

[0003] Existing energy economic scenario analysis methods have significant limitations in complex system modeling, qualitative and quantitative integration, and scenario consistency assurance, which affects the scientific rigor, transparency, and adaptability of subsequent policy formulation based on the generated low-carbon energy socio-economic future scenarios. Summary of the Invention

[0004] In view of this, this application provides a method, apparatus and system for generating a low-carbon energy socio-economic future scenario.

[0005] Specifically, this application is implemented through the following technical solution: In a first aspect, embodiments of this disclosure provide a method for generating a low-carbon energy socio-economic future scenario, the method comprising: Acquire interactive data between socioeconomic and energy systems; Based on the interactive data, key driving factor identification and state definition processing are performed to obtain multiple core driving factors affecting low-carbon transformation, as well as state definition information corresponding to each core driving factor. Generate multiple cross-influence matrices corresponding to the core driving factors; wherein each element in the cross-influence matrix is ​​used to characterize the influence intensity information of the first core driving factor on the second core driving factor among the two core driving factors corresponding to that element. Based on the state definition information of multiple core driving factors, multiple candidate scenarios are generated. Stability scores are performed on each of the multiple candidate scenarios based on the cross-influence matrix. Based on the stability score results, multiple target scenarios are determined from the multiple candidate scenarios. Based on the target scenario, quantitative modeling is performed to generate a future scenario model for low-carbon energy socio-economic development.

[0006] Optionally, the interactive data includes: structured data and unstructured data; The structured data includes at least one of the following: energy consumption data, carbon emission data, economic indicator data, and basic parameter data; The unstructured data includes: policy text data, technical document data, public opinion data, and enterprise operation data.

[0007] Optionally, acquiring the interaction data between the socio-economic and energy systems includes: Acquire raw interaction data of socio-economic and energy systems; the raw interaction data includes: raw structured data and raw unstructured data; The original structured data is sequentially cleaned and normalized to obtain the structured data. as well as, The original unstructured data is cleaned, and a pre-trained natural language model is used to perform semantic understanding processing on the cleaned unstructured data to obtain semantic feature information corresponding to the unstructured data. The semantic feature information is then used as the unstructured data.

[0008] Optionally, based on the interactive data, key driving factor identification and state definition processing are performed to obtain multiple core driving factors affecting the low-carbon transformation, including: With the goal of extracting candidate factors influencing low-carbon transformation, literature mining and policy text analysis were performed based on the interactive data. Based on the results of the literature mining and policy text analysis, several first candidate factors influencing low-carbon transformation were obtained. Through the expert consultation interface, expert consultation information is obtained, and based on the expert consultation information, several second candidate factors affecting the low-carbon transformation are obtained. Clustering is performed on the first candidate factor and the second candidate factor to obtain multiple candidate factor categories; By ranking the importance of multiple candidate factor categories, core driving factors covering socio-economic and energy technologies are obtained.

[0009] Optionally, generating the cross-influence matrix corresponding to the multiple core driving factors includes: Through multiple rounds of anonymous expert consultation, using structured questionnaires and scoring interfaces, we collected scores from different experts on the influence relationship between each pair of core driving factors, and generated candidate cross-influence matrices. The candidate cross-influence matrices are subjected to consistency verification and matrix filling optimization to obtain the cross-influence matrices corresponding to the multiple core driving factors.

[0010] Optionally, the stability scoring of the various candidate scenarios based on the cross-influence matrix includes: For each candidate scenario, based on the cross-influence matrix, the cross-influence direction and state gradient difference between every two core driving factors in the candidate scenario are determined, and based on the cross-influence direction and the state gradient difference, the state compatibility coefficient of every two core driving factors in the corresponding candidate scenario is determined. And, based on the cross-influence matrix, determine the influence weight between each pair of core driving factors; Based on the state compatibility coefficient between each pair of core driving factors and the corresponding influence weight, the stability score corresponding to each candidate scenario is determined.

[0011] Optionally, based on the stability score results, the determination of multiple target scenarios from the multiple candidate scenarios includes: The candidate scenarios are sorted in descending order of their respective stability scores. According to the sorting, a preset number of candidate scenarios are determined from the candidate scenarios as the target scenarios; or, The stability scores and preset score thresholds corresponding to the multiple candidate scenarios are compared respectively; Candidate scenarios with stability scores greater than the preset score threshold are selected as target scenarios.

[0012] Optionally, the step of generating a low-carbon energy socio-economic future scenario model based on the target scenario through quantitative modeling includes: Density clustering is performed on the various target scenarios to obtain multiple scenario patterns, and semantic labels and policy interpretations corresponding to each scenario pattern are generated; Based on the scenario patterns, corresponding semantic tags, and policy interpretations, multiple scenario patterns are modeled to obtain low-carbon energy socio-economic future scenario models corresponding to each scenario pattern.

[0013] Secondly, embodiments of this disclosure also provide a device for generating a low-carbon energy socio-economic future scenario, the device comprising: The acquisition module is used to acquire interactive data between socio-economic and energy systems. The factor identification module is used to perform key driving factor identification and state definition processing based on the interactive data, so as to obtain multiple core driving factors affecting low-carbon transformation and state definition information corresponding to each core driving factor. A matrix construction module is used to generate cross-influence matrices corresponding to multiple core driving factors; wherein each element in the cross-influence matrix is ​​used to characterize the influence intensity information of the first core driving factor on the second core driving factor among the two core driving factors corresponding to that element. The scenario screening module is used to generate multiple candidate scenarios based on the state definition information of multiple core driving factors, perform stability scoring on the multiple candidate scenarios based on the cross-influence matrix, and determine multiple target scenarios from the multiple candidate scenarios based on the stability scoring results. The quantitative modeling module is used to perform quantitative modeling based on the target scenario and generate a future scenario model of low-carbon energy socio-economic development.

[0014] Optionally, the interactive data includes: structured data and unstructured data; The structured data includes at least one of the following: energy consumption data, carbon emission data, economic indicator data, and basic parameter data; The unstructured data includes: policy text data, technical document data, public opinion data, and enterprise operation data.

[0015] Optionally, the acquisition module, when acquiring interaction data between the socio-economic and energy systems, is used to: Acquire raw interaction data of socio-economic and energy systems; the raw interaction data includes: raw structured data and raw unstructured data; The original structured data is sequentially cleaned and normalized to obtain the structured data. as well as, The original unstructured data is cleaned, and a pre-trained natural language model is used to perform semantic understanding processing on the cleaned unstructured data to obtain semantic feature information corresponding to the unstructured data. The semantic feature information is then used as the unstructured data.

[0016] Optionally, the factor identification module, when performing key driving factor identification processing and state definition processing based on the interactive data to obtain multiple core driving factors affecting low-carbon transformation, is used for: With the goal of extracting candidate factors influencing low-carbon transformation, literature mining and policy text analysis were performed based on the interactive data. Based on the results of the literature mining and policy text analysis, several first candidate factors influencing low-carbon transformation were obtained. Through the expert consultation interface, expert consultation information is obtained, and based on the expert consultation information, several second candidate factors affecting the low-carbon transformation are obtained. Clustering is performed on the first candidate factor and the second candidate factor to obtain multiple candidate factor categories; By ranking the importance of multiple candidate factor categories, core driving factors covering socio-economic and energy technologies are obtained.

[0017] Optionally, the matrix construction module, when generating the cross-influence matrix corresponding to multiple core driving factors, is used to: Through multiple rounds of anonymous expert consultation, using structured questionnaires and scoring interfaces, we collected scores from different experts on the influence relationship between each pair of core driving factors, and generated candidate cross-influence matrices. The candidate cross-influence matrices are subjected to consistency verification and matrix filling optimization to obtain the cross-influence matrices corresponding to the multiple core driving factors.

[0018] Optionally, the matrix construction module performs stability scoring on various candidate scenarios based on the cross-influence matrix, including: For each candidate scenario, based on the cross-influence matrix, the cross-influence direction and state gradient difference between every two core driving factors in the candidate scenario are determined, and based on the cross-influence direction and the state gradient difference, the state compatibility coefficient of every two core driving factors in the corresponding candidate scenario is determined. And, based on the cross-influence matrix, determine the influence weight between each pair of core driving factors; Based on the state compatibility coefficient between each pair of core driving factors and the corresponding influence weight, the stability score corresponding to each candidate scenario is determined.

[0019] Optionally, the scenario filtering module, when determining multiple target scenarios from multiple candidate scenarios based on stability scoring results, is used to: The candidate scenarios are sorted in descending order of their respective stability scores. According to the sorting, a preset number of candidate scenarios are determined from the candidate scenarios as the target scenarios; or, The stability scores and preset score thresholds corresponding to the multiple candidate scenarios are compared respectively; Candidate scenarios with stability scores greater than the preset score threshold are selected as target scenarios.

[0020] Optionally, the quantitative modeling module, when generating a low-carbon energy socio-economic future scenario model based on the target scenario through quantitative modeling, is used for: Density clustering is performed on the various target scenarios to obtain multiple scenario patterns, and semantic labels and policy interpretations corresponding to each scenario pattern are generated; Based on the scenario patterns, corresponding semantic tags, and policy interpretations, multiple scenario patterns are modeled to obtain low-carbon energy socio-economic future scenario models corresponding to each scenario pattern.

[0021] Thirdly, this disclosure also provides a low-carbon energy socio-economic future scenario generation system, including a memory, a processor, an expert evaluation terminal, a policy decision-making terminal, and an energy system model interface. The memory stores a low-carbon energy socio-economic future scenario model generated based on the low-carbon energy socio-economic future scenario generation method described in the first aspect or any one of the first aspects. The energy system model interface is used to receive energy system models; the expert evaluation terminal is used to receive expert evaluation information. The processor is further configured to couple at least one of the received energy system model and the expert evaluation information with the low-carbon energy socio-economic future scenario model to obtain a policy decision, and output the policy decision to the policy decision terminal.

[0022] Fourthly, an optional implementation of this disclosure also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the first aspect above, or any possible implementation of the first aspect.

[0023] Fifthly, an optional implementation of this disclosure also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the first aspect or any possible implementation of the first aspect.

[0024] In a sixth aspect, an optional implementation of this disclosure also provides a computer program product carrying program code, the program code including instructions that can be used to perform steps as described in the first aspect or any possible implementation of the first aspect.

[0025] This embodiment of the disclosure acquires interaction data between socio-economic and energy systems, and performs key driver factor identification and state definition processing based on this interaction data to obtain multiple core driver factors affecting low-carbon transformation and their corresponding state definition information. Then, a cross-influence matrix is ​​generated for these core driver factors, where each element characterizes the influence intensity between the two core driver factors corresponding to that element. Next, based on the state definition information of the multiple core driver factors, various candidate scenarios are generated. Stability scores are then performed on each candidate scenario based on the cross-influence matrix, and multiple target scenarios are determined from the candidate scenarios based on the stability score results. Finally, the target scenarios are quantitatively modeled to obtain a low-carbon energy socio-economic future scenario model corresponding to the target scenarios. In this process, the complex interaction relationship between the socio-economic and energy systems is established by generating the cross-influence matrix. By integrating cross-influence balance analysis, a complete and scalable low-carbon energy scenario generation system is constructed, effectively solving the limitations of traditional methods in complex system modeling, qualitative and quantitative fusion, and scenario consistency assurance, and significantly improving the scientific rigor, transparency, and adaptability of energy and climate policy formulation.

[0026] To make the above-mentioned objects, features and advantages of this disclosure more apparent and understandable, preferred embodiments are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating an exemplary embodiment of the present disclosure of a method for generating a low-carbon energy socio-economic future scenario; Figure 2 This is a schematic diagram of the structure of a system shown in an exemplary embodiment of this disclosure; Figure 3 This is a schematic diagram of a low-carbon energy socio-economic future scenario generation device shown in an exemplary embodiment of the present disclosure; Figure 4 This is a schematic diagram of a low-carbon energy socio-economic future scenario generation system, as illustrated in an exemplary embodiment of this disclosure. Detailed Implementation

[0028] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0029] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0030] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0031] The core of energy economic scenario analysis is to construct multiple differentiated scenario schemes by setting different future development assumptions, simulating the interaction between the energy system and the macroeconomic system, and then assessing the impact of different energy development paths on economic growth, industrial structure, energy security, carbon emissions, and other dimensions, so as to provide a scientific basis for energy policy formulation, industrial planning and investment decisions.

[0032] When conducting energy economic scenario analysis, various different energy economic scenarios can be generated, and the following objectives can be achieved based on the generated economic scenarios: Revealing the linkage mechanism: clarifying how factors such as energy prices, energy structure, and energy technology progress affect economic growth, industrial competitiveness, and residents' consumption costs; at the same time, analyzing the driving effect of economic growth models (such as extensive and intensive) on energy demand.

[0033] Evaluate policy effectiveness: Simulate the economic and energy effects of different energy policies (such as carbon tax, new energy subsidies, and total energy consumption control) to determine the feasibility and cost-effectiveness of the policies.

[0034] Anticipate risks and opportunities: Identify potential risks in the energy transition process (such as employment pressure caused by the decline of traditional energy industries and excessively high costs due to the immaturity of new energy technologies), as well as new development opportunities (such as economic growth points in the new energy industry chain).

[0035] Optimize development path: Compare energy and economic performance under different scenarios to select the optimal or second-best development path that balances energy security, economic growth, and low-carbon emission reduction.

[0036] The generated energy economic scenarios typically include the following centralizations: Baseline scenario: This assumes no new low-carbon energy policies are introduced, and the current energy development model and economic growth path continue. It is typically used as a reference point to measure the relative magnitude of change in other policy scenarios.

[0037] Policy reinforcement scenario: Strictly implement certain policy objectives and increase policy support. This is often used to assess the impact of radical policies on the energy system and the economy.

[0038] Technological Breakthrough Scenario: This scenario assumes that a certain key energy technology (such as high-efficiency energy storage or green hydrogen) achieves a breakthrough and is applied on a large scale. It is usually used to analyze the disruptive effect of technological innovation on the energy economic system.

[0039] Risk scenarios: These assume extreme external conditions and are typically used to predict economic resilience and response strategies in the face of risks such as energy supply disruptions.

[0040] In related technologies, energy economic scenario analysis methods mostly rely on top-down model-driven or parameter extrapolation approaches. While these methods can provide quantitative outputs, they have the following limitations: 1. It is difficult to handle the complex nonlinear interactions between socioeconomic systems and energy systems; 2. Lack of systematic integration of qualitative factors (such as policy intentions and changes in social behavior) and quantitative factors (such as technology costs and carbon emissions); 3. The scenario generation process lacks an automated evaluation mechanism for internal consistency and logical self-consistency; 4. Traditional methods often rely on a single model or a limited number of parameters, making it difficult to reflect multiple uncertainties and cross-system feedback.

[0041] Therefore, a scenario generation system is needed that can integrate qualitative expert knowledge with quantitative model simulation, and has the ability to verify scenario consistency and output multiple dimensions, in order to improve the scientific nature and operability of low-carbon energy policy formulation.

[0042] The shortcomings of the above solutions are the result of the inventor's practical experience and careful research. Therefore, the discovery process of the above problems and the solutions proposed in this disclosure below should be considered as the inventor's contribution to this disclosure.

[0043] To facilitate understanding of this embodiment, a detailed description of the method for generating a low-carbon energy socio-economic future scenario disclosed in this disclosure will be provided first. The execution entity of the method for generating a low-carbon energy socio-economic future scenario provided in this disclosure is generally a computer device with a certain computing capability. This computer device may include, for example, a terminal device, a server, or other processing equipment. The terminal device may be a user equipment (UE), a mobile device, a user terminal, a terminal, an in-vehicle device, a wearable device, etc. In some possible implementations, this method for generating a low-carbon energy socio-economic future scenario can be implemented by a processor calling computer-readable instructions stored in memory.

[0044] The following describes the method for generating future socio-economic scenarios for low-carbon energy provided in the embodiments of this disclosure.

[0045] See Figure 1 The diagram shows a flowchart of a method for generating a low-carbon energy socio-economic future scenario according to an embodiment of this disclosure. The method includes steps S101 to S105, wherein: S101: Acquire interactive data between socioeconomic and energy systems; S102: Based on the interactive data, perform key driving factor identification and state definition processing to obtain multiple core driving factors affecting low-carbon transformation and state definition information corresponding to each core driving factor. S103: Generate a cross-influence matrix corresponding to multiple core driving factors; wherein, each element in the cross-influence matrix is ​​used to characterize the influence intensity information between the two core driving factors corresponding to the element; S104: Based on the state definition information of multiple core driving factors, generate multiple candidate scenarios, perform stability scoring on each of the multiple candidate scenarios based on the cross-influence matrix, and determine multiple target scenarios from the multiple candidate scenarios based on the stability scoring results; S105: Based on the target scenario, perform quantitative modeling to generate a future scenario model for low-carbon energy socio-economic development.

[0046] This embodiment of the disclosure acquires interaction data between socio-economic and energy systems, and performs key driver factor identification and state definition processing based on this interaction data to obtain multiple core driver factors affecting low-carbon transformation and their corresponding state definition information. Then, a cross-influence matrix is ​​generated for these core driver factors, where each element characterizes the influence intensity between the two core driver factors corresponding to that element. Next, based on the state definition information of the multiple core driver factors, various candidate scenarios are generated. Stability scores are then performed on each candidate scenario based on the cross-influence matrix, and multiple target scenarios are determined from the candidate scenarios based on the stability score results. Finally, the target scenarios are quantitatively modeled to obtain a low-carbon energy socio-economic future scenario model corresponding to the target scenarios. In this process, the complex interaction relationship between the socio-economic and energy systems is established by generating the cross-influence matrix. By integrating cross-influence balance analysis, a complete and scalable low-carbon energy scenario generation system is constructed, effectively solving the limitations of traditional methods in complex system modeling, qualitative and quantitative fusion, and scenario consistency assurance, and significantly improving the scientific rigor, transparency, and adaptability of energy and climate policy formulation.

[0047] The following provides a detailed explanation of S101 to S105.

[0048] Regarding the above S101: In practice, the interaction data between the socio-economic and energy systems is the core link between energy transition and economic development. This includes structured data that can be directly quantified and stored in a standardized manner, as well as unstructured data that has no fixed format and requires further processing and extraction.

[0049] Structured data includes: a1: Energy consumption data, which includes, for example: ① Energy consumption by type, such as coal, coal, natural gas, hydropower, wind power, photovoltaic, etc.; ② Energy consumption by sector, such as energy consumption corresponding to industry, construction, transportation, and residential use; ③ Total energy production and import / export volume; ④ Energy consumption per unit of GDP and energy consumption per unit of industrial added value; ⑤ Energy consumption growth rate and per capita energy consumption, etc.

[0050] a2: Carbon emission data, which includes, for example: ① carbon emissions by sector, such as carbon emission data corresponding to industry, energy production, transportation, and construction; ② carbon emission factors by energy type; ③ total regional carbon emissions and carbon emission intensity; ④ carbon sink volume; ⑤ corporate carbon emission quotas, carbon trading volume, and prices.

[0051] A3: Economic indicators and data, including: ① Macroeconomic data: GDP total / growth rate, output value and proportion of the three industries; ② Industrial economic data: output value of high energy-consuming industries, output value of new energy industries, scale of energy investment; ③ Employment data: number of employees in traditional energy industries, number of employees in new energy industries; ④ Price data: energy product prices (such as electricity prices, oil prices), PPI / CPI, enterprise production costs; ⑤ Trade data: import and export value of energy products, import and export volume of new energy equipment, etc.

[0052] a4: Basic parameter data, which include, for example: ① Energy technology efficiency: coal consumption for thermal power supply, conversion efficiency of photovoltaic power plants, and charging and discharging efficiency of energy storage systems; ② Technology cost: LCOE of photovoltaic / wind power, unit investment cost of energy storage, and CCS technology cost; ③ Installed capacity parameters: installed capacity of renewable energy and utilization hours of units; ④ Technology penetration rate: proportion of new energy vehicles and installation rate of smart meters.

[0053] The structured data mentioned above may come from reports, statements, announcements, etc., from relevant departments.

[0054] Unstructured data includes: b1: Policy text data, which includes, for example: energy policies, low-carbon policy documents (such as dual-carbon action plans, new energy subsidy policies, and draft carbon tax policies), government work reports, planning outlines, etc.

[0055] The policy text data can come from sources such as government websites and policy platforms.

[0056] b2: Technical literature data, which includes, for example, energy technology patents, academic papers, technology research and development reports, industry technical standards, etc.

[0057] The technical literature data can come from relevant technical websites, databases, etc.

[0058] b3: Social sentiment data, which includes, for example: surveys on residents' acceptance of new energy sources (such as the NIMBY effect of wind and solar power), interviews with companies on their willingness to reduce emissions, media reports, and social media comments.

[0059] The aforementioned public opinion data may come from sources such as surveys and news media platforms.

[0060] b4: Enterprise operational data, which includes, for example, energy company annual reports, social responsibility reports, production logs, and equipment operation and maintenance records.

[0061] The aforementioned enterprise operation data can be obtained from sources such as the company's official website, listed company announcements, and industry research materials.

[0062] Specifically, this disclosure also provides a specific method for acquiring interactive data of socio-economic and energy systems, including: Acquire raw interaction data of socio-economic and energy systems; the raw interaction data includes: raw structured data and raw unstructured data; The original structured data is sequentially cleaned and normalized to obtain the structured data. as well as, The original unstructured data is cleaned, and a pre-trained natural language model is used to perform semantic understanding processing on the cleaned unstructured data to obtain semantic feature information corresponding to the unstructured data. The semantic feature information is then used as the unstructured data.

[0063] In practice, data cleaning is the core step that connects the original data input with subsequent analysis. Its purpose is to remove invalid data, correct erroneous data, standardize data definitions, handle missing values, and ensure the accuracy of structured data and the usability of unstructured data.

[0064] For structured data, which typically has a clear format and defined metrics, data cleaning aims to address issues of data integrity, consistency, and accuracy. The data cleaning process for raw structured data includes, for example, the following steps: c1: Data integrity check and missing value handling.

[0065] Here, for raw structured data, for example, indicator fields can be identified to determine problems such as missing fields, missing records, and time series breakpoints.

[0066] To address the issue of missing data, the mean or median can be used to fill in the missing data. For example, the average renewable energy penetration rate of surrounding years can be used to fill in the missing values.

[0067] To address the issue of missing time series data, linear interpolation or trend extrapolation can be used to supplement the missing data. For example, based on data from 2019 and 2021, interpolation can be used to calculate the energy consumption per unit of GDP in 2020.

[0068] To address the issue of missing key indicators, a method such as removing samples and adding annotations can be used. For example, if a province has a large number of missing carbon sequestration data, the sample can be removed directly, and this can be noted in the data.

[0069] For cases where there is a clear logical relationship, logical deduction can be used to fill in the gaps, such as using "total energy consumption × carbon emission factor" to derive the missing carbon emissions.

[0070] c2: Data consistency check: The purpose is to standardize data definitions and eliminate inconsistencies caused by differences in statistical standards, scope, and units. Examples include: ① Consistency of statistical standards: Unified spatial scope: For example, "regional energy consumption" needs to be clearly defined as either "administrative jurisdiction" or "industrial park" to avoid mixing city and province data; Standardize the time range: For example, "annual data" needs to be confirmed as either a calendar year (January-December) or a statistical year (October of the previous year to September of the current year). Unified indicator definition: For example, "renewable energy penetration rate" needs to be clarified as either "installed capacity penetration rate" or "power generation penetration rate" and should be consistent throughout the entire process.

[0071] ② Data logical consistency: Verify the logical relationships between indicators and eliminate contradictory data.

[0072] Example: If "total industrial energy consumption" < "energy consumption of the steel industry", it is considered a logical error and needs to be corrected.

[0073] ③Unit consistency: Standardize the units of measurement, such as "ten thousand tons of standard coal", "ten thousand kilowatt-hours", and "tons of CO2" according to conversion formulas, to avoid calculation errors caused by the misuse of units.

[0074] c3: Outlier identification and handling: Outliers refer to extreme values ​​that deviate from the normal data distribution. They typically include: true outliers and erroneous outliers.

[0075] Among these methods, statistical methods and business methods can be used to filter out outliers from structured data.

[0076] Errors and outliers can be corrected directly. For example, the entry "wind power cost 1000 yuan / kWh" is obviously incorrect and should be corrected to "0.3 yuan / kWh".

[0077] For genuine outliers, retain the data and provide annotations to avoid their inclusion in subsequent trend analysis.

[0078] c4: Data Redundancy Removal Remove duplicate records: If a company's carbon emission data is entered multiple times, retain only one valid record; Remove irrelevant fields: For example, when filtering data related to "energy technology costs", remove fields such as "number of employees in the enterprise" that are not relevant to the analysis to simplify the dataset.

[0079] After cleaning the original structured data, normalization can be performed on the cleaned original structured data to obtain structured data.

[0080] When performing normalization, for example, normalization can be carried out according to time for data of the same type. For example, according to months, normalize data of the same type every year, or normalize data of the same type together.

[0081] The present disclosure embodiments do not limit the granularity of normalization, which can be selected according to actual needs.

[0082] After that, the structured data that has undergone data cleaning and normalization processing is stored in the knowledge base in chronological order according to the time information corresponding to the structured data.

[0083] Here, when storing, for the same type of structured data, the chronological relationship between data can be retained in the knowledge base to facilitate the subsequent specific data processing and analysis process. For example, for energy consumption by year, store the energy consumption of different years in the knowledge base and retain the chronological relationship corresponding to the energy consumption of different years respectively.

[0084] For the original unstructured data, since the unstructured data has no fixed format, the key points of cleaning are to remove noise information, extract effective content, and unify the text format to prepare for subsequent text mining and semantic analysis. The data cleaning process for the original unstructured data includes the following steps: (1) Text preprocessing: For example, it can include encoding the text uniformly, such as converting it to DTF-8, removing irrelevant format information, such as headers, footers, watermarks, table borders, etc. in policy texts, and retaining only the pure text content. Unify the text length, and truncate or remove process literature abstracts, short and ineffective comments, etc. In addition, it can also include noise removal, such as removing meaningless characters: such as punctuation marks, special symbols (@, #), numbers (numbers with no actual meaning); removing stop words: such as "de, di, de", "guanyu, genju" and other words with no substantial meaning to reduce text redundancy.

[0085] (2): Text content purification: For example, it can include core information extraction and ambiguity elimination.

[0086] Among them, core information extraction, for example, includes: for policy texts: extracting core elements such as "policy objectives, regulatory objects, implementation time, policy tools", and removing publicity and general expressions; for public opinion data: filtering advertisements and irrelevant comments (such as when discussing energy policies, removing comments on home appliance promotions); for technical literature: extracting key contents such as "technical parameters, R & D progress, cost data", and removing duplicate expressions in the literature review.

[0087] Ambiguity resolution: This addresses the issue of polysemy, such as the fact that "energy storage" may refer to "electrochemical energy storage" or "pumped hydro storage" in different contexts. It is necessary to specify the specific meaning based on the context to avoid bias in subsequent cluster analysis.

[0088] (3): Text structuring and transformation preparation: The cleaned text is segmented into words, and the continuous text is split into independent words; professional terms are standardized, such as classifying "wind-solar-storage integration" and "source-grid-load-storage integration" into "new power system model", laying the foundation for subsequent candidate factor extraction.

[0089] After cleaning the unstructured data, a pre-trained natural language model is used to perform semantic understanding processing on the cleaned unstructured data, mapping it to a high-dimensional semantic space to obtain the semantic feature information corresponding to the unstructured data. In specific implementation, the following method can be used to identify key driving factors based on the interactive data: With the goal of extracting candidate factors influencing low-carbon transformation, literature mining and policy text analysis were performed based on the interactive data. Based on the results of the literature mining and policy text analysis, several first candidate factors influencing low-carbon transformation were obtained. Through the expert consultation interface, expert consultation information is obtained, and based on the expert consultation information, several second candidate factors affecting the low-carbon transformation are obtained. Clustering is performed on the first candidate factor and the second candidate factor to obtain multiple candidate factor categories; By ranking the importance of multiple candidate factor categories, core driving factors covering socio-economic and energy technologies are obtained.

[0090] In practice, literature mining based on interactive data is usually achieved by using technical literature data from unstructured data; policy text analysis based on interactive data is usually achieved by using policy text data from unstructured data.

[0091] When mining documents based on interactive data, a method combining core keywords and extended keywords is specifically adopted. From the interactive data, the first target interactive data corresponding to the core keywords and extended keywords is selected. Then, based on the first target interactive data, factor extraction is performed to obtain the first candidate factors influencing low-carbon transformation.

[0092] For example, suppose the core keywords include: low-carbon transition, energy transition, and dual-carbon goals. The corresponding extended keywords would then include driving factors, influencing factors, key variables, and core elements. When using the aforementioned core keywords and extended keywords to filter the primary target interaction data from the interaction data, the following methods can be used, for example: Preliminary screening: Filter by "title + abstract" and remove documents that are irrelevant to the topic (such as purely engineering and technical documents or purely macroeconomic documents with no overlap). Precise screening: Screening is based on "publication time (e.g., the last 10 years) + literature quality (core journals, SCI / SSCI, highly cited literature)" to prioritize the retention of authoritative research results; Final Literature: A certain number of high-quality documents are generally retained to ensure coverage of dimensions such as socio-economics, energy technology, and policy mechanisms. These retained final literature documents are used as the primary target interactive data.

[0093] After obtaining the initial target interaction data, factor extraction can be performed using either of the following two methods: (1) Factor extraction based on word frequency-co-occurrence analysis: Statistical analysis is conducted to identify the frequency of words and extract high-frequency keywords (such as "renewable energy penetration rate", "industrial structure upgrading", "carbon price", "electrification level", etc.). These words are usually the core factors of academic research. Co-occurrence analysis is performed to construct a keyword co-occurrence network (such as using tools like CiteSpace and VOSviewer) to identify closely related keyword clusters (such as the "carbon capture-technology cost-demonstration project" cluster) and uncover hidden combination factors.

[0094] (2) Factor induction based on content analysis: Extract the "low-carbon transition driving factors" explicitly mentioned in the literature and record the mechanism of action of each factor (e.g., "carbon prices force companies to reduce emissions by increasing fossil energy costs"). Encode and classify the extracted factors, and merge synonymous or near-synonymous factors (e.g., "photovoltaic electricity cost" and "wind power electricity cost" are merged into "renewable energy technology cost"). Calculate the frequency of each factor mentioned in the literature. The higher the frequency, the higher the academic consensus of the factor, and the higher it is included in the candidate factor pool.

[0095] Subsequently, a list of candidate driving factors from an academic perspective was compiled, including factor name, mechanism of action, literature support (frequency of mention), and the dimension to which the factor belongs (socio-economic / energy-technology).

[0096] When analyzing policy documents based on interactive data, one can first classify the policy text data in the interactive data to obtain multiple policy text data categories. Then, for each category, driving factors are extracted to obtain the first candidate factors from the policy text data.

[0097] Specifically, when classifying policy text data, it can be divided into the following four categories: Planning category: Energy development plan, dual-carbon action plan. Policy text data of this category are used to indicate long-term goals; Control category: Energy consumption dual control policies and capacity control of high energy-consuming industries. The policy text data of this category is used to indicate and constrain indicators. Incentive category: New energy subsidy policies and green finance policies. The policy text data in this category is used to indicate incentive measures. Market-related: Carbon market policies, electricity price reform policies. The policy text data in this category is used to indicate market-based tools.

[0098] Then, the various policy text data can be analyzed and processed using the following methods (1) or (2): (1) Factor extraction based on policy instrument coding: Step 1: Construct a policy tool classification framework, mapping policy text content to specific tool types, as shown in the example below: Administrative control tools include energy intensity targets, access standards for high-energy-consuming industries, and new energy installation targets. The corresponding driving factors include: energy consumption intensity and renewable energy penetration rate.

[0099] Market incentive tools include carbon price levels, new energy subsidies, and green taxes. Corresponding driving factors include the strength of carbon pricing mechanisms and the level of support for low-carbon industries.

[0100] Technology promotion tools include CCS demonstration projects, smart grid construction, and charging pile planning. The corresponding driving factors include the maturity of CCS technology and the flexibility of the energy system.

[0101] Step 2: Encode each policy text, extract the core indicators corresponding to the policy tools (such as "the proportion of wind power and photovoltaic installed capacity will exceed 40% by 2030"), and convert these indicators into candidate factors.

[0102] (2) Factor priority judgment based on policy strength index: Step 1: Construct a policy strength evaluation system from three dimensions: policy authority, coverage, and binding strength (1-5 points for each dimension). For example, the evaluation system may include: Authority: National policies (5 points) > Provincial policies (3 points) > Municipal policies (1 point); Coverage: All industries (5 points) > Key industries (3 points) > Pilot industries (1 point); Constraint strength: Mandatory policies (5 points) > Guiding policies (3 points) > Recommendational policies (1 point).

[0103] Step 2: Calculate the strength index of each policy tool. The higher the index, the higher the priority of the corresponding driving factor in the scenario analysis (e.g., if the strength index of the "dual control of energy consumption" policy is high, the corresponding "upgrading of industrial structure" factor should be given special attention).

[0104] (3) Hidden factor mining based on text topic modeling: For factors not explicitly mentioned in policy texts but implicit in policy guidance, unsupervised probabilistic generation (Latent Dirichlet Allocation, LDA) topic models can be used for mining: Step 1: Input the policy text data into the LDA model and set the number of topics (e.g., 5-8 topics). Step 2: Output the core themes of the policy text (such as "construction of new power systems", "guidance of green consumption", and "regional energy synergy"). Step 3: Extract hidden driving factors from the theme (such as "regional energy coordination mechanism" and "residents' willingness to consume low-carbon products").

[0105] Subsequently, using the results of the above-mentioned driving factors extraction, a list of candidate driving factors from a policy perspective is formed, which is the first candidate factor obtained based on policy text data, including factor name, corresponding policy tool, policy strength index, and clarifying the policy controllability of the factor.

[0106] Furthermore, embodiments of this disclosure may also provide an expert consultation interface. Through this interface, expert consultation information can be received.

[0107] Among them, expert information can be obtained through questionnaires to obtain the driving factors affecting low-carbon transformation from experts in the fields of energy economics, technology research and development, and policy formulation. Alternatively, various information or documents formed in expert seminars can be used as expert consultation information to obtain the driving factors affecting low-carbon transformation. These driving factors can serve as practical factors to fill the gaps in literature and policy, which are the second candidate factors in this disclosure embodiment.

[0108] Using the above method, after obtaining the first and second candidate factors affecting the low-carbon transformation, the first and second candidate factors are clustered.

[0109] In the clustering process, the overlapping and intersecting factors in the first and second candidate factors are usually merged and simplified through clustering algorithms to ensure the independence and representativeness of the core driving factors.

[0110] Specifically, when clustering the first and second candidate factors, we can rely on the similarity of their connotations and the consistency of their mechanisms of action. Specifically, we can map different candidate factors to a high-dimensional feature space to obtain feature data for each first and second candidate factor. Then, we use the similarity between the feature data to cluster the different candidate factors, ultimately forming multiple candidate factor categories. After forming multiple candidate factor categories, we further fuse and reduce the dimensionality of each candidate factor category to combine multiple candidate factors into a single candidate factor.

[0111] For example, "cost per kilowatt-hour of photovoltaic electricity," "cost per kilowatt-hour of wind power," and "unit cost of energy storage" can be clustered into "cost of renewable energy and energy storage technology"; "industrial energy-saving renovation efforts," "building energy efficiency improvement standards," and "transportation electrification rate" can be clustered into "energy efficiency and decarbonization levels of key sectors." Subsequently, a ranking method combining quantitative analysis and qualitative assessment can be used to rank the importance of multiple candidate factor categories, and based on the ranking results, the core driving factors covering socio-economic and energy technologies can be obtained.

[0112] When conducting quantitative analysis on each candidate factor category, methods such as grey relational analysis based on historical data and Pearson correlation coefficient calculation can be used to measure the correlation between factors and low-carbon indicators such as carbon emission intensity and energy consumption per unit of GDP. When conducting qualitative evaluation on each candidate factor category, expert scoring can be used to score each factor on different evaluation dimensions, such as scoring each factor on three evaluation dimensions: policy controllability, technical feasibility, and economic impact significance. Finally, the scores of each expert are combined to rank the importance of multiple candidate factor categories according to the scores of each factor.

[0113] Subsequently, based on the ranking of importance, several core driving factors that can cover socio-economic and energy technology were identified from each candidate factor category.

[0114] After obtaining the core driving factors, state definition information associated with each core driving factor can be generated. The state definition information includes multiple states corresponding to the core driving factors.

[0115] In generating state definition information, semantic descriptions and policy implications corresponding to various state values ​​can also be generated.

[0116] Multiple states, their corresponding semantic descriptions, and policy implications constitute the state definition information for each core driving factor.

[0117] Multiple core driving factors and corresponding state definition information form a structured factor system.

[0118] When generating state definition information, the general principles of gradient, differentiation, and quantifiability are followed, covering typical paths such as "conservative-benchmark-radical" and matching the direction of policy tools.

[0119] For example, regarding the socio-economic dimension, the corresponding core driving factors include: Economic growth patterns, whose state definition information includes, for example: (1) Extensive growth, the corresponding semantic descriptions include: high energy-consuming industries drive GDP growth of more than 5%; the policy implications include: no new low-carbon policy constraints; (2) Intensive growth, the corresponding semantic descriptions include: industrial structure optimization, and a 10% decrease in the proportion of high energy consumption; the policy implications include: implementing dual control of energy consumption and eliminating outdated production capacity; (3) Green and low-carbon growth, the corresponding semantic descriptions include: the output value of green industries accounts for more than 30%, and GDP growth is coordinated with carbon emission reduction; the policy implications include: implementing green GDP assessment and increasing subsidies for low-carbon industries.

[0120] Regarding the temperature of energy technologies, the core driving factors include, for example: Renewable energy penetration rate, whose status definition information includes, for example: (1) Low penetration (<20%), the corresponding semantic descriptions include: wind power and photovoltaic installed capacity account for <15%, relying on thermal power as a backup; the policy implications include: new energy subsidies are phased out, and grid connection constraints are strict; (2) Penetration (20%-40%), the corresponding semantic description includes: wind and solar installed capacity accounts for 25%, with a small amount of energy storage; the policy implications include: increasing grid transformation and improving the consumption mechanism; (3) High penetration (40%-60%), the corresponding semantic descriptions include: wind and solar installed capacity accounts for 45%, and energy storage configuration ratio exceeds 15%; the policy implications include: mandatory energy storage, promoting the integration of source, grid, load and storage; (4) Ultra-high penetration (>60%), the corresponding semantic descriptions include: wind and solar installed capacity accounts for 65%; the policy implications include: the cross-provincial and cross-regional power transmission channels have been fully completed.

[0121] Regarding the above S103: In practical implementation, for example, the following method can be used to generate a cross-influence matrix corresponding to multiple core driving factors: Through multiple rounds of anonymous expert consultation, using structured questionnaires and scoring interfaces, we collected scores from different experts on the influence relationship between each pair of core driving factors, and generated candidate cross-influence matrices. The candidate cross-influence matrices are subjected to consistency verification and matrix filling optimization to obtain the cross-influence matrices corresponding to the multiple core driving factors.

[0122] In this way, through multiple rounds of anonymous expert consultation combined with structured scoring, consistency verification, and matrix optimization, a cross-influence matrix corresponding to multiple core driving factors is constructed. Assuming there are n core driving factors, the constructed cross-influence matrix is ​​n×n dimensional.

[0123] For example, assuming there are n core driving factors, the corresponding cross-influence matrix is ​​represented as follows: The cross-influence matrix consists of n rows and n columns; each of the n rows corresponds to one of the n core driving factors, and each of the n columns also corresponds to one of the n core driving factors. This indicates the influence strength information between the first core driving factor located in the i-th row and the second core driving factor located in the j-th column.

[0124] In one possible implementation, the influence strength information includes, for example, the influence strength information of the first core driving factor on the second core driving factor, which can be represented by a score. Assuming the score ranges from [-k, k], when the influence strength information is negative, the first core driving factor has a suppressive effect on the second core driving factor; and when the influence strength information is at the minimum value of the aforementioned range, it indicates a strong suppressive effect. As the influence strength information increases, the degree of suppression of the first core driving factor on the second core driving factor gradually decreases. When the influence strength information is positive, the first core driving factor has a promoting effect on the second core driving factor; and the larger the value, the greater the promoting effect of the first core driving factor on the second core driving factor.

[0125] For example, the range of the above score can be [-3, 3]. Specifically, a value of -3 indicates strong inhibition; -2 indicates moderate inhibition; -1 indicates slight inhibition; 0 indicates no inhibition or facilitation; 1 indicates slight facilitation; 2 indicates moderate facilitation; and 3 indicates strong facilitation.

[0126] Specifically, when collecting scores from different experts on the influence relationship between each pair of core driving factors through multiple rounds of anonymous expert consultation, using structured questionnaires and scoring interfaces, and generating candidate cross-influence matrices, for example, the scoring rules and matrix dimensions can be determined first; then, the Delphi method can be used to conduct multiple rounds of consultation, gradually converge expert opinions, and obtain candidate cross-influence matrices.

[0127] When conducting multiple rounds of consultation using the Delphi method, the following process can be employed, for example: First round of consultation: Initial assessment: A structured questionnaire was distributed to experts, consisting of three parts: ① A list of key driving factors and their status definitions are provided to clarify the meaning of the factors and avoid misunderstandings; ② Blank rating matrix; ③ Scoring rules and instructions (Example: Please assess the impact of "carbon pricing mechanism strength" on "renewable energy penetration rate" and fill in the score value in the corresponding cell). Experts independently fill out the matrix, give a score for each element, and indicate the basis for the score (such as policy logic, technical mechanism, practical experience). The questionnaires were collected, and the mean, standard deviation, and coefficient of variation of the first round of scores were statistically analyzed to identify elements with high dispersion, that is, the influence relationship of large differences in expert opinions.

[0128] Second round of consultation: convergence of opinions: Feedback on the first round of scoring results to experts, focusing on the scoring distribution of elements with high dispersion and the scoring basis of different experts; Invite experts to revise or maintain their ratings based on feedback, and explain the reasons for any revisions. The second round of questionnaires was collected, and the dispersion of the scores was analyzed again. If some elements still had high dispersion, a third round of consultation was conducted.

[0129] Termination conditions: When the coefficient of dispersion of matrix elements exceeding a preset proportion is less than a preset threshold, such as the coefficient of dispersion of all matrix elements being less than 0.3, or the coefficient of dispersion of more than 80% of matrix elements being less than 0.3, it is determined that the expert opinions have converged, the consultation is stopped, and the expert score data is obtained, which is the candidate cross-influence matrix.

[0130] Afterwards, the converged expert scoring data can be cleaned and its consistency checked to remove outliers and ensure the rationality of the matrix data.

[0131] When performing consistency checks on candidate cross-influence matrices, for example, the causal logic of the matrix can be checked to remove obviously contradictory element values. For example, if the score for "strength of carbon pricing mechanism → renewable energy penetration rate" in the candidate cross matrix is ​​+3 (strong promotion), but the score for "renewable energy penetration rate → strength of carbon pricing mechanism" is +2, it is necessary to confirm whether it conforms to the logic of reality. Usually, the latter has a weaker impact and can be corrected to +1 or 0.

[0132] Furthermore, when performing matrix imputation optimization on the candidate cross-influence matrix, it can handle the fuzzy relationship between missing data and different core driving factors: For example, for "weak influence / no influence" relationships that are difficult for some experts to determine, the system's built-in optimization mechanism is used to fill in the matrix and ensure its integrity.

[0133] For elements not filled in by experts (missing values), if the corresponding influence relationship has no clear logical logic in theory, they are directly assigned a preset value, such as 0. If there is theoretical logic but expert opinions are ambiguous, the method of analogy between adjacent factors can be used to fill in the gaps: refer to the average score of similar influence relationships, and adjust the assigned value in combination with the factor attributes (such as the influence strength between technical factors is usually higher than that between social factors).

[0134] In the process of optimizing fuzzy relationships, for fuzzy influence relationships with scores concentrated in the range of -1 to +1, a literature support weight is introduced for correction: if an influence relationship has been verified multiple times in academic literature, the absolute value of the score is appropriately increased (e.g., adjusted from +1 to +2); if the literature support is low, the original score is maintained or adjusted to 0.

[0135] After performing consistency verification and matrix filling optimization on the candidate cross-influence matrices, all element values ​​are standardized to integer scores of [-k, k], ultimately forming an n×n cross-influence matrix.

[0136] Alternatively, other methods can be used to generate the aforementioned cross-influence matrix. For example, for any two core driving factors, based on the interaction data corresponding to the two core driving factors, the inhibition and promotion relationships between the two core driving factors can be mined, and the cross-influence matrix can be obtained based on the mined inhibition and promotion relationships. Simultaneously, the above two methods can also be combined to obtain the cross-influence matrix. Specific embodiments disclosed herein are not limited.

[0137] Regarding S104 above: In practice, a candidate scenario is a specific development plan formed by combining and matching the discrete states of all core driving factors. That is, each core driving factor selects a discrete state, and combining them together forms a candidate scenario.

[0138] Specifically, in the above steps, multiple core driving factors and discrete state definition information corresponding to each core driving factor have been generated.

[0139] For example, several core driving factors include: Socioeconomic dimensions: strength of carbon pricing mechanism, economic growth model, level of industrial structure sophistication, and residents' willingness to consume low-carbon products.

[0140] Energy technology dimensions: renewable energy penetration rate, terminal electrification level, CCS technology maturity, and energy system flexibility.

[0141] Each core driver is defined with multiple discrete states, each with clear quantitative boundaries, semantic descriptions, and policy implications. For example, for "carbon pricing mechanism strength," the state definition information includes: 1. No carbon price; 2. Medium carbon price (150 yuan / ton); 3. High carbon price (300 yuan / ton). For "renewable energy penetration rate," the state definition information includes: 1. Low penetration (<20%); 2. Medium penetration (20%-40%); 3. High penetration (>40%).

[0142] When generating candidate scenarios, one state is selected from the discrete states of each core driving factor, and the selected states are combined to form a complete candidate scenario.

[0143] If there are s core driving factors and the j-th factor has p discrete states, then the total number of theoretical scenario combinations is the product of the number of states of each factor.

[0144] After obtaining multiple candidate scenarios, for example, the stability scores of each candidate scenario can be calculated based on the cross-influence matrix in the following manner: For each candidate scenario, based on the cross-influence matrix, the cross-influence direction and state gradient difference between every two core driving factors in the candidate scenario are determined, and based on the cross-influence direction and the state gradient difference, the state compatibility coefficient of every two core driving factors in the corresponding candidate scenario is determined. And, based on the cross-influence matrix, determine the influence weight between each pair of core driving factors; Based on the state compatibility coefficient between each pair of core driving factors and the corresponding influence weight, the stability score corresponding to each candidate scenario is determined.

[0145] In practice, during the stability scoring of each candidate scenario, the higher the logical matching degree between the states corresponding to different candidate factors, the more stable the candidate scenario is, and the higher its corresponding stability score will be.

[0146] The core quantitative carriers of logical matching degree include, for example, the state compatibility coefficient, which directly measures the degree of logical fit between the states of two factors in a scenario combination. This coefficient is ultimately incorporated into the scoring system through its product with the influence weights. The logical matching degree is obtained based on the prior laws of the cross-influence matrix and the gradient relationship between the states of the core driving factors, specifically including the following process: The cross-influence matrix describes the influence relationships between different core driving factors.

[0147] For elements in the cross-influence matrix The influence strength of the first core driving factor (referred to as factor i) located in the i-th row on the second core driving factor (referred to as factor j) in the j-th column is defined, and the direction of the influence is defined by the positive or negative value.

[0148] The direction of this influence determines whether the combination of specific states of different core driving factors matches.

[0149] like The more "aggressive" factor i is, the more "aggressive" factor j should also be. Only such a combination makes sense, such as "high carbon price" promoting "high renewable energy penetration rate".

[0150] like The more "aggressive" factor i is, the more "conservative" factor j should be. Such a combination is logical, such as "high traditional energy subsidies" suppressing "high renewable energy penetration".

[0151] like The states of factors i and j have no logical connection, and the matching degree is 0.

[0152] The specific value of the logical matching degree is achieved by defining a compatibility coefficient for each pair of factor states. The coefficient ranges from [-1, 1], corresponding to the gradient from completely incompatible to completely compatible, which is the state gradient difference.

[0153] For any factor Its discrete states are assigned gradient labels in the order of "conservative, baseline, radical": such as Indicates conservatism Indicates the reference, Indicates radicalism; State gradient difference : Indicates the state of factor i With the state of factor j The degree of radicalism varies.

[0154] (1) Under the promoting relationship, the states of the two factors are required to change in the same direction (both conservative or both radical), and the smaller the gradient difference, the higher the matching degree.

[0155] If the state combinations of factors i and j are completely in the same direction, such as the state gradient difference ,like The compatibility coefficient between these two factors is... A value of +1 indicates that factor i and factor j are a perfect match, which conforms to the facilitation logic.

[0156] If the state combinations of factors i and j are semi-same, such as the state gradient difference... ,like The compatibility coefficient between these two factors is... The value of +0.5 indicates that factor i and factor j are moderately matched, and the logic is basically self-consistent.

[0157] If the state combinations of factors i and j are completely opposite, such as the state gradient difference ,like The compatibility coefficient between these two factors is... The value of -1 indicates that factor i and factor j are completely mismatched, violating the facilitation logic.

[0158] (2) Under the inhibition relationship, the states of the two factors are required to change in opposite directions (if factor i is aggressive, then factor j is conservative). The greater the gradient difference, the higher the matching degree.

[0159] If the state combinations of factors i and j are completely opposite, such as the state gradient difference ,like The compatibility coefficient between these two factors is... The value of +1 indicates that factor i and factor j are a perfect match, which conforms to the inhibition logic.

[0160] If the state combination of factor i and factor j is semi-inverse, such as the state gradient difference... ,like The compatibility coefficient between these two factors is... The value of +0.5 indicates that factor i and factor j are moderately matched, and the logic is basically self-consistent.

[0161] If the state combinations of factors i and j are completely in the same direction, such as the state gradient difference ,like The compatibility coefficient between these two factors is... The value of -1 indicates that factor i and factor j are completely mismatched, violating the inhibition logic.

[0162] (3) In the absence of an influencing relationship, no matter how the states of the two factors are combined, they are logically unrelated, therefore the compatibility coefficient is low. It is always 0.

[0163] Furthermore, for every two core driving factors, the weight of factor i on factor j is: That is, the absolute value of the influence intensity.

[0164] Next, using the influence weights corresponding to each pair of core driving factors, the state compatibility coefficients between these two core driving factors are weighted to obtain their contribution scores. The contribution scores of all core driving factor pairs are then summed to obtain the stability score of the corresponding candidate scenario.

[0165] After obtaining the stability scores corresponding to multiple candidate scenarios, the multiple candidate scenarios can be sorted in descending order of their respective stability scores. According to the sorting, a preset number of candidate scenarios are determined from the candidate scenarios as the target scenarios; Alternatively, the stability scores corresponding to multiple candidate scenarios are compared with a preset score threshold; the candidate scenario with a stability score greater than the preset score threshold is selected as the target scenario.

[0166] Regarding the above S105: After generating the target scenario based on the above process, the following methods can be used to perform quantitative modeling based on the target scenario to generate a future socio-economic scenario model for low-carbon energy: Density clustering is performed on the various target scenarios to obtain multiple scenario patterns, and semantic labels and policy interpretations corresponding to each scenario pattern are generated; Based on the scenario patterns, corresponding semantic tags, and policy interpretations, multiple scenario patterns are modeled to obtain low-carbon energy socio-economic future scenario models corresponding to each scenario pattern.

[0167] Specifically, when performing density clustering on multiple target scenarios, for example, by aggregating different target scenarios based on the similarity of their state features, the system automatically identifies scenario clusters with similar features and ultimately extracts 3-5 typical scenario patterns.

[0168] When performing density clustering, for example, each target scenario can first be transformed into a numerical feature vector, which serves as the input to the clustering algorithm. Then, a specific clustering algorithm, such as DBSCAN density clustering, is used to perform density clustering on different target scenarios based on the numerical feature vectors, resulting in multiple scenario patterns. Simultaneously, based on the semantic description and policy information of the target scenarios, corresponding semantic labels and policy interpretations are generated for each scenario pattern.

[0169] Subsequently, based on the extracted typical scenario patterns, and by coupling the energy system model with the macroeconomic model, a sub-model low-carbon energy socio-economic future scenario model is constructed, realizing the transformation from qualitative characteristics to quantitative deduction.

[0170] In the scenario-based quantitative modeling phase, qualitative scenarios are transformed into input parameter sets for energy system and macroeconomic models through a parameter mapping interface. Parameters include technology learning rate, carbon price trajectory, energy efficiency assumptions, and investment preferences. Through model coupling and parallel computing, key indicators such as carbon emission pathways, energy structure evolution, and economic cost-effectiveness are output for each scenario.

[0171] To enhance the interpretability and decision support capabilities of the results, a visualization output module can be integrated, providing multi-scenario comparison dashboards, sensitivity analysis tools, and policy lever identification functions. Users can adjust the status or influence weight of driving factors through an interactive interface, generate new scenarios in real time, and view their quantitative impact, supporting dynamic strategy exploration and uncertainty management.

[0172] In addition, it can provide scenario library management, version control and collaborative editing functions, support multi-user online evaluation and scenario iteration optimization, and is suitable for the long-term energy and climate strategy formulation needs of various users such as institutions, units and enterprises.

[0173] Corresponding to the aforementioned embodiments of the method for generating future socio-economic scenarios of low-carbon energy, this application also provides embodiments of a device for generating future socio-economic scenarios of low-carbon energy.

[0174] The embodiments of the low-carbon energy socio-economic future scenario generation device of this application can be applied to computer equipment. The device embodiments can be implemented through software, hardware, or a combination of both. Taking software implementation as an example, as a logical device, it is formed by the processor of the computer equipment loading the corresponding computer program instructions from non-volatile memory into memory for execution. From a hardware perspective, such as... Figure 2 The diagram shown is a hardware structure diagram of the computer equipment housing the low-carbon energy socio-economic future scenario generation device of this application. (Except for...) Figure 2 In addition to the processor, memory, network interface, and non-volatile memory shown, the computer device in which the device is located in the embodiment may also include other hardware depending on the actual function of the low-carbon energy socio-economic future scenario generation device, which will not be described in detail here.

[0175] Please refer to Figure 3 The low-carbon energy socio-economic future scenario generation device provided in this disclosure includes: Module 31 is used to acquire interactive data between the socio-economic and energy systems; The factor identification module 32 is used to perform key driving factor identification processing and state definition processing based on the interactive data to obtain multiple core driving factors affecting low-carbon transformation and state definition information corresponding to each core driving factor. The matrix construction module 33 is used to generate a cross-influence matrix corresponding to multiple core driving factors; wherein, each element in the cross-influence matrix is ​​used to characterize the influence intensity information of the first core driving factor on the second core driving factor among the two core driving factors corresponding to the element. The scenario screening module 34 is used to generate multiple candidate scenarios based on the state definition information of multiple core driving factors, perform stability scoring on the multiple candidate scenarios based on the cross-influence matrix, and determine multiple target scenarios from the multiple candidate scenarios based on the stability scoring results. The quantitative modeling module 35 is used to perform quantitative modeling based on the target scenario to generate a future scenario model of low-carbon energy socio-economic conditions.

[0176] Optionally, the interactive data includes: structured data and unstructured data; The structured data includes at least one of the following: energy consumption data, carbon emission data, economic indicator data, and basic parameter data; The unstructured data includes: policy text data, technical document data, public opinion data, and enterprise operation data.

[0177] Optionally, the acquisition module 31, when acquiring interaction data of the socio-economic and energy systems, is used for: Acquire raw interaction data of socio-economic and energy systems; the raw interaction data includes: raw structured data and raw unstructured data; The original structured data is sequentially cleaned and normalized to obtain the structured data. as well as, The original unstructured data is cleaned, and a pre-trained natural language model is used to perform semantic understanding processing on the cleaned unstructured data to obtain semantic feature information corresponding to the unstructured data. The semantic feature information is then used as the unstructured data.

[0178] Optionally, the factor identification module 32, when performing key driving factor identification processing and state definition processing based on the interactive data to obtain multiple core driving factors affecting low-carbon transformation, is used for: With the goal of extracting candidate factors influencing low-carbon transformation, literature mining and policy text analysis were performed based on the interactive data. Based on the results of the literature mining and policy text analysis, several first candidate factors influencing low-carbon transformation were obtained. Through the expert consultation interface, expert consultation information is obtained, and based on the expert consultation information, several second candidate factors affecting the low-carbon transformation are obtained. Clustering is performed on the first candidate factor and the second candidate factor to obtain multiple candidate factor categories; By ranking the importance of multiple candidate factor categories, core driving factors covering socio-economic and energy technologies are obtained.

[0179] Optionally, the matrix construction module 33, when generating the cross-influence matrix corresponding to the multiple core driving factors, is used to: Through multiple rounds of anonymous expert consultation, using structured questionnaires and scoring interfaces, we collected scores from different experts on the influence relationship between each pair of core driving factors, and generated candidate cross-influence matrices. The candidate cross-influence matrices are subjected to consistency verification and matrix filling optimization to obtain the cross-influence matrices corresponding to the multiple core driving factors.

[0180] Optionally, the matrix construction module 33 performs stability scoring on various candidate scenarios based on the cross-influence matrix, including: For each candidate scenario, based on the cross-influence matrix, the cross-influence direction and state gradient difference between every two core driving factors in the candidate scenario are determined, and based on the cross-influence direction and the state gradient difference, the state compatibility coefficient of every two core driving factors in the corresponding candidate scenario is determined. And, based on the cross-influence matrix, determine the influence weight between each pair of core driving factors; Based on the state compatibility coefficient between each pair of core driving factors and the corresponding influence weight, the stability score corresponding to each candidate scenario is determined.

[0181] Optionally, the scenario filtering module 34, when determining multiple target scenarios from multiple candidate scenarios based on stability scoring results, is used to: The candidate scenarios are sorted in descending order of their respective stability scores. According to the sorting, a preset number of candidate scenarios are determined from the candidate scenarios as the target scenarios; or, The stability scores and preset score thresholds corresponding to the multiple candidate scenarios are compared respectively; Candidate scenarios with stability scores greater than the preset score threshold are selected as target scenarios.

[0182] Optionally, the quantitative modeling module 35, when performing quantitative modeling based on the target scenario to generate a low-carbon energy socio-economic future scenario model, is used for: Density clustering is performed on the various target scenarios to obtain multiple scenario patterns, and semantic labels and policy interpretations corresponding to each scenario pattern are generated; Based on the scenario patterns, corresponding semantic tags, and policy interpretations, multiple scenario patterns are modeled to obtain low-carbon energy socio-economic future scenario models corresponding to each scenario pattern.

[0183] The specific implementation process of the functions and roles of each unit in the above device can be found in the implementation process of the corresponding steps in the above method, and will not be repeated here.

[0184] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0185] See Figure 4 As shown in the embodiments of this disclosure, a decision generation system based on a low-carbon energy socio-economic future scenario model is also provided, including: a processor 41, a memory 42, an expert evaluation terminal 43, a policy decision terminal 44, and an energy system model interface 45.

[0186] The memory stores a low-carbon energy socio-economic future scenario model generated based on the low-carbon energy socio-economic future scenario generation method described in any embodiment of this disclosure. The energy system model interface is used to receive energy system models; the expert evaluation terminal is used to receive expert evaluation information. The processor is further configured to couple at least one of the received energy system model and the expert evaluation information with the low-carbon energy socio-economic future scenario model to obtain a policy decision, and output the policy decision to the policy decision terminal.

[0187] This disclosure also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program performs the steps of the low-carbon energy socio-economic future scenario generation method described in the above-described method embodiments. The storage medium can be a volatile or non-volatile computer-readable storage medium.

[0188] This disclosure also provides a computer program product carrying program code. The program code includes instructions that can be used to execute the steps of the method for generating low-carbon energy socio-economic future scenarios described in the above method embodiments. For details, please refer to the above method embodiments, which will not be repeated here.

[0189] The aforementioned computer program product can be implemented through hardware, software, or a combination thereof. In one optional embodiment, the computer program product is specifically embodied in a computer storage medium; in another optional embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0190] The computer program or instructions may be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another. For example, the computer program or instructions may be transferred from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. The computer-readable storage medium may be any available medium that a computer can access, or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; or an optical medium, such as a digital video optical disc; or a semiconductor medium, such as a solid-state drive. The computer-readable storage medium may be a volatile or non-volatile storage medium, or may include both volatile and non-volatile types of storage media.

[0191] The embodiments of the subject matter and functional operation described in this specification can be implemented in the following ways: digital electronic circuits, tangibly embodied computer software or firmware, computer hardware including the structures disclosed in this specification and their structural equivalents, or combinations thereof. Embodiments of the subject matter described in this specification can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions encoded on a tangible, non-transitory program carrier for execution by a data processing apparatus or for controlling the operation of a data processing apparatus. Alternatively or additionally, the program instructions may be encoded on artificially generated propagation signals, such as machine-generated electrical, optical, or electromagnetic signals, which are generated to encode information and transmit it to a suitable receiving device for execution by the data processing apparatus. The computer storage medium may be a machine-readable storage device, a machine-readable storage substrate, a random or serial access memory device, or combinations thereof.

[0192] The processing and logic flow described in this specification can be executed by one or more programmable computers that execute one or more computer programs to perform corresponding functions by operating on input data and generating output. The processing and logic flow can also be executed by dedicated logic circuitry—such as FPGAs (Field-Programmable Gate Arrays) or ASICs (Application-Specific Integrated Circuits), and the device can also be implemented as dedicated logic circuitry.

[0193] Suitable computers for executing computer programs include, for example, general-purpose and / or special-purpose microprocessors, or any other type of central processing unit. Typically, the central processing unit receives instructions and data from read-only memory and / or random access memory. The basic components of a computer include a central processing unit for implementing or executing instructions and one or more memory devices for storing instructions and data. Typically, a computer will also include one or more mass storage devices for storing data, such as disks, magneto-optical disks, or optical disks, or the computer will be operatively coupled to such mass storage devices to receive data from or transfer data to them, or both. However, a computer is not required to have such devices. Furthermore, a computer can be embedded in another device, such as a mobile phone, a personal digital assistant (PDA), a mobile audio or video player, a game console, a global positioning system (GPS) receiver, or a portable storage device such as a universal serial bus (USB) flash drive, to name a few.

[0194] Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media, and memory devices, such as semiconductor memory devices (e.g., EPROM, EEPROM, and flash memory devices), magnetic disks (e.g., internal hard disks or removable disks), magneto-optical disks, and CD-ROM and DVD-ROM disks. Processors and memory may be supplemented by or incorporated into dedicated logic circuitry.

[0195] While this specification contains numerous specific implementation details, these should not be construed as limiting the scope of any invention or the scope of the claims, but rather are primarily intended to describe features of specific embodiments of a particular invention. Certain features described in the various embodiments herein may also be implemented in combination in a single embodiment. Conversely, various features described in a single embodiment may also be implemented separately in various embodiments or in any suitable sub-combination. Furthermore, while features may function in certain combinations as described above and even initially claimed in this way, one or more features from a claimed combination may be removed from that combination in some cases, and a claimed combination may refer to a sub-combination or a variation thereof.

[0196] Similarly, although the operations are depicted in a specific order in the accompanying drawings, this should not be construed as requiring these operations to be performed in the specific order shown or sequentially, or requiring all illustrated operations to be performed to achieve the desired result. In some cases, multitasking and parallel processing may be advantageous. Furthermore, the separation of various system modules and components in the above embodiments should not be construed as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

[0197] Thus, specific embodiments of the subject matter have been described. Other embodiments are within the scope of the appended claims. In some cases, the actions recited in the claims may be performed in a different order and still achieve the desired result. Furthermore, the processes depicted in the drawings are not necessarily shown in a specific order or sequence to achieve the desired result. In some implementations, multitasking and parallel processing may be advantageous.

[0198] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for generating socio-economic future scenarios based on low-carbon energy, characterized in that, The method includes: Acquire interactive data between socioeconomic and energy systems; Based on the interactive data, key driving factor identification and state definition processing are performed to obtain multiple core driving factors affecting low-carbon transformation, as well as state definition information corresponding to each core driving factor. Generate multiple cross-influence matrices corresponding to the core driving factors; wherein each element in the cross-influence matrix is ​​used to characterize the influence intensity information of the first core driving factor on the second core driving factor among the two core driving factors corresponding to that element. Based on the state definition information of multiple core driving factors, multiple candidate scenarios are generated. Stability scores are performed on each of the multiple candidate scenarios based on the cross-influence matrix. Based on the stability score results, multiple target scenarios are determined from the multiple candidate scenarios. Based on the target scenario, quantitative modeling is performed to generate a future scenario model for low-carbon energy socio-economic development.

2. The method according to claim 1, characterized in that, The interactive data includes: structured data and unstructured data; The structured data includes at least one of the following: energy consumption data, carbon emission data, economic indicator data, and basic parameter data; The unstructured data includes: policy text data, technical document data, public opinion data, and enterprise operation data.

3. The method according to claim 2, characterized in that, The acquisition of interactive data between socioeconomic and energy systems includes: Acquire raw interaction data of socio-economic and energy systems; the raw interaction data includes: raw structured data and raw unstructured data; The original structured data is sequentially cleaned and normalized to obtain the structured data. as well as, The original unstructured data is cleaned, and a pre-trained natural language model is used to perform semantic understanding processing on the cleaned unstructured data to obtain semantic feature information corresponding to the unstructured data. The semantic feature information is then used as the unstructured data.

4. The method according to claim 1, characterized in that, Based on the interactive data, key driving factor identification and state definition processes are performed to obtain multiple core driving factors affecting the low-carbon transformation, including: With the goal of extracting candidate factors influencing low-carbon transformation, literature mining and policy text analysis were performed based on the interactive data. Based on the results of the literature mining and policy text analysis, several first candidate factors influencing low-carbon transformation were obtained. Through the expert consultation interface, expert consultation information is obtained, and based on the expert consultation information, several second candidate factors affecting the low-carbon transformation are obtained. Clustering is performed on the first candidate factor and the second candidate factor to obtain multiple candidate factor categories; By ranking the importance of multiple candidate factor categories, core driving factors covering socio-economic and energy technologies are obtained.

5. The method according to claim 1, characterized in that, The generation of the cross-influence matrix corresponding to the multiple core driving factors includes: Through multiple rounds of anonymous expert consultation, using structured questionnaires and scoring interfaces, we collected scores from different experts on the influence relationship between each pair of core driving factors, and generated candidate cross-influence matrices. The candidate cross-influence matrices are subjected to consistency verification and matrix filling optimization to obtain the cross-influence matrices corresponding to the multiple core driving factors.

6. The method according to claim 1, characterized in that, The stability scoring of various candidate scenarios based on the cross-influence matrix includes: For each candidate scenario, based on the cross-influence matrix, the cross-influence direction and state gradient difference between every two core driving factors in the candidate scenario are determined, and based on the cross-influence direction and the state gradient difference, the state compatibility coefficient of every two core driving factors in the corresponding candidate scenario is determined. And, based on the cross-influence matrix, determine the influence weight between each pair of core driving factors; Based on the state compatibility coefficient and the corresponding influence weight between each pair of core driving factors, the stability score corresponding to each candidate scenario is determined.

7. The method according to claim 6, characterized in that, Based on the stability score results, multiple target scenarios are determined from the multiple candidate scenarios, including: The candidate scenarios are sorted in descending order of their respective stability scores. According to the sorting, a preset number of candidate scenarios are determined from the candidate scenarios as the target scenarios; or, The stability scores and preset score thresholds corresponding to the multiple candidate scenarios are compared respectively; Candidate scenarios with stability scores greater than the preset score threshold are selected as target scenarios.

8. The method according to claim 1, characterized in that, The quantitative modeling based on the target scenario to generate a low-carbon energy socio-economic future scenario model includes: Density clustering is performed on the various target scenarios to obtain multiple scenario patterns, and semantic labels and policy interpretations corresponding to each scenario pattern are generated; Based on the scenario patterns, corresponding semantic tags, and policy interpretations, multiple scenario patterns are modeled to obtain low-carbon energy socio-economic future scenario models corresponding to each scenario pattern.

9. A device for generating future socio-economic scenarios based on low-carbon energy, characterized in that, The device includes: The acquisition module is used to acquire interactive data between socio-economic and energy systems. The factor identification module is used to perform key driving factor identification processing and state definition processing based on the interactive data, so as to obtain multiple core driving factors affecting low-carbon transformation and state definition information corresponding to each core driving factor. A matrix construction module is used to generate cross-influence matrices corresponding to multiple core driving factors; wherein each element in the cross-influence matrix is ​​used to characterize the influence intensity information of the first core driving factor on the second core driving factor among the two core driving factors corresponding to that element. The scenario screening module is used to generate multiple candidate scenarios based on the state definition information of multiple core driving factors, perform stability scoring on the multiple candidate scenarios based on the cross-influence matrix, and determine multiple target scenarios from the multiple candidate scenarios based on the stability scoring results. The quantitative modeling module is used to perform quantitative modeling based on the target scenario and generate a future scenario model of low-carbon energy socio-economic development.

10. A decision generation system based on a low-carbon energy socio-economic future scenario model, characterized in that, This includes memory, processor, expert evaluation terminal, policy decision-making terminal, and energy system model interface; The memory stores a low-carbon energy socio-economic future scenario model generated based on the low-carbon energy socio-economic future scenario generation method according to any one of claims 1-8. The energy system model interface is used to receive energy system models; the expert evaluation terminal is used to receive expert evaluation information. The processor is further configured to couple at least one of the received energy system model and the expert evaluation information with the low-carbon energy socio-economic future scenario model to obtain a policy decision, and output the policy decision to the policy decision terminal.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, it implements the steps of the method according to any one of claims 1-8.

12. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method according to any one of claims 1-8.

Images