CN-121981415-A - Dynamic prediction method, system, equipment and storage medium for medium-long term planning of power grid enterprise carbon business

CN121981415ACN 121981415 ACN121981415 ACN 121981415ACN-121981415-A

Abstract

The invention relates to the technical field of power grid carbon business planning, in particular to a dynamic prediction method, a system, equipment and a storage medium for long-term planning in power grid enterprise carbon business. The method comprises the steps of collecting multi-source heterogeneous data, carrying out grid business labeling processing to construct a dedicated database of the carbon business, quantitatively analyzing the synergistic effect of international rules, domestic policies and regional rules on the grid carbon business based on a hierarchical policy analysis system in combination with four-level evaluation indexes, establishing a scene suitability evaluation system, carrying out multi-dimensional evaluation scoring on carbon business cases outside the region, screening high-quality cases and refining reusable experience, carrying out basic trend fitting through a time sequence model, integrating dynamic factors into a machine learning model to construct a hybrid prediction model, setting a fluctuation threshold value based on a dynamic factor library, triggering correction when the factor change exceeds the threshold value, and carrying out dynamic adjustment on a prediction result. The problems of insufficient policy analysis, unscientific case reference, inaccurate prediction method and closed loop of floor execution are solved.

Inventors

HAN ZIJIAO
JIANG HAIWEI
ZHANG LI
SANG SHENGYING
WANG YANG
ZHANG YANNI
WANG ZHENG
YU JIA
ZHANG NA
HU JINGWEI
HU JINJING
LIU YUTONG

Assignees

国网辽宁省电力有限公司经济技术研究院
国网英大碳资产管理(上海)有限公司

Dates

Publication Date: 20260505
Application Date: 20251125

Claims (10)

1. A dynamic prediction method for medium-long term planning of power grid enterprise carbon business is characterized by comprising the steps of collecting multi-source heterogeneous data and carrying out power grid business labeling treatment, constructing a dedicated database of the carbon business, constructing four-level evaluation indexes by combining power grid business characteristics based on a hierarchical policy analysis system, and quantitatively analyzing the cooperative influence of international rules, domestic policies and regional rules on the power grid carbon business to obtain policy influence evaluation results; Establishing a scene suitability evaluation system, performing multidimensional evaluation scoring on the carbon service cases outside the area, screening high-quality cases, refining reusable experience, outputting a prediction result, setting a fluctuation threshold based on a dynamic factor library, triggering a correction mode when factor change exceeds the threshold, and dynamically adjusting the prediction result to obtain a medium-long-term trend prediction value of the carbon service; the method comprises the steps of combining a prediction result and policy influence evaluation, constructing a special floor closed-loop system of a power grid scene, including current situation diagnosis, layered target setting, annual task decomposition, execution monitoring and iterative optimization, and outputting a standardized planning report and a customized operation manual.
2. The method for dynamically predicting medium-long term planning of power grid enterprise carbon business of claim 1, wherein the method for acquiring multi-source heterogeneous data and carrying out power grid business labeling processing comprises the steps of acquiring international carbon rule files, national double-carbon policy files, provincial implementation rules, industry standard pole cases, enterprise internal business data and technical innovation dynamic data, extracting core constraint and support terms of policy files through natural language processing technology, classifying according to policy types, business types, technical types and customer types, and adding power grid exclusive labels such as green electricity transaction, customer types and the like; and constructing a unified carbon service database.
3. The method for dynamic prediction of medium-term planning of power grid enterprise carbon business as set forth in claim 2, wherein the analyzing the collaborative impact of international rules, national policies and regional rules on the power grid carbon business based on the hierarchical policy resolution system comprises, Constructing a joint word stock, extracting core constraint terms and support fields of a three-level policy by a text mining technology, constructing a four-level evaluation index system based on policy level, service association degree, influence degree and power grid suitability, and determining each index weight; The compliance requirements of international carbon rules on export enterprises, the support strength of domestic policies on carbon transaction and green electric service and the incentive measures of regional rules on local service expansion are respectively evaluated; and combining four-level indexes to carry out weighted scoring, quantifying the synergistic effect of three-level policies, and determining policy opportunities and compliance boundaries of the power grid carbon business.
4. The method for dynamically predicting medium-long term planning of power grid enterprise carbon business of claim 3, wherein the establishing of a scene suitability evaluation system performs multidimensional evaluation scoring on the out-of-area carbon business cases, and comprises setting five evaluation dimensions of policy environment similarity, business mode mobility, resource condition matching degree, technical difficulty controllability and implementation cost rationality; Setting a refinement index and determining weights for each dimension to form a complete suitability evaluation index system, scoring the case outside the region by adopting a fuzzy comprehensive evaluation method, and setting a screening threshold value; The method comprises the steps of refining a core mechanism and an implementation method for a high-quality case reaching a threshold value, and analyzing reusable key points of a policy docking mechanism, a business combination mode and a customer service system; And combining the local policy environment, the power grid business characteristics and the client requirements, adaptively adjusting the extracted experience to form a localization implementation path.
5. The method for dynamically predicting medium-long term planning of power grid enterprise carbon business of claim 4, wherein the steps of screening high-quality cases and refining reusable experience, and outputting a prediction result comprise performing differential stabilization processing on historical carbon business scale data by adopting an autoregressive integral moving average model to eliminate trending influence; the method comprises the steps of performing historical trend fitting based on a stable sequence, obtaining a basic predicted value by combining an autoregressive term and a moving average term, constructing a feature vector comprising the basic predicted value, a policy change factor, a technical innovation factor and a market demand factor, training the feature vector, and fusing the influence of a multidimensional dynamic factor to obtain an optimized predicted value.
6. The method for dynamically predicting medium-term planning in power grid enterprise carbon business of claim 5, wherein the method is characterized in that a fluctuation threshold is set based on a dynamic factor library, and a correction mode is triggered when factor changes exceed the threshold, and the method comprises the steps of constructing a policy type, a technical type and a market type dynamic factor library, setting a reference threshold and a fluctuation threshold for each factor; The method comprises the steps of carrying out weighted summation on single factor fluctuation according to factor category weights to obtain a comprehensive fluctuation index, calculating a correction coefficient based on the fluctuation index when the comprehensive fluctuation index exceeds a preset correction trigger value, and applying the correction coefficient to an output value of a mixed prediction model to obtain a final prediction result after dynamic correction.
7. The dynamic prediction method for medium-long term planning of power grid enterprise carbon business of claim 6, wherein the construction of a dedicated floor closed loop system of a power grid scene comprises diagnosing the current situation of the power grid enterprise carbon business from four dimensions of service scale, profit level, execution efficiency and customer coverage, defining advantages and short boards, setting medium-long term targets including financial targets, customer targets, execution targets and technical targets by combining trend prediction results; establishing a task progress visual monitoring platform, and evaluating the task completion condition according to a set period; And combining the dynamic factor change and the evaluation result, and timely adjusting the task target and the implementation path to form closed-loop optimization.
8. The dynamic prediction system for medium-long term planning of the power grid enterprise carbon business is based on the dynamic prediction method for medium-long term planning of the power grid enterprise carbon business according to any one of claims 1-7, and is characterized by further comprising a data acquisition and processing module, a data processing module and a data processing module, wherein the data acquisition and processing module is used for acquiring multi-source heterogeneous data and carrying out labeling processing on the power grid business to construct a dedicated database of the carbon business; The policy collaborative analysis module is used for quantitatively analyzing the collaborative influence of international rules, domestic policies and regional rules on the power grid carbon business based on a hierarchical policy analysis system; the case evaluation and conversion module is used for establishing a scene suitability evaluation system, evaluating and screening the out-of-area carbon business cases and refining localized experience; the dynamic prediction module is used for constructing a hybrid prediction model and correcting based on a dynamic factor library to obtain a predicted value of the medium-long-term trend of the carbon service; the floor closed-loop management module is used for constructing a dedicated floor closed-loop system of a power grid scene and realizing target setting, task decomposition, execution monitoring and iterative optimization; and the achievement output module is used for outputting a standardized planning report and a customized operation manual.
9. The computer equipment comprises a memory and a processor, wherein the memory stores a computer program, and the computer program is characterized in that the processor realizes the steps of the dynamic prediction method for long-term planning in the power grid enterprise carbon business according to any one of claims 1-7 when executing the computer program.
10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the dynamic prediction method for long-term planning in a power grid enterprise carbon business according to any one of claims 1 to 7.

Description

Dynamic prediction method, system, equipment and storage medium for medium-long term planning of power grid enterprise carbon business Technical Field The invention relates to the technical field of power grid carbon business planning, in particular to a dynamic prediction method, a system, equipment and a storage medium for long-term planning in power grid enterprise carbon business. Background The power grid enterprise carbon business is a key grip for realizing a double-carbon target, and comprises diversified business forms such as carbon transaction, green electricity transaction, carbon asset management and the like. At present, domestic power grid enterprises have continuously developed carbon business practices, but the problems of policy analysis fragmentation, blind reference, static prediction model and the like generally exist in the middle-long term planning and making process, the existing planning method mostly depends on manual experience judgment, and lacks of deep collaborative analysis on international rules, national policies and regional rules, so that planning results and policy environment suitability are insufficient, and meanwhile, the introduction of success cases in other regions lacks of a scientific evaluation system, and direct copying is easy to cause problems. At present, in a prediction link, the traditional method mainly adopts a single time sequence model or linear regression, cannot capture the complex influence of dynamic factors such as policy mutation, market fluctuation and the like on business development, has low prediction precision and lacks dynamic correction capability, and in a planning floor, the common method is to formulate static targets and task decomposition, lacks execution monitoring and iterative optimization links, and cannot adjust strategies in time when external environment changes, so that planning and actual execution are disjointed. In addition, the existing method is insufficient in consideration of power grid business characteristics, lacks data organization modes aiming at exclusive labels such as green electricity transaction, customer types and the like, and is difficult to form a data base for supporting accurate decisions. Disclosure of Invention The present invention has been made in view of the problems occurring in the prior art. Therefore, the problems to be solved by the invention are how to solve the problems of inaccurate policy analysis, unscientific case reference, inaccurate prediction method and non-closed loop of floor execution in long-term planning of the carbon business of the power grid enterprise. In order to solve the technical problems, the invention provides the following technical scheme: In a first aspect, an embodiment of the present invention provides a dynamic prediction method for long-term planning in a carbon service of a power grid enterprise, which includes collecting multi-source heterogeneous data and performing power grid service labeling processing, constructing a dedicated database of the carbon service, constructing four-level evaluation indexes based on a hierarchical policy analysis system and combining power grid service characteristics, and quantitatively analyzing synergistic effects of international rules, domestic policies and regional rules on the power grid carbon service to obtain policy impact evaluation results; Establishing a scene suitability evaluation system, performing multidimensional evaluation scoring on the carbon service cases outside the area, screening high-quality cases, refining reusable experience, outputting a prediction result, setting a fluctuation threshold based on a dynamic factor library, triggering a correction mode when factor change exceeds the threshold, and dynamically adjusting the prediction result to obtain a medium-long-term trend prediction value of the carbon service; the method comprises the steps of combining a prediction result and policy influence evaluation, constructing a special floor closed-loop system of a power grid scene, including current situation diagnosis, layered target setting, annual task decomposition, execution monitoring and iterative optimization, and outputting a standardized planning report and a customized operation manual. The invention relates to a dynamic prediction method for medium-and-long-term planning of power grid enterprise carbon business, which comprises the steps of collecting multi-source heterogeneous data and carrying out power grid business labeling treatment, wherein the method comprises the steps of collecting international carbon rule files, national double-carbon policy files, provincial implementation rules, industry standard pole cases, enterprise internal business data and technical innovation dynamic data; and constructing a unified carbon service database. The optimization method has the advantages that core constraint terms and support fields can be accurately extracted from massive policy files through cons