Carbon emission intelligent analysis method and system based on knowledge graph

Abstract

The invention relates to the technical field of carbon emission intelligent analysis, and particularly discloses a carbon emission intelligent analysis method and system based on a knowledge graph, and the method specifically comprises the following steps: 1, constructing a knowledge graph about carbon emission; 2, analyzing the constructed knowledge graph, judging whether abnormity exists or not, and generating an abnormal signal; and step 3, based on the abnormal signal, obtaining an entity group 2 corresponding to the unreasonable entity 1 in the knowledge graph, and adding a new triple to improve the knowledge graph. The method comprises the following steps: firstly, acquiring data, establishing a knowledge graph, analyzing triples in the knowledge graph, acquiring the number of unreasonable entities 1, calculating the ratio of the unreasonable entities 1, comparing the ratio of the unreasonable entities 1 with a threshold value, and judging whether the established knowledge graph is abnormal or not; therefore, abnormal conditions in the knowledge graph can be found in time, and powerful and accurate quantity support is provided for subsequent carbon emission analysis.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent analysis of carbon emissions, and in particular to a method and system for intelligent analysis of carbon emissions based on a knowledge graph. Background Art

[0002] As global climate change becomes increasingly severe, the monitoring and management of carbon emissions has become a focus of widespread international attention. Traditional carbon emission analysis methods mainly rely on a single data source, and have low integration and analysis capabilities for data from different sources. At the same time, existing carbon emission analysis methods are often not in-depth enough in terms of time dimension, which in turn affects the formulation of emission reduction strategies. In view of this, we proposed a carbon emissions intelligent analysis method and system based on knowledge graph. Summary of the invention

[0003] The purpose of the present invention is to provide a carbon emission intelligent analysis method and system based on knowledge graph to solve the technical problems in the above background.

[0004] The purpose of the present invention can be achieved by the following technical solutions: In a first aspect, the present invention provides a carbon emission intelligent analysis method based on a knowledge graph, which specifically comprises the following steps: Step 1: Build a knowledge graph about carbon emissions; Step 2: Analyze the constructed knowledge graph to determine whether there are any anomalies and generate anomaly signals; Step 3: Based on the abnormal signal, obtain the entity group 2 corresponding to the unreasonable entity 1 in the knowledge graph, and add new triples to improve the knowledge graph; Step 4: Analyze the trend of carbon emissions based on all triples; Among them, the changing trend includes stable trend and unstable trend; Step 5: Further analysis of carbon emissions based on the marked unstable trend.

[0005] As a further solution of the present invention: the process of generating an abnormal signal: Obtaining an unreasonable entity 1 ratio value, and comparing the unreasonable entity 1 ratio value with an unreasonable entity 1 ratio threshold; If the unreasonable entity 1 ratio value is greater than the unreasonable entity 1 ratio threshold, an abnormal signal is generated.

[0006] As a further solution of the present invention: the process of obtaining the proportion of the unreasonable entity 1 is as follows: Count the number of entity 2s in each triple; Compare the number of entity 2 in the triples with the expected number of entity 2 respectively; If the number of entity 2 is greater than the expected number of entity 2, it is marked as unreasonable entity 1; Count the number of unreasonable entities 1, and ratio the number of unreasonable entities 1 to the total number of entities 1 in the knowledge graph to obtain the proportion of unreasonable entities 1.

[0007] As a further solution of the present invention: the process of step three is: Based on the generated abnormal signal, the unreasonable entity 1 and the corresponding entity 2 group in the knowledge graph are obtained; For each entity 1, obtain the time period and record timestamp corresponding to the entity 2 value in the entity 2 group; Sort the time periods in ascending order and number them, calculate the difference between adjacent time periods, and obtain the time period difference; The time period difference is compared with the time period difference threshold, and if the time period difference is greater than the time period difference threshold, a large deviation signal is generated; Get the number of the time period corresponding to the generation of the large deviation signal and classify the time period groups; Sum and average the time periods in each group to get the time period mean, and add a new triplet based on the time period mean and the record timestamp; Basic triplet: company name - carbon emissions - carbon emissions value; Time period triplet: enterprise name - record timestamp - record time value; Classification triplet: company name - time period average - including recording time value.

[0008] As a further solution of the present invention: the process of step 4 is: For the same enterprise name, extract the record timestamps contained in the mean of the same time period and obtain the carbon emissions corresponding to these record timestamps; All carbon emissions are summed up and averaged to obtain the mean carbon emissions, all carbon emissions are calculated with the mean carbon emissions, the difference is taken as the absolute value of the difference to obtain the carbon emissions deviation value, all carbon emissions deviation values ​​are summed up and averaged to obtain the carbon emissions deviation mean; Compare the carbon emission deviation value with the carbon emission deviation mean; If the carbon emission deviation value is less than or equal to the carbon emission deviation mean value, a stable value is generated; If the carbon emission deviation value is greater than the carbon emission deviation mean, an unstable value is generated; Count the number of stable values ​​and the number of unstable values, and sum them up to get the total number. Ratio the number of stable values ​​to the total number to get the proportion of stable values. comparing the stable value ratio with a stable value ratio threshold; If the stable value ratio is greater than the stable value ratio threshold, it is marked as a stable trend; If the stable value ratio is less than or equal to the stable value ratio threshold, it is marked as an unstable trend.

[0009] As a further solution of the present invention: the process of step 5 is: Obtain the proportion of growth sub-curve segments and the mean value of growth slope deviation ratio; Substitute into the formula , the analysis and judgment value FX is calculated, where ZB represents the proportion of the growth sub-curve segment, PC represents the mean value of the growth slope deviation ratio, and a1 and a2 are preset proportional coefficients.

[0010] As a further solution of the present invention: the process of obtaining the proportion of the growth sub-curve segment is: With time as the horizontal axis and carbon emissions as the vertical axis, a carbon emissions variation curve is drawn in a plane coordinate system; Mark the curve between adjacent coordinate points in the carbon emission change curve as a sub-curve segment; Calculate the slope of each sub-curve segment, count the number of sub-curve segments with positive slopes, and mark them as growing sub-curve segments; It should be noted that the slope of the sub-curve segment is calculated from two adjacent coordinate points; The ratio of the number of growing sub-curve segments to the total number of sub-curve segments is processed to obtain the proportion of growing sub-curve segments.

[0011] As a further solution of the present invention: the process of obtaining the mean value of the growth slope deviation ratio is: Extract the slope of the growth sub-curve segment, calculate the difference between the slope of the growth sub-curve segment and the slope limit, take the absolute value of the difference and ratio it with the slope limit to obtain the growth slope deviation ratio, sum and average all the growth slope deviation ratios to obtain the mean of the growth slope deviation ratios.

[0012] As a further solution of the present invention: the process of step 5 also includes: Obtaining an analysis judgment value, comparing the analysis judgment value with an analysis judgment threshold, If the analysis judgment value is less than or equal to the analysis judgment threshold, a no-analysis signal is generated; If the analysis judgment value is greater than the analysis judgment threshold, an analysis signal is generated.

[0013] In a second aspect, the present invention provides a carbon emission intelligent analysis system based on a knowledge graph, the system comprising: Knowledge graph construction module: construct a knowledge graph about carbon emissions; Knowledge graph anomaly judgment module: analyzes the constructed knowledge graph, determines whether there are anomalies, and generates anomaly signals; Knowledge graph optimization module: Based on abnormal signals, obtain entity 2 groups corresponding to unreasonable entity 1 in the knowledge graph, and add new triples to improve the knowledge graph; Carbon emission trend judgment module: analyzes the trend of carbon emission changes based on all triples; Among them, the changing trend includes stable trend and unstable trend; Carbon emissions analysis module: Further analysis of carbon emissions based on marked unstable trends.

[0014] Beneficial effects of the present invention: (1) The present invention first acquires data and establishes a knowledge graph, analyzes the triples in the knowledge graph, obtains the number of unreasonable entities 1, calculates the proportion of unreasonable entities 1, compares the proportion of unreasonable entities 1 with a threshold, and determines whether the constructed knowledge graph has anomalies, thereby timely discovering anomalies in the knowledge graph and providing powerful and accurate quantitative support for subsequent carbon emission analysis; (2) The present invention obtains the entity 2 group corresponding to the unreasonable entity 1, and obtains the time period corresponding to the value of entity 2 and the recording timestamp, analyzes the time period corresponding to the value of entity 2, divides the time periods with similar time periods into a group, and sums and averages the time periods of the same group to obtain the time period average, and adds a new triplet according to the time period average and the recording timestamp, so as to more accurately reflect the carbon emissions of the enterprise in different time periods, which helps the enterprise to more accurately analyze the carbon emission level, formulate more effective emission reduction strategies, and monitor the emission reduction effect; (3) The present invention obtains the record timestamps contained in the average of the same time period, obtains the carbon emissions corresponding to these record timestamps, and performs trend analysis on these carbon emissions. Thus, the knowledge graph can be used to integrate carbon emission-related data. By calculating the mean and analyzing the deviation of carbon emissions, the changing trend of carbon emissions can be accurately determined, providing data support for enterprises to formulate emission reduction plans; (4) The present invention counts the sub-curve segments of the growth trend in the carbon emission change curve, analyzes the sub-curve segments of the carbon emission growth trend, calculates the proportion of the growth sub-curve segments and the average value of the growth slope deviation ratio, calculates and analyzes the judgment based on the proportion of the growth sub-curve segments and the average value of the growth slope deviation ratio, and compares them with the threshold value to generate an analysis signal or a non-analysis signal. Based on the non-analysis signal, the carbon emissions are continuously monitored, but no in-depth analysis or intervention measures are taken immediately. Based on the analysis signal, a detailed carbon emission analysis report is generated for the corresponding enterprise, and targeted emission reduction suggestions or measures are put forward to help the enterprise optimize the carbon emission management strategy, thereby providing a quantitative carbon emission growth trend analysis, so that the enterprise can more accurately grasp its own carbon emission situation and take corresponding management measures in time. At the same time, by providing specific emission reduction suggestions for the enterprise, it can also help the enterprise achieve more environmentally friendly and sustainable development, which not only improves the accuracy and efficiency of carbon emission analysis, but also provides the enterprise with a scientific and feasible emission reduction strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention will be further described below in conjunction with the accompanying drawings.

[0016] Figure 1 It is a flowchart of a carbon emission intelligent analysis method based on knowledge graph of the present invention; Figure 2 It is a flowchart of the process of obtaining analysis and judgment values ​​in a carbon emission intelligent analysis method based on knowledge graph of the present invention; Figure 3 It is a block diagram of a carbon emission intelligent analysis system based on knowledge graph in the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] Embodiment 1: See also Figure 1 As shown, a carbon emission intelligent analysis method based on a knowledge graph according to an embodiment of the present invention specifically includes the following steps: Step 1: Build a knowledge graph about carbon emissions; In some embodiments, carbon emission data is obtained, wherein the carbon emission data includes but is not limited to: carbon emission amount, carbon emission intensity, and carbon emission trend; It should be noted that the sources of carbon emission data include but are not limited to: government reports, international agency data, professional research, and corporate reports; Preprocessing the acquired carbon emission data, wherein the preprocessing process includes but is not limited to: cleaning, removing duplicate, erroneous or invalid data; According to the characteristics of the carbon emission field, the entity types in the map are defined, including but not limited to: enterprises, regions, industries, energy types, and emission sources; Define the carbon emission relationships between entities, including but not limited to: emission volume, emission trend, emission intensity, carbon footprint, and energy efficiency; Extracting knowledge related to carbon emissions from raw data, where the extraction methods include but are not limited to: text parsing and natural language processing; The extracted knowledge is stored in a graph database, a data layer is built in the graph database, and the extracted knowledge is stored in the form of triples (entity 1-relationship-entity 2) or (entity-attribute-value); It should be noted that the data layer should contain data of all entities and relationships; Build a model layer on top of the data layer. The model layer should contain refined knowledge ontology and model information to standardize and manage entities, relationships, and entity types and attributes. Step 2: Analyze the constructed knowledge graph to determine whether there are any anomalies; In some embodiments, based on the constructed knowledge graph, traverse the triples in the knowledge graph; In a possible embodiment, taking entity 1 - relation - entity 2 triples as an example, the number of entity 2s in each triple is counted; Among them, the entity 1-relationship-entity 2 triple can be set as: company name-carbon emissions-carbon emissions value; For example, if entity 1 is enterprise A and the relationship is carbon emissions, if the matching entity 2 is 100 tons, 300 tons, 500 tons, or 1000 tons, then the number of entity 2 in the triples counted is 4; if the matching entity 2 is 100 tons, then the number of entity 2 in the triples counted is 1; Compare the number of entity 2 in the triples with the expected number of entity 2 respectively; If the number of entity 2 is less than or equal to the expected number of entity 2, it is marked as reasonable entity 1; if the number of entity 2 is greater than the expected number of entity 2, it is marked as unreasonable entity 1; It needs to be explained that unreasonable entity 1 means that when the corresponding entity 2 exceeds the expected number, it is necessary to construct a new triplet knowledge extraction for entity 2; Count the number of unreasonable entities 1, and perform a ratio process on the number of unreasonable entities 1 and the total number of entities 1 in the knowledge graph to obtain the ratio of unreasonable entities 1; Compare the unreasonable entity 1 ratio value with the unreasonable entity 1 ratio threshold, wherein the unreasonable entity 1 ratio threshold is set by a person skilled in the art based on experience and multiple historical experiments; If the unreasonable entity 1 ratio is less than or equal to the unreasonable entity 1 ratio threshold, the knowledge graph contains fewer unreasonable entities 1, and a normal signal is generated; If the unreasonable entity 1 ratio is greater than the unreasonable entity 1 ratio threshold, the knowledge graph contains more unreasonable entities 1, and an abnormal signal is generated; The technical solution of the embodiment of the present invention is mainly as follows: first, a knowledge graph is established by acquiring data, and then the triples in the knowledge graph are analyzed to obtain the number of unreasonable entities 1, calculate the proportion of unreasonable entities 1, and compare the proportion of unreasonable entities 1 with the threshold value to determine whether there are anomalies in the constructed knowledge graph, so as to timely discover anomalies in the knowledge graph and provide strong and accurate quantitative support for subsequent carbon emission analysis.

[0019] Embodiment 2: Based on Example 1, please refer to Figure 1 As shown, the carbon emission intelligent analysis method based on knowledge graph described in the embodiment of the present invention further includes the following steps: Step 3: Based on the abnormal signal, obtain the entity group 2 corresponding to the unreasonable entity 1 in the knowledge graph, and add new triples to improve the knowledge graph; In some embodiments, based on the generated abnormal signal, an unreasonable entity 1 and a corresponding entity 2 group in the knowledge graph are obtained; In some possible embodiments, for each entity 1, a time period corresponding to the entity 2 value in the entity 2 group and a recording timestamp are obtained; It should be noted that the time period is in days. For example, the values ​​in entity group 2 may include 100 tons, 300 tons, 500 tons, and 1,000 tons. The time period corresponding to 100 tons may be 30 days, the time period corresponding to 300 tons may also be 30 days, the time period corresponding to 500 tons may be 60 days, and the time period corresponding to 1,000 tons may be 90 days. Sort the time periods in ascending order and number them, calculate the difference between adjacent time periods, and obtain the time period difference; Comparing the time period difference with a time period difference threshold, where the time period difference threshold is set by a person skilled in the art based on experience; If the time period difference is less than or equal to the time period difference threshold, a small deviation signal is generated; if the time period difference is greater than the time period difference threshold, a large deviation signal is generated; Get the number of the time period corresponding to the generation of the large deviation signal and classify the time period groups; For example, if the numbers of the time periods corresponding to the generation of the large deviation signal are 10, 20, and 30, respectively, the time periods are divided into four groups, namely, numbers 1-9, numbers 10-19, numbers 20-29, and the remaining numbers; Sum and average the time periods in each group to get the time period mean, and add a new triplet based on the time period mean and the record timestamp; Basic triplet: company name - carbon emissions - carbon emissions value; Time period triplet: enterprise name - record timestamp - record time value; Classification triples: company name - time period mean - including recording time value; For example, assuming that the total carbon emissions recorded by enterprise A on April 1, 2023 is 100 tons, and the average time period is 30 days, then we can construct the following triples: Basic triplet: Company A – Carbon emissions – 100 tons; Time period triplet: Enterprise A - Carbon emissions recording time value - April 1, 2023; Classification triplet: Company A-30-day carbon emissions - including April 1, 2023; The technical solution of the embodiment of the present invention is mainly: by obtaining the entity group 2 corresponding to the unreasonable entity 1, and obtaining the time period corresponding to the value of entity 2 and the recording timestamp, the time period corresponding to the value of entity 2 is analyzed, and the time periods with similar time periods are divided into a group, and the time periods of the same group are summed and averaged to obtain the time period mean, and new triples are added according to the time period mean and the recording timestamp, so as to more accurately reflect the carbon emissions of the enterprise in different time periods, which helps the enterprise to more accurately analyze the carbon emission level, formulate more effective emission reduction strategies, and monitor the emission reduction effects.

[0020] Embodiment 3: Based on Example 1 and Example 2, please refer to Figure 1 As shown, the carbon emission intelligent analysis method based on knowledge graph described in the embodiment of the present invention further includes the following steps: Step 4: Analyze the trend of carbon emissions based on all triples; Among them, the changing trend includes stable trend and unstable trend; In some embodiments, for the same enterprise name, the record timestamps contained in the mean of the same time period are extracted to obtain the carbon emissions corresponding to these record timestamps; All carbon emissions are summed up and averaged to obtain the mean carbon emissions, all carbon emissions are calculated with the mean carbon emissions, the difference is taken as the absolute value of the difference to obtain the carbon emissions deviation value, all carbon emissions deviation values ​​are summed up and averaged to obtain the carbon emissions deviation mean; Compare the carbon emission deviation value with the carbon emission deviation mean; If the carbon emission deviation value is less than or equal to the carbon emission deviation mean, it means that the deviation between the carbon emission and the carbon emission mean is small, and a stable value is generated; If the carbon emission deviation value is greater than the carbon emission deviation mean, it means that the deviation between the carbon emission and the carbon emission mean is large, generating an unstable value; Count the number of stable values ​​and the number of unstable values, and sum them up to get the total number. Ratio the number of stable values ​​to the total number to get the proportion of stable values. The stable value ratio is compared with the stable value ratio threshold, wherein the stable value ratio threshold is a critical value used to determine whether the change trend of carbon emissions is stable, and is summarized and set by technicians in this field based on multiple historical experimental data; If the proportion of stable values ​​is greater than the threshold of stable value proportion, it means that the carbon emissions corresponding to the mean value of the same time period are relatively stable, which is marked as a stable trend; If the proportion of stable values ​​is less than or equal to the threshold of the proportion of stable values, it means that the carbon emissions corresponding to the mean value of the same time period are unstable and marked as an unstable trend; The technical solution of the embodiment of the present invention is mainly: by obtaining the record timestamps contained in the average of the same time period, obtaining the carbon emissions corresponding to these record timestamps, and performing trend analysis on these carbon emissions, the knowledge graph can be used to integrate carbon emission related data, and the mean calculation and deviation analysis of carbon emissions can be performed to accurately judge the changing trend of carbon emissions, thereby providing support for enterprises to formulate emission reduction plans.

[0021] Embodiment 4: Based on Example 1, Example 2 and Example 3, please refer to Figure 1 and Figure 2 As shown, the carbon emission intelligent analysis method based on knowledge graph described in the embodiment of the present invention further includes the following steps: Step 5: Further analysis of carbon emissions based on the marked unstable trend; In some embodiments, a carbon emission variation curve is plotted in a plane coordinate system with time as the abscissa and carbon emission as the ordinate; Mark the curve between adjacent coordinate points in the carbon emission change curve as a sub-curve segment; Calculate the slope of each sub-curve segment, count the number of sub-curve segments with positive slopes, and mark them as growing sub-curve segments; It should be noted that the slope of the sub-curve segment is calculated from two adjacent coordinate points; The ratio of the number of growing sub-curve segments to the total number of sub-curve segments is processed to obtain the proportion of growing sub-curve segments; Extract the slope of the growth sub-curve segment, calculate the difference between the slope of the growth sub-curve segment and the slope limit, take the absolute value of the difference and perform ratio processing with the slope limit to obtain the growth slope deviation ratio, sum and average all the growth slope deviation ratios to obtain the mean of the growth slope deviation ratio; Among them, the slope limit is set by those skilled in the art based on the summary of multiple historical experimental data; Substitute into the formula , the analysis and judgment value FX is calculated, where ZB represents the proportion of the growth sub-curve segment, PC represents the mean value of the growth slope deviation ratio, a1 and a2 are preset proportional coefficients, where a1 is 2.52 and a2 is 1.48; It should be explained that the meaning reflected by the analysis and judgment value is: the analysis and judgment value is calculated by the proportion of the growth sub-curve segment and the mean value of the growth slope deviation ratio, wherein the proportion of the growth sub-curve segment represents the proportion of the number of growth sub-curve segments in the carbon emission change curve. The larger the proportion of the growth sub-curve segment, the more sub-curve segments with a growth trend, and the more carbon emissions show a growth trend. The larger the mean value of the growth slope deviation ratio, the more sub-curve segments with a larger slope in the growth trend, and the faster the growth rate of carbon emissions. Compare the analysis judgment value with the analysis judgment threshold, wherein the analysis judgment threshold is set by a person skilled in the art based on a summary of multiple historical experimental data; If the analysis judgment value is less than or equal to the analysis judgment threshold, it means that the sub-curve segments of the carbon emission change curve showing an increasing trend are relatively small, and the slope of the growth sub-curve is relatively small, and a non-analysis signal is generated; Based on non-analytical signals, carbon emissions are continuously monitored without immediate in-depth analysis or intervention measures; If the analysis judgment value is greater than the analysis judgment threshold, it means that the carbon emission change curve has more sub-curve segments with an increasing trend and more sub-curve segments with a large slope of the growth sub-curve, and an analysis signal is generated; Based on the analysis signals, a detailed carbon emission analysis report is generated for the corresponding enterprise, and targeted emission reduction suggestions or measures are put forward to help the enterprise optimize its carbon emission management strategy; It should be explained that the number of sub-curve segments of the growth trend is large and the slope is large, which means that carbon emissions are continuing to grow rapidly, thus providing scientific carbon emission management suggestions for enterprises; The technical solution of the embodiment of the present invention is mainly as follows: by counting the sub-curve segments of the growth trend in the carbon emission change curve, and analyzing the sub-curve segments of the carbon emission growth trend, the proportion of the growth sub-curve segments and the mean of the growth slope deviation ratio are calculated, and the analysis and judgment are calculated according to the proportion of the growth sub-curve segments and the mean of the growth slope deviation ratio, and compared with the threshold, an analysis signal or a non-analysis signal is generated, and based on the non-analysis signal, the carbon emissions are continuously monitored, but no in-depth analysis or intervention measures are taken immediately, and based on the analysis signal, a detailed carbon emission analysis report is generated for the corresponding enterprise, and targeted emission reduction suggestions or measures are put forward to help the enterprise optimize the carbon emission management strategy, thereby providing a quantitative carbon emission growth trend analysis, so that the enterprise can more accurately grasp its own carbon emissions and take corresponding management measures in time. At the same time, by providing specific emission reduction suggestions for enterprises, it can also help enterprises achieve more environmentally friendly and sustainable development, which not only improves the accuracy and efficiency of carbon emission analysis, but also provides scientific and feasible emission reduction strategies for enterprises.

[0022] Embodiment 5: Based on Example 1, Example 2 and Example 3, please refer to Figure 3 As shown, a carbon emission intelligent analysis system based on knowledge graph according to an embodiment of the present invention includes: Knowledge graph construction module: construct a knowledge graph about carbon emissions; Knowledge graph anomaly judgment module: analyzes the constructed knowledge graph to determine whether there are any anomalies; Knowledge graph optimization module: Based on abnormal signals, obtain entity 2 groups corresponding to unreasonable entity 1 in the knowledge graph, and add new triples to improve the knowledge graph; Carbon emission trend judgment module: analyzes the trend of carbon emission changes based on all triples; Carbon emissions analysis module: Further analysis of carbon emissions based on marked unstable trends.

[0023] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A carbon emission intelligent analysis method based on knowledge graph, characterized in that: The specific steps include: Step 1: Build a knowledge graph about carbon emissions; Step 2: Analyze the constructed knowledge graph to determine whether there are any anomalies and generate anomaly signals; Step 3: Based on the abnormal signal, obtain the entity group 2 corresponding to the unreasonable entity 1 in the knowledge graph, and add new triples to improve the knowledge graph; Step 4: Analyze the trend of carbon emissions based on all triples; Among them, the changing trend includes stable trend and unstable trend; Step 5: Further analysis of carbon emissions based on the marked unstable trend.

2. According to claim 1, a carbon emission intelligent analysis method based on knowledge graph is characterized in that: The process of generating abnormal signals: Obtaining an unreasonable entity 1 ratio value, and comparing the unreasonable entity 1 ratio value with an unreasonable entity 1 ratio threshold; If the unreasonable entity 1 ratio value is greater than the unreasonable entity 1 ratio threshold, an abnormal signal is generated.

3. According to claim 1, a carbon emission intelligent analysis method based on knowledge graph is characterized in that: The process of obtaining the proportion of unreasonable entity 1 is as follows: Count the number of entity 2s in each triple; Compare the number of entity 2 in the triples with the expected number of entity 2 respectively; If the number of entity 2 is greater than the expected number of entity 2, it is marked as unreasonable entity 1; Count the number of unreasonable entities 1, and ratio the number of unreasonable entities 1 to the total number of entities 1 in the knowledge graph to obtain the proportion of unreasonable entities 1.

4. According to claim 1, a carbon emission intelligent analysis method based on knowledge graph is characterized in that: The process of step three is: Based on the generated abnormal signal, the unreasonable entity 1 and the corresponding entity 2 group in the knowledge graph are obtained; For each entity 1, obtain the time period and record timestamp corresponding to the entity 2 value in the entity 2 group; Sort the time periods in ascending order and number them, calculate the difference between adjacent time periods, and obtain the time period difference; The time period difference is compared with the time period difference threshold, and if the time period difference is greater than the time period difference threshold, a large deviation signal is generated; Get the number of the time period corresponding to the generation of the large deviation signal and classify the time period groups; Sum and average the time periods in each group to get the time period mean, and add a new triplet based on the time period mean and the record timestamp; Basic triplet: company name - carbon emissions - carbon emissions value; Time period triplet: enterprise name - record timestamp - record time value; Classification triplet: company name - time period average - including recording time value.

5. According to claim 1, a carbon emission intelligent analysis method based on knowledge graph is characterized in that: The process of step 4 is as follows: For the same enterprise name, extract the record timestamps contained in the mean of the same time period and obtain the carbon emissions corresponding to these record timestamps; All carbon emissions are summed up and averaged to obtain the mean carbon emissions, all carbon emissions are calculated with the mean carbon emissions, the difference is taken as the absolute value of the difference to obtain the carbon emissions deviation value, all carbon emissions deviation values ​​are summed up and averaged to obtain the carbon emissions deviation mean; Compare the carbon emission deviation value with the carbon emission deviation mean; If the carbon emission deviation value is less than or equal to the carbon emission deviation mean value, a stable value is generated; If the carbon emission deviation value is greater than the carbon emission deviation mean, an unstable value is generated; Count the number of stable values ​​and the number of unstable values, and sum them up to get the total number. Ratio the number of stable values ​​to the total number to get the proportion of stable values. comparing the stable value ratio with a stable value ratio threshold; If the stable value ratio is greater than the stable value ratio threshold, it is marked as a stable trend; If the stable value ratio is less than or equal to the stable value ratio threshold, it is marked as an unstable trend.

6. According to claim 1, a carbon emission intelligent analysis method based on knowledge graph is characterized in that: The process of step five is: Obtain the proportion of growth sub-curve segments and the mean value of growth slope deviation ratio; Substitute into the formula , the analysis and judgment value FX is calculated, where ZB represents the proportion of the growth sub-curve segment, PC represents the mean value of the growth slope deviation ratio, and a1 and a2 are preset proportional coefficients.

7. According to claim 6, a carbon emission intelligent analysis method based on knowledge graph is characterized in that: The process of obtaining the proportion of the growth sub-curve segment is as follows: With time as the horizontal axis and carbon emissions as the vertical axis, a carbon emissions variation curve is drawn in a plane coordinate system; Mark the curve between adjacent coordinate points in the carbon emission change curve as a sub-curve segment; Calculate the slope of each sub-curve segment, count the number of sub-curve segments with positive slopes, and mark them as growing sub-curve segments; The ratio of the number of growing sub-curve segments to the total number of sub-curve segments is processed to obtain the proportion of growing sub-curve segments.

8. The carbon emission intelligent analysis method based on knowledge graph according to claim 7 is characterized in that: The process of obtaining the mean value of the growth slope deviation ratio is as follows: Extract the slope of the growth sub-curve segment, calculate the difference between the slope of the growth sub-curve segment and the slope limit, take the absolute value of the difference and ratio it with the slope limit to obtain the growth slope deviation ratio, sum and average all the growth slope deviation ratios to obtain the mean of the growth slope deviation ratios.

9. The carbon emission intelligent analysis method based on knowledge graph according to claim 6 is characterized in that: The process of step five also includes: Obtaining an analysis judgment value, comparing the analysis judgment value with an analysis judgment threshold, If the analysis judgment value is less than or equal to the analysis judgment threshold, a no-analysis signal is generated; If the analysis judgment value is greater than the analysis judgment threshold, an analysis signal is generated.

10. A carbon emission intelligent analysis system based on knowledge graph, characterized in that: The system is used to execute the method described in any one of claims 1 to 9, and the system comprises: Knowledge graph construction module: construct a knowledge graph about carbon emissions; Knowledge graph anomaly judgment module: analyzes the constructed knowledge graph, determines whether there are anomalies, and generates anomaly signals; Knowledge graph optimization module: Based on abnormal signals, obtain entity 2 groups corresponding to unreasonable entity 1 in the knowledge graph, and add new triples to improve the knowledge graph; Carbon emission trend judgment module: analyzes the trend of carbon emission changes based on all triples; Among them, the changing trend includes stable trend and unstable trend; Carbon emissions analysis module: Further analysis of carbon emissions based on marked unstable trends.

Images