Enterprise carbon emission reduction data intelligent monitoring and analysis system

Abstract

The invention relates to the technical field of environmental protection and climate change, in particular to an enterprise carbon emission reduction data intelligent monitoring and analysis system, which comprises a data acquisition and preprocessing module, a mechanism model library module, a carbon emission accounting and monitoring module, an abnormality diagnosis and traceability module, an emission reduction potential analysis module and a visualization and report module. Through multi-source data access, cleaning and standardization processing, high-quality and uniform-format basic data are provided for the system, the accuracy and reliability of subsequent carbon accounting and analysis are guaranteed, accurate positioning and source tracing of carbon emission abnormity are achieved by means of causal reasoning and a path optimization algorithm, and the carbon emission abnormity analysis method has the advantages of being high in practicability and easy to popularize. A scientific basis is provided for an enterprise to quickly solve the problem that emission exceeds the standard, and through reinforcement learning simulation, multi-target optimization and knowledge graph matching, an optimal strategy considering the emission reduction effect, cost and productivity is dynamically recommended, and the enterprise is assisted to achieve collaborative improvement of carbon emission reduction and economic benefits.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of environmental protection and climate change, in particular to an enterprise carbon emission reduction data intelligent monitoring and analysis system. BACKGROUND

[0002] Enterprise carbon emission reduction refers to the process and action of reducing the emission of carbon dioxide and other greenhouse gases (such as methane, nitrous oxide, etc.) generated in production and operation activities through a series of active measures such as technology, management and operation, but its effective implementation is highly dependent on the accurate monitoring of carbon emission data, the rapid positioning of abnormal problems and the scientific decision-making of emission reduction schemes.

[0003] In the prior art, enterprise carbon emission reduction management still faces many bottlenecks: on the one hand, the production process of the enterprise involves multi-source heterogeneous data, the data format is not unified, and there is much noise interference, which leads to low quality of basic data and directly affects the accuracy of carbon emission accounting; on the other hand, when the carbon emission appears abnormal fluctuation, the traditional method is difficult to distinguish the cause and effect relationship in the data correlation and irrelevant interference, and cannot quickly trace the abnormal source, delaying the adjustment opportunity of emission reduction; in addition, in the development of emission reduction scheme, the existing technology mainly focuses on a single target, and it is difficult to balance the relationship between emission reduction effect, cost investment and capacity loss, resulting in the lack of economy and feasibility of the scheme.

[0004] Based on this, the present application provides an enterprise carbon emission reduction data intelligent monitoring and analysis system to solve the above technical problems. SUMMARY

[0005] The purpose of the present application is to provide an enterprise carbon emission reduction data intelligent monitoring and analysis system, which provides high-quality, unified format basic data for the system through multi-source data access, cleaning and standardization processing, guarantees the accuracy and reliability of subsequent carbon accounting and analysis, and realizes the accurate positioning and root tracing of carbon emission abnormality by means of causal reasoning and path optimization algorithm, provides scientific basis for the enterprise to quickly solve the problem of emission exceeding standard, dynamically recommends the optimal strategy considering emission reduction effect, cost and capacity through reinforcement learning simulation, multi-objective optimization and knowledge graph matching, and helps the enterprise to realize the coordinated improvement of carbon emission reduction and economic benefit.

[0006] To achieve the above purpose, the present application provides the following technical scheme:

[0007] The present application provides an enterprise carbon emission reduction data intelligent monitoring and analysis system, which comprises a data acquisition and preprocessing module, a mechanism model library module, a carbon emission accounting and monitoring module, an abnormal diagnosis and tracing module, an emission reduction potential analysis module and a visualization and report module, wherein:

[0008] The data acquisition and preprocessing module is used for collecting raw data of enterprise production energy consumption and emission sources in real time through Internet of Things devices, and cleaning, standardizing and removing outliers of the raw data.

[0009] The mechanism model library module is used for integrating modular and customizable industry carbon metabolism mechanism models, and providing a high-precision physical / chemical calculation framework for carbon emission accounting through dynamic parameter optimization driven by digital twinning.

[0010] The carbon emission accounting and monitoring module is used for calling the mechanism model to dynamically calculate the preprocessed data, and generating indexes of carbon emissions and emission intensity of each link of the enterprise in real time, and displaying the monitoring results through a visual interface.

[0011] The anomaly diagnosis and tracing module is based on causal reasoning and path optimization algorithm, and compares the actual data with the mechanism model threshold to accurately diagnose and trace the carbon emission anomaly.

[0012] The emission reduction potential analysis module is based on reinforcement learning and multi-objective optimization technology to construct an intelligent analysis model, and dynamically recommends the optimal emission reduction path and strategy considering the emission reduction effect, cost and production capacity influence in combination with the emission reduction scheme knowledge graph.

[0013] The visualization and reporting module is used for converting the accounting results, anomaly diagnosis, potential analysis and other data into charts and dynamic dashboards, and automatically generating carbon emission reports in accordance with national standards to support the internal management and external disclosure needs of the enterprise.

[0014] The data acquisition and preprocessing module includes a multi-source data access unit, a data cleaning unit and a data standardization unit.

[0015] The multi-source data access unit is used for accessing multi-type raw data of energy consumption, emission sources and production processes through Internet of Things sensors, industrial buses and third-party system APIs.

[0016] The data cleaning unit is used for removing noise data, filling missing values and correcting abnormal readings caused by equipment failure by using statistical analysis and rule engine.

[0017] The data standardization unit is used for converting data in different formats and units into a unified format that meets the input requirements of the mechanism model.

[0018] The mechanism model library module includes an industry model adaptation unit, a digital twinning parameter optimization unit and a model management unit.

[0019] The industry model adaptation unit is used for storing the basic mechanism model of carbon metabolism of the chemical and manufacturing industries, and supports quick calling according to enterprise type.

[0020] The digital twin parameter optimization unit is configured to dynamically calibrate key parameters in the mechanism model through production scene digital twin simulation.

[0021] The model management unit is configured to implement model version iteration, permission control and call log recording, and support user-defined model parameter configuration.

[0022] The digital twin parameter optimization unit dynamically calibrates key parameters in the mechanism model through production scene digital twin simulation, and the specific operation is as follows:

[0023] A1: Data mapping: used for dynamically synchronizing real-time production data with digital twin virtual scene;

[0024] A2: Parameter sensitivity analysis: based on the variance decomposition method to determine the key parameters in the mechanism model that need to be optimized first;

[0025] A3: Dynamic calibration: minimize the deviation between digital twin output and actual monitoring data through gradient descent algorithm, and iteratively update model parameters;

[0026] A4: Verification feedback: use historical data to verify the error rate of the calibrated model, and feed back the optimization results to the model management unit.

[0027] The carbon emission accounting and monitoring module includes a dynamic accounting unit, a real-time monitoring unit, and a visualization display unit, wherein:

[0028] The dynamic accounting unit is configured to call the mechanism model to perform real-time calculation on preprocessed data, and generate instantaneous / cumulative carbon emissions and emission intensity of each process and device;

[0029] The real-time monitoring unit is configured to compare the carbon emission indicators with the threshold to determine whether the indicators exceed the threshold and trigger an over-limit warning;

[0030] The visualization display unit is configured to convert the accounting results into real-time dashboards and trend curves, and intuitively present the overall and each link of the enterprise's carbon emission dynamics.

[0031] The abnormal diagnosis and tracing module includes a causal relationship analysis unit, an abnormal path searching unit, and a root cause tracing unit, wherein:

[0032] The causal relationship analysis unit is configured to distinguish the causal relationship and irrelevant association in the carbon emission data based on a causal diagram network model through Bayesian inference;

[0033] The abnormal path searching unit is configured to use a path optimization algorithm combined with a production process flowchart to quickly locate the propagation path of abnormal emissions;

[0034] The root cause tracing unit is used to combine the abnormal path with the mechanism model threshold, accurately identify the cause of the exceeding standard, and generate a traceability report.

[0035] The cause-effect analysis unit distinguishes the cause-effect relationship and irrelevant association in the carbon emission data based on the cause-effect diagram network model through Bayesian inference, and the specific operation is as follows:

[0036] B1: Construct a cause-effect diagram network model of carbon emission related variables to represent the cause-effect dependent relationship between variables;

[0037] B2: Train and parameter learn the Bayesian network based on historical and real-time carbon emission data;

[0038] B3: Calculate the conditional probability distribution between variables by using Bayesian inference algorithm;

[0039] B4: Identify and distinguish the cause-effect relationship and related interference in the abnormal fluctuation of carbon emission;

[0040] B5: Output the cause-effect relationship diagram to provide basis for subsequent abnormal traceability.

[0041] The emission reduction potential analysis module includes a scheme simulation unit, a multi-objective optimization unit, and a knowledge graph matching unit, wherein:

[0042] The scheme simulation unit simulates the carbon emission change effect of different emission reduction schemes based on a reinforcement learning model;

[0043] The multi-objective optimization unit is used to balance the emission reduction amount, cost input and capacity loss by genetic algorithm to screen the optimal scheme combination;

[0044] The knowledge graph matching unit is used to associate the emission reduction technology database and recommend the emission reduction path suitable for the enterprise scene.

[0045] The multi-objective optimization unit balances the emission reduction amount, cost input and capacity loss by genetic algorithm to screen the optimal scheme combination, and the specific operation is as follows:

[0046] C1: Construct a multi-objective optimization model:

[0047] Define the objective function as:

[0048]

[0049] Wherein: = carbon emission reduction amount; = total cost input; = capacity loss rate;

[0050] Set the constraint condition:

[0051]

[0052]

[0053] C2: Encoding and initializing population:

[0054] Encode the emission reduction scheme parameters such as equipment replacement ratio, process improvement degree into chromosomes, and randomly generate an initial population;

[0055] C3: Design fitness function:

[0056] Calculate individual fitness by Pareto optimal sorting, and the fitness function is based on the following formula:

[0057]

[0058] Wherein: 、 、 is the weight coefficient, is the emission reduction amount target function value corresponding to the ith individual; is the cost input target function value corresponding to the ith individual; is the capacity loss target function value corresponding to the ith individual; is the maximum value of the emission reduction amount target function; is the maximum value of the cost input target function; is the maximum value of the capacity loss target function;

[0059] C4: Iterative optimization of genetic operation:

[0060] Selection: Adopt tournament selection method to reserve elite individuals;

[0061] Crossover: Perform arithmetic crossover operation on selected individuals;

[0062] Mutation: Perform Gaussian mutation on specific gene sites of chromosomes;

[0063] Iterate for N generations until the convergence condition is met;

[0064] C5: Output Pareto optimal solution set:

[0065] Select non-dominated solutions to form the optimal scheme set, and each solution corresponds to a different weight-based emission reduction-cost-capacity balance scheme.

[0066] The visualization and reporting module includes a multi-dimensional visualization unit, a report generation unit, and a permission management unit, wherein:

[0067] The multi-dimensional visualization unit is used to convert accounting, diagnosis, and analysis data into three-dimensional heat maps and dynamic Sankey diagrams to visually display carbon emission distribution and emission reduction potential.

[0068] The report generation unit is used for automatically extracting key data, generating a standardized report meeting the standard of the Greenhouse Gas Accounting and Reporting Requirements, and supporting PDF and Excel format export;

[0069] The permission management unit is used for displaying a differentiated data view according to a user role, and meeting the permission isolation requirement of internal management and external disclosure.

[0070] Compared with the prior art, the present application has the following beneficial effects:

[0071] The present application provides high-quality and unified format basic data for the system through multi-source data access, cleaning and standardization processing, guarantees the accuracy and reliability of subsequent carbon accounting and analysis, and realizes the accurate positioning and root cause tracing of carbon emission abnormalities by means of causal reasoning and path optimization algorithm, provides a scientific basis for enterprises to quickly solve the problem of emission exceeding the standard, dynamically recommends the optimal strategy considering emission reduction effect, cost and production capacity through reinforcement learning simulation, multi-objective optimization and knowledge graph matching, and helps enterprises to realize the coordinated improvement of carbon emission reduction and economic benefits. BRIEF DESCRIPTION OF DRAWINGS

[0072] Fig. 1 It is a system diagram of the enterprise carbon emission reduction data intelligent monitoring and analysis system of the present application.

[0073] Fig. 2 It is a system architecture diagram of the enterprise carbon emission reduction data intelligent monitoring and analysis system of the present application.

[0074] Fig. 3 It is a digital twin parameter optimization flowchart of the enterprise carbon emission reduction data intelligent monitoring and analysis system of the present application.

[0075] Explanation of reference numerals:

[0076] 1, data acquisition and preprocessing module; 11, multi-source data access unit; 12, data cleaning unit; 13, data standardization unit; 2, mechanism model library module; 21, industry model adaptation unit; 22, digital twin parameter optimization unit; 23, model management unit; 3, carbon emission accounting and monitoring module; 31, dynamic accounting unit; 32, real-time monitoring unit; 33, visual display unit; 4, abnormal diagnosis and tracing module; 41, causal relationship analysis unit; 42, abnormal path searching unit; 43, root cause tracing unit; 5, emission reduction potential analysis module; 51, scheme simulation unit; 52, multi-objective optimization unit; 53, knowledge graph matching unit; 6, visualization and report module; 61, multi-dimensional visualization unit; 62, report generation unit; 63, permission management unit. DETAILED DESCRIPTION

[0077] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0078] Example:

[0079] like Figs. 1-3 As shown, this embodiment provides an intelligent monitoring and analysis system for enterprise carbon emission reduction data, including a data acquisition and preprocessing module 1, a mechanism model library module 2, a carbon emission accounting and monitoring module 3, an anomaly diagnosis and source tracing module 4, an emission reduction potential analysis module 5, and a visualization and reporting module 6. Specifically: the data acquisition and preprocessing module 1 is used to collect raw data on enterprise production energy consumption and emission sources in real time through IoT devices, and to clean, standardize, and remove outliers from the raw data; the mechanism model library module 2 is used to integrate modular and customizable industry carbon metabolism mechanism models, and to provide a high-precision physical / chemical calculation framework for carbon emission accounting through dynamic parameter optimization driven by digital twins; the carbon emission accounting and monitoring module 3 is used to call the mechanism model to preprocess... The system dynamically calculates subsequent data to generate real-time indicators of carbon emissions and emission intensity for each stage of the enterprise, and displays the monitoring results through a visual interface; Module 4: Based on causal reasoning and path optimization algorithms, it compares actual data with thresholds from mechanistic models to accurately diagnose and trace the source of carbon emission anomalies; Module 5: Based on reinforcement learning and multi-objective optimization technology, it constructs an intelligent analysis model and combines it with a knowledge graph of emission reduction schemes to dynamically recommend the optimal emission reduction path and strategy that takes into account the emission reduction effect, cost, and capacity impact; Module 6: It is used to transform data such as accounting results, anomaly diagnosis, and potential analysis into charts and dynamic dashboards, and automatically generates carbon emission reports that meet national standards, supporting the enterprise's internal management and external disclosure needs.

[0080] It should be noted that the data acquisition and preprocessing module 1 collects and processes raw data in real time, the mechanism model library module 2 provides a dynamically optimized calculation framework, the carbon emission accounting and monitoring module 3 generates real-time indicators based on the model, the anomaly diagnosis and source tracing module 4 locates problems through causal analysis, the emission reduction potential analysis module 5 recommends the optimal strategy, and finally the visualization and reporting module 6 outputs multi-dimensional analysis results, realizing intelligent carbon management across the entire chain from data acquisition to decision support.

[0081] In the present embodiment, it is also necessary to point out that the data acquisition and preprocessing module 1 comprises a multi-source data access unit 11, a data cleaning unit 12, a data standardization unit 13, wherein: the multi-source data access unit 11 is used to access multi-type raw data of energy consumption, emission sources and production processes through Internet of Things sensors, industrial buses and third-party system APIs; the data cleaning unit 12 is used to remove noise data, fill in missing values and correct abnormal readings caused by equipment failure by using statistical analysis and rule engine; the data standardization unit 13 is used to convert data of different formats and units into a unified format meeting the input requirements of the mechanism model.

[0082] It should be noted that the multi-source data access unit 11 collects multi-type raw data, which is processed by the data cleaning unit 12 to remove noise and correct abnormalities, and then converted into a unified format and unit by the data standardization unit 13.

[0083] Further, it should be noted that the types of Internet of Things sensors include temperature and humidity sensors, gas sensors and pressure sensors; the types of industrial buses include Profinet and EtherNet / IP; the types of third-party system APIs include enterprise resource planning (ERP) systems and manufacturing execution systems (MES); the statistical analysis method is Z-score standardization, which can convert data into a standard normal distribution and identify outliers by setting a threshold (e.g., Z-score greater than 3 or less than -3); the working principle of the rule engine and the definition of the rules are based on business logic and expert experience to write rules for identifying and correcting abnormal readings caused by equipment failure. For example, when the energy consumption data of a certain device suddenly increases significantly, the rule engine can determine whether it is a device failure according to the pre-set rules and make the corresponding correction.

[0084] In this embodiment, it also needs to be explained that the mechanism model library module 2 comprises an industry model adaptation unit 21, a digital twin parameter optimization unit 22, and a model management unit 23, wherein: the industry model adaptation unit 21 is used for storing the carbon metabolism basic mechanism model of the chemical industry and the manufacturing industry, and supports quick calling according to the enterprise type; the digital twin parameter optimization unit 22 is used for dynamically calibrating the key parameters in the mechanism model through production scene digital twin simulation; the specific operation is as follows: A1: data mapping, which is used for dynamically synchronizing real-time production data with a digital twin virtual scene; A2: parameter sensitivity analysis, which is used for determining the key parameters that need to be optimized in the mechanism model based on the variance decomposition method; A3: dynamic calibration, which is used for minimizing the deviation between the digital twin output and the actual monitoring data through the gradient descent algorithm, and iteratively updating the model parameters; A4: verification feedback, which is used for verifying the error rate of the calibrated model by using historical data, and feeding back the optimization results to the model management unit 23. The model management unit 23 is used for realizing model version iteration, permission control and calling log recording, and supports user-defined model parameter configuration.

[0085] It needs to be explained that the industry model adaptation unit 21 provides a basic model library, the digital twin parameter optimization unit 22 completes the closed-loop parameter optimization of “data mapping-parameter analysis-dynamic calibration-verification feedback”, and finally realizes the whole life cycle management through the model management unit 23.

[0086] Further, it needs to be explained that the update formula of the gradient descent algorithm is: wherein, is the model parameter; is the learning rate; is the loss function; is the gradient of the loss function at .

[0087] In this embodiment, it also needs to be explained that the carbon emission accounting and monitoring module 3 comprises a dynamic accounting unit 31, a real-time monitoring unit 32, and a visual display unit 33, wherein: the dynamic accounting unit 31 is used for calling the mechanism model to perform real-time calculation on the pretreated data, and generates the instantaneous / cumulative carbon emission and emission intensity of each process and device; the real-time monitoring unit 32 is used for monitoring whether the carbon emission index exceeds the threshold value through threshold comparison, and triggering an over-limit early warning; the visual display unit 33 is used for converting the accounting results into real-time dashboards and trend curves, and intuitively presenting the carbon emission dynamics of the whole enterprise and each link.

[0088] It needs to be explained that the dynamic accounting unit 31 performs real-time calculation on the carbon emission data, the threshold comparison and early warning judgment are performed by the real-time monitoring unit 32, and finally the multi-dimensional data presentation is realized by the visual display unit 33.

[0089] Further, it needs to be explained that the formula for calculating carbon emissions is: Wherein, A is the activity data, and EF is the emission factor.

[0090] In this embodiment, it also needs to be explained that the abnormal diagnosis and tracing module 4 includes a causal relationship analysis unit 41, an abnormal path searching unit 42, and a root cause tracing unit 43, wherein: the causal relationship analysis unit 41: based on the causal graph network model, the causal relationship and irrelevant association in the carbon emission data are distinguished by Bayesian inference; the specific operation is as follows: B1: build a causal graph network model of carbon emission related variables, representing the causal dependence relationship between variables; B2: train and parameter learning on Bayesian network based on historical and real-time carbon emission data; B3: calculate the conditional probability distribution between variables using Bayesian inference algorithm; B4: identify and distinguish the causal relationship and related interference in the abnormal fluctuation of carbon emissions; B5: output the causal relationship atlas, which provides the basis for subsequent abnormal tracing. The abnormal path searching unit 42: used to adopt path optimization algorithm combined with production process flow chart to quickly locate the propagation path of abnormal emissions; the root cause tracing unit 43: used to combine abnormal path and mechanism model threshold to accurately identify the over-standard reason and generate a tracing report.

[0091] It needs to be explained that the causal relationship analysis unit 41 establishes the carbon emission causal network and identifies the key association, quickly locates the abnormal propagation path through the abnormal path searching unit 42, and finally accurately locks the over-standard reason and generates a report through the root cause tracing unit 43.

[0092] Further, it needs to be explained that the formula of Bayes theorem is: Wherein, P(A|B) is the posterior probability, P(B|A) is the likelihood function, P(A) is the prior probability, and P(B) is the evidence. The cost function of path optimization algorithm: .

[0093] In this embodiment, it also needs to be explained that the emission reduction potential analysis module 5 includes a scheme simulation unit 51, a multi-objective optimization unit 52, and a knowledge graph matching unit 53, wherein: the scheme simulation unit 51: simulates the carbon emission change effect of different emission reduction schemes based on the reinforcement learning model; the multi-objective optimization unit 52: used to balance the emission reduction amount, cost input and capacity loss through genetic algorithm to select the optimal scheme combination; the specific operation is as follows: C1: construct a multi-objective optimization model: define the objective function as:

[0094]

[0095] Wherein: = carbon emission reduction amount; = total cost input; = capacity loss rate; set the constraint condition:

[0096]

[0097]

[0098] C2: Encoding and initializing population: encode the abatement scheme parameters such as equipment replacement ratio, process improvement degree into chromosomes, and randomly generate the initial population; C3: Designing fitness function: calculate individual fitness by Pareto optimal sorting, and the fitness function is based on the following formula:

[0099]

[0100] wherein: 、 、 is the weight coefficient, is the target function value of the abatement amount corresponding to the i-th individual; is the target function value of the cost input corresponding to the i-th individual; is the target function value of the capacity loss corresponding to the i-th individual; is the maximum value of the abatement amount target function; is the maximum value of the cost input target function; is the maximum value of the capacity loss target function; C4: Iterative optimization of genetic operation: selection: adopt tournament selection method to reserve elite individuals; crossover: perform arithmetic crossover operation on selected individuals; mutation: perform Gaussian mutation on specific gene sites of the chromosome; iterate for N generations until the convergence condition is met; C5: Output Pareto optimal solution set: select non-dominated solutions to form the optimal scheme set, and each solution corresponds to a different weight-based abatement-cost-capacity balance scheme. Knowledge graph matching unit 53: used for associating the abatement technology database and recommending the abatement path adapted to the enterprise scene.

[0101] It should be noted that the scheme simulation unit 51 preplays the abatement scenario, the multi-objective optimization unit 52 solves the "abatement-cost-capacity" Pareto optimal solution set by using the improved genetic algorithm, and finally the knowledge graph matching unit 53 realizes intelligent matching of the technical scheme.

[0102] Further, it should be noted that the training data sources of the reinforcement learning model (such as historical abatement cases, simulation data), the evaluation indicators of the model (such as accuracy, recall rate), and the visualization methods of the simulation results (such as heat map, scatter plot). The parameter settings of the genetic algorithm (such as population size, crossover rate, mutation rate), the screening method of the Pareto optimal solution set (such as ε-dominance), and the decision support tool of the optimal scheme (such as analytic hierarchy process).

[0103] In the present embodiment, it is also necessary to point out that the visualization and reporting module 6 comprises a multidimensional visualization unit 61, a report generation unit 62, a permission management unit 63, wherein: the multidimensional visualization unit 61 is used to convert accounting, diagnosis, and analysis data into three-dimensional heat maps and dynamic Sankey diagrams, and intuitively display carbon emission distribution and emission reduction potential; the report generation unit 62 is used to automatically extract key data, generate standardized reports in accordance with the Greenhouse Gas Emission Accounting and Reporting Requirements, and support PDF and Excel format export; and the permission management unit 63 is used to display differentiated data views according to user roles, and meet the permission isolation needs of internal management and external disclosure.

[0104] It is necessary to point out that the multidimensional visualization unit 61 realizes interactive graphical display of carbon emission data, automatically generates standardized compliance reports via the report generation unit 62, and implements hierarchical data control by the permission management unit 63.

[0105] Further, it is necessary to point out that the rendering engine of the three-dimensional heat map (such as Three.js), the drawing tool of the dynamic Sankey diagram (such as D3.js), and the interactive function of the visualization result (such as zooming, panning, and filtering), the definition of user roles (such as administrator, analyst, and decision maker), the configuration method of data views (such as SQL view and API permission control), and the audit log of permissions.

[0106] In the description of the present specification, the description of the terms "one embodiment", "example", "specific example", and the like means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are contained in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0107] The preferred embodiments of the application disclosed above are only used to help explain the application. The preferred embodiments do not describe all the details, nor limit the application to the specific embodiments described. Obviously, many modifications and variations can be made according to the content of the present specification. The present specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the application, so that those skilled in the art can well understand and utilize the application. The application is limited only by the claims and their full scope and equivalents.

Claims

1. An enterprise carbon emission reduction data intelligent monitoring and analysis system, characterized in that, The system comprises a data acquisition and preprocessing module (1), a mechanism model library module (2), a carbon emission accounting and monitoring module (3), an anomaly diagnosis and tracing module (4), an emission reduction potential analysis module (5), and a visualization and reporting module (6), wherein: The data acquisition and preprocessing module (1) is configured to acquire raw data of enterprise production energy consumption and emission sources in real time through Internet of Things devices, and clean, standardize and remove outliers of the raw data; The mechanism model library module (2) is configured to integrate modular and customizable industry carbon metabolism mechanism models, and provide a high-precision physical / chemical calculation framework for carbon emission accounting through dynamic parameter optimization driven by digital twinning; The carbon emission accounting and monitoring module (3) is configured to call the mechanism model to perform dynamic calculation on the preprocessed data, and generate indexes of carbon emissions and emission intensity of each link of the enterprise in real time, and display the monitoring results through a visual interface; The anomaly diagnosis and tracing module (4) is configured to perform accurate diagnosis and tracing of carbon emission anomalies based on causal reasoning and path optimization algorithms, and compare the actual data with the mechanism model threshold values; The emission reduction potential analysis module (5) is configured to construct an intelligent analysis model based on reinforcement learning and multi-objective optimization techniques, and dynamically recommend optimal emission reduction paths and strategies that take into account emission reduction effects, costs and production capacity impacts in combination with an emission reduction scheme knowledge graph; The visualization and reporting module (6) is configured to convert the accounting results, anomaly diagnosis and potential analysis data into charts and dynamic dashboards, and automatically generate carbon emission reports in accordance with national standards, thereby supporting internal management and external disclosure needs of the enterprise.

2. The enterprise carbon emission reduction data intelligent monitoring and analysis system according to claim 1, characterized in that, The data acquisition and preprocessing module (1) comprises a multi-source data access unit (11), a data cleaning unit (12) and a data standardization unit (13), wherein: The multi-source data access unit (11) is configured to access multi-type raw data of energy consumption, emission sources and production processes through Internet of Things sensors, industrial buses and third-party system APIs; The data cleaning unit (12) is configured to remove noise data, fill in missing values and correct abnormal readings caused by equipment failure by using statistical analysis and rule engines; The data standardization unit (13) is configured to convert data in different formats and units into a unified format that meets the input requirements of the mechanism model.

3. The enterprise carbon emission reduction data intelligent monitoring and analysis system according to claim 1, characterized in that, The mechanism model library module (2) comprises an industry model adaptation unit (21), a digital twinning parameter optimization unit (22) and a model management unit (23), wherein: The industry model adaptation unit (21) is configured to store basic mechanism models of carbon metabolism of the chemical and manufacturing industries, and support quick calling according to enterprise types; The digital twinning parameter optimization unit (22) is configured to dynamically calibrate key parameters in the mechanism model through production scene digital twinning simulation; The model management unit (23) is configured to realize model version iteration, permission control and calling log recording, and support user-defined model parameter configuration.

4. The enterprise carbon emission reduction data intelligent monitoring and analysis system according to claim 3, characterized in that, In the digital twinning parameter optimization unit (22), key parameters in the mechanism model are dynamically calibrated through production scene digital twinning simulation, and the specific operation is as follows: A1: Data mapping: for dynamically synchronizing real-time production data with digital twin virtual scene; A2: Parameter sensitivity analysis: based on variance decomposition method to determine the key parameters in the mechanism model that need to be optimized first; A3: Dynamic calibration: minimize the deviation between digital twin output and actual monitoring data by gradient descent algorithm, and iteratively update model parameters; A4: Verification feedback: use historical data to verify the error rate of the calibrated model, and feed back the optimization results to the model management unit (23).

5. The enterprise carbon emission reduction data intelligent monitoring and analysis system according to claim 4, characterized in that, The carbon emission accounting and monitoring module (3) includes a dynamic accounting unit (31), a real-time monitoring unit (32), and a visualization display unit (33), wherein: The dynamic accounting unit (31) is used to call the mechanism model to calculate the preprocessed data in real time, and generate the instantaneous / cumulative carbon emissions and emission intensity of each process and device; The real-time monitoring unit (32) is used to compare the monitoring carbon emission indicators with the threshold to determine whether they exceed the standard and trigger an early warning; The visualization display unit (33) is used to convert the accounting results into real-time dashboards and trend curves to intuitively present the overall and each link of the enterprise's carbon emission dynamics.

6. The enterprise carbon emission reduction data intelligent monitoring and analysis system according to claim 1, characterized in that, The abnormal diagnosis and tracing module (4) includes a causal relationship analysis unit (41), an abnormal path searching unit (42), and a root cause tracing unit (43), wherein: The causal relationship analysis unit (41) is based on a causal diagram network model and uses Bayesian inference to distinguish the causal relationship from irrelevant association in carbon emission data; The abnormal path searching unit (42) is used to quickly locate the propagation path of abnormal emissions by using a path optimization algorithm combined with a production process flowchart; The root cause tracing unit (43) is used to accurately identify the cause of the over-standard and generate a traceability report by combining the abnormal path and the mechanism model threshold.

7. The enterprise carbon emission reduction data intelligent monitoring and analysis system according to claim 6, characterized in that, The causal relationship analysis unit (41) is based on a causal diagram network model and uses Bayesian inference to distinguish the causal relationship from irrelevant association in carbon emission data, and the specific operation is as follows: B1: Construct a causal diagram network model of carbon emission related variables to represent the causal dependence relationship between variables; B2: Train and learn parameters of the Bayesian network based on historical and real-time carbon emission data; B3: Calculate the conditional probability distribution between variables using Bayesian inference algorithm; B4: Identify and distinguish the causal relationship and related interference in carbon emission abnormal fluctuations; B5: Output the causal relationship map to provide a basis for subsequent abnormal tracing.

8. The enterprise carbon emission reduction data intelligent monitoring and analysis system according to claim 1, characterized in that, The emission reduction potential analysis module (5) includes a scenario simulation unit (51), a multi-objective optimization unit (52), and a knowledge graph matching unit (53), wherein: The scenario simulation unit (51) simulates the carbon emission change effect of different emission reduction schemes based on a reinforcement learning model; The multi-objective optimization unit (52) is used to balance the emission reduction amount, cost input, and capacity loss by genetic algorithm to select the optimal scheme combination; The knowledge graph matching unit (53) is used to associate the emission reduction technology database and recommend the emission reduction path suitable for the enterprise scene.

9. The enterprise carbon emission reduction data intelligent monitoring and analysis system according to claim 8, characterized in that, In the multi-objective optimization unit (52), the emission reduction amount, cost input, and capacity loss are balanced by genetic algorithm to select the optimal scheme combination, and the specific operation is as follows: C1: Construct a multi-objective optimization model: Define the objective function as: wherein: = amount of carbon emission reduction; = total cost input; = capacity loss rate; Set the constraint conditions: C2: Encoding and initialization of population: Encode the emission reduction scheme parameters such as equipment replacement ratio and process improvement degree into chromosomes, and randomly generate the initial population; C3: Design fitness function: Calculate individual fitness by Pareto optimal sorting, and the fitness function is based on the following formula: wherein: , , is a weight coefficient, is a value of the emission reduction target function corresponding to the ith individual; is a value of the cost input target function corresponding to the ith individual; is a value of the capacity loss target function corresponding to the ith individual; is a maximum value of the emission reduction target function; is a maximum value of the cost input target function; is a maximum value of the capacity loss target function; C4: Genetic operation iteration optimization: Selection: Adopt tournament selection method to reserve elite individuals; Crossover: Perform arithmetic crossover operation on selected individuals; Mutation: Perform Gaussian mutation on specific gene sites of chromosomes; Iterate for N generations until the convergence condition is met; C5: Output the Pareto optimal solution set: Select non-dominated solutions to form the optimal solution set, and each solution corresponds to a different weight-based emission reduction-cost-capacity balance scheme.

10. The enterprise carbon emission reduction data intelligent monitoring and analysis system according to claim 1, characterized in that, The visualization and reporting module (6) includes a multi-dimensional visualization unit (61), a report generation unit (62), and a permission management unit (63), wherein: The multi-dimensional visualization unit (61) is used to convert accounting, diagnosis, and analysis data into three-dimensional heat maps and dynamic Sankey diagrams to visually display carbon emission distribution and emission reduction potential; The report generation unit (62) is used to automatically extract key data, generate standardized reports in accordance with the "Greenhouse Gas Accounting and Reporting Requirements", and support PDF and Excel format export; The permission management unit (63) displays differentiated data views according to user roles to meet the permission isolation needs of internal management and external disclosure.

Images