An analysis management system based on double-carbon multi-source data

Abstract

The application discloses a kind of based on double carbon multi-source data analysis management system, including multi-source data acquisition module, acquisition enterprise in park, energy facility, building multi-source data, wherein, multi-source data includes energy consumption data, production activity data, environmental monitoring data, meteorological data and enterprise carbon quota data;Data preprocessing module, the multi-source data collected are cleaned, converted and standardized, remove duplicate, error and missing data, and different format data is converted into uniform format;The beneficial effects of the application are: break through the limitation that data source is single in traditional carbon emission management system, data integration is difficult, can comprehensively integrate multi-source data of each level of enterprise in park, energy facility, building etc., including energy consumption data, production activity data, environmental monitoring data etc., provide abundant data support for accurately accounting park carbon emission.

Description

Technical Field

[0001] This invention belongs to the field of carbon emission management technology, specifically relating to an analysis and management system based on dual-carbon multi-source data. Background Technology

[0002] With the growing severity of global climate change, reducing carbon emissions has become a global consensus. Currently, data acquisition is difficult and fragmented, and there is a lack of effective integration methods, making it difficult to fully and accurately grasp the carbon emission situation of the park and to integrate multi-dimensional data, resulting in a high carbon accounting deviation rate. For example, enterprises' energy consumption data, production activity data and environmental monitoring data are often stored in different systems, making it impossible to achieve automatic correlation and integration. On the other hand, traditional carbon emission management methods lack intelligent analysis and management functions, making it difficult to monitor the carbon reduction effect in real time, which is not conducive to the park taking targeted emission reduction measures.

[0003] The patent publication number CN117557220A discloses a carbon emission analysis and management system, management method, and computer-readable storage medium based on dual-carbon multi-source data. This patent includes: a data chart analysis module, a data upload module, a carbon emission flow analysis module, a knowledge visualization module, and a carbon emission situation awareness module. The data chart analysis module supports dynamic data, enabling automatic chart updates on the front end when data is changed in the background. The data upload module includes two sub-modules: one for uploading text data and the other for uploading digital data. The carbon emission flow analysis module provides visualization analysis of carbon emission flows. The knowledge visualization module is used for 3D graph visualization, supporting the visualization of knowledge graphs in three-dimensional space. The carbon emission situation awareness module provides daily carbon emission data for an entire year. While this patent possesses certain data processing and analysis capabilities, it still falls short in terms of the comprehensiveness of data integration and the depth of intelligent analysis. Summary of the Invention

[0004] The purpose of this invention is to provide an analysis and management system based on dual-carbon multi-source data, which solves the problems of data integration difficulties and low level of intelligence in carbon emission management in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an analysis and management system based on dual-carbon multi-source data, comprising... The multi-source data acquisition module collects multi-source data from enterprises, energy facilities, and buildings within the park. This multi-source data includes energy consumption data, production activity data, environmental monitoring data, meteorological data, and enterprise carbon quota data. The data preprocessing module cleans, transforms, and standardizes the collected multi-source data, removing duplicate, erroneous, and missing data, and converting data of different formats into a unified format. The intelligent data analysis module uses big data analysis and artificial intelligence algorithms to deeply mine and analyze pre-processed multi-source data, including carbon emission accounting, carbon emission trend prediction, analysis of carbon emission influencing factors, and assessment of carbon emission reduction potential. The carbon management decision-making module generates carbon management decision-making recommendations based on the analysis results of the intelligent data analysis module and the actual situation of the park. The data visualization module displays multi-source data, intelligent data analysis results, and carbon management decision-making recommendations in intuitive charts, maps, and 3D models, supporting users to perform data interaction queries and visual analysis.

[0006] As a preferred technical solution of the present invention, the multi-source data acquisition module further includes a data acquisition scheduling unit, which arranges the execution time and priority of data acquisition tasks according to the data update frequency and importance.

[0007] As a preferred technical solution of the present invention, the data preprocessing module further includes a data quality assessment unit, which assesses the quality of the preprocessed data and generates a data quality report. The assessment indicators include data integrity, accuracy, consistency and timeliness.

[0008] As a preferred embodiment of the present invention, the intelligent data analysis module is implemented as follows: A carbon emission accounting model is constructed based on a deep learning-based long short-term memory network, using energy consumption data, production activity data, and meteorological data as input features. The model is trained on historical carbon emission data of the park and corresponding input features to learn the nonlinear relationship between the data. During the model training process, the backpropagation algorithm is used to optimize the parameters, with the mean squared error as the loss function, and the model parameters are adjusted. At the same time, the carbon emission factor is used to calibrate the model accounting results. Using the seasonal decomposition algorithm in time series analysis, carbon emission time series data is decomposed into trend, seasonal, and random components. The Prophet algorithm is used to predict the trend component, and combined with historical seasonal data, a comprehensive carbon emission trend prediction result is generated. In the prediction process, external factors such as policy changes and adjustments to enterprise production plans are considered, and the model is corrected by introducing external variables. Principal component analysis was used to reduce the dimensionality of multi-source data and extract the main components affecting carbon emissions. The random forest algorithm was combined to calculate the importance score of each component to carbon emissions and identify key influencing factors. A causal analysis model was constructed to analyze the causal relationship between the influencing factors and the impact path on carbon emissions. Establish a carbon emission reduction potential assessment index system, use the analytic hierarchy process to determine the weight of each index, and combine the fuzzy comprehensive evaluation method to assess the carbon emission reduction potential of each enterprise or facility in the park.

[0009] As a preferred technical solution of the present invention, a carbon emission reduction potential assessment index system is established, including the dimensions of technological emission reduction potential, management emission reduction potential, and structural emission reduction potential.

[0010] As a preferred technical solution of the present invention, the intelligent data analysis module further includes an abnormal data detection unit. This unit monitors carbon emission data in real time by constructing an abnormal detection model. When abnormal data is detected, it automatically issues an early warning message and performs a preliminary analysis of the cause of the abnormality.

[0011] As a preferred embodiment of the present invention, the carbon management decision-making module is implemented as follows: A multi-objective mixed integer programming model is constructed, with the objective functions being to maximize carbon emission reduction, minimize economic cost, and meet the "3060" requirement within the timeframe. The carbon emission calculation results, carbon emission trend prediction data, enterprise production plans, and actual energy consumption provided by the intelligent data analysis module are used as constraints. A genetic algorithm is employed to solve the model, using selection, crossover, and mutation operations to search for the optimal path planning scheme and determine the emission reduction targets, emission reduction measures, and implementation time nodes for different stages. Based on conditional generative adversarial networks (GANs) in generative adversarial networks, specific carbon management decision recommendations are generated using intelligent data analysis results and planning schemes as conditions. The rationality of the generated decision recommendations is verified and optimized by combining an expert knowledge base.

[0012] As a preferred technical solution of the present invention, the carbon management decision module further includes a decision scheme evaluation unit, which simulates and evaluates the generated carbon management decision suggestions and carbon peaking and carbon neutrality path planning schemes, and predicts the emission reduction effect and economic benefits after the schemes are implemented.

[0013] As a preferred technical solution of the present invention, it also includes a data storage module, which stores the collected multi-source data, preprocessed data, intelligent data analysis results, and carbon management decision data; and adopts distributed storage technology to store the data on multiple storage nodes.

[0014] As a preferred technical solution of the present invention, the data storage module further includes a data backup and recovery unit, which backs up the stored data and recovers the data when the data is damaged or lost.

[0015] Compared with the prior art, the beneficial effects of the present invention are: It breaks through the limitations of traditional carbon emission management systems, which rely on a single data source and have difficulty in data integration. It can comprehensively integrate multi-source data from various levels of enterprises, energy facilities, and buildings within the park, including energy consumption data, production activity data, and environmental monitoring data, providing rich data support for accurately calculating the park's carbon emissions. The system is designed closely around the "3060" goal. It can not only realize basic functions such as carbon inventory, carbon analysis and carbon management in the park, but also plan carbon peaking and carbon neutrality paths according to the actual situation of the park, and monitor the carbon reduction effect in real time. It provides comprehensive technical support for the park to implement the "3060" goal, and is forward-looking and innovative. By utilizing data analysis and processing technologies, such as big data analytics and artificial intelligence algorithms, we can deeply mine and analyze carbon emission data in the park, providing the park with scientific and reasonable carbon management decision-making suggestions. Compared with traditional carbon emission management methods, it has a higher level of intelligence and scientific decision-making, and can effectively improve the efficiency and effectiveness of carbon emission management in the park. The system adopts a modular design, with each module having a clearly defined function, which facilitates system maintenance and expansion. Attached Figure Description

[0016] Figure 1 This is a block diagram illustrating the principle of the analysis and management system of the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] Example 1 Please see Figure 1 This is the first embodiment of the present invention, which provides an analysis and management system based on dual-carbon multi-source data, including... The multi-source data acquisition module collects multi-source data from enterprises, energy facilities, and buildings within the park. This multi-source data includes energy consumption data, production activity data, environmental monitoring data, meteorological data, and enterprise carbon quota data. The implementation method of the multi-source data acquisition module is as follows: The system establishes communication with enterprise internal management systems, energy metering devices (smart meters, gas meters, etc.), environmental monitoring instruments (CO2 concentration detectors, PM2.5 monitors, etc.), meteorological monitoring stations, and carbon trading platforms using industrial IoT protocols. For devices that support the MQTT protocol, real-time data subscription and push are achieved by setting up an MQTT server. For energy metering devices that comply with the Modbus protocol, a Modbus gateway is used for protocol conversion to connect the device data to the system network. The design employs a tiered data acquisition mechanism. For data with high real-time requirements, such as energy consumption data and production activity data, high-frequency acquisition is carried out at the minute level. For data with relatively slow changes, such as meteorological data and enterprise carbon quota data, hourly or daily acquisition frequencies are used. At the same time, edge computing nodes are introduced to perform preliminary filtering and aggregation of the collected data at the data source, reducing invalid data transmission and reducing network bandwidth pressure. During the data acquisition process, the transmitted data is verified using the CRC algorithm. If a data error is detected, a retransmission mechanism is immediately triggered. For energy metering equipment, manual calibration is performed regularly, and the calibration data is fed back to the system to correct the acquired data and ensure data accuracy. The data preprocessing module cleans, transforms, and standardizes the collected multi-source data, removing duplicate, erroneous, and missing data, and converting data of different formats into a unified format. The implementation method of the data preprocessing module is as follows: Data cleaning: Build a rule engine and pre-set a data cleaning rule base; for example, for energy consumption data, set a threshold range (e.g., the hourly electricity consumption of industrial equipment should not exceed 120% of the maximum theoretical power consumption of the enterprise's equipment), and mark data exceeding the range as erroneous data; for timestamp data, check for duplicates or jumps, and correct abnormal timestamps; use the isolated forest algorithm in machine learning to detect outliers in the data, identify them as erroneous data and process them; for missing data, use a variational autoencoder model based on deep learning to predict and fill missing values ​​based on historical data and relevant feature data; Data conversion: For structured data (such as tabular data in a database), field type conversion (such as converting string-type dates to timestamp format) and encoding conversion (such as converting UTF-8 encoding to the system's unified GBK encoding) are performed according to the system's data storage format requirements; for unstructured data (such as environmental monitoring report documents), natural language processing technology is used to extract information and convert it into a structured data format; Data standardization: The Z-score standardization method is used to normalize numerical data, mapping the data to a distribution interval with a mean of 0 and a standard deviation of 1; for categorical data (such as enterprise industry type), one-hot encoding is used for standardization, converting each category into a binary vector representation. The intelligent data analysis module uses big data analysis and artificial intelligence algorithms to deeply mine and analyze pre-processed multi-source data, including carbon emission accounting, carbon emission trend prediction, analysis of carbon emission influencing factors, and assessment of carbon emission reduction potential. The carbon management decision-making module, based on the analysis results of the intelligent data analysis module and combined with the actual situation of the park, formulates carbon peaking and carbon neutrality path planning and generates carbon management decision-making recommendations. The data visualization module displays multi-source data, intelligent data analysis results, and carbon management decision-making recommendations in intuitive charts, maps, and 3D models, supporting users to perform data interaction queries and visual analysis.

[0019] In this embodiment, preferably, the multi-source data acquisition module further includes a data acquisition scheduling unit. The data acquisition scheduling unit is used to reasonably arrange the execution time and priority of data acquisition tasks according to the data update frequency and importance, so as to ensure that key data can be acquired in a timely and accurate manner.

[0020] In this embodiment, preferably, the data preprocessing module further includes a data quality assessment unit, which is used to assess the quality of the preprocessed data and generate a data quality report. The assessment indicators include data integrity, accuracy, consistency and timeliness.

[0021] In this embodiment, the preferred method for implementing the intelligent data analysis module is as follows: A carbon emission calculation model is constructed based on a deep learning-based Long Short-Term Memory (LSTM) network, using energy consumption data, production activity data, and meteorological data as input features. The model is trained on historical carbon emission data of the industrial park and corresponding input features to learn the nonlinear relationships between the data. During model training, the backpropagation (BP) algorithm is used for parameter optimization, with mean squared error (MSE) as the loss function, continuously adjusting the model parameters to improve the accuracy of carbon emission calculation. Simultaneously, the model calculation results are calibrated by incorporating carbon emission factors. Using seasonal decomposition algorithms (such as STL decomposition) in time series analysis, carbon emission time series data is decomposed into trend, seasonal, and random terms. Then, the Prophet algorithm is used to predict the trend term, and combined with historical seasonal data, a comprehensive carbon emission trend prediction result for the future period is generated. In the prediction process, external factors such as policy changes and adjustments to enterprise production plans are considered, and the reliability of the prediction is improved by introducing external variables to correct the model. Principal component analysis (PCA) was used to reduce the dimensionality of multi-source data and extract the main components affecting carbon emissions. The random forest algorithm was combined to calculate the importance score of each component to carbon emissions and identify key influencing factors (such as the operating time of high-energy-consuming equipment and the proportion of energy structure). Causal analysis models (such as structural causal model SCM) were constructed to analyze the causal relationships between various influencing factors and the impact paths on carbon emissions. Establish a carbon emission reduction potential assessment index system, including dimensions such as technological emission reduction potential, management emission reduction potential, and structural emission reduction potential; use the analytic hierarchy process (AHP) to determine the weight of each index, and combine it with the fuzzy comprehensive evaluation method to assess the carbon emission reduction potential of each enterprise or facility in the park; based on the assessment results, highlight areas or links with greater emission reduction potential to provide a basis for carbon emission reduction decision-making.

[0022] Example 2 Please see Figure 1 This is the second embodiment of the present invention, which is based on the previous embodiment, but differs in that: The intelligent data analysis module also includes an anomaly detection unit. This unit monitors carbon emission data in real time by building an anomaly detection model. When anomaly data is detected, it automatically issues an early warning and conducts a preliminary analysis of the cause of the anomaly.

[0023] The implementation method of the carbon management decision-making module is as follows: A multi-objective mixed integer programming model is constructed, with the objective functions being to maximize carbon emission reduction, minimize economic cost, and meet the "3060" requirement within the timeframe. The carbon emission calculation results, carbon emission trend prediction data, and actual conditions such as enterprise production plans and energy consumption provided by the intelligent data analysis module are used as constraints. A genetic algorithm is employed to solve the model, and through operations such as selection, crossover, and mutation, the optimal carbon peaking and carbon neutrality path planning scheme is searched to determine the emission reduction targets, emission reduction measures, and implementation time nodes at different stages. Based on Conditional Generative Adversarial Networks (cGANs) within Generative Adversarial Networks (GANs), this method generates specific carbon management decision recommendations using intelligent data analysis results and carbon peaking and carbon neutrality path planning schemes as conditions. For example, based on a company's energy structure and carbon emissions, it generates decision recommendations such as equipment upgrades, energy substitution, and production process optimization, and quantitatively analyzes the implementation cost and expected emission reduction effect of each recommendation. Simultaneously, it utilizes an expert knowledge base to verify the rationality of the generated decision recommendations and optimize them accordingly. The generated carbon management decision-making recommendations and path planning schemes are simulated and evaluated using a system dynamics model. Different scenario parameters (such as the intensity of policy changes and fluctuations in market energy prices) are set to simulate the implementation effects of the schemes under different scenarios, including changes in carbon emissions, economic benefits, and social impacts. Based on the simulation results, the schemes are optimized using a multi-objective particle swarm optimization algorithm (MOPSO) to adjust the implementation sequence and intensity of emission reduction measures, thereby improving the feasibility and effectiveness of the schemes.

[0024] It also includes a data storage module, which stores collected multi-source data, preprocessed data, intelligent data analysis results, and carbon management decision-making data. The data storage module employs distributed storage technology, storing data across multiple storage nodes to improve data storage reliability and scalability. The data storage module also includes a data backup and recovery unit, which periodically backs up the stored data and can quickly recover data in case of corruption or loss, ensuring data security. The data storage module is implemented as follows: The underlying storage architecture is built using the Ceph distributed storage system, leveraging Ceph's object storage, block storage, and file storage capabilities to adapt to the storage needs of different types of data. A storage cluster is formed by deploying multiple Ceph nodes, each equipped with a high-performance CPU, large-capacity memory, and a high-speed disk array. Nodes communicate at high speed via 10 Gigabit Ethernet to ensure efficient data read and write operations. Monitor nodes are set up in the cluster to maintain the cluster's metadata and status information, OSD (Object Storage Device) nodes are responsible for the actual data storage and management, and MDS (Metadata Server) nodes are used to handle metadata operations for file storage, forming a stable and reliable distributed storage environment. For structured data such as energy consumption data and production activity data, Ceph's block storage method is used to store the data in a relational database (such as a MySQL cluster). The database's indexing and query optimization functions are used to achieve fast data retrieval. For unstructured data such as environmental monitoring reports and meteorological data, Ceph's object storage is used to store the data in the form of objects and assign a unique identifier and metadata tag to each object to facilitate data classification, management and retrieval. The preprocessed data is stored in a hierarchical manner based on its subsequent usage frequency and importance; high-frequency access and important data are stored on high-performance solid-state storage devices (SSDs) to speed up data reading; low-frequency access data is stored on large-capacity hard disk drives (HDDs) to reduce storage costs; at the same time, a version management mechanism is established for the preprocessed data to record the data processing process and modification history, which facilitates data backtracking and problem investigation. Storing carbon emission accounting results, carbon emission trend predictions, and other analytical results in a columnar database (such as ClickHouse) leverages the characteristics of columnar storage to improve the efficiency of aggregate queries and data analysis; creating a time-series index for the analytical results data and partitioning the data according to the time dimension facilitates quick querying and analysis of data by time range; Carbon management decision-making recommendations, pathway planning schemes, and other data are stored in a document-oriented database (such as MongoDB) in JSON format to facilitate flexible data expansion and modification. Access control is set for decision-related data, and different read and write permissions are assigned according to user roles and permissions to ensure data security. A combination of full and incremental backups is used; a full backup is performed weekly to completely back up all data to an off-site backup server; incremental backups are performed daily, backing up only the data that changes on that day to reduce backup time and storage space usage; a backup time window is set to perform backup tasks during off-peak hours to avoid affecting normal system operation; at the same time, backup data is encrypted using the AES-256 encryption algorithm to ensure the security of backup data during storage and transmission; When data corruption or loss is detected, the data recovery unit is automatically activated. First, the recovery priority is determined based on the importance and frequency of use of the data. For critical business data, recovery is prioritized from the most recent backup. For non-critical data, recovery can be performed when the system is idle. During the recovery process, Ceph's distributed storage features are utilized to read backup data in parallel, accelerating the data recovery speed. After the recovery is completed, the integrity and consistency of the recovered data are verified to ensure that the data recovery is successful and usable. Establish a backup data lifecycle management mechanism to regularly clean up expired backup data and free up storage space; set backup data retention periods, such as retaining full backup data for 6 months and incremental backup data for 1 month; and conduct regular readability tests on backup data to verify its validity and prevent situations where backup data cannot be recovered.

[0025] The data visualization module is implemented as follows: The ECharts charting library is used to create two-dimensional charts (such as line charts, bar charts, and pie charts). For displaying the trend of carbon emissions over time, a line chart is used, with the time axis as the horizontal axis and carbon emissions as the vertical axis. Different colors and line styles are used to distinguish carbon emission data from different companies or regions. For displaying the proportion of energy structure, a pie chart is used to visually present the proportion of various energy sources in total energy consumption. The D3.js library is used to implement dynamic chart effects, such as automatically updating the chart to display data changes within a specific time period when the user selects it. Based on Baidu Maps API or Gaode Maps API, the park's geographical information is combined with carbon emission data; the locations of enterprises, energy facilities, etc. are marked on the map, and different marker colors and sizes are set according to their carbon emission levels to achieve spatial visualization of carbon emission data; a heat map is used to display the carbon emission density distribution in different areas of the park, and users can view detailed carbon emission information for each area by zooming and dragging the map. The Three.js 3D graphics library is used to build a 3D model of the park, creating 3D models of buildings, energy facilities, etc. Carbon emission data is mapped onto the 3D model, and the carbon emission intensity of different areas or facilities is displayed through color gradients, lighting effects, etc. Users can rotate, translate, and zoom in the 3D scene using a mouse or keyboard to view the carbon emission distribution of the park from different angles. At the same time, WebGL technology is used to achieve high-performance rendering of the 3D model, ensuring smooth display on the browser. Add interactive event listeners to charts, maps, and 3D models. When a user clicks on a data point in a chart, a marker on a map, or an object in a 3D model, a detailed information window pops up, displaying multi-source data (such as energy consumption data, production activity data, etc.), intelligent data analysis results (such as carbon emission calculation values, carbon emission trend prediction values, etc.), and related carbon management decision-making suggestions for that object. Support users to filter and query data through input boxes, such as by company name, time range, carbon emission index, etc. The system dynamically updates the visualization display content based on user input.

[0026] It also includes a system management module, which manages user permissions, roles, logs, and system configuration to ensure the secure and stable operation of the system. The system management module adopts a role-based access control mechanism to assign different operation permissions according to the user's role. It also includes an interface module for external systems, which is used to enable data interaction and functional integration with other relevant systems, including government carbon emission monitoring platforms, enterprise resource planning systems, and energy management systems.

[0027] Although embodiments of the invention have been shown and described in detail above, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An analysis and management system based on dual-carbon multi-source data, characterized in that: include The multi-source data acquisition module collects multi-source data from enterprises, energy facilities, and buildings within the park. This multi-source data includes energy consumption data, production activity data, environmental monitoring data, meteorological data, and enterprise carbon quota data. The data preprocessing module cleans, transforms, and standardizes the collected multi-source data, removing duplicate, erroneous, and missing data, and converting data of different formats into a unified format. The intelligent data analysis module uses big data analysis and artificial intelligence algorithms to deeply mine and analyze pre-processed multi-source data, including carbon emission accounting, carbon emission trend prediction, analysis of carbon emission influencing factors, and assessment of carbon emission reduction potential. The carbon management decision-making module, based on the analysis results of the intelligent data analysis module and combined with the actual situation of the park, formulates carbon peaking and carbon neutrality path planning and generates carbon management decision-making recommendations. The data visualization module displays multi-source data, intelligent data analysis results, and carbon management decision-making recommendations in intuitive charts, maps, and 3D models, supporting users to perform data interaction queries and visual analysis.

2. The analysis and management system based on dual-carbon multi-source data according to claim 1, characterized in that: The multi-source data acquisition module also includes a data acquisition scheduling unit, which arranges the execution time and priority of data acquisition tasks according to the data update frequency and importance.

3. The analysis and management system based on dual-carbon multi-source data according to claim 1, characterized in that: The data preprocessing module also includes a data quality assessment unit, which assesses the quality of the preprocessed data and generates a data quality report. The assessment indicators include data integrity, accuracy, consistency, and timeliness.

4. The analysis and management system based on dual-carbon multi-source data according to claim 1, characterized in that: The implementation method of the intelligent data analysis module is as follows: A carbon emission accounting model is constructed based on a deep learning-based long short-term memory network, using energy consumption data, production activity data, and meteorological data as input features. The model is trained on historical carbon emission data of the industrial park and corresponding input features to learn the nonlinear relationships between the data. During model training, a backpropagation algorithm is used for parameter optimization, with mean squared error as the loss function to adjust the model parameters. Simultaneously, the model's accounting results are calibrated using carbon emission factors provided by the Intergovernmental Panel on Climate Change (IPCC). Using the seasonal decomposition algorithm in time series analysis, carbon emission time series data is decomposed into trend, seasonal, and random components. The Prophet algorithm is used to predict the trend component, and combined with historical seasonal data, a comprehensive carbon emission trend prediction result is generated. In the prediction process, external factors such as policy changes and adjustments to enterprise production plans are considered, and the model is corrected by introducing external variables. Principal component analysis was used to reduce the dimensionality of multi-source data and extract the main components affecting carbon emissions. The random forest algorithm was combined to calculate the importance score of each component to carbon emissions and identify key influencing factors. A causal analysis model was constructed to analyze the causal relationship between the influencing factors and the impact path on carbon emissions. Establish a carbon emission reduction potential assessment index system, use the analytic hierarchy process to determine the weight of each index, and combine the fuzzy comprehensive evaluation method to assess the carbon emission reduction potential of each enterprise or facility in the park.

5. The analysis and management system based on dual-carbon multi-source data according to claim 4, characterized in that: Establish a carbon emission reduction potential assessment index system, including dimensions of technological emission reduction potential, management emission reduction potential, and structural emission reduction potential.

6. The analysis and management system based on dual-carbon multi-source data according to claim 1, characterized in that: The intelligent data analysis module also includes an abnormal data detection unit. This unit monitors carbon emission data in real time by constructing an abnormal detection model. When abnormal data is detected, it automatically issues an early warning and conducts a preliminary analysis of the cause of the abnormality.

7. The analysis and management system based on dual-carbon multi-source data according to claim 1, characterized in that: The carbon management decision-making module is implemented as follows: A multi-objective mixed integer programming model is constructed, with the objective functions being maximizing carbon emission reduction, minimizing economic costs, and meeting the "3060" strategic requirements within the specified timeframe. The carbon emission calculation results, carbon emission trend prediction data, enterprise production plans, and actual energy consumption provided by the intelligent data analysis module are used as constraints. A genetic algorithm is employed to solve the model, and through selection, crossover, and mutation operations, the optimal carbon peaking and carbon neutrality path planning scheme is searched to determine the emission reduction targets, emission reduction measures, and implementation timelines for different stages. Based on conditional generative adversarial networks (GANs) in generative adversarial networks, specific carbon management decision recommendations are generated using intelligent data analysis results and carbon peaking and carbon neutrality pathway planning schemes as conditions. The rationality of the generated decision recommendations is verified and optimized by combining an expert knowledge base.

8. The analysis and management system based on dual-carbon multi-source data according to claim 1, characterized in that: The carbon management decision-making module also includes a decision-making scheme evaluation unit, which simulates and evaluates the generated carbon management decision-making recommendations and carbon peaking and carbon neutrality path planning schemes, and predicts the emission reduction effects and economic benefits after the schemes are implemented.

9. The analysis and management system based on dual-carbon multi-source data according to claim 1, characterized in that: It also includes a data storage module, which stores the collected multi-source data, preprocessed data, intelligent data analysis results, and carbon management decision data; it uses distributed storage technology to store data on multiple storage nodes.

10. The analysis and management system based on dual-carbon multi-source data according to claim 9, characterized in that: The data storage module also includes a data backup and recovery unit, which backs up the stored data and recovers the data when it is damaged or lost.

Images