Engineering carbon emission factor dynamic calculation and traceability analysis method and system

Abstract

The invention provides an engineering carbon emission factor dynamic calculation and traceability analysis method and system, and belongs to the field of engineering construction. The method comprises the following steps: acquiring multi-source heterogeneous data of the whole engineering construction process, and preprocessing the multi-source heterogeneous data to form a data set supporting knowledge modeling and computational analysis; constructing a semantic knowledge structure in the field of carbon emission factors in the whole engineering construction process by utilizing a knowledge graph technology; carrying out dynamic calculation on the carbon emission factor and correcting the weight; and carrying out visual analysis on a dynamic calculation result and a correction result to complete dynamic calculation and traceability analysis on the engineering carbon emission factor. According to the method, the defects in a carbon emission accounting system in the existing engineering construction field are overcome.

Description

Technical Field

[0001] This invention belongs to the field of engineering construction, and in particular relates to a method and system for dynamic calculation and source tracing analysis of carbon emission factors in engineering projects. Background Technology

[0002] As a significant sector of energy consumption and greenhouse gas emissions, the engineering and construction industry has become a crucial link in low-carbon transformation and energy conservation and emission reduction. Engineering and construction activities encompass multiple stages, including material production, transportation, construction, and operation. Carbon emission sources are numerous, structurally complex, and data is scattered, necessitating the establishment of a scientific, systematic, and dynamically updatable carbon emission factor calculation and management system.

[0003] Currently, commonly used carbon emission accounting methods include emission factor method, material balance method and measurement method. However, these methods are mostly designed for industrial and operational phases, and are still insufficient for handling multi-source dynamic data and regional differences during the construction phase, resulting in a lack of uniformity and real-time performance in the accounting results.

[0004] In terms of carbon emission management, digital functions for "viewing, analyzing, and reducing carbon emissions" have been achieved, but these are mainly concentrated at the energy industry or enterprise level. A systematic solution for monitoring, calculating, and tracing carbon emissions throughout the entire engineering construction process is still lacking. Furthermore, although low-carbon technology research is relatively mature in the building and energy sectors, the engineering construction stage still faces challenges such as the complexity of low-carbon technology types, inconsistent evaluation systems, and a lack of knowledge-based, structured management and correlation mechanisms, making it difficult to achieve dynamic linkage between low-carbon technologies and carbon emission factor calculation results. Therefore, it is necessary to construct a knowledge graph-based dynamic calculation and traceability analysis system for engineering carbon emission factors. This system would combine multi-source data fusion with knowledge-based management to establish a dynamic update, knowledge association, and full-process traceability mechanism for carbon emission factors, supporting accurate carbon emission accounting and intelligent decision-making during the engineering construction phase. Summary of the Invention

[0005] To address the aforementioned shortcomings in existing technologies, this invention provides a method and system for dynamic calculation and source tracing analysis of engineering carbon emission factors, which solves the deficiencies in the existing carbon emission accounting system in the field of engineering construction.

[0006] To achieve the above objectives, the technical solution adopted by this invention is: a method for dynamic calculation and source tracing analysis of engineering carbon emission factors, comprising the following steps: S1. Acquire multi-source heterogeneous data throughout the entire engineering construction process, and preprocess the multi-source heterogeneous data to form a dataset that supports knowledge modeling and computational analysis; S2. Based on the dataset, a semantic knowledge structure for carbon emission factors in the entire process of engineering construction is constructed using knowledge graph technology. The semantic knowledge structure for carbon emission factors is used for intelligent completion, semantic reasoning and causal tracing of carbon emission factors. S3. Based on the semantic knowledge structure of the carbon emission factor domain, the carbon emission factor is dynamically calculated and the weights are corrected. S4. Visualize and analyze the dynamic calculation results and correction results to complete the dynamic calculation and source analysis of the carbon emission factors of the project.

[0007] The beneficial effects of this invention are: by using multi-source heterogeneous data fusion, semantic modeling and knowledge reasoning technology, this invention realizes dynamic updating of carbon emission factors, causal tracing and full-process visual management, which significantly improves the scientificity, accuracy and intelligence of carbon accounting in the engineering construction stage and solves the shortcomings of the existing carbon emission accounting system in the field of engineering construction.

[0008] Further, S1 includes the following steps: S101. Acquire multi-source heterogeneous data throughout the entire engineering construction process and preprocess the multi-source heterogeneous data. The multi-source heterogeneous data includes basic engineering data, material and energy data, monitoring and environmental data, and standards and specifications data. S102. Write the preprocessed multi-source heterogeneous data into a graph database and a time-series database to form a dataset that supports knowledge modeling and computational analysis.

[0009] The beneficial effects of the above-mentioned further solutions are as follows: By standardizing and preprocessing multi-source heterogeneous data throughout the entire engineering construction process and using graph databases and time-series databases for structured storage, the present invention achieves unified management of data in both semantic and temporal dimensions, significantly improving the consistency, integrity, and traceability of data. It also provides a high-quality, fusionable data foundation for subsequent knowledge modeling and dynamic computation, thereby enhancing the system's real-time response capability and computational accuracy, and ensuring the dynamic updability and highly reliable traceability of carbon emission factors in both spatial and temporal dimensions.

[0010] Furthermore, S2 includes the following steps: S201. Domain ontology construction and semantic definition: Based on the dataset, define entity categories and their attributes, and use an ontology description language to define the hierarchical relationships and semantic constraints between entities, so as to model the domain ontology model covering the entire life cycle of engineering construction. S202 Knowledge Extraction and Fusion: Extract knowledge elements related to carbon emissions, and based on the semantic definition results, transform the knowledge into a structured form through named entity recognition, attribute extraction and triple generation, and process the semantics between knowledge to align and express them in a consistent manner; S203. Semantic Alignment and Integration: Based on the fusion results, semantic similarity calculation methods based on word embedding and graph embedding are used to perform entity-level and probability-level alignment processing, so as to form a unified knowledge structure through semantic merging and concept aggregation, and store it in a graph database. S204. Knowledge Graph Storage and Indexing: In the integrated knowledge system, multiple semantic relationships are constructed. Among them, multiple semantic relationships constitute the semantic skeleton of the semantic knowledge structure in the field of carbon emission factors, forming a dynamic knowledge grid. S205. Knowledge Reasoning and Completion Mechanism: Based on the knowledge graph, rule-based reasoning and probabilistic reasoning mechanisms are integrated to mine implicit relationships and automatically complete knowledge, thereby constructing a semantic knowledge structure for the carbon emission factor domain. Among them, rule-based reasoning is used to deduce causal chains according to predefined logical rules; and probabilistic reasoning mechanisms are used to infer unknown relationships, identify potential associations, and dynamically update them.

[0011] The beneficial effects of the aforementioned further approach are as follows: By constructing an ontology model covering the entire process of engineering construction in the field of carbon emissions, and combining knowledge extraction, semantic alignment, and reasoning completion technologies, semantic fusion and intelligent modeling of multi-source knowledge are achieved. This step not only transforms scattered engineering data and standard documents into a structured, reasonable knowledge system, but also automatically identifies and completes the potential correlations between carbon emission factors, forming a dynamically evolving semantic knowledge structure. This improves the semantic accuracy and intelligence level of carbon emission factor calculation, providing logical support and a data foundation for subsequent dynamic calculation, causal tracing, and low-carbon technology recommendations.

[0012] Furthermore, the multiple semantic relationships include: Activity-factor relationship: characterizing the direct impact of engineering activities on carbon emission factors; Factor-Energy Relationship: Describes the correspondence between energy consumption and carbon emission factors in terms of energy conversion and emissions; Energy-equipment relationship: revealing the dependency between energy type and equipment operating characteristics; Factor-technology relationship: indicates the effect of low-carbon technologies on the reduction, control, or substitution of carbon emission factors; Causal-temporal relationship: reflects the evolution logic and causal chain of carbon emission factors across time and process stages.

[0013] The beneficial effects of the aforementioned further solutions are as follows: By constructing multi-dimensional semantic relationships such as activity-factor, factor-energy, energy-equipment, factor-technology, and causality-time series, a core semantic framework for the semantic knowledge structure of carbon emission factors is formed. This achieves logical connections between engineering activities, energy consumption, equipment operation, and low-carbon technologies. Furthermore, it establishes causal chains for carbon emission factors in the time and process dimensions. The system can achieve dynamic correlation, causal tracing, and technology matching of carbon emission data at the semantic level, significantly improving the accuracy, interpretability, and intelligence level of carbon emission factor calculation.

[0014] Furthermore, step S3 includes the following steps: S301. Construct a carbon emission calculation formula based on the semantic knowledge structure of the carbon emission factor domain; S302. Within the framework of the carbon emission calculation formula, the carbon emission factor is dynamically corrected based on multidimensional semantic relationships to generate a dynamic emission factor. S303. Introduce a multi-source data fusion mechanism, use dynamic emission factors as prior information, use Bayesian methods to update the prior information, and use weighted correction of carbon emission factors to generate dynamic emission coefficients. The dynamic emission coefficients are fed back into the carbon emission formula in S301 for real-time calculation of engineering carbon emissions. S304. Perform trend identification and anomaly detection on carbon emission data. Based on the detection results, locate potential causes by combining semantic relationships in the knowledge graph. Based on the location results, perform self-correction processing using dynamic emission coefficients.

[0015] The beneficial effects of the aforementioned further solutions are as follows: By constructing a carbon emission calculation formula based on a semantic knowledge structure and combining dynamic correction of multi-dimensional semantic relationships with weighted fusion of multi-source data, adaptive updates of carbon emission factors are achieved. The system can automatically adjust emission factors and emission reduction parameters when new monitoring data arrives, ensuring that the calculation results continuously align with actual engineering conditions. Simultaneously, utilizing trend recognition and anomaly detection mechanisms, the system can locate potential anomalies based on a knowledge graph and perform self-correction operations, forming a dynamic closed loop of "semantic reasoning—data correction—model feedback." This improves the real-time performance, accuracy, and intelligence level of carbon emission accounting, ensuring high robustness and reliability even in complex and ever-changing engineering scenarios.

[0016] Furthermore, the carbon emission calculation formula is as follows: ; ; ; ; ; in, Indicates carbon emissions. Indicates activity level data, Indicates carbon emission factor, Indicates the emission reduction rate. This indicates the effective emission percentage that was not reduced. , and All of these represent weight coefficients determined through Bayesian learning. This represents the theoretical emission reduction percentage. This indicates the emission reduction efficiency obtained based on real-time monitoring data from the engineering site. This represents the emission reduction percentage predicted through knowledge graph reasoning and machine learning. This represents the baseline emissions without the adoption of low-carbon technologies. This represents the measured emissions after adopting low-carbon technologies. This represents the theoretical activity level under conditions without energy-saving measures. This indicates the current measured activity level.

[0017] The beneficial effect of the above-mentioned further measures is that they will reduce emissions by [percentage missing]. The model is a multi-source weighted fusion model composed of technological emission reduction rate, observed emission reduction rate and predicted emission reduction rate. Bayesian learning is used to determine the weight coefficients, enabling the system to comprehensively utilize low-carbon technology parameters, measured monitoring data and knowledge reasoning results to achieve dynamic correction and adaptive updating of carbon emission factors. This improves the real-time performance, accuracy and stability of carbon emission calculation, maintains calculation consistency under different engineering stages and data conditions, and has automatic learning and self-correction capabilities, thus providing a highly reliable calculation basis for engineering carbon emission accounting and emission reduction decisions.

[0018] Furthermore, S303 specifically refers to: A multi-source data fusion mechanism is introduced, using dynamic emission factors as prior information, and Bayesian methods are used to update the prior information. Based on the updated results, the carbon emission factors are weighted and corrected by comprehensively utilizing project ledgers, energy bills, sensor monitoring data, and external database information. When weighting and correcting, new measured data are treated as posterior information when they appear, and the prior carbon emission factor distribution is updated to obtain the dynamic carbon emission coefficient that best meets the engineering conditions.

[0019] The beneficial effects of the aforementioned further solutions are as follows: By introducing multi-source data fusion and a Bayesian update mechanism, the system can comprehensively utilize multi-source data such as project ledgers, energy bills, sensor monitoring, and external databases to dynamically weight and correct carbon emission factors. When new measured data appears, the system automatically uses it as posterior information to update the prior distribution, thereby obtaining the dynamic carbon emission coefficient that best matches the engineering conditions. This improves the calculation accuracy and real-time performance of carbon emission factors and also possesses continuous learning and adaptive adjustment capabilities, enabling the system to remain stable and reliable during long-term operation. This provides technical support for accurate carbon accounting and intelligent decision-making in complex engineering scenarios.

[0020] This invention also provides a system for dynamic calculation and source tracing analysis of engineering carbon emission factors, comprising: The data acquisition and fusion layer is used to acquire multi-source heterogeneous data throughout the entire engineering construction process, and to preprocess the multi-source heterogeneous data to form a dataset that supports knowledge modeling and computational analysis. The knowledge modeling and reasoning layer is used to construct semantic knowledge structures in the field of carbon emission factors based on datasets; The dynamic calculation and analysis layer is used to dynamically calculate carbon emission factors and correct their weights based on the semantic knowledge structure of the carbon emission factor domain. The visualization and service layer is used to visualize and analyze the dynamic calculation results and correction results, and to complete the dynamic calculation and source analysis of the carbon emission factors of the project.

[0021] Compared with existing engineering carbon emission accounting and management systems, this invention has the following significant advantages: 1. Dynamic updates and high-precision calculations: By linking engineering activities, energy types and emission factors through knowledge graphs, dynamic calculation and automatic correction of emission factors are achieved, reflecting changes in regions, processes and materials in real time, and improving the accuracy and timeliness of the calculations; 2. Traceable and highly reliable data management: Utilize knowledge graph path reasoning mechanisms to establish a full-chain traceability system for carbon emission factors, enabling traceability of data sources, calculation models, and reference standards, thereby enhancing the verifiability and transparency of results; 3. Data fusion and resource saving: By adopting a hierarchical storage and graph database structure, semantic fusion and efficient querying of multi-source data are achieved, reducing redundant storage and network transmission, and saving computing and bandwidth resources; 4. Intelligent analysis and decision support: Combining semantic reasoning and knowledge matching mechanisms, it automatically identifies high-emission links and recommends low-carbon technology solutions, providing intelligent carbon emission reduction decision support for engineering projects; 5. Safe, reliable and highly compatible: Through multi-level access control, data encryption and access auditing mechanisms, it ensures data security and operation traceability, while supporting system integration with existing carbon management platforms, and has good compatibility and deployment flexibility. Attached Figure Description

[0022] Figure 1 This is a flowchart of the method of the present invention.

[0023] Figure 2 This is a schematic diagram of the system structure of the present invention. Detailed Implementation

[0024] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0025] Example 1 To address the shortcomings of existing carbon emission accounting systems in the engineering construction field, this invention mainly solves the following problems: 1. The carbon emission factors lack a dynamic update mechanism, making it difficult to reflect the differences in carbon emissions caused by changes in different times, regions, and construction processes; 2. Carbon emission data sources are scattered and have complex structures. A semantic link and traceability mechanism between engineering activities and carbon emission factors have not yet been established, resulting in insufficient consistency in accounting results. 3. Existing carbon accounting and management platform models are static and lack dynamic perception, automatic reasoning and knowledge learning capabilities, which cannot support real-time analysis and decision-making in complex scenarios; 4. The low-carbon technology system is fragmented and fails to achieve knowledge linkage with carbon emission factors, making it difficult to support the intelligent identification and recommendation of carbon emission reduction paths.

[0026] To address the aforementioned issues, this invention provides a method for dynamic calculation and source tracing analysis of engineering carbon emission factors based on knowledge graphs. This invention utilizes multi-source heterogeneous data fusion, semantic modeling, and knowledge reasoning techniques to achieve dynamic updates of carbon emission factors, causal source tracing, and full-process visualized management, significantly improving the scientific rigor, accuracy, and intelligence of carbon accounting during the engineering construction phase.

[0027] like Figure 1 As shown, this invention provides a method for dynamic calculation and source tracing analysis of engineering carbon emission factors, the implementation method of which is as follows: S1. Acquire multi-source heterogeneous data from the entire engineering construction process, and preprocess the multi-source heterogeneous data to form a dataset that supports knowledge modeling and computational analysis. The implementation method is as follows: S101. Acquire multi-source heterogeneous data throughout the entire engineering construction process and preprocess the multi-source heterogeneous data. The multi-source heterogeneous data includes basic engineering data, material and energy data, monitoring and environmental data, and standards and specifications data. S102. Write the preprocessed multi-source heterogeneous data into a graph database and a time-series database to form a dataset that supports knowledge modeling and computational analysis.

[0028] In this embodiment, S1 is implemented using a data acquisition and fusion layer. This layer is used to access multi-source heterogeneous data from the entire engineering construction process. The multi-source heterogeneous data includes: Basic engineering data, including project location, engineering type, construction stage, process flow, equipment type, construction period, etc.; Material and energy data, including energy consumption in building material production, transportation distance and mode, fuel type and usage, electricity consumption, etc. Monitoring and environmental data, including on-site temperature and humidity, climate conditions, environmental monitoring parameters, and real-time energy consumption monitoring data; Standards and specifications data, including national and local carbon emission standards, industry technical specifications, and policy indicators.

[0029] The data acquisition and fusion layer enables real-time access and format unification of multi-source heterogeneous data through interface modules. It employs methods such as data cleaning, standardization, unit conversion, and time-series resampling to ensure structural consistency and semantic uniformity across the multi-source heterogeneous data. The processed multi-source heterogeneous data is then written into graph and time-series databases, forming a foundational dataset that supports knowledge modeling and computational analysis.

[0030] S2. Based on the dataset, a semantic knowledge structure for carbon emission factors throughout the entire engineering construction process is constructed using knowledge graph technology. This semantic knowledge structure is used for intelligent completion, semantic reasoning, and causal tracing of carbon emission factors. The implementation method is as follows: S201. Domain ontology construction and semantic definition: Based on the dataset, define entity categories and their attributes, and use an ontology description language to define the hierarchical relationships and semantic constraints between entities, so as to model the domain ontology model covering the entire life cycle of engineering construction. S202 Knowledge Extraction and Fusion: Extract knowledge elements related to carbon emissions, and based on the semantic definition results, transform the knowledge into a structured form through named entity recognition, attribute extraction and triple generation, and process the semantics between knowledge to align and express them in a consistent manner; S203. Semantic Alignment and Integration: Based on the fusion results, semantic similarity calculation methods based on word embedding and graph embedding are used to perform entity-level and probability-level alignment processing, so as to form a unified knowledge structure through semantic merging and concept aggregation, and store it in a graph database. S204. Knowledge Graph Storage and Indexing: In the integrated knowledge system, multiple semantic relationships are constructed. Among them, multiple semantic relationships constitute the semantic skeleton of the semantic knowledge structure in the field of carbon emission factors, forming a dynamic knowledge grid. S205. Knowledge Reasoning and Completion Mechanism: Based on the knowledge graph, rule-based reasoning and probabilistic reasoning mechanisms are integrated to mine implicit relationships and automatically complete knowledge, thereby constructing a semantic knowledge structure for the carbon emission factor domain. Among them, rule-based reasoning is used to deduce causal chains according to predefined logical rules; and probabilistic reasoning mechanisms are used to infer unknown relationships, identify potential associations, and dynamically update them.

[0031] In this embodiment, S2 is implemented using a knowledge modeling and reasoning layer.

[0032] The knowledge modeling and reasoning layer is the core of this invention. It is responsible for constructing a semantic knowledge structure for carbon emission factors throughout the entire engineering construction process, enabling semantic modeling, logical connections, and knowledge reasoning among multiple sources of elements such as engineering activities, energy types, equipment, materials, carbon emission factors, and low-carbon technologies. This layer establishes a semantic knowledge structure for carbon emission factors throughout the entire engineering construction process using knowledge graph technology, enabling intelligent completion, semantic reasoning, and causal tracing of carbon emission factors, providing semantic support and a logical foundation for subsequent dynamic calculations and intelligent analysis. Specifically, it includes the following steps: This invention constructs a domain ontology and defines semantics, establishing a semantic knowledge structure covering the entire process of engineering construction for carbon emission factors. It defines core entity categories and their attributes, primarily including engineering activities (design, construction, transportation, operation, etc.), energy types (electricity, diesel, natural gas, etc.), materials (steel, cement, concrete, etc.), equipment types, carbon emission factors, carbon emission sources, and low-carbon technologies. Based on this, the invention uses an ontology description language (OWL) to define the hierarchical relationships and semantic constraints between entities, clarifying logical paths such as "consumption—generation—control—optimization," thus forming a semantic knowledge structure with reasoning capabilities. This invention employs knowledge extraction and fusion based on Natural Language Processing (NLP) and relation extraction techniques to extract carbon emission-related knowledge elements from standards, industry guidelines, databases, and historical engineering projects. Through named entity recognition, attribute extraction, and triple generation, the knowledge is transformed into a structured form, such as: (Engineering Activity → Consumption → Energy Type), (Energy Type → Generation → Emission Factor), (Emission Factor → Controlled by → Low-Carbon Technology), etc. To address the heterogeneity issue caused by diverse data sources, this invention adopts a semantic fusion method based on a rule-based template model to achieve semantic alignment and consistent expression among knowledge sources. To address the naming differences, varying granularities, and heterogeneous units among knowledge sources, this invention employs a semantic similarity calculation method based on word embedding and graph embedding to achieve entity-level and concept-level alignment, eliminating semantic conflicts. A unified knowledge structure is formed through semantic merging and concept aggregation, and persistently stored in a graph database (Neo4j), thereby constructing a knowledge network with multi-layered semantic relationships. In the knowledge graph storage and indexing, within the integrated knowledge system, this invention focuses on constructing the following five types of semantic relationships: Activity-Factor Relationship: Characterizes the direct impact of engineering activities on carbon emission factors; Factor-Energy: Describes the relationship between energy consumption and carbon emission factors in terms of energy conversion and emissions; Energy–Equipment: Reveals the dependency between energy type and equipment operating characteristics; Factor-Technology: This refers to the effect of low-carbon technologies on reducing, controlling, or replacing emission factors. Causal-Temporal relationship: reflects the evolution logic and causal chain of carbon emission factors across time and process stages.

[0033] The above relationships constitute the core semantic framework of the system, forming a dynamic knowledge grid of "engineering activities - energy consumption - emission factors - low-carbon technologies", which can support multi-dimensional queries, semantic associations and path reasoning; This invention integrates rule-based reasoning and probabilistic reasoning mechanisms on the basis of knowledge graphs to achieve implicit relationship mining and automatic knowledge completion. On the one hand, based on rule-based reasoning, causal chain deduction is performed according to predefined logical rules, such as: "If diesel machinery is used during the construction phase → high emission factor → electric drive equipment is recommended". On the other hand, a probabilistic graphical model (Bayesian network) is used to infer unknown or uncertain relationships, realizing potential association identification and dynamic updates. Simultaneously, the path reasoning algorithm can trace the formation process and impact path of any carbon emission factor node, achieving multi-level tracing of emission sources.

[0034] Through the aforementioned modeling and reasoning mechanisms, the knowledge modeling and reasoning layer constructs a multi-dimensional, evolvable, and interpretable carbon emission knowledge network (i.e., a semantic knowledge structure in the domain of carbon emission factors). This carbon emission knowledge network not only realizes the semantic association between carbon emission factors and engineering behavior, energy consumption, equipment operation, and low-carbon technologies, but also dynamically infers causal chains and technology optimization paths, thereby providing a knowledge foundation and logical support for dynamic calculation, carbon traceability, and low-carbon decision recommendation.

[0035] S3. Based on the semantic knowledge structure of the carbon emission factor domain, the carbon emission factor is dynamically calculated and its weights are corrected. The implementation method is as follows: S301. Construct a carbon emission calculation formula based on the semantic knowledge structure of the carbon emission factor domain; S302. Within the framework of the carbon emission calculation formula, the carbon emission factor is dynamically corrected based on multidimensional semantic relationships to generate a dynamic emission factor. S303. A multi-source data fusion mechanism is introduced, using dynamic emission factors as prior information. Bayesian methods are used to update the prior information to weight and correct the carbon emission factors, generating dynamic emission coefficients. These dynamic emission coefficients are then fed back into the carbon emission formula in S301 for real-time calculation of engineering carbon emissions. The implementation method is as follows: A multi-source data fusion mechanism is introduced, using dynamic emission factors as prior information and updating the prior information using Bayesian methods. Based on the update results, the carbon emission factors are weighted and corrected by comprehensively utilizing project ledgers, energy bills, sensor monitoring data, and external database information. When weighting and correcting, new measured data are treated as posterior information when they appear, and the distribution of prior carbon emission factors is updated to obtain the dynamic carbon emission coefficient that best meets the engineering conditions. S304. Perform trend identification and anomaly detection on carbon emission data. Based on the detection results, locate potential causes by combining semantic relationships in the knowledge graph. Based on the location results, perform self-correction processing using dynamic emission coefficients.

[0036] In this embodiment, the correction result generated in S302 (dynamic emission factor EF′: semantic layer correction result) is used as prior information in S303 and further updated by Bayesian algorithm to generate the final dynamic emission coefficient EF* (fusion layer correction result). This dynamic emission coefficient EF* is then directly fed back into the carbon emission calculation formula E = A × EF × (1 – η) in S301 for real-time calculation of engineering carbon emissions. Subsequently, it is continuously verified and optimized in the anomaly detection and self-correction stage of S304, forming a dynamic and cyclical adaptive calculation system. That is, S301: Construct the formula E = A × EF × (1 – η) η); S302: Correct emission factors based on semantic relationships → obtain EF′; S303: Fusion of multi-source data and Bayesian weighted update → obtain EF*; S304: Calculate E(A, EF*, η), where EF′ represents the dynamic emission factor and EF* represents the dynamic emission coefficient.

[0037] In this embodiment, the measured data refers to the actual operating data directly collected at the engineering site through monitoring equipment or energy consumption sensing systems, including electricity meters, fuel flow meters, water meters, carbon emission sensors, flue gas detectors, environmental monitoring devices, construction machinery, generators, transportation equipment, etc.

[0038] In this embodiment, S3 is implemented using a dynamic calculation and analysis layer. The dynamic calculation and analysis layer performs dynamic calculation and weight correction of the carbon emission factor, mainly including: Constructing a carbon emission calculation formula: ; Where: E represents carbon emissions (unit: tCO2-e), A represents activity data, i.e. the actual amount of energy, fuel or materials used in a certain engineering activity, EF represents emission factor, which is derived by comprehensive reasoning from semantic relationships such as energy type, process conditions, and equipment characteristics associated in the knowledge graph, η (Eta) represents emission reduction efficiency or carbon capture rate, i.e. the emission reduction ratio achieved through low-carbon technologies, energy-saving equipment or carbon capture measures, and (1 - η) represents the effective emission ratio that was not reduced, which is used to correct the actual emissions of activity data after the implementation of emission reduction measures.

[0039] The physical meaning of the carbon emission calculation formula is: carbon emissions equal the theoretical emissions generated by the activity multiplied by the actual emission rate. Taking into account factors such as carbon capture and energy-saving retrofits, it can dynamically reflect the real emission levels under different stages and technological conditions. This invention achieves a quantitative description of the effectiveness of carbon emission reduction measures by dynamically estimating the emission reduction rate η.

[0040] In this embodiment, the emission reduction rate η is not a statically set parameter, but is dynamically inferred and calculated based on multidimensional relationships in the knowledge graph. Its value mainly comes from the following three categories: 1. Calculation based on low-carbon technology parameters: A semantic chain of "low-carbon technology - energy consumption reduction rate - carbon capture efficiency" is established in the knowledge modeling layer. When a project adopts a specific energy-saving or carbon capture technology (waste heat recovery system, renewable energy substitution, carbon capture device, etc.), the emission reduction coefficient corresponding to that technology is retrieved through the knowledge graph, and the calculation formula is as follows: ; in, This represents the baseline emissions without the adoption of low-carbon technologies. This represents the measured emissions or model predictions after adopting low-carbon technologies. This result is used to correct the emission reduction rate η in the dynamic calculation model.

[0041] 2. Estimation based on measured energy efficiency and monitoring data: This invention collects real-time energy efficiency change data by connecting to on-site energy consumption monitoring equipment (electricity meters, fuel flow meters, carbon emission sensors, etc.), and dynamically estimates the emission reduction rate η by comparing the energy consumption levels before and after the technical upgrade. ; in, This indicates the theoretical emission reduction percentage achieved by the adopted low-carbon technologies or carbon capture technologies. This indicates the emission reduction efficiency obtained based on real-time monitoring data from the engineering site. This represents the emission reduction rate obtained through knowledge graph reasoning and machine learning prediction models. This represents the theoretical activity level under conditions without energy-saving measures. This represents the current measured activity level. If the energy consumption is lower than the theoretical energy consumption, it indicates that there is an energy-saving effect, and the value of the emission reduction rate η will increase accordingly.

[0042] 3. Comprehensive Prediction Based on Knowledge Reasoning and Historical Samples: When complete measured data is lacking, the system utilizes the semantic reasoning function of the knowledge graph and historical project samples to estimate the value of parameter η. Finally, the comprehensive emission reduction rate is obtained by weighted fusion of three sources (technology coefficients, monitoring data, and inference predictions). ; in, , and This indicates that the weight coefficients determined through Bayesian learning satisfy... .

[0043] In this embodiment, the carbon emission calculation formula is applied dynamically. In actual operation, when new engineering data or monitoring information is received, the following calculation logic is automatically executed: By retrieving relevant nodes from the knowledge graph, the energy type, equipment type, and low-carbon technology involved in this phase of the project are determined; based on the "technology-energy consumption-emission reduction rate" relationship between nodes, the corresponding emission reduction rate is deduced. The value; the dynamically calculated emission reduction rate Input into carbon emission calculation formula In the middle; output the corrected carbon emissions E in real time, and include the calculation path (including emission reduction rate). Source and parameters The correlation relationships are recorded in the graph database to support subsequent source tracing analysis. Therefore, the emission reduction rate η plays a "dual adjustment" role in the model: vertically, it reflects the dynamic effect of carbon capture and energy-saving technologies; horizontally, it adjusts the emission differences at different stages and in different regions.

[0044] In this embodiment, the dynamic correction mechanism for carbon emission factors, within the framework of the carbon emission calculation formula, utilizes the multi-dimensional semantic relationships of "engineering activities—energy types—equipment—emission factors" in the knowledge graph to achieve dynamic correction of carbon emission factors. Specifically, when input data (such as energy structure, construction technology, or equipment operating parameters) changes, the system automatically identifies the affected node relationships through an inference mechanism and updates the corresponding emission factor EF value in real time.

[0045] In this embodiment, to improve the robustness and timeliness of the calculation results, this invention introduces a multi-source data fusion mechanism, comprehensively utilizing project ledgers, energy bills, sensor monitoring data, and external database information to perform weighted correction of carbon emission factors. The correction algorithm employs a Bayesian Update Model to achieve dynamic probabilistic correction: when new measured data appears, the system treats it as posterior information and updates the prior emission factor distribution, thereby obtaining the dynamic emission coefficient that best matches the engineering conditions. This method can effectively reduce the bias caused by a single data source, enabling the carbon emission factor to have adaptive adjustment capabilities in both spatial and temporal dimensions.

[0046] In this embodiment, anomaly identification and model self-correction are performed. This invention utilizes time series analysis to identify trends and detect anomalies in emission data. When carbon emission factors or calculation results deviate from historical patterns or expected ranges, an early warning mechanism is automatically triggered. The mechanism then uses semantic relationships from a knowledge graph to pinpoint potential causes (such as equipment failure, changes in energy structure, or missing data). Subsequently, a model self-correction strategy is executed to restore the model's stability and computational accuracy.

[0047] S4. Visualize and analyze the dynamic calculation results and correction results to complete the dynamic calculation and source analysis of the carbon emission factors of the project.

[0048] In this embodiment, S4 is implemented using a visualization and service layer. This layer provides a dynamic visualization analysis of carbon emission factors and project carbon emissions through a web-based interface, primarily including: Carbon emission monitoring view: Displays real-time carbon emission data for each stage of the project and its subsystems; Low-carbon technology recommendation interface: Based on knowledge reasoning results and project characteristics, it recommends matching energy-saving and carbon-reduction technology solutions.

[0049] In summary, the data acquisition and fusion layer, knowledge modeling and reasoning layer, dynamic calculation and analysis layer, and visualization and service layer of this invention work together to form a closed-loop architecture of "data-driven - knowledge association - intelligent reasoning - visual decision-making". This invention realizes intelligent calculation and traceability analysis of carbon emission factors through knowledge graph modeling, graph database storage, multi-source data fusion and dynamic calculation algorithms.

[0050] Example 2 like Figure 2 As shown, this invention provides a dynamic calculation and source tracing analysis system for engineering carbon emission factors, used to execute the dynamic calculation and source tracing analysis method for engineering carbon emission factors described in Example 1, including: The data acquisition and fusion layer is used to acquire multi-source heterogeneous data throughout the entire engineering construction process, and to preprocess the multi-source heterogeneous data to form a dataset that supports knowledge modeling and computational analysis. The knowledge modeling and reasoning layer is used to construct semantic knowledge structures in the field of carbon emission factors based on datasets; The dynamic calculation and analysis layer is used to dynamically calculate carbon emission factors and correct their weights based on the semantic knowledge structure of the carbon emission factor domain. The visualization and service layer is used to visualize and analyze the dynamic calculation results and correction results, and to complete the dynamic calculation and source analysis of the carbon emission factors of the project.

[0051] In this embodiment, the engineering carbon emission factor dynamic calculation and source tracing analysis system, in order to realize the principle and beneficial effects of the engineering carbon emission factor dynamic calculation and source tracing analysis method, includes hardware structures and / or software modules corresponding to each function. Those skilled in the art should readily recognize that, in conjunction with the illustrative units and algorithm steps described in the embodiments disclosed in this invention, the present invention can be implemented in hardware and / or a combination of hardware and computer software. Whether a function is executed by hardware or computer software depends on the specific application and design constraints of the technical solution. Different methods can be used to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

Claims

1. A method for dynamic calculation and source tracing analysis of engineering carbon emission factors, characterized in that, Includes the following steps: S1. Acquire multi-source heterogeneous data throughout the entire engineering construction process, and preprocess the multi-source heterogeneous data to form a dataset that supports knowledge modeling and computational analysis; S2. Based on the dataset, a semantic knowledge structure for carbon emission factors in the entire process of engineering construction is constructed using knowledge graph technology. The semantic knowledge structure for carbon emission factors is used for intelligent completion, semantic reasoning and causal tracing of carbon emission factors. S3. Based on the semantic knowledge structure of the carbon emission factor domain, the carbon emission factor is dynamically calculated and the weights are corrected. S4. Visualize and analyze the dynamic calculation results and correction results to complete the dynamic calculation and source analysis of the carbon emission factors of the project.

2. The method for dynamic calculation and source tracing analysis of engineering carbon emission factors according to claim 1, characterized in that, S1 includes the following steps: S101. Acquire multi-source heterogeneous data throughout the entire engineering construction process and preprocess the multi-source heterogeneous data. The multi-source heterogeneous data includes basic engineering data, material and energy data, monitoring and environmental data, and standards and specifications data. S102. Write the preprocessed multi-source heterogeneous data into a graph database and a time-series database to form a dataset that supports knowledge modeling and computational analysis.

3. The method for dynamic calculation and source tracing analysis of engineering carbon emission factors according to claim 1, characterized in that, S2 includes the following steps: S201. Domain ontology construction and semantic definition: Based on the dataset, define entity categories and their attributes, and use an ontology description language to define the hierarchical relationships and semantic constraints between entities, so as to model the domain ontology model covering the entire life cycle of engineering construction. S202 Knowledge Extraction and Fusion: Extract knowledge elements related to carbon emissions, and based on the semantic definition results, transform the knowledge into a structured form through named entity recognition, attribute extraction and triple generation, and process the semantics between knowledge to align and express them in a consistent manner; S203. Semantic Alignment and Integration: Based on the fusion results, semantic similarity calculation methods based on word embedding and graph embedding are used to perform entity-level and probability-level alignment processing, so as to form a unified knowledge structure through semantic merging and concept aggregation, and store it in a graph database. S204. Knowledge Graph Storage and Indexing: In the integrated knowledge system, multiple semantic relationships are constructed. Among them, multiple semantic relationships constitute the semantic skeleton of the semantic knowledge structure in the field of carbon emission factors, forming a dynamic knowledge grid. S205. Knowledge Reasoning and Completion Mechanism: Based on the knowledge graph, rule-based reasoning and probabilistic reasoning mechanisms are integrated to mine implicit relationships and automatically complete knowledge, thereby constructing a semantic knowledge structure for the carbon emission factor domain. Among them, rule-based reasoning is used to deduce causal chains according to predefined logical rules; and probabilistic reasoning mechanisms are used to infer unknown relationships, identify potential associations, and dynamically update them.

4. The method for dynamic calculation and source tracing analysis of engineering carbon emission factors according to claim 3, characterized in that, The multiple semantic relationships include: Activity-factor relationship: characterizing the direct impact of engineering activities on carbon emission factors; Factor-Energy Relationship: Describes the correspondence between energy consumption and carbon emission factors in terms of energy conversion and emissions; Energy-equipment relationship: revealing the dependency between energy type and equipment operating characteristics; Factor-technology relationship: indicates the effect of low-carbon technologies on the reduction, control, or substitution of carbon emission factors; Causal-temporal relationship: reflects the evolution logic and causal chain of carbon emission factors across time and process stages.

5. The method for dynamic calculation and source tracing analysis of engineering carbon emission factors according to claim 1, characterized in that, S3 includes the following steps: S301. Construct a carbon emission calculation formula based on the semantic knowledge structure of the carbon emission factor domain; S302. Within the framework of the carbon emission calculation formula, the carbon emission factor is dynamically corrected based on multidimensional semantic relationships to generate a dynamic emission factor. S303. Introduce a multi-source data fusion mechanism, use dynamic emission factors as prior information, use Bayesian methods to update the prior information, and use weighted correction of carbon emission factors to generate dynamic emission coefficients. The dynamic emission coefficients are fed back into the carbon emission formula in S301 for real-time calculation of engineering carbon emissions. S304. Perform trend identification and anomaly detection on carbon emission data. Based on the detection results, locate potential causes by combining semantic relationships in the knowledge graph. Based on the location results, perform self-correction processing using dynamic emission coefficients.

6. The method for dynamic calculation and source tracing analysis of engineering carbon emission factors according to claim 5, characterized in that, The carbon emission calculation formula is as follows: ； ； ； ； ； in, Indicates carbon emissions. Indicates activity level data, Indicates carbon emission factor, Indicates the emission reduction rate. This indicates the effective emission percentage that was not reduced. , and All of these represent weight coefficients determined through Bayesian learning. This represents the theoretical emission reduction percentage. This indicates the emission reduction efficiency obtained based on real-time monitoring data from the engineering site. This represents the emission reduction percentage predicted through knowledge graph reasoning and machine learning. This represents the baseline emissions without the adoption of low-carbon technologies. This represents the measured emissions after adopting low-carbon technologies. This represents the theoretical activity level under conditions without energy-saving measures. This indicates the current measured activity level.

7. The method for dynamic calculation and source tracing analysis of engineering carbon emission factors according to claim 5, characterized in that, Specifically, S303 is: A multi-source data fusion mechanism is introduced, using dynamic emission factors as prior information, and Bayesian methods are used to update the prior information. Based on the updated results, the carbon emission factors are weighted and corrected by comprehensively utilizing project ledgers, energy bills, sensor monitoring data, and external database information. When weighting and correcting, new measured data are treated as posterior information when they appear, and the prior carbon emission factor distribution is updated to obtain the dynamic carbon emission coefficient that best meets the engineering conditions.

8. A system for dynamic calculation and source tracing analysis of engineering carbon emission factors, used to execute the method for dynamic calculation and source tracing analysis of engineering carbon emission factors according to any one of claims 1-7, characterized in that, include: The data acquisition and fusion layer is used to acquire multi-source heterogeneous data throughout the entire engineering construction process, and to preprocess the multi-source heterogeneous data to form a dataset that supports knowledge modeling and computational analysis. The knowledge modeling and reasoning layer is used to construct semantic knowledge structures in the field of carbon emission factors based on datasets; The dynamic calculation and analysis layer is used to dynamically calculate carbon emission factors and correct their weights based on the semantic knowledge structure of the carbon emission factor domain. The visualization and service layer is used to visualize and analyze the dynamic calculation results and correction results, and to complete the dynamic calculation and source analysis of the carbon emission factors of the project.

Images