Carbon emission reduction optimization system based on knowledge graph

Abstract

The application discloses a carbon emission reduction optimization system based on a knowledge graph, relates to the technical field of carbon emission reduction optimization, and comprises the following steps: after recognizing emission points of a current process flow, detecting and acquiring corresponding emission detection data, and performing pollution risk assessment, generating an emission coefficient from emission risk data of the emission points, screening out risk points in the emission points according to the emission coefficient, matching corresponding treatment strategies for the emission of the risk points, and sequentially treating the emission of each risk point; after collecting and detecting abnormal emission, if a reduction proportion of abnormal density is lower than an expectation, generating an optimization degree from carbon emission data of each emission point, optimizing control parameters of a production and processing process, and executing the optimized production and processing process. Through combination and selection of multiple optimization modes, carbon emission is reduced, and a carbon emission reduction process in a current stage is fed back.

Description

Technical Field

[0001] This invention relates to the field of carbon emission reduction optimization technology, specifically a carbon emission reduction optimization system based on knowledge graphs. Background Technology

[0002] Carbon emission reduction refers to a series of measures, including saving energy, improving energy efficiency, developing clean energy, promoting low-carbon technologies and products, and strengthening carbon sink construction (such as afforestation), to reduce greenhouse gas emissions such as carbon dioxide generated by human socio-economic activities. These actions aim to mitigate the trend of global warming, reduce the occurrence of extreme weather events, and protect the ecological environment, so as to achieve a win-win situation for environmental protection and socio-economic development.

[0003] Chinese invention patent application CN114580759A discloses an urban low-carbon emission reduction assessment system, comprising: a data acquisition module for acquiring and collecting parameter indicators of a target city; wherein the parameter indicators include at least driving force indicators, pressure indicators, response indicators, state indicators, and impact indicators; an assessment indicator system module for selecting assessment indicators based on preset low-carbon emission reduction standards and establishing an assessment indicator system; and an assessment module for assessing the parameter indicators based on the assessment indicator system and generating assessment results.

[0004] Combining the above applications and existing technologies:

[0005] The production and processing of industrial products generate a large amount of carbon emissions due to the direct or indirect use of various fossil energy sources, which in turn causes environmental pollution or triggers climate problems. Therefore, in order to control and optimize excessive carbon emissions, it is necessary to assess the carbon emissions generated during the production process after implementing corresponding emission reduction strategies.

[0006] Existing methods for assessing the effectiveness of carbon emission reduction typically collect carbon emission data before and after emission reduction optimization. The data is used to determine whether the emission reduction in the current stage has achieved the expected results. However, this assessment is relatively one-dimensional. Especially when there are still many abnormal emission behaviors, a single-level assessment is unlikely to provide the expected positive feedback, which leads to the inability to sustain subsequent carbon emission reduction effects and to contribute to continuous carbon emission reduction.

[0007] To address this, the present invention provides a carbon emission reduction optimization system based on knowledge graphs. Summary of the Invention

[0008] (a) Technical problems to be solved

[0009] To address the shortcomings of existing technologies, this invention provides a carbon emission reduction optimization system based on a knowledge graph. After identifying emission points in the current process flow, it detects and acquires corresponding emission data for pollution risk assessment. Emission coefficients are generated from the emission risk data of each emission point. Based on these coefficients, risk points are selected from several emission points, and appropriate treatment strategies are matched to the emissions of these risk points. The emissions from each risk point are then treated sequentially. After collecting and detecting abnormal emissions, if the reduction rate of the abnormal intensity is lower than expected, an optimization degree is generated from the carbon emission data at each emission point. This optimization is then applied to the control parameters of the production process, and the optimized production process is executed. By combining and selecting multiple optimization methods, carbon emission reduction is achieved, thus solving the technical problems in the background technology.

[0010] (II) Technical Solution

[0011] To achieve the above objectives, the present invention provides the following technical solution: a carbon emission reduction optimization system based on knowledge graphs, comprising an anomaly analysis unit that identifies abnormal emission data during the production and processing process and generates anomaly density from the abnormal emission data. If the abnormal density If the emission exceeds the preset expectations, an anomaly identification command is issued externally; among these, anomaly density is generated from the anomaly emission data within the anomaly emission data set. ,in,

[0012]

[0013] Weighting coefficients: , , This represents the number of abnormal emission events. This represents the severity of the abnormal emission event. This is the corresponding mean; Let be the time interval between the i-th and j-th abnormal emission events. Let be the mean of the i-th time interval;

[0014] The risk assessment unit identifies the emission points of the current process flow, detects and acquires the corresponding emission detection data, and conducts a pollution risk assessment. If the obtained risk score exceeds the preset risk threshold, it sends an alarm command to the outside.

[0015] The emission point screening unit generates emission coefficients from emission risk data of emission points. Based on emission coefficients Risk points are selected from several emission points, and corresponding treatment strategies are matched for the emissions of risk points. Then, the emissions of each risk point are treated in turn.

[0016] The fault detection unit detects faults in the production and processing equipment. If an operational fault is found, the fault maintenance knowledge graph provides a corresponding maintenance plan, and the faulty equipment is maintained after a maintenance interval that meets the constraints.

[0017] The optimization unit collects and detects abnormal emissions; if the abnormal density... The reduction rate was lower than expected, and the optimization was generated from carbon emission data at each emission point. The control parameters of the production and processing technology are optimized, and the optimized production and processing technology is executed.

[0018] Furthermore, carbon emission data during the production and processing process is collected. Using the carbon emission data as input, an anomaly analysis is performed using a trained anomaly emission identification model to identify and obtain corresponding anomaly emission data, including the type, severity, and time of the anomaly emission, and then a set of anomaly emission data is generated.

[0019] Furthermore, upon receiving the anomaly identification instruction, the system queries and retrieves the current production and processing technology, performs process analysis, and identifies the emission points at each stage. The sensor network detects and acquires the corresponding emission detection data at the emission points, obtains the carbon emission data for each emission point, uses the detected carbon emission data as input, and performs pollution risk assessment using the trained emission risk assessment model to obtain the corresponding pollution risk score.

[0020] Furthermore, upon receiving an alarm command, sensitivity analysis is used to determine the sensitivity of the emission point to the production process and obtain the corresponding sensitivity coefficient. The time node at which the alarm command is received is taken as the alarm node, based on the risk score of the risk point, the distribution of alarm nodes, and the sensitivity coefficient. Generation emission factor If the obtained emission coefficient If the emission threshold is exceeded, the corresponding emission point will be designated as a risk point.

[0021] Furthermore, based on the risk scores of risk points, the distribution of alarm nodes, and sensitivity coefficients... Generation emission factor The method is as follows:

[0022]

[0023] In the formula: This is the risk score of the i-th risk point at time t. For carbon emission monitoring cycles, As a risk threshold, The value is 2.718. Where μ is the time decay coefficient and μ is the distance decay coefficient; Let j be the alarm node located at time t. Let be the time interval between the i-th and j-th alarm nodes, and n be the number of alarm nodes. For the risk score at the j-th alarm node; weighting coefficient, , , .

[0024] Furthermore, identify carbon emission data and other emission data at risk points, obtain emission categories such as by-products and waste, pre-obtain several emission treatment strategies, and match the corresponding emission treatment strategies for the emission from the emission treatment strategy library;

[0025] Construct an electronic map covering the production and processing area, and mark each risk point on the electronic map; based on the location of the risk points and their emission coefficients... The inspection path is planned by a pre-trained path planning algorithm, and the emissions at each risk point are processed in sequence according to the inspection path.

[0026] Furthermore, if the number of risk points exceeds expectations, the operational status of production and processing equipment related to the current emission process is monitored to obtain corresponding operational status data; using the operational status data as input, the trained fault detection model is used to perform fault detection and output fault detection data.

[0027] Furthermore, after feature extraction from the fault detection data, fault features are obtained. Using processing equipment fault maintenance as the target term, a fault maintenance knowledge graph is pre-constructed; the maintenance interval of the faulty equipment is then defined. Constraints are applied, and maintenance is performed on the faulty equipment after a maintenance interval that meets the constraints. The maintenance interval is... The constraint methods are as follows:

[0028]

[0029] in, The number of carbon emission monitoring cycles, It is the difference in anomalous density between the i-th and j-th carbon emission monitoring cycle maps. The average of the differences, weighted by: , ,and .

[0030] Furthermore, emission data is continuously collected to detect abnormal emission events, and abnormal intensity is regenerated from the detection data. If abnormal density The reduction rate was lower than expected;

[0031] Using carbon emission data as input, the trained emission risk assessment model outputs a corresponding pollution risk score. Pollution risk scores for each emission point before and after optimization. Optimization degree is generated after alignment If the optimization degree is obtained If the preset optimization threshold is exceeded, an optimization command is sent to the outside.

[0032] Furthermore, upon receiving the optimization instruction, various control parameters of the production and processing technology are collected. With reducing the frequency of abnormal emissions as the optimization objective, a pre-trained multi-objective optimization algorithm is used to optimize the control parameters of the production and processing technology. After executing the optimized production and processing technology, carbon emission data is continuously collected and a carbon emission reduction report is generated.

[0033] (III) Beneficial Effects

[0034] This invention provides a knowledge graph-based carbon emission reduction optimization system, which has the following beneficial effects:

[0035] 1. Real-time carbon emission data at various time points; anomaly identification of the carbon emission process; identification of abnormal emission events; timely handling of abnormal carbon emissions.

[0036] 2. Construct corresponding anomaly density from abnormal state data. Based on abnormal density Assessing the overall degree of abnormality in current carbon emissions is an extension of anomaly detection. When abnormal carbon emissions persist, timely intervention can be implemented to reduce or mitigate the current carbon emission process.

[0037] 3. By combining the current production and processing technology, carbon emission points can be screened out. When adjusting and optimizing carbon emissions during the production and processing process, the optimization and adjustment efficiency can be improved. By evaluating each emission point, it can be determined whether the carbon emissions of the current emission point exceed the standard, and real-time monitoring can be achieved when the standard is exceeded.

[0038] 4. Emission coefficients are generated based on risk scores for risk points, the time of receiving alarm commands, and sensitivity coefficients. Based on emission coefficients By identifying key emission points from a pool of sources, targeted measures can be taken when time and cost are limited.

[0039] 5. After screening the risk points that need to be addressed, by planning the corresponding inspection routes, each risk point can be addressed sequentially, thereby improving the overall efficiency of emission treatment; by performing fault detection or fault prediction, potential operational faults in production and processing equipment can be detected or predicted in a timely manner, and if operational faults exist, they can be addressed promptly.

[0040] 6. By providing corresponding maintenance solutions through a fault maintenance knowledge graph, targeted handling of equipment faults in production and processing equipment can be achieved. By adjusting the maintenance frequency of production and processing equipment to be compatible with production status and carbon emission status, it is possible to avoid excessive maintenance frequency of multiple production and processing equipment, which could affect the production process.

[0041] 7. Generate optimization degree after processing the feedback data. To optimize The effectiveness of carbon emission optimization in the current stage is evaluated to determine whether the optimization measures have achieved the expected results. If the expected results are not achieved, further actions can be taken to improve the effectiveness of the optimization.

[0042] 8. Optimize and automate the current production and processing technology to respond to optimization methods. Based on the assessment of carbon emissions, reduce carbon emissions by combining and selecting multiple optimization methods. Provide feedback on the carbon emission reduction process in the current stage by periodically generating carbon emission reduction reports. Attached Figure Description

[0043] Figure 1 This is a schematic diagram of the carbon emission reduction optimization method based on knowledge graphs according to the present invention;

[0044] Figure 2 This is a schematic diagram of the carbon emission reduction optimization system based on knowledge graphs according to the present invention. Detailed Implementation

[0045] 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.

[0046] Please see Figure 1 This invention provides a knowledge graph-based carbon emission reduction optimization method, including:

[0047] Step 1: After identifying abnormal emission data during the production and processing process, generate abnormal density from the abnormal emission data. If the abnormal density If the error exceeds the preset expectation, an anomaly identification command will be issued to the outside.

[0048] Step one includes the following:

[0049] Step 101: Collect carbon emission data in the current production and processing process, such as fossil fuel consumption, electricity use, and waste data. Train a convolutional network using the labeled sample data to obtain a trained abnormal emission identification model.

[0050] Using carbon emission data as input, an anomaly analysis is performed using a trained anomaly emission identification model to identify and obtain corresponding anomaly emission data, including the type, severity, and time of the anomaly emission, and to generate an anomaly emission data set.

[0051] When in use, during the production and processing process, in order to reduce carbon emissions during production, carbon emission data at each time point is monitored in real time. By identifying abnormal emission events in the carbon emission process, abnormal carbon emissions can be dealt with in a timely manner.

[0052] Step 102: Under dimensionless conditions, generate anomaly density from the anomaly emission data within the anomaly emission dataset. ,in,

[0053]

[0054] Weighting coefficients: , , ; The number of abnormal emission events. This represents the severity of the abnormal emission event. This is the corresponding mean; Let be the time interval between the i-th and j-th abnormal emission events. Let be the mean of the i-th time interval;

[0055] Based on historical data and management expectations for abnormal emissions, a density threshold is pre-set; if the obtained abnormal density... If the frequency exceeds the preset threshold, it indicates that there are many abnormal emission events and a large amount of carbon emissions, requiring timely intervention and handling, and issuing an anomaly identification command to the outside.

[0056] When using it, refer to the content in steps 101 and 102:

[0057] As a further development, when abnormal carbon emissions occur continuously, corresponding abnormality density will be constructed from relevant abnormal state data. Based on abnormal density Assessing the overall degree of abnormality in current carbon emissions is an extension of anomaly detection. When abnormal carbon emissions persist, timely intervention can be implemented to reduce or mitigate the current carbon emission process.

[0058] Combining the above applications and existing technologies:

[0059] The production and processing of industrial products generate a large amount of carbon emissions due to the direct or indirect use of various fossil energy sources, which in turn causes environmental pollution or triggers climate problems. Therefore, in order to control and optimize excessive carbon emissions, it is necessary to assess the carbon emissions generated during the production process after implementing corresponding emission reduction strategies.

[0060] Existing methods for assessing the effectiveness of carbon emission reduction typically collect carbon emission data before and after emission reduction optimization. The data is used to determine whether the emission reduction in the current stage has achieved the expected results. However, this assessment is relatively one-dimensional. Especially when there are still many abnormal emission behaviors, a single-level assessment is unlikely to provide the expected positive feedback, which leads to the inability to sustain subsequent carbon emission reduction effects and to contribute to continuous carbon emission reduction.

[0061] Step 2: After identifying the emission points of the current process, detect and obtain the corresponding emission detection data and conduct a pollution risk assessment. If the obtained risk score exceeds the preset risk threshold, issue an alarm command to the outside.

[0062] Step two includes the following:

[0063] Step 201: Upon receiving the anomaly identification command, query and obtain the current production and processing technology, perform process analysis on the current production process, and identify the emission points at each stage of the current process flow; deploy a sensor network at the emission points, such as air quality sensors and carbon emission detection sensors, etc.

[0064] The sensor network detects and acquires the corresponding emission detection data at the emission point, and acquires the carbon emission data of each emission point, including gaseous emissions, liquid emissions and solid waste, such as carbon dioxide or waste organic matter, and then aggregates them to generate a carbon emission data set.

[0065] When used, after decomposing and identifying the current production and processing process, and combining it with the current production and processing technology, carbon emission points are screened out. When adjusting and optimizing carbon emissions in the production and processing process, the optimization and adjustment efficiency can be improved.

[0066] Step 202: Train a neural convolutional network using the labeled sample data to obtain the trained emission risk assessment model;

[0067] Using the carbon emission data obtained from the detection as input, the trained emission risk assessment model is used to conduct a pollution risk assessment and obtain a corresponding pollution risk score. If the risk score exceeds the preset risk threshold, it indicates that the current carbon emission risk is high and needs to be adjusted. At this time, an alarm command is issued to the outside.

[0068] When using it, refer to steps 201 and 202:

[0069] As a further step, after determining the reference standards for carbon emissions and pollution risks at each carbon emission point, the system assesses each emission point to determine whether its carbon emissions exceed the standards, and enables real-time monitoring when they do.

[0070] Step 3: Generate emission coefficients from emission risk data at emission points. Based on emission coefficients Risk points are selected from several emission points, and corresponding treatment strategies are matched for the emissions of risk points. Then, the emissions of each risk point are treated in turn.

[0071] Step three includes the following:

[0072] Step 301: Upon receiving an alarm command, sensitivity analysis is used to determine the sensitivity of the emission point to the production process, and the corresponding sensitivity coefficient is obtained. The time node at which the alarm command is received is taken as the alarm node. Under dimensionless conditions, based on the risk score of the risk point, the distribution of the alarm node, and the sensitivity coefficient... Generation emission factor The method is as follows:

[0073]

[0074] In the formula: This is the risk score of the i-th risk point at time t. For carbon emission monitoring cycles, As a risk threshold, The value is 2.718. is the time decay coefficient, with a value between 0 and 1; μ is the distance decay coefficient, with a value between 0.01 and 0.1. Let j be the alarm node located at time t. Let be the time interval between the i-th and j-th alarm nodes, and n be the number of alarm nodes. For the risk score at the j-th alarm node; weighting coefficient, , , ;

[0075] Based on historical data and expected emission management at emission points, emission thresholds are pre-set; if the obtained emission coefficients... If the emission threshold is exceeded, it indicates that the carbon emissions at the current emission point are too high and the pollution risk is too great. The corresponding emission point will be regarded as a risk point, and the emission points that need to be improved will be selected based on the risk value.

[0076] In use, based on the assessment of each emission point, emission coefficients are generated according to the risk score of the risk point, the time of receiving the alarm command, and the sensitivity coefficient. It can be based on emission coefficients By identifying key emission points from a pool of sources, targeted measures can be taken when time and cost are limited.

[0077] Step 302: Identify carbon emission data and other emission data at risk points, and obtain emission categories such as by-products and waste; through online retrieval and offline construction, pre-acquire several emission treatment strategies for reducing pollution and carbon emissions, such as waste recycling, landfill, or decomposition, and generate an emission treatment strategy library after summarizing them; based on the correspondence between emission data and treatment strategies, the emission treatment strategy library matches the corresponding emission treatment strategies for the emissions;

[0078] As a further step, by identifying the emissions at the emission points and matching them with targeted treatment methods, carbon emissions can be reduced.

[0079] Step 303: Construct an electronic map covering the production and processing area, and mark each risk point on the electronic map; based on the location of the risk points and emission coefficients... The inspection path is planned by a pre-trained path planning algorithm, and the emissions at each risk point are processed sequentially according to the inspection path.

[0080] When using this method, refer to steps 301 to 303:

[0081] After identifying the risk points that need to be addressed, by planning the corresponding inspection routes, each risk point can be addressed sequentially, thereby improving the overall efficiency of emission treatment.

[0082] Step 4: Conduct fault detection on the production and processing equipment. If an operational fault is found, the fault maintenance knowledge graph will provide a corresponding maintenance plan. After a maintenance interval that meets the constraints, the faulty equipment will be maintained.

[0083] Step four includes the following:

[0084] Step 401: If the number of risk points exceeds expectations, conduct operational status monitoring on the production and processing equipment related to the current emission process and obtain relevant operational status data, including equipment energy consumption, operating power, and power factor.

[0085] A machine learning algorithm is trained using labeled sample data to obtain a trained fault detection model; the running status data is used as input to perform fault detection using the trained fault detection model, and the fault detection data is output and aggregated to generate a fault detection data set;

[0086] When in use, based on the collected production data of the production and processing equipment, fault detection or fault prediction can be performed to promptly identify or predict potential operational faults in the production and processing equipment. If an operational fault is found, it can be dealt with in a timely manner.

[0087] Step 402: After extracting features from the fault detection data, obtain fault features. Using processing equipment fault maintenance as the target term, after deep retrieval and entity relationship construction, pre-construct a fault maintenance knowledge graph. Based on the correspondence between fault maintenance solutions and fault features, the fault maintenance knowledge graph provides the corresponding maintenance solutions.

[0088] and maintenance intervals for faulty equipment Constraints are applied, and after a maintenance interval that meets the constraints, maintenance is performed on the faulty equipment. The maintenance interval is... The constraint methods are as follows:

[0089]

[0090] in, The number of carbon emission monitoring cycles, It is the difference in anomalous density between the i-th and j-th carbon emission monitoring cycle maps. The average of the differences, weighted by: , ,and ;

[0091] When using this method, refer to steps 401 and 402:

[0092] After fault detection of production and processing equipment, corresponding maintenance solutions are provided through fault maintenance knowledge graphs. This allows for targeted handling of equipment faults, targeted reduction of energy consumption, and reduction of operating load, thereby reducing overall carbon emissions during the production and processing process. By adjusting the maintenance frequency of production and processing equipment to match production status and carbon emission status, it is possible to avoid excessive maintenance frequency of multiple production and processing equipment, which could affect the production process.

[0093] Step 5: After collecting and detecting abnormal emissions, if the abnormal density... The reduction rate was lower than expected, and the optimization was generated from carbon emission data at each emission point. The control parameters of the production and processing technology are optimized, and the optimized production and processing technology is executed.

[0094] Step five includes the following:

[0095] Step 501: Continuously collect emission data within the preset observation period, detect abnormal emission events, and regenerate abnormal density from the detection data. If abnormal density The reduction rate was lower than expected, triggering the carbon emission control feedback mechanism and collecting carbon emission data.

[0096] Step 502: Using carbon emission data as input, output the corresponding pollution risk score using the trained emission risk assessment model. Pollution risk scores for each emission point before and after optimization Alignment, and optimization degree generated as follows: :

[0097]

[0098] in, As the median value of the optimization degree, Its mean, where n is the number of emission points. and These are the pollution risk scores for the i-th emission point before and after optimization, respectively. and This is the corresponding mean;

[0099] Based on historical data and expectations for optimized carbon emission management, optimization thresholds are pre-set;

[0100] If the optimization degree is obtained If the optimization threshold is exceeded, it means that the carbon emissions in the current stage have achieved a certain effect. On the contrary, if the degree of optimization obtained does not meet expectations, it means that the current production and processing technology still needs to be optimized. At this time, an optimization command is issued to the outside.

[0101] During use, feedback data is collected after adaptive treatment of carbon emissions during the production process, and the feedback data is processed to generate an optimization degree. To optimize The effectiveness of carbon emission optimization in the current stage is evaluated to determine whether the optimization measures have achieved the expected results. If the expected results are not achieved, further actions can be taken to improve the effectiveness of the optimization.

[0102] Step 503: After receiving the optimization instruction, collect various control parameters of the production and processing technology, take reducing the frequency of abnormal emissions as the optimization goal, optimize the control parameters of the production and processing technology with a pre-trained multi-objective optimization algorithm, execute the optimized production and processing technology, continuously collect carbon emission data and generate a carbon emission reduction report.

[0103] When using this method, refer to steps 501 to 503:

[0104] As a further development, when existing optimization methods fail to achieve the expected results, the current production and processing technology will be optimized and automatically controlled to respond to the optimization methods. Based on the assessment of carbon emissions, carbon emissions will be reduced by combining and selecting multiple optimization methods. Furthermore, by periodically generating carbon emission reduction reports, feedback on the carbon emission reduction process in the current stage can be achieved.

[0105] Building a knowledge graph for the fault maintenance of processing and production equipment is a systematic project, involving multiple stages such as information collection, organization, relationship extraction, and graph construction. The following is a detailed construction method:

[0106] Before building a knowledge graph for the fault maintenance of processing and production equipment, it is first necessary to clarify the graph's goals and application scenarios. For example, is the graph intended to help maintenance personnel quickly locate and resolve equipment faults, or to provide early warnings and preventative maintenance recommendations? Clarifying the goals and application scenarios will facilitate subsequent information collection and graph design.

[0107] Basic Equipment Information: Includes equipment model, specifications, manufacturer, production date, etc. Failure Cases: Collects historical equipment failure cases, including failure symptoms, causes, solutions, and repair processes.

[0108] Maintenance Manual: Obtain the equipment's maintenance manual and operating instructions to understand the equipment's daily maintenance and repair requirements.

[0109] Expert experience: Invite experts in the field of equipment maintenance to conduct interviews and obtain their maintenance experience and advice.

[0110] Named entity recognition: Identifying key entities from collected information, such as equipment components, fault types, and maintenance tools. Relationship extraction: Determining the relationships between entities, such as the compositional relationships between components, the correspondence between faults and components, and the correspondence between maintenance methods and fault types.

[0111] Data standardization: Standardize the identified entities and relationships to ensure data consistency and accuracy.

[0112] Choose a knowledge graph building tool: Select a suitable knowledge graph building tool based on actual needs, such as Neo4j or Apache Jena. Create nodes and relationships: Create entity nodes and relationship edges in the building tool, and input the collected information into the graph. Graph optimization: Optimize the graph, such as merging duplicate nodes and adjusting relationship weights, to improve the accuracy and usability of the graph.

[0113] Fault diagnosis: When equipment malfunctions, maintenance personnel can quickly locate the faulty component and possible causes by consulting the knowledge graph, and then refer to the solutions in the graph for repair.

[0114] Preventive maintenance: Based on the equipment maintenance manual in the diagram and expert experience, develop a preventive maintenance plan for the equipment to reduce the occurrence of equipment failures.

[0115] Knowledge graph updates: As equipment is continuously updated and maintenance experience is accumulated, the knowledge graph needs to be updated and improved regularly to ensure its timeliness and accuracy.

[0116] Data Privacy: When collecting and processing equipment malfunction and maintenance information, it is essential to protect the company's data privacy and intellectual property rights. Data Quality: Ensure that the collected information is accurate, complete, and timely to improve the practicality and reliability of the knowledge graph. Technology Selection: Choose appropriate knowledge graph construction tools and technology solutions based on the company's actual needs and resources.

[0117] In summary, building a knowledge graph for the maintenance of processing and production equipment requires multiple steps, including defining objectives, collecting information, organizing relationships, constructing the graph, and application and maintenance. By building a knowledge graph, enterprises can quickly locate and resolve equipment faults, improving equipment reliability and production efficiency.

[0118] Please see Figure 2 This invention provides a knowledge graph-based carbon reduction optimization system, comprising:

[0119] The anomaly analysis unit identifies abnormal emission data during the production and processing process, and then generates anomaly density from the abnormal emission data. If the abnormal density If the error exceeds the preset expectation, an anomaly identification command will be issued to the outside.

[0120] The risk assessment unit identifies the emission points of the current process flow, detects and acquires the corresponding emission detection data, and conducts a pollution risk assessment. If the obtained risk score exceeds the preset risk threshold, it sends an alarm command to the outside.

[0121] The emission point screening unit generates emission coefficients from emission risk data of emission points. Based on emission coefficients Risk points are selected from several emission points, and corresponding treatment strategies are matched for the emissions of risk points. Then, the emissions of each risk point are treated in turn.

[0122] The fault detection unit detects faults in the production and processing equipment. If an operational fault is found, the fault maintenance knowledge graph provides a corresponding maintenance plan, and the faulty equipment is maintained after a maintenance interval that meets the constraints.

[0123] The optimization unit collects and detects abnormal emissions; if the abnormal density... The reduction rate was lower than expected, and the optimization was generated from carbon emission data at each emission point. The control parameters of the production and processing technology are optimized, and the optimized production and processing technology is executed.

[0124] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0125] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0126] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0127] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0128] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A carbon emission reduction optimization system based on knowledge graphs, characterized in that: include, The anomaly analysis unit identifies abnormal emission data during the production and processing process, and then generates anomaly density from the abnormal emission data. If the abnormal density If the emission exceeds the preset expectations, an anomaly identification command is issued externally; among these, anomaly density is generated from the anomaly emission data within the anomaly emission data set. : Weighting coefficients: , , This represents the number of abnormal emission events. This represents the severity of the abnormal emission event. This is the corresponding mean; Let be the time interval between the i-th and j-th abnormal emission events. Let be the mean of the i-th time interval; The risk assessment unit identifies the emission points of the current process flow, detects and acquires the corresponding emission detection data, and conducts a pollution risk assessment. If the obtained risk score exceeds the preset risk threshold, it sends an alarm command to the outside. The emission point screening unit generates emission coefficients from emission risk data of emission points. Based on emission coefficients Risk points are selected from several emission points, and corresponding treatment strategies are matched for the emissions of risk points. Then, the emissions of each risk point are treated in turn. The fault detection unit detects faults in the production and processing equipment. If an operational fault is found, the fault maintenance knowledge graph provides a corresponding maintenance plan, and the faulty equipment is maintained after a maintenance interval that meets the constraints. The optimization unit collects and detects abnormal emissions; if the abnormal density... The reduction rate was lower than expected, and the optimization was generated from carbon emission data at each emission point. The control parameters of the production and processing technology are optimized, and the optimized production and processing technology is executed.

2. The carbon emission reduction optimization system based on knowledge graphs according to claim 1, characterized in that: Carbon emission data is collected during the production and processing process. Using the carbon emission data as input, an anomaly analysis is performed using a trained anomaly emission identification model to identify and obtain corresponding anomaly emission data, including the type, severity, and time of the anomaly emission, and then a set of anomaly emission data is generated.

3. The carbon emission reduction optimization system based on knowledge graphs according to claim 1, characterized in that: Upon receiving the anomaly identification instruction, the system queries and retrieves the current production and processing technology, performs process analysis, and identifies the emission points at each stage. The sensor network detects and acquires the corresponding emission detection data at the emission points, obtains the carbon emission data for each emission point, uses the detected carbon emission data as input, and performs a pollution risk assessment using the trained emission risk assessment model to obtain the corresponding pollution risk score.

4. The knowledge graph-based carbon reduction optimization system according to claim 3, characterized in that: Upon receiving an alarm command, sensitivity analysis is used to determine the sensitivity of the emission point to the production process and obtain the corresponding sensitivity coefficient. The time node at which the alarm command is received is taken as the alarm node, and the alarm is determined based on the risk score of the risk point, the distribution of alarm nodes, and the sensitivity coefficient. Generation emission factor If the obtained emission coefficient If the emission threshold is exceeded, the corresponding emission point will be designated as a risk point.

5. The knowledge graph-based carbon reduction optimization system according to claim 4, characterized in that: Based on the risk score of the risk point, the distribution of alarm nodes, and the sensitivity coefficient. Generation emission factor The method is as follows: In the formula: This is the risk score of the i-th risk point at time t. For carbon emission monitoring cycles, As a risk threshold, The value is 2.

718. Where μ is the time decay coefficient and μ is the distance decay coefficient; Let j be the alarm node located at time t. Let be the time interval between the i-th and j-th alarm nodes, and n be the number of alarm nodes. For the risk score at the j-th alarm node; weighting coefficient, , , .

6. The knowledge graph-based carbon reduction optimization system according to claim 5, characterized in that: Identify carbon emission data and other emission data at risk points, obtain the categories of by-product and waste emissions, pre-acquire several emission treatment strategies, and match the corresponding emission treatment strategies for the emissions using the emission treatment strategy library. Construct an electronic map covering the production and processing area, and mark each risk point on the electronic map; based on the location of the risk points and their emission coefficients... The inspection path is planned by a pre-trained path planning algorithm, and the emissions at each risk point are processed in sequence according to the inspection path.

7. The knowledge graph-based carbon emission reduction optimization system according to claim 6, characterized in that: If the number of risk points exceeds expectations, conduct operational status monitoring on production and processing equipment related to the current emission process and obtain the corresponding operational status data; Using operational status data as input, the trained fault detection model is used to detect faults and output fault detection data.

8. The knowledge graph-based carbon reduction optimization system according to claim 7, characterized in that: After extracting features from the fault detection data, fault features are obtained, and a fault maintenance knowledge graph is pre-constructed using processing equipment fault maintenance as the target term. Maintenance intervals for faulty equipment Constraints are applied, and maintenance is performed on the faulty equipment after a maintenance interval that meets the constraints. The maintenance interval is... The constraint methods are as follows: ;in, The number of carbon emission monitoring cycles, It is the difference in anomalous density between the i-th and j-th carbon emission monitoring cycle maps. The average of the differences, weighted by: , ,and .

9. The knowledge graph-based carbon reduction optimization system according to claim 8, characterized in that: Continuously collect emission data, detect abnormal emission events, and regenerate abnormal intensity from the detected data. If abnormal density The reduction rate was lower than expected; Using carbon emission data as input, the trained emission risk assessment model outputs a corresponding pollution risk score. Pollution risk scores for each emission point before and after optimization. Optimization degree is generated after alignment If the optimization degree is obtained If the preset optimization threshold is exceeded, an optimization command is sent to the outside.

10. The knowledge graph-based carbon reduction optimization system according to claim 9, characterized in that: Upon receiving the optimization instruction, the system collects various control parameters of the production and processing technology, takes reducing the frequency of abnormal emissions as the optimization objective, optimizes the control parameters of the production and processing technology using a pre-trained multi-objective optimization algorithm, and continuously collects carbon emission data and generates a carbon emission reduction report after executing the optimized production and processing technology.

Images