Association analysis method and system for electricity demand and carbon emission of emission control enterprise based on knowledge graph

Abstract

The invention provides a mapping knowledge domain-based electricity demand and carbon emission correlation analysis method and system for a high-voltage electric control and emission enterprise. According to the method and the system, through collection and preprocessing of multi-source heterogeneous data, association rule mining and causal relationship modeling technologies are combined, and the influence of the electrification degree on the power consumption demand and the carbon emission of an enterprise is quantified. By constructing the knowledge graph of the electricity-carbon demand relationship, the system realizes systematic management of the power demand and carbon emission of the enterprise, provides intelligent decision support, helps the enterprise to optimize an electrification transformation scheme, and reduces carbon emission and energy consumption cost.

Description

Technical Field

[0001] The present invention belongs to the field of energy management and carbon emission control, and specifically relates to a knowledge graph-based analysis method for the correlation between electricity demand and carbon emissions of highly electrified emission-controlled enterprises. Background Art

[0002] The intensification of global climate change has prompted countries around the world to introduce policies to encourage businesses to reduce carbon emissions. Highly electrified, emission-controlled enterprises, due to their high energy consumption, have become a key focus of carbon emission control efforts. However, existing enterprise energy management systems (EMS) are unable to comprehensively process multi-source, heterogeneous data, making it difficult to systematically analyze the impact of electrification levels on electricity demand and carbon emissions. Furthermore, businesses lack scientific decision-making support, making it difficult to effectively optimize electricity usage plans and reduce carbon emissions during electrification transformations.

[0003] However, some existing analytical methods estimate energy consumption and carbon emissions based solely on single factors or linear models, failing to reveal the complex relationships between multiple factors. This makes it difficult for companies to accurately assess the impact of different electrification scenarios on carbon emissions and electricity costs, and they lack decision-making support tools. Therefore, there is an urgent need for a system that can efficiently integrate data and analyze the relationship between electrification levels and corporate carbon emissions, helping companies achieve intelligent energy management. Summary of the Invention

[0004] Based on the above background, this paper proposes a knowledge graph-based method for analyzing the correlation between electricity demand and carbon emissions of emission-controlled enterprises. The specific implementation scheme is as follows:

[0005] The first aspect of the present invention provides a method for analyzing the correlation between electricity demand and carbon emissions of emission-controlled enterprises based on a knowledge graph, comprising the following steps:

[0006] Collect energy consumption, production operations, policies and regulations, and environmental monitoring data from the company's multi-source heterogeneous data and perform data preprocessing to form a comprehensive data set;

[0007] By using association rule mining algorithms, we analyze the correlation between electrification transformation progress, electricity demand, and carbon emissions, and identify key factors.

[0008] Use the RDF model to build a knowledge graph related to electrification transformation and carbon emissions, representing related entities and their relationships;

[0009] Based on the constructed knowledge graph, it provides optimization suggestions for electrification transformation and evaluates the impact of the transformation on energy costs and carbon emissions.

[0010] Furthermore, the data preprocessing includes:

[0011] Data cleaning, removing outliers and redundant data;

[0012] Missing value filling, using interpolation or historical data to fill missing values;

[0013] Data standardization: unify the units of data from different sources and use the following formula for standardization:

[0014]

[0015] Among them, x is the original data, μ is the mean, σ is the standard deviation, and x′ is the standardized data.

[0016] Furthermore, the association rule mining algorithm is an Apriori algorithm, and analyzing the correlation between the electrification transformation process and electricity demand and carbon emissions includes the following steps:

[0017] The support of the Apriori algorithm calculates the frequency relationship in the association rules:

[0018]

[0019] Among them, A represents the electrification transformation process of the enterprise, and B represents electricity demand or carbon emissions;

[0020] Confidence, which measures the probability of B occurring in all cases containing A:

[0021]

[0022] Lift, which measures the independence between A and B:

[0023]

[0024] If the lift is greater than 1, it means there is a strong correlation between A and B.

[0025] Furthermore, the identification of key factors adopts principal component analysis method, which specifically includes:

[0026] Standardized data: Standardize the original data to obtain a data matrix X with zero mean and unit variance;

[0027] Covariance matrix calculation: Calculate the covariance matrix of the data \C, the formula is:

[0028]

[0029] Eigenvalue decomposition: Perform eigenvalue decomposition on the covariance matrix C to find the eigenvalue λ and the corresponding eigenvector W i ;

[0030] Principal component projection: Select the eigenvectors corresponding to the first k largest eigenvalues to form the projection matrix W k , project the original data into a new low-dimensional space:

[0031] Z=XW k

[0032] Among them, Z is the data after dimensionality reduction, W k is the eigenvector matrix of the first k principal components.

[0033] Furthermore, the use of the RDF model to construct a knowledge graph related to electrification transformation and carbon emissions includes:

[0034] Extract entities from unstructured data through natural language processing techniques;

[0035] Use the RDF model to uniformly represent enterprise production activities, electrification transformation, electricity demand, carbon emission entities and their relationships, build a knowledge graph, and store it in a graph database.

[0036] Furthermore, the results based on the knowledge graph provide optimization suggestions for electrification transformation and evaluate the impact of the transformation on energy costs and carbon emissions, including:

[0037] Forecast electricity demand and carbon emissions based on different electrification scenarios;

[0038] A differential analysis of carbon emissions from different electricity sources is conducted, and the carbon emissions calculation formula is as follows:

[0039]

[0040] Among them, C is the total carbon emissions, E i is the energy consumption of i, F i is the corresponding carbon emission coefficient.

[0041] Furthermore, the method further comprises:

[0042] The effects of the electrification transformation plan are displayed through a visual interactive interface, including changes in electricity demand, carbon emissions and energy costs. Different scenarios are displayed based on user instructions for decision simulation and effect comparison.

[0043] Furthermore, the method further comprises:

[0044] Simulate electricity demand, carbon emissions, and energy costs of enterprises under different electrification transformation scenarios;

[0045] Verify the accuracy and feasibility of the optimization plan based on historical data and provide decision support.

[0046] The second aspect of the present invention provides a knowledge graph-based analysis system for the correlation between electricity demand and carbon emissions of emission-controlled enterprises, including:

[0047] The data collection module is used to collect energy consumption, production operations, policies and regulations, and environmental monitoring data from the enterprise's multi-source heterogeneous data to form a comprehensive data set;

[0048] The data preprocessing module is used to clean, convert and standardize the collected data to ensure its consistency, integrity and availability;

[0049] The correlation analysis module is used to analyze the correlation between the degree of electrification and corporate electricity demand and carbon emissions through association rule mining algorithms, and identify key factors;

[0050] A knowledge graph construction module is used to construct a knowledge graph related to electrification transformation and carbon emissions using an RDF model, representing related entities and their relationships;

[0051] The decision support module is used to provide optimization suggestions for electrification transformation based on the results of the knowledge graph and evaluate the impact of the transformation on energy costs and carbon emissions.

[0052] Furthermore, the knowledge graph construction module is also used to automatically update the changing data of the enterprise's electricity demand and carbon emissions, and incorporate the data into the graph structure to ensure dynamic adjustment and real-time updating of the electricity-carbon demand relationship.

[0053] The beneficial effects of the present invention are as follows:

[0054] The knowledge graph-based analysis method and system for the correlation between electricity demand and carbon emissions of emission-controlled enterprises provided by the present invention can provide enterprises with intelligent decision-making support functions based on the data and correlation relationships in the knowledge graph. By analyzing the impact of electrification transformation on the electricity demand and carbon emissions of enterprises, a detailed electrification transformation recommendation report is generated, providing analysis results based on multiple aspects such as the company's current energy consumption, equipment efficiency, and policy incentives. Customized carbon emission reduction strategies can also be generated to help companies formulate optimal emission reduction plans and implementation paths by simulating changes in carbon emissions under different scenarios. Finally, the method and system of the present invention can generate decision support reports to help corporate management make scientific long-term strategic decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a flow chart of an embodiment of the knowledge graph-based method for analyzing the correlation between electricity demand and carbon emissions of highly electrified emission-controlled enterprises of the present invention.

[0056] Figure 2 It is a flowchart of data collection, data preprocessing, correlation analysis, knowledge graph construction and decision support in the embodiment of the method of the present invention.

[0057] Figure 3It is a flow chart of the process from data collection to final decision support in the embodiment of the method of the present invention, which goes through the steps of data collection, preprocessing, correlation analysis, knowledge graph construction and decision support in sequence. Specific implementation plan

[0058] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0059] Unless otherwise defined, technical or scientific terms used in the present invention shall have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs.

[0060] See also Figure 1-3 The first embodiment of the present invention is a method for analyzing electricity demand of emission-controlled enterprises based on a knowledge graph, comprising the following steps:

[0061] S1. Extract various types of data from the enterprise's multi-source data through the data acquisition module, and clean and standardize the data in the data preprocessing module to ensure data consistency.

[0062] As a preferred embodiment, step S1 specifically includes:

[0063] S11. Extract historical energy consumption data from the enterprise energy management system (EMS). The system connects with the EMS to collect the following data:

[0064] ① Daily, weekly, and monthly electricity consumption: The system extracts historical electricity consumption records from the EMS and breaks them down to hourly granularity to analyze the company's energy demand fluctuations over different time periods.

[0065] ② Electricity bills and peak and valley electricity price information: Energy bills record a company's electricity expenses over different time periods, while peak and valley electricity prices reflect the distribution of a company's energy consumption during peak and off-peak periods, helping to identify the company's electricity cost structure.

[0066] ③Energy consumption distribution: The system also extracts energy consumption data from different departments or production lines to help companies analyze the energy consumption proportions of different departments and provide a reference for optimizing production electricity consumption decisions.

[0067] S12. Obtain real-time data on power consumption during the production process from the production line control system (SCADA). Production control systems (SCADA) are widely used in the industrial field to monitor and manage real-time data in the enterprise production line. By integrating with the SCADA system, the system can collect real-time power consumption data during the enterprise production process, including:

[0068] ① Equipment-level electricity consumption data: The system can extract the electricity consumption of each production equipment and analyze the energy consumption of the equipment at different production stages. This helps to evaluate the energy efficiency of different equipment, identify high-energy-consuming equipment, and provide data support for electrification transformation;

[0069] ② Real-time production power demand: The SCADA system can provide power load fluctuation data of the production line, monitor the power demand changes of each production line in real time during the production process, and ensure the timeliness and stability of energy supply;

[0070] ③ Production events and energy consumption fluctuations: During the production process, energy consumption may fluctuate significantly due to failures, shutdowns, or process adjustments. The system helps optimize production energy consumption strategies and reduce energy waste by capturing the correlation between these events and power consumption fluctuations.

[0071] S13. Extract data related to corporate compliance from policy documents and regulatory documents. To ensure that companies comply with policy and regulatory requirements during electrification and carbon emission management, the system needs to extract data related to carbon emission limits, subsidy policies, and compliance requirements from policy documents and regulations. This type of data usually exists in the form of unstructured documents (such as PDFs and Word files) or web pages. The system uses natural language processing (NLP) and text parsing technology to process these files, including:

[0072] ① Policy and Regulation Analysis: The system uses text parsing and natural language processing technologies to extract clauses related to enterprise electrification and carbon emissions from national and local policy documents, such as carbon emission limits and electrification incentive policies.

[0073] ② Subsidies and incentives: Extract relevant national policy support and subsidy information for enterprise electrification transformation from documents to provide reference for enterprises to obtain policy benefits during the electrification upgrade process;

[0074] ③ Environmental standards and carbon emission targets: Analyze the environmental standards and carbon emission targets that companies must follow to ensure that companies comply with national and local regulations during the electrification transformation process, and use system warnings to avoid penalties for violations.

[0075] S14. Collect climate data and environmental monitoring data.

[0076] S15, a unified data collection interface, supports multiple data collection modes, including API integration, database query, file parsing, and text mining. Through this unified interface, the system can efficiently and comprehensively collect data from various sources, providing a sufficient data foundation for subsequent data preprocessing, analysis, and optimization.

[0077] S16, clean up the data. The system first cleans up all the data, including: removing duplicate data, handling error values, and filtering incomplete data.

[0078] S17. Missing value filling: To ensure the continuity and integrity of the analysis data, the system provides a variety of missing value filling strategies:

[0079] ① Interpolation: For data from continuous time periods (such as electricity consumption), linear interpolation and polynomial interpolation are used to fill in missing values. Interpolation methods use existing time series data to infer data from missing time periods, maintaining data continuity and rationality.

[0080] ② Filling with historical average values: For some non-continuous data (such as the power consumption of a specific device), the system uses the average value of historical data to fill in the gaps.

[0081] ③Filling based on upstream and downstream data: The system infers missing data based on the upstream and downstream data of related links.

[0082] S18. Data format conversion

[0083] S19, standardization processing, including: measurement unit standardization, numerical normalization, time synchronization

[0084] S2. Utilize association rule mining algorithms to analyze the complex relationships between electrification levels, electricity demand, and carbon emissions. Dimensionality reduction techniques reduce data complexity and enhance the effectiveness of association analysis. This step utilizes a series of data mining and machine learning algorithms to reveal the quantitative impact of electrification on enterprise energy consumption and carbon emissions.

[0085] As a preferred embodiment, step S2 specifically includes:

[0086] S21, Association Rule Mining Algorithm: The system first uses the Apriori algorithm to identify association patterns between corporate production activities and energy consumption. Based on large-scale data sets, this algorithm uncovers the core factors that influence corporate electricity demand and carbon emissions.

[0087] ① Correlation between enterprise production activities and electricity demand: The Apriori algorithm analyzes the correlation between the operating time, production volume, and production efficiency of enterprise production equipment and electricity demand, helping to identify production equipment or production processes that are high in energy consumption. For example, the system may discover that the energy consumption of a specific piece of equipment increases significantly during a specific production period or stage, and thus correlate this with the need for electrification of the equipment.

[0088] ② The correlation between electrification progress and carbon emissions: During the electrification process, the system can mine data to identify the actual impact of increased electrification on carbon emissions. For example, through association rule analysis, the system may discover that companies using clean energy at higher electrification levels significantly reduce carbon emissions, thereby identifying the most valuable electrification transformation links.

[0089] ③ Interaction Analysis of Multidimensional Factors: In addition to single-factor correlation analysis, the system combines multiple factors (such as equipment energy efficiency, power source, and production time) to identify the combined impact of multidimensional interactions on a company's power demand and carbon emissions. These correlation patterns will help companies determine under what circumstances electrification can maximize carbon reductions.

[0090] S22. Principal Component Analysis (PCA) technology: The system uses PCA to reduce data dimensionality. By combining related variables in high-dimensional data into principal components, the data structure is simplified and the key factors influencing a company's electricity demand and carbon emissions, such as production scale, equipment energy efficiency, and production process, are extracted. The system also quantitatively analyzes the contribution of each variable to electricity demand and carbon emissions by evaluating the weights of different principal components, helping managers focus on optimizing the most important electrification transformation links or process adjustments.

[0091] S23. Quantitative prediction of different electrification transformation scenarios. The system predicts the company's future electricity demand and carbon emissions by simulating different scenarios.

[0092] ① Scenario Design and Simulation: The system sets several electrification transformation scenarios, including full electrification, partial electrification, and traditional energy. Under each scenario, the system simulates the company's production process, equipment operation, and changes in energy sources, quantifying the power demand and carbon emissions under each scenario.

[0093] ② Electricity Demand Forecast: For each scenario, the system predicts the company's electricity demand at different levels of electrification. For example, in a fully electrified scenario, the company may require more electricity to replace traditional energy sources, while also improving energy efficiency and reducing overall energy consumption. In a partially electrified scenario, the system analyzes which production processes are most suitable for transformation first to optimize electricity use.

[0094] ③ Carbon Emissions Forecast: The system calculates the carbon emissions for each electrification scenario based on the carbon emission coefficient of the electricity source (e.g., renewable energy or traditional fossil fuels). Scenario analysis compares the carbon emissions of various conversion options, helping companies identify those with the greatest potential for carbon reduction. For example, the system may find that while full electrification offers the greatest carbon emission reduction, partial electrification offers a more cost-effective carbon reduction.

[0095] ④ Contribution calculation: The system further calculates the contribution of electrification transformation to the company's total carbon emissions. By comparing carbon emission data before and after the transformation, the system can quantitatively demonstrate the carbon emission reduction effect of electrification transformation under different scenarios.

[0096] S24, data visualization and report generation, the system provides a correlation diagram between electrification progress and electricity demand, a carbon emission reduction comparison chart and an optimization suggestion report

[0097] S25. Causal relationship modeling scheme uses structural equation modeling (SEM) to establish causal relationships, including: selecting key variables such as "electrification level", "energy consumption", "production equipment energy efficiency" and "carbon emissions" based on previous correlation analysis, constructing hypothetical causal paths between variables, quantifying the direct and indirect effects of variables through path analysis, using multi-path modeling to handle complex variable interactions, and finally optimizing the causal structure through goodness of fit evaluation and model modification to ensure that the model accurately reflects the impact of electrification transformation on enterprise energy consumption and carbon emissions.

[0098] S26. Use Bayesian networks (BN) to determine the comprehensive impact of the interaction of multiple factors, including constructing conditional probability networks to demonstrate the dependency relationship between variables such as "electrification level", "energy price", and "carbon emission policy", handling the complex interaction of multiple variables, flexibly predicting corporate carbon emissions and electricity demand under different factor combinations, handling uncertainty and simulating different energy price and policy scenarios, performing causal inference to optimize electrification transformation decisions, and providing enterprises with reliability assessments of electricity demand and carbon emissions under electrification transformation scenarios through scenario simulation and probabilistic prediction.

[0099] S27, data visualization and model output, the system uses visualization tools to intuitively present the model results of SEM and BN. The main visualization outputs include: causal relationship path diagram, conditional probability network diagram, scenario simulation chart

[0100] S3. Using natural language processing technology to extract entities and relationships from policy documents, constructing a knowledge graph of the electricity-carbon demand relationship, and implementing storage and query through a graph database is a preferred implementation. Step S3 specifically includes:

[0101] S31. Entity recognition: Through natural language processing (NLP) technology, the system can automatically identify entities related to energy consumption, carbon emissions, and electrification transformation. These entities will serve as nodes in the knowledge graph, reflecting the core content of the company's electricity-carbon needs. Specifically, it includes key entity extraction, combined with rule recognition and deep learning models, to identify key entities such as "enterprise" (the subject of electrification transformation and carbon emission management), "power consumption" (the electricity demand of the enterprise in the production stage), "carbon emissions" (carbon emission data in the production process), "electrification transformation" (transformation measures for low-carbon goals), "energy prices", "subsidy policies", "production equipment", etc., and classify these entities into categories such as "physical entity" or "environmental entity" through entity classification.

[0102] S32. Identify the positive correlation between electricity consumption and carbon emissions, the emission reduction path of electrification and carbon reduction, the dependence of electricity consumption on energy prices, and the compliance constraints of enterprises being restricted by policies and regulations.

[0103] S33, RDF model representation, uses the RDF (Resource Description Framework) model to perform structured representation of entities and their relationships in the knowledge graph, and uses triples (subject-predicate-object) to represent the relationships between an enterprise's electricity consumption, carbon emissions, and policies and regulations, such as "Enterprise A - Consumption - Electricity 1000kWh" or "Electrification transformation - Reduction - Carbon emissions 20%". Each triple is an edge connecting entity nodes in the graph and is stored in the graph database Neo4j.

[0104] S34. Graph visualization and query. Using graph visualization tools, the system graphically displays the relationships between energy consumption, carbon emissions, and policy compliance during a company's electrification process. Entities and relationships are presented as a network diagram of nodes and edges. Furthermore, the system supports SPARQL queries, allowing users to customize complex relationship searches and generate customized analytical reports on electrification progress, electricity demand, and carbon emissions.

[0105] S4. Based on the associated information in the knowledge graph, a visual analysis interface is provided to help corporate decision makers optimize electricity usage plans and reduce carbon emissions and energy costs.

[0106] As a preferred embodiment, step S4 specifically includes:

[0107] S41. Provide feasibility analysis for electrification upgrades. The system evaluates the potential and economic benefits of electrification by analyzing the company's energy consumption, equipment energy efficiency, policy incentives, and the input-output ratio of transformation. Transformation potential assessment identifies high-energy-consuming links, cost-benefit analysis quantifies investment returns, and policy support analysis combines policy incentives to help companies optimize transformation decisions and obtain subsidies.

[0108] S42. Provides carbon emission reduction optimization strategies. Based on correlation analysis and causal models, the system develops practical emission reduction plans in conjunction with enterprise operations and policy requirements. Emission reduction path analysis identifies the carbon reduction potential under different electrification scenarios and finds the optimal path. Long-term emission reduction planning sets future emission reduction targets and equipment upgrade schedules based on production plans to ensure policy compliance. Scenario simulation and risk assessment help companies simulate uncertain market and policy environments, optimize decisions, and mitigate potential risks.

[0109] S43. Data visualization and optimization report generation. The system uses visual charts and optimization reports to help enterprise management clearly understand the impact of electrification transformation on electricity demand and carbon emissions. A graph linking electrification level and production cost illustrates the changes in electricity demand and costs at different electrification levels. A carbon reduction target and progress chart compares a company's current carbon emissions status with its target, enabling assessment of the progress of transformation projects. The system also automatically generates customized optimization reports, providing electrification transformation recommendations, electricity consumption forecasts, carbon emission strategies, and financial return analysis to support decision-making. Furthermore, real-time data tracking and alarm functions ensure that enterprises can monitor and optimize their energy management in real time.

[0110] S44, Multi-Scenario Simulation, systematically constructs various electrification transformation scenarios to help companies predict energy consumption and carbon emissions under different strategies, ensuring optimized implementation plans. The Full Electrification Transformation Scenario simulates the electricity demand, carbon emission reduction, and long-term cost changes of a company after full electrification; the Partial Electrification Transformation Scenario assesses the impact of gradual electrification on overall benefits; the Combination of Traditional Energy and Electrification Scenario demonstrates the optimal strategy for achieving carbon emission reduction while retaining some traditional energy sources; and the Renewable Energy Utilization Scenario simulates the impact of using different proportions of clean energy on energy demand and carbon emissions, supporting companies in formulating long-term energy plans.

[0111] S45. Optimization feedback of simulation results. The system provides adjustment suggestions for the electrification transformation of enterprises based on the simulation and verification results through optimization feedback reports to ensure the realization of expected benefits. Model parameter optimization adjusts model parameters according to actual data to improve prediction accuracy. For example, adjustments are made when the energy efficiency performance of equipment is inconsistent. Optimization suggestions for transformation plans are based on scenario simulation feedback, providing suggestions for prioritizing the transformation of inefficient links or increasing the use of clean energy. Long-term strategic adjustments help enterprises optimize the path of electrification transformation through backtesting feedback, achieve carbon emission reduction targets and improve energy efficiency.

[0112] S46. Data visualization and report generation: The system presents simulation scenarios and historical verification results in graphical form. The simulation scenario comparison chart shows the power demand and carbon emissions performance under different electrification transformation scenarios, visually demonstrating the impact of each scenario on operations. The historical data verification chart visualizes the comparison results between forecasts and historical data, helping companies evaluate model accuracy and adjust solutions. The optimization recommendation report automatically generates recommendations for path optimization, carbon emission reduction adjustments, and future planning based on simulation and verification results, providing data support and implementation planning guidance for electrification transformation.

[0113] The second embodiment of the present invention provides a knowledge graph-based analysis system for the correlation between electricity demand and carbon emissions of emission-controlled enterprises, including:

[0114] The data collection module is used to collect energy consumption, production operations, policies and regulations, and environmental monitoring data from the enterprise's multi-source heterogeneous data to form a comprehensive data set;

[0115] The data preprocessing module is used to clean, convert and standardize the collected data to ensure its consistency, integrity and availability;

[0116] The correlation analysis module is used to analyze the correlation between the degree of electrification and corporate electricity demand and carbon emissions through association rule mining algorithms, and identify key factors;

[0117] A knowledge graph construction module is used to construct a knowledge graph related to electrification transformation and carbon emissions using an RDF model, representing related entities and their relationships;

[0118] The decision support module is used to provide optimization suggestions for electrification transformation based on the results of the knowledge graph and evaluate the impact of the transformation on energy costs and carbon emissions.

[0119] The system in this embodiment is further described below.

[0120] First, the system in this embodiment builds a unified interface through the data acquisition module to extract multi-dimensional data from multiple data sources inside and outside the enterprise. After connecting with the enterprise energy management system (EMS), the system regularly obtains historical energy consumption data from the EMS, which is refined to the hourly level to capture daily, weekly and monthly energy consumption fluctuations. Electricity bill data and peak and valley electricity price information are automatically extracted from the bill data stream through bill format parsing, so that the subsequent analysis can accurately reflect the energy cost distribution of the enterprise in different time periods. The energy consumption data of different departments or production lines of the enterprise are decomposed and collected by querying and processing some data tables in the EMS system.

[0121] Next, the system in this embodiment is integrated with the production control system (SCADA) to obtain real-time power consumption data during the production process. To achieve real-time monitoring and efficient data transmission, the system connects to the data stream of the SCADA system and uses a combination of timed polling and event-driven methods to capture device-level power consumption data. The production line power load data in SCADA is associated with the device ID and production plan to capture power load fluctuations in the production process. Key event data such as equipment start-up, shutdown, and failure during the production process are automatically collected through an event-triggered mechanism to ensure that energy consumption changes during production fluctuations can be recorded and analyzed in real time.

[0122] In order to ensure that the system can parse and process the policy and regulatory requirements faced by enterprises in electrification transformation and carbon emission management, the system further implements the processing of policy documents and regulatory documents. Through natural language processing (NLP) technology, the system can automatically parse policy documents in various formats, such as PDF, Word documents or web page text, and extract data related to carbon emission limits, electrification incentive policies and compliance requirements. The specific processing method is to first use a text parsing tool to convert unstructured documents into a parseable text format, and then use the NLP model to identify the legal terms and data points involved in the text, and convert this data into structured information for subsequent analysis.

[0123] The system in this embodiment also integrates meteorological and environmental monitoring data via external APIs to obtain information on climate conditions and environmental pollution that are closely related to a company's energy consumption. Meteorological data such as temperature, humidity, and air pressure are recorded by the system by time period and correlated with the company's production and operation data to help analyze the impact of climate conditions on energy consumption. Environmental monitoring data, through an integrated data interface with air quality monitoring stations, regularly obtains information on the concentrations of pollutants such as PM2.5 and sulfur dioxide. This data is used to assess a company's carbon emissions performance under different environmental conditions.

[0124] After completing data collection, the system in this embodiment preprocesses all collected data to ensure data consistency and accuracy. Data cleaning is the first step in preprocessing. The system automatically removes duplicate data, marks and processes outliers by setting thresholds and rules. Different strategies are used to handle missing values: for continuous energy consumption data, the system uses linear interpolation or polynomial interpolation to fill in the gaps, while for discrete categorical data, the system fills in the gaps based on business rules or historical averages. In addition, the system standardizes all data from different sources into the same unit of measurement through standardization, such as converting all electricity consumption data into kilowatt-hours (kWh), ensuring that the data in subsequent calculations and analyses can be effectively compared and processed. The format of the time field is also unified into the ISO standard timestamp to ensure that all time series data are aligned on the time axis.

[0125] After data preprocessing is completed, the system in this embodiment performs correlation analysis based on these standardized data to explore the complex relationship between the company's electrification level, electricity demand and carbon emissions. First, the system uses the Apriori algorithm to mine association rules to identify hidden patterns and associations from the company's production equipment operation data and energy consumption data. For example, the system can find that the electricity demand of certain specific equipment increases significantly during specific production periods, helping companies identify high-energy-consuming links. Through this pattern analysis, companies can clearly identify which production equipment and processes have the greatest impact on total energy consumption, thereby providing optimization directions for subsequent electrification transformation.

[0126] To further simplify the data structure and focus on core influencing factors, the system in this embodiment uses principal component analysis (PCA). By reducing the dimensionality of high-dimensional data, the system extracts the principal components that most influence a company's electricity demand and carbon emissions. For example, the system can identify key variables such as production scale, equipment energy efficiency, and production process flow, and quantitatively analyze their contributions to electricity demand and carbon emissions by quantifying the weights of each principal component.

[0127] After completing the correlation analysis, the system in this embodiment uses causal modeling technology to quantify the specific impact of electrification transformation on the enterprise's electricity demand and carbon emissions. The system constructs a causal path analysis framework based on the structural equation model (SEM). Through data training, the system can clearly understand how different electrification transformation plans affect the enterprise's electricity consumption and carbon emissions. Next, the system conducts multi-scenario simulations to simulate the enterprise's electricity demand and carbon emissions under full electrification, partial electrification, and traditional energy scenarios. Through this multi-scenario prediction, the system can provide enterprises with scientific transformation plan recommendations to help them achieve optimal energy management and carbon emission control under different levels of electrification.

[0128] To further support the intelligent decision-making of enterprises, the system in this embodiment constructs a knowledge graph to display the above-mentioned correlation analysis results in the form of a graph. The first step in building a knowledge graph is entity recognition. The system automatically identifies key entities such as the company's electricity consumption, carbon emissions, and electrification transformation through natural language processing technology, and stores these entities and their relationships in the form of triples in the graph database. For example, a triple such as "Company A-Consumption-Electricity 1000kWh" represents the company's electricity demand. Next, the system uses an automatically generated knowledge graph to display the association path between electrification transformation and carbon emissions, helping companies to intuitively analyze the relationship between each link.

[0129] Based on the data and relationships in the knowledge graph, the system in this embodiment provides intelligent decision-making support for enterprises. By analyzing the impact of electrification on a company's electricity demand and carbon emissions, the system generates a detailed electrification recommendation report, providing analysis results based on multiple factors such as the company's current energy consumption, equipment efficiency, and policy incentives. The system also generates customized carbon reduction strategies, simulating carbon emission changes under different scenarios to help companies develop optimal emission reduction plans and implementation paths. Finally, the system generates a decision support report to help company management make scientific, long-term strategic decisions.

[0130] It should be noted that the method of the embodiment of the present invention can be performed by a single device, such as a computer or server. The method of this embodiment can also be applied in a distributed scenario, where multiple devices cooperate to perform the method. In such a distributed scenario, one of the multiple devices may only perform one or more steps of the method of the embodiment of the present invention, and the multiple devices will interact with each other to complete the method.

[0131] It should be noted that the above description is limited to some embodiments of the present invention. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in an order different from that described in the above embodiments and still achieve the desired results. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0132] The embodiments of the present invention are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the embodiments of the present invention should be included in the scope of protection of the present invention.

Claims

1. A knowledge graph-based method for analyzing the correlation between electricity demand and carbon emissions of emission-controlled enterprises, characterized by: The following steps are involved: Collect energy consumption, production operations, policies and regulations, and environmental monitoring data from the company's multi-source heterogeneous data and perform data preprocessing to form a comprehensive data set; By using association rule mining algorithms, we analyze the correlation between electrification transformation progress, electricity demand, and carbon emissions, and identify key factors. Use the RDF model to build a knowledge graph related to electrification transformation and carbon emissions, representing related entities and their relationships; Based on the constructed knowledge graph, it provides optimization suggestions for electrification transformation and evaluates the impact of the transformation on energy costs and carbon emissions.

2. The method according to claim 1, characterized in that The data preprocessing includes: Data cleaning, removing outliers and redundant data; Missing value filling, using interpolation or historical data to fill missing values; Data standardization: unify the units of data from different sources and use the following formula for standardization: Among them, x is the original data, μ is the mean, σ is the standard deviation, x ′ The data are standardized.

3. The method according to claim 1, characterized in that The association rule mining algorithm is the Apriori algorithm. The analysis of the correlation between the electrification transformation process and electricity demand and carbon emissions includes the following steps: The support of the Apriori algorithm calculates the frequency relationship in the association rules: Among them, A represents the electrification transformation process of the enterprise, and B represents electricity demand or carbon emissions; Confidence, which measures the probability of B occurring in all cases containing A: Lift, which measures the independence between A and B: If the lift is greater than 1, it means there is a strong correlation between A and B.

4. The method according to claim 3, characterized in that The identification of key factors adopts principal component analysis method, which specifically includes: Standardized data: Standardize the original data to obtain a data matrix X with zero mean and unit variance; Covariance matrix calculation: Calculate the covariance matrix of the data \C, the formula is: Eigenvalue decomposition: Perform eigenvalue decomposition on the covariance matrix C to find the eigenvalue λ and the corresponding eigenvector W i ; Principal component projection: Select the eigenvectors corresponding to the first k largest eigenvalues to form the projection matrix W k , project the original data into a new low-dimensional space: Z=XW k Among them, Z is the data after dimensionality reduction, W k is the eigenvector matrix of the first k principal components.

5. The method according to claim 4, characterized in that The knowledge graph related to electrification transformation and carbon emissions constructed using the RDF model includes: Extract entities from unstructured data through natural language processing techniques; Use the RDF model to uniformly represent enterprise production activities, electrification transformation, electricity demand, carbon emission entities and their relationships, build a knowledge graph, and store it in a graph database.

6. The method according to any one of claims 1 to 5, characterized in that The results based on the knowledge graph provide optimization suggestions for electrification transformation and evaluate the impact of the transformation on energy costs and carbon emissions, including: Forecast electricity demand and carbon emissions based on different electrification scenarios; A differential analysis of carbon emissions from different electricity sources is conducted, and the carbon emissions calculation formula is as follows: Among them, C is the total carbon emissions, E i is the energy consumption of i, F i is the corresponding carbon emission coefficient.

7. The method according to claim 6, characterized in that Also includes: The effects of the electrification transformation plan are displayed through a visual interactive interface, including changes in electricity demand, carbon emissions and energy costs. Different scenarios are displayed based on user instructions for decision simulation and effect comparison.

8. The method according to claim 6, characterized in that Also includes: Simulate electricity demand, carbon emissions, and energy costs of enterprises under different electrification transformation scenarios; Verify the accuracy and feasibility of the optimization plan based on historical data and provide decision support.

9. A knowledge graph-based analysis system for the correlation between electricity demand and carbon emissions of emission-controlled enterprises, characterized by: include: The data collection module is used to collect energy consumption, production operations, policies and regulations, and environmental monitoring data from the enterprise's multi-source heterogeneous data to form a comprehensive data set; The data preprocessing module is used to clean, convert and standardize the collected data to ensure its consistency, integrity and availability; The correlation analysis module is used to analyze the correlation between the degree of electrification and corporate electricity demand and carbon emissions through association rule mining algorithms, and identify key factors; A knowledge graph construction module is used to construct a knowledge graph related to electrification transformation and carbon emissions using an RDF model, representing related entities and their relationships; The decision support module is used to provide optimization suggestions for electrification transformation based on the results of the knowledge graph and evaluate the impact of the transformation on energy costs and carbon emissions.

10. The system according to claim 9, characterized in that The knowledge graph construction module is also used to automatically update the changing data of enterprise electricity demand and carbon emissions, and incorporate the data into the graph structure to ensure dynamic adjustment and real-time update of the electricity-carbon demand relationship.

Images