A big data analysis intelligent building carbon emission early warning method

Abstract

This invention relates to a smart building carbon emission early warning method based on big data analysis, belonging to the field of carbon emission early warning. The method includes: finding related buildings and relationships between buildings; classifying similar and related buildings; acquiring real-time and historical carbon emission data of the target building, similar real-time and similar historical carbon emission data, and related real-time and related historical carbon emission data; obtaining a fitted future carbon emission curve for the target building, a fitted future carbon emission curve for similar buildings, and a fitted future carbon emission curve for related buildings; determining a correlation parameter; outputting a numerical anomaly early warning signal when the correlation parameter is below a correlation threshold; correcting the fitted future carbon emission curve for the target building when it is above the correlation threshold; and issuing an early warning signal based on the corrected future carbon emission curve. This application utilizes the real-time carbon emissions of related buildings as a reference to dynamically correct prediction biases, making carbon emission predictions more accurate.

Description

Technical Field

[0001] This invention relates to the field of carbon emission early warning, and in particular to a smart building carbon emission early warning method based on big data analysis. Background Technology

[0002] Currently, using big data technology to accurately acquire building carbon emission data has become one of the key technologies for achieving green transformation in building operations. By utilizing big data analysis and establishing a carbon emission threshold alarm mechanism, the system can trigger multi-level warnings at the initial stage of equipment failure, energy waste, or excessive emissions, enabling maintenance personnel to respond quickly and avoid carbon emission exceedances due to machine malfunctions.

[0003] Existing carbon emission early warning methods integrate multi-source data such as building energy consumption, equipment operation, and environmental parameters, clean and model the data, and identify key influencing factors to construct a carbon emission model. Based on this model and machine learning, they predict carbon emission trends in real time and issue early warnings. Current technologies allow for more accurate and detailed predictions of carbon emissions in different seasons by setting different prediction models for the same building. Furthermore, to more accurately predict carbon emissions from HVAC equipment during seasonal transitions, different models are combined to capture the unique characteristics of different HVAC systems, resulting in more precise carbon emission predictions.

[0004] The aforementioned technologies rely on historical building operation parameters for predicting building carbon emissions. However, the lack of real-time carbon emissions from related buildings makes it difficult to dynamically correct prediction biases, which can lead to difficulty in detecting abnormal fluctuations in carbon emissions. Summary of the Invention

[0005] To utilize real-time carbon emissions from associated buildings as a reference to dynamically correct prediction biases and achieve more accurate carbon emission forecasts, this invention provides a smart building carbon emission early warning method based on big data analysis.

[0006] This invention provides a smart building carbon emission early warning method based on big data analysis, employing the following technical solution:

[0007] A smart building carbon emission early warning method based on big data analysis includes:

[0008] Step 1: In response to the received target building, find the corresponding related buildings and the relationships between buildings according to the preset building carbon emission database;

[0009] Step 2: Based on the relationships between the buildings, classify the associated buildings into similar buildings and related buildings;

[0010] Step 3: Obtain the target real-time carbon emission data and target historical carbon emission data of the target building, obtain the similar real-time carbon emission data and similar historical carbon emission data of the similar buildings, and obtain the relevant real-time carbon emission data and relevant historical carbon emission data of the related buildings;

[0011] Step 4: Obtain the target fitted carbon emission future curve based on the target real-time carbon emission data and the target historical carbon emission data; obtain the similar fitted carbon emission future curve based on the similar real-time carbon emission data and the similar historical carbon emission data; obtain the relevant fitted carbon emission future curve based on the relevant real-time carbon emission data and the relevant historical carbon emission data.

[0012] Step 5: Determine the correlation parameters based on the target fitted future carbon emission curve and the similar fitted future carbon emission curve;

[0013] Step 60: When the correlation parameter is lower than the preset correlation threshold, a preset numerical anomaly warning signal is output.

[0014] Step 61: When the correlation parameter is higher than the correlation threshold, the target fitted carbon emission future curve is corrected based on the correlation fitted carbon emission future curve to obtain the corrected carbon emission future curve;

[0015] Step 62: Based on the revised future carbon emission curve, issue a preset warning signal for excessive carbon emissions and a preset warning signal for abnormal growth in carbon emissions.

[0016] Optionally, the method for establishing the building carbon emissions database includes:

[0017] Step 7: In response to the received sample building and sample historical carbon emission data, input them into the building carbon emission database;

[0018] Step 71: Perform correlation analysis based on all the sample buildings and their historical carbon emission data to obtain building correlation parameters;

[0019] Step 72: If the absolute value of the building correlation parameter between the sample buildings is greater than the preset correlation parameter, then an association mapping relationship is constructed for the corresponding sample buildings;

[0020] Step 73: Based on the historical carbon emission data of the sample buildings, classify the sample buildings into different preset building carbon emission ranges;

[0021] Step 74: Identify the building type and carbon emission type of all the sample buildings;

[0022] Step 75: Based on the building type and the carbon emission type, classify the sample buildings in each building carbon emission range into similar building groups, wherein the building type, the carbon emission type and the building carbon emission range to which all the sample buildings in the similar building groups belong are the same;

[0023] Step 76: Extract the common features of the similar building groups;

[0024] Step 77: Construct the inter-building relationships among the sample buildings based on the common features of the similar building groups;

[0025] Step 78: Establish the building carbon emission database based on the association mapping relationship and the relationship between buildings.

[0026] Optional, also includes:

[0027] Step 50: Assemble the similar buildings into a verification building group;

[0028] Step 51: Perform a correlation test on the fitted future carbon emission curves of the similar buildings to obtain the similarity correlation parameters between the similar buildings and all other similar buildings;

[0029] Step 52: When the similarity correlation parameter is less than a preset correlation threshold, the similar building corresponding to the similarity correlation parameter is removed from the verification building group to form a corrected verification building group;

[0030] Step 53: Obtain the group-fitted carbon emission future curve based on the modified verification building complex, wherein the group-fitted carbon emission future curve is the average curve of the similar fitted carbon emission future curves of all similar buildings in the modified verification building complex.

[0031] Step 54: Perform a correlation test on the target fitted future carbon emission curve and the group fitted future carbon emission curve to obtain the target correlation parameter;

[0032] Step 540: When the target correlation parameter is less than the preset target correlation threshold, the numerical anomaly warning signal is issued;

[0033] Step 541: When the target correlation parameter is greater than the target correlation threshold, the numerical abnormality warning signal will not be issued.

[0034] Optionally, a method for issuing a numerical anomaly warning signal when the target correlation parameter is greater than the target correlation threshold also includes:

[0035] Step 5410: Sort the corrected verification building groups from high to low based on the correlation parameters to obtain the ordered corrected verification building groups;

[0036] Step 5411: Select a preset number of similar buildings from the building group in descending order of the ordered correction verification, define them as highly similar buildings, define the similar real-time carbon emission data of highly similar buildings as highly similar real-time carbon emission data, and define the similar historical carbon emission data of highly similar buildings as highly similar historical carbon emission data.

[0037] Step 5412: Perform historical data correlation analysis on the target historical carbon emission data and the highly similar historical carbon emission data to obtain historical data correlation parameters;

[0038] Step 5413: Obtain a highly similar fitted future carbon emission curve based on the highly similar real-time carbon emission data and the highly similar historical carbon emission data;

[0039] Step 5414: Perform correlation analysis on the target fitted future carbon emission curve and the highly similar fitted future carbon emission curve to obtain the correlation parameters of the predicted curve;

[0040] Step 5415: Based on the predicted curve correlation parameters, find the corresponding predicted curve correlation parameter range from the preset correlation database;

[0041] Step 5416: If the predicted correlation parameter does not fall within the range of the predicted curve correlation parameter, then issue the numerical anomaly warning signal.

[0042] Optionally, a method for issuing the numerical anomaly warning signal even when the predicted correlation parameter does not fall within the predicted curve correlation parameter range, the method comprising:

[0043] Step 54160: Obtain the target carbon emission types and target carbon emission percentages of the target building, and obtain the highly similar carbon emission types and highly similar carbon emission percentages of the highly similar buildings;

[0044] Step 54161: Perform a significance test on the target carbon emission percentage and the highly similar carbon emission percentage to obtain the significant difference value;

[0045] Step 54162: If the significant difference value is greater than the preset difference value, then issue the abnormal value warning signal.

[0046] Optionally, when the correlation parameter is higher than the correlation threshold, the target fitted carbon emission future curve is corrected based on the correlation-fitted carbon emission future curve, and the method for obtaining the corrected carbon emission future curve includes:

[0047] Step 610: Construct a network of relationships between the related buildings based on the relationships between the buildings;

[0048] Step 611: Analyze the network to obtain the association attributes and association strength between the associated buildings;

[0049] Step 612: Based on the correlation attributes, the correlation strength, and the correlation fitted future carbon emission curve, obtain the fitting carbon emission future curve correction method from the preset building data correction library;

[0050] Step 613: Based on the fitted carbon emission future curve correction method, correct the target fitted carbon emission future curve of the target building to obtain the corrected carbon emission future curve;

[0051] Step 614: Based on the modified carbon emission future curve, obtain the target carbon emission intensity and the target carbon emission growth trend intensity within the future time interval;

[0052] Step 6140: If the target carbon emission intensity is less than the carbon emission threshold and the target carbon emission growth trend intensity is less than the growth trend threshold, then the carbon emission exceeding warning signal and the carbon emission abnormal growth warning signal will not be issued.

[0053] Step 6141: If the target carbon emission intensity is greater than the carbon emission threshold, then issue the carbon emission exceedance warning signal;

[0054] Step 6142: If the intensity of the target carbon emission growth trend is greater than the growth trend threshold, then issue the abnormal carbon emission growth warning signal.

[0055] Optionally, if the carbon emission intensity is less than the carbon emission threshold and the carbon emission growth trend intensity is less than the growth trend threshold, the abnormal carbon emission growth warning signal is still issued. This method includes:

[0056] Step 61400: Based on the inter-building relationship between the target building and the related buildings, classify the related buildings into buildings with the same trend and buildings with different trends;

[0057] Step 61401: Obtain the future carbon emission curves of the same-trend buildings;

[0058] Step 61402: Perform correlation analysis between the modified future carbon emission curve and the fitted future carbon emission curve with the same trend to obtain the trend correlation value;

[0059] Step 61403: When the trend correlation value is greater than the preset trend threshold, and the abnormal carbon emission growth warning signal is received from the building with the same trend, the abnormal carbon emission growth warning signal is also issued to the target building.

[0060] Optionally, when the trend correlation value is greater than the trend threshold, and a warning signal for abnormal carbon emission growth is received from a building exhibiting the same trend, a method may be used to prevent the issuance of the warning signal. This method includes:

[0061] Step 614030: Obtain the trend growth reasons and growth trend lines corresponding to the target carbon emission growth trend intensity from the preset trend database;

[0062] Step 614031: Perform a credibility analysis on the modified future carbon emission curve and the growth trend line to obtain the credibility.

[0063] Step 6140310: If the credibility is lower than a preset credibility threshold, then issue the abnormal carbon emission growth warning signal;

[0064] Step 6140311: If the credibility is higher than the credibility threshold and the target carbon emission intensity is higher than the carbon emission threshold, then issue a carbon emission exceeding warning signal.

[0065] Step 6140312: If the credibility is higher than the credibility threshold and the target carbon emission intensity is less than the carbon emission threshold, then the carbon emission exceeding warning signal will not be issued.

[0066] Optionally, a method for issuing the numerical anomaly warning signal when the target carbon emission intensity is less than the carbon emission threshold and the target carbon emission growth trend intensity is less than the growth trend threshold, the method comprising:

[0067] Step 61404: Confirm carbon emission pathways based on the relationships between the buildings;

[0068] Step 61405: Determine the nodal buildings based on the carbon emission pathway;

[0069] Step 61406: Obtain the total energy level input for each of the carbon emission pathways;

[0070] Step 61407: Obtain the real-time carbon emission data of the node building;

[0071] Step 61408: Determine the total carbon emission level based on the real-time carbon emission data of all node buildings on the carbon emission path to which the target building belongs;

[0072] Step 61409: When the total carbon emission level does not match the total energy level, an abnormal value warning signal is issued to all buildings along the carbon emission path.

[0073] Optional, also includes:

[0074] Step 8: When the target building lacks the historical target carbon emission data, obtain the highly similar historical carbon emission data of the highly similar building;

[0075] Step 80: Replace the historical target carbon emission data of the target building with the highly similar historical carbon emission data.

[0076] In summary, this application includes at least one of the following beneficial technical effects:

[0077] 1. By integrating real-time and historical data of similar and related buildings, a prediction model with multiple data sources is constructed. Compared with historical data of a single building, this cross-building collaborative analysis can more comprehensively capture the common patterns and individual differences in carbon emissions, reduce the risk of misjudgment caused by local data bias, and improve the accuracy of dynamic carbon emission prediction.

[0078] 2. By introducing a correlation parameter threshold judgment, when the carbon emission trends of the target building and similar buildings deviate from the correlation, an abnormal warning is automatically triggered; if the trends are consistent, the prediction results are optimized by combining the relevant buildings. This not only avoids the lag caused by the single data in traditional methods, but also dynamically adjusts the warning strategy through the real-time feedback of relevant buildings, thereby improving the credibility of decision-making in complex scenarios.

[0079] 3. The solution distinguishes between numerical anomaly warning signals, carbon emission exceedance warning signals, and abnormal carbon emission growth warning signals, thus handling data reliability and the risk of substantial carbon emission exceedances separately. The former indicates data anomalies, while the latter focuses on the exceedance problem, helping managers quickly locate the root cause of the problem and take targeted measures such as equipment maintenance and load control, avoiding the waste of resources caused by a "one-size-fits-all" emergency response. Attached Figure Description

[0080] Figure 1 This is a flowchart of a smart building carbon emission early warning method based on big data analysis, as described in an embodiment of this application.

[0081] Figure 2 This is a flowchart of the method for obtaining the modified future carbon emission curve in the embodiments of this application. Detailed Implementation

[0082] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments.

[0083] This application discloses a method for early warning of carbon emissions from smart buildings based on big data analysis. (Refer to...) Figure 1 A smart building carbon emission early warning method based on big data analysis includes:

[0084] Step 1: In response to the received target building, find the corresponding related buildings and the relationships between buildings based on the preset building carbon emission database.

[0085] The target building is a specific building selected by the user or system that requires carbon emission analysis and prediction. It is received by entering the building's unique identifier, such as number, name, or geographical location, through the front-end interface.

[0086] Related buildings refer to a group of buildings that are identified as having a carbon emission correlation with the target building through data analysis of the building carbon emission database.

[0087] Inter-building relationships refer to the carbon emission relationships between buildings obtained through correlation analysis of multi-dimensional data in the building carbon emission database, such as similar emission trends, opposite emission trends, and similar fluctuation cycles.

[0088] A building carbon emissions database is an intelligent analytical database built by collecting data on real-time and historical carbon emissions, building types, and carbon emission categories. The database includes real-time and historical carbon emissions of buildings, building types, and correlation mappings between buildings. The database was created through experiments by staff who performed correlation analysis on the real-time and historical carbon emissions of all buildings. Buildings with high correlations were linked and their relationships were further analyzed based on their function, carbon emission types, and emission intensity. Finally, the staff stored the obtained correlation mappings, inter-building relationships, and the corresponding real-time and historical carbon emissions of each building into the building carbon emissions database.

[0089] Step 2: Based on the relationships between buildings, classify related buildings into similar buildings and related buildings.

[0090] Similar buildings refer to a group of buildings whose carbon emission trends are synchronized with those of the target building in real time, and whose carbon emission magnitude, type, and functional attributes are highly matched. Related buildings refer to a group of buildings whose carbon emissions are inversely linked to, delayed in growth, or complementary to those of the target building.

[0091] The determination of similar and related buildings can be based on the relationship between buildings. For example, if the relationship between buildings is the same or the same fluctuation period, based on the building carbon emission database, the two buildings are identified as similar buildings. If the relationship between buildings is the opposite or complementary emission trends, based on the building carbon emission database, the two buildings are identified as related buildings.

[0092] For example, in an industrial park, production workshops and warehouse buildings share the same power system. When the production line is running at full load, the warehouse area reduces carbon emissions due to off-peak electricity use, and the two are identified as related buildings. On the other hand, two office buildings of the same size and purpose are classified as similar buildings because their real-time energy consumption curves are highly similar.

[0093] Step 3: Obtain the target real-time carbon emission data and target historical carbon emission data for the target building, similar real-time carbon emission data and similar historical carbon emission data for similar buildings, and relevant real-time carbon emission data and relevant historical carbon emission data for related buildings.

[0094] Real-time carbon emission data for a target building refers to the current amount and types of carbon emissions collected at any given moment, reflecting the building's current energy consumption status and emission intensity. This data can be collected in real-time from IoT sensors, environmental monitoring terminals, and other devices deployed within the target building.

[0095] Historical carbon emission data for a target building refers to the long-term accumulated carbon emission records of that building. This data is used to analyze the periodic patterns and trends of building carbon emissions, or as a benchmark for model training and anomaly detection. Historical carbon emission data can be extracted from the target building's energy management system, utility system, and other data records. Similarly, real-time carbon emission data for similar buildings refers to the current carbon emission values ​​and types collected at that moment, while historical carbon emission data for similar buildings refers to the long-term accumulated carbon emission records of those buildings. Both are obtained using the same methods as for the target building. Real-time carbon emission data for related buildings refers to the current carbon emission values ​​and types collected at that moment, while historical carbon emission data for related buildings refers to the long-term accumulated carbon emission records of those buildings. Both are obtained using the same methods as for the target building.

[0096] Step 4: Obtain the target fitted carbon emission future curve based on real-time target carbon emission data and real-time historical target carbon emission data; obtain the similar fitted carbon emission future curve based on similar real-time carbon emission data and similar historical carbon emission data; and obtain the relevant fitted carbon emission future curve based on relevant real-time carbon emission data and relevant historical carbon emission data.

[0097] The target-fitted carbon emission future curve refers to a curve generated by fitting real-time and historical carbon emission data of a target building through time series analysis or machine learning models, and is used to predict the future carbon emission trend of that building.

[0098] The similarity-fitted future carbon emission curve refers to a curve generated by selecting a group of buildings highly similar to the target building in terms of function, scale, and energy consumption patterns from a building carbon emission database, such as similar office buildings, and fitting their real-time carbon emission data with historical carbon emission data. The fitting method for the similarity-fitted future carbon emission curve is the same as that for the target building. This curve can be used as a benchmark to determine the reasonableness of the target building's carbon emission data.

[0099] The correlation-fitted future carbon emission curve refers to a curve generated from the real-time and historical carbon emission data of "related buildings" that have energy or process connections with the target building, such as the living and production areas in a factory. The carbon emissions of similar buildings show an inverse or complementary relationship with the target building, and their prediction results can correct deviations from the target curve. The fitting method for the correlation-fitted future carbon emission curve is the same as that for the target-fitted future carbon emission curve.

[0100] Step 5: Determine the correlation parameters based on the target fitted future carbon emission curve and the similar fitted future carbon emission curve.

[0101] The correlation coefficient is a quantitative indicator that measures the synergy between the future carbon emission curve fitted by the target building and the fitted curve of similar buildings. Its value ranges from -1 to 1. The larger the value, the closer the trend, magnitude and periodicity characteristics of the two are.

[0102] For example, the Pearson correlation coefficient can be used to calculate the degree of matching for linearly increasing carbon emissions:

[0103] The points on the carbon emission curve for the target building over the next 5 hours are [100, 120, 110, 130, 120].

[0104] The points on the mean prediction curve for similar building groups are [105, 118, 112, 128, 115].

[0105] The average predicted value of the target building = (100 + 120 + 110 + 130 + 120) / 5 = 116;

[0106] The average predicted value of similar buildings = (105 + 118 + 112 + 128 + 115) / 5 = 115.6;

[0107] Covariance between the target building and similar buildings: ;

[0108] Standard deviation of the target building: ;

[0109] Covariance of similar buildings: ;

[0110] Correlation coefficient: ;

[0111] In the formula: Let be the predicted carbon emissions of the target building at time point i. Let be the predicted carbon emissions of similar buildings at the i-th time point; The average of the predicted values ​​for the target building; is the average of the predicted values ​​for similar buildings; n is the number of time points.

[0112] In addition to the Pearson correlation coefficient calculation method mentioned above, when the data exhibits a non-linear relationship, such as a parabola, exponential, or logarithmic curve, the correlation index method and other calculation methods can also be used to fit the curve, which will not be elaborated here.

[0113] Step 60: When the correlation parameter is lower than the preset correlation threshold, a preset numerical anomaly warning signal is output.

[0114] The correlation threshold is a threshold used to determine whether the carbon emissions of a target building and related buildings are significantly correlated. The correlation threshold is determined through extensive experimentation. Researchers analyze the correlation parameters of carbon emission curves from a large number of buildings to find the lowest threshold that demonstrates the similar trend in the predicted carbon emissions of both buildings; this threshold is considered the correlation threshold.

[0115] The numerical anomaly warning signal is an automatic alert triggered by the system when the correlation parameter between the target fitted future carbon emission curve and the similar fitted future carbon emission curve of the target building falls below a preset threshold. When the correlation parameter falls below the preset correlation threshold, the system flashes red on the terminal screen and issues the text "Numerical Anomaly Warning". (Refer to...) Figure 2 Under condition ②, if the target building's fitted future carbon emission curve is uncorrelated with the similar fitted future carbon emission curves, a numerical anomaly warning signal should be issued. This signal indicates a significant deviation between the target building's carbon emission prediction data and the common patterns of similar building groups. This deviation may be caused by equipment failure, data acquisition errors, model bias, or sudden abnormal events, suggesting insufficient reliability of the current prediction results. Immediate verification of the data source, calibration of the prediction model, or investigation of the building's operational status are necessary to avoid decision-making risks caused by data anomalies.

[0116] Step 61: When the correlation parameter is higher than the correlation threshold, the target fitted carbon emission future curve is corrected based on the correlation fitted carbon emission future curve to obtain the corrected carbon emission future curve.

[0117] When the correlation parameter is higher than the correlation threshold, it means that the predicted carbon emission trend of the target building highly matches the overall pattern of similar building groups. At this point, the overall carbon emission prediction of the target building can be considered accurate. To further improve the accuracy of carbon emission prediction, the future carbon emission curve fitted with the correlation parameter is used to correct the future carbon emission curve fitted to the target building. (Refer to...) Figure 2 Under condition ①, the target fitted carbon emission future curve is corrected by fitting carbon emission curves with opposite, the same, or complementary carbon emission trends, thus obtaining the corrected carbon emission future curve.

[0118] Step 62: Based on the revised future carbon emission curve, issue a preset warning signal for excessive carbon emissions and a preset warning signal for abnormal growth in carbon emissions.

[0119] A carbon emission exceedance warning signal is an alert triggered when the revised future carbon emission curve of a target building shows that its carbon emission intensity will exceed a preset threshold within a set time period. This indicates that the building is at risk of exceeding the permissible carbon emission limit. When this risk occurs, a red light flashes on the system terminal screen, and the text "Carbon Emission Exceedance Warning" is displayed.

[0120] The abnormal carbon emission growth warning signal is an alert triggered when the carbon emission growth trend of a target building exceeds a preset rate threshold. This indicates that while the current carbon emission intensity of the target building has not yet exceeded the limit, the predicted value shows a sharp upward trend, prompting early intervention to avoid the risk of exceeding the limit. When this risk occurs, a red light flashes on the system terminal screen, and the text "Abnormal Carbon Emission Growth Warning" is displayed.

[0121] Methods for establishing a building carbon emissions database include:

[0122] Step 7: In response to the received sample building and sample historical carbon emission data, input them into the building carbon emission database.

[0123] Step 71: Conduct correlation analysis based on all sample buildings and their historical carbon emission data to obtain building correlation parameters.

[0124] Building correlation parameters are statistically quantified indicators used to characterize the correlation between carbon emission behaviors of different buildings. The core of this approach is to analyze the correlation coefficients, such as the Pearson correlation coefficient, between historical carbon emission data of sample buildings. A higher building correlation parameter indicates a stronger correlation between the sample buildings. Specific correlation analysis methods have been introduced in step 5 and will not be repeated here.

[0125] Step 72: If the absolute value of the building correlation parameter between sample buildings is greater than the preset correlation parameter, then the corresponding sample buildings will be associated with a mapping relationship.

[0126] The correlation mapping relationship refers to the relationship between buildings that has a strong correlation, obtained through correlation analysis.

[0127] The correlation parameter is a preset threshold used to determine whether a significant correlation exists between sample buildings. This parameter was obtained by staff through extensive experimentation. Staff calculated the correlation parameter between sample buildings.

[0128] Step 73: Based on the sample building's historical carbon emission data, classify the sample building into different preset building carbon emission ranges.

[0129] Building carbon emission intervals are discrete classification intervals based on historical carbon emission intensity ranges, used to group buildings with similar carbon emission levels into the same analysis group. These intervals are obtained through experiments by staff who adjust the upper and lower limits of each interval to ensure coverage of all carbon emission intensities for the sample buildings while maintaining an equal number of sample buildings within each interval.

[0130] Step 74: Identify the building type and carbon emission type of all sample buildings.

[0131] Building type refers to the classification of sample buildings based on their functional attributes and spatial characteristics, such as residential buildings, public buildings, commercial buildings, and industrial buildings. Carbon emission type refers to the category of carbon emission sources generated during building operation, mainly including energy consumption emissions, equipment operation emissions, and ancillary service emissions.

[0132] Step 75: Based on building type and carbon emission type, classify the sample buildings in each building carbon emission range into similar building groups. All sample buildings in the similar building groups have the same building type, carbon emission type and building carbon emission range.

[0133] Similar building clusters refer to a collection of homogeneous buildings categorized based on three dimensions: building functional attributes, energy consumption characteristics, and carbon emission levels. Their core commonalities are: all buildings have the exact same building type, completely identical carbon emission types, and belong to the same preset carbon emission range.

[0134] Step 76: Extract common features of similar building groups.

[0135] Common characteristics refer to the core attributes of all buildings in a similar building group that are highly consistent in terms of functional attributes, structural parameters, energy consumption patterns, and carbon emission patterns. These characteristics, such as building functional attributes, energy consumption characteristics, and carbon emission levels, can reflect the essential laws of the group.

[0136] Taking the extraction of building functional attributes as an example:

[0137] Collect all functional attributes of similar building groups, extract the functional attribute with the highest proportion, and if its proportion exceeds 80%, then the functional attribute is taken as the common feature of similar building groups.

[0138] Taking a hospital complex as an example:

[0139] Functional attributes: Grade III Class A general hospital. Carbon emission level: Annual average carbon emission intensity approximately 120 kgCO2 / m³. 2 Carbon emissions during the summer cooling season account for 60% of the annual total, and the annual average carbon emission variation curve is shown. Energy consumption: The system operates 24 hours a day throughout the year, with peak electricity load concentrated at midday and night.

[0140] Specific common features include, but are not limited to, the examples above, and can be adjusted according to the building type.

[0141] Step 77: Construct inter-building relationships between sample buildings based on the common features of similar building groups.

[0142] Its specific construction method consists of two aspects:

[0143] ① Building relationships within similar building clusters: Buildings within similar building clusters are all of the same building type, have the same carbon emission type, and belong to the same preset carbon emission range. Therefore, correlation analysis is used to obtain the degree of correlation between the carbon emission change trends of each building, and the similarity of buildings within similar building clusters is determined based on the correlation calculation results.

[0144] ② Architectural Relationships Among Similar Building Clusters: After extracting common characteristics of similar building clusters, correlation analysis is performed on the common characteristics of different similar building clusters to obtain the relationships between them. A correlation parameter close to 1 indicates a positive correlation in carbon emission changes among different similar building clusters, while a correlation parameter close to -1 indicates a negative correlation in carbon emission changes among different similar building clusters.

[0145] Step 78: Establish a building carbon emission database based on the association mapping relationship and the relationship between buildings.

[0146] The building carbon emissions database is an intelligent analytical database built by collecting data on real-time carbon emissions, historical carbon emissions, building type, and carbon emission types of buildings. Each building has a unique identifier, such as a building ID, which is stored as an independent record in the database. Then, based on the association mapping and relationships between buildings, related buildings are marked in the corresponding records as "related objects".

[0147] For example, add tags such as "Similar Building: Hospital B (same trend)" and "Related Building: Cold Storage Warehouse (complementary cycle)" to the independent records in Hospital A. This ultimately forms a dual information database containing building attributes and dynamic relationships, supporting subsequent carbon emission analysis and optimization.

[0148] Optional, also includes:

[0149] Step 50: Group similar buildings into a verification building group.

[0150] A verification building complex refers to a collection of buildings obtained by integrating similar buildings.

[0151] Step 51: Perform a correlation test on the fitted carbon emission future curves of similar buildings to obtain the similarity correlation parameters between similar buildings and all other similar buildings.

[0152] The correlation test is an analytical process that uses statistical methods to quantify the degree of correlation between the future carbon emission curves fitted by different buildings. The specific correlation analysis method has been introduced in step 5 and will not be repeated here.

[0153] Step 52: When the similarity correlation parameter is less than the preset correlation threshold, the similar buildings corresponding to the similarity correlation parameter are removed from the verification building group to form a corrected verification building group.

[0154] The modified validation building cluster refers to the optimized subset formed after removing abnormal buildings that significantly differ from the predicted carbon emission trend of the group from the initial validation building cluster. Its members all meet the condition that the similarity correlation parameter with other buildings in the group is higher than the preset correlation threshold.

[0155] When the similarity correlation parameter of a building is less than the preset correlation threshold, it indicates that the fitted future carbon emission curve of the building deviates significantly from the overall trend of the group. Removing such buildings can eliminate the interference of abnormal data on the group analysis and ensure the homogeneity of the building group and the reliability of the prediction model.

[0156] Step 53: Obtain the group-fitted carbon emission future curve based on the modified verification building cluster. The group-fitted carbon emission future curve is the average curve of the similar fitted carbon emission future curves of all similar buildings in the modified verification building cluster.

[0157] The cluster-fitted carbon emission future curve is a comprehensive baseline curve formed by aggregating and correcting the individual prediction results of all similar buildings within the building cluster. Specifically, it can be obtained by directly calculating the arithmetic mean of the similar fitted carbon emission future curves of each similar building, or by weighting the average based on similarity correlation parameters.

[0158] Step 54: Perform a correlation test between the target fitted future carbon emission curve and the group fitted future carbon emission curve to obtain the target correlation parameters.

[0159] The target relevance parameter is a quantitative indicator calculated using statistical methods to measure the linear or monotonic correlation between the target-fitted future carbon emission curve and the group-fitted future carbon emission curve. The specific correlation analysis methods have been introduced in step 5 and will not be repeated here.

[0160] Step 540: When the target correlation parameter is less than the preset target correlation threshold, a numerical anomaly warning signal is issued.

[0161] If the target relevance parameter is less than the preset target relevance threshold, it means that the carbon emission data of the target building is damaged due to data source failure, such as sensor malfunction, transmission interference or sudden environmental changes, resulting in the loss of correlation between the target data and the associated benchmark data. The reliability of the carbon emission data of the target building needs to be checked.

[0162] Step 541: If the target correlation parameter is greater than the target correlation threshold, no numerical anomaly warning signal will be issued.

[0163] When the target correlation parameter is greater than the target correlation threshold, it indicates that there is a correlation between the target fitted future carbon emission curve and the similar fitted future carbon emission curve of the target building, indicating that the carbon emission data of the target building is reliable and the subsequent correction steps can be carried out.

[0164] It also includes a method for issuing a numerical anomaly warning signal when the target correlation parameter is greater than the target correlation threshold, the method comprising:

[0165] Step 5410: Sort the corrected verification building groups from high to low based on the correlation parameters to obtain the ordered corrected verification building groups.

[0166] The ordered correction validation building group refers to an ordered list of buildings that have passed the correlation test, sorted from highest to lowest according to their correlation parameters with the target building. This sorting result guides the prioritization of subsequent analyses, ensuring that highly similar buildings are included in the validation scope first, thereby improving the reliability of data correction and the efficiency of analysis.

[0167] Step 5411: Select a preset number of similar buildings from the sequentially corrected verification building group in descending order, define them as highly similar buildings, define the similar real-time carbon emission data of highly similar buildings as highly similar real-time carbon emission data, and define the similar historical carbon emission data of highly similar buildings as highly similar historical carbon emission data.

[0168] The selection quantity refers to a fixed number of buildings selected from the top-ranked buildings in the sequential correction verification building group based on relevance, used to define the range of highly similar buildings.

[0169] Highly similar buildings are selected from the corrected and verified building cluster based on their correlation scores with the target building's carbon emission characteristics, ranked from highest to lowest relevance, and the top N buildings with the strongest correlation are chosen. Simultaneously, the real-time and historical carbon emission data of these buildings are labeled as "highly similar real-time carbon emission data" and "highly similar historical carbon emission data," respectively.

[0170] Step 5412: Perform historical data correlation analysis on the target historical carbon emission data and highly similar historical carbon emission data to obtain historical data correlation parameters.

[0171] Historical data correlation parameters are quantitative indicators of the correlation between historical carbon emission data of a target building and historical carbon emission data of highly similar buildings, calculated through statistical analysis methods. They measure the similarity of their historical emission trends. A higher parameter value indicates stronger synchronicity or regularity in the historical emission patterns of the two buildings, allowing for a more reliable assessment of the accuracy of the target building's prediction model using similar building data. Conversely, a lower or abnormal parameter value may indicate abnormalities in the target carbon emission data, a weakened correlation between the target building and similar buildings, or a need to reassess the building cluster classification.

[0172] The specific correlation analysis method has been introduced in step 5 and will not be repeated here.

[0173] Step 5413: Obtain a highly similar fitted future carbon emission curve based on highly similar real-time carbon emission data and highly similar historical carbon emission data.

[0174] The highly similarity-fitted future carbon emission curve refers to the future emission prediction curve generated by fitting highly similar real-time carbon emission data and highly similar historical carbon emission data of highly similar buildings. The specific method has been introduced in step 4 and will not be repeated here.

[0175] Step 5414: Perform correlation analysis between the target fitted future carbon emission curve and the highly similar fitted future carbon emission curve to obtain the correlation parameters of the predicted curve.

[0176] The correlation parameter of the prediction curve refers to the degree of correlation between the future carbon emission curve fitted by the target building and the fitted curve of highly similar buildings through statistical analysis. It is usually quantified by the correlation coefficient, such as the Pearson coefficient, with a value range of [-1, 1]. The closer the absolute value of the correlation parameter of the prediction curve is to 1, the stronger the synchronicity or inverse linkage of the future emission trends of the two, indicating that the prediction data of similar buildings has a higher reference value for correcting the target building.

[0177] The specific correlation analysis method has been introduced in step 5 and will not be repeated here.

[0178] Step 5415: Based on the correlation parameters of the predicted curve, find the corresponding correlation parameter range of the predicted curve from the preset correlation database.

[0179] A correlation database refers to a pre-built database that stores the mapping relationship between the correlation parameters of prediction curves and the corresponding parameter intervals under different scenarios.

[0180] The predicted curve correlation parameter range refers to the range obtained by broadening the range of predicted curve correlation parameters according to preset rules in the correlation database. If the predicted correlation parameters fall within this range, it can be considered that the carbon emission change trend of the target building and similar buildings is similar to historical data.

[0181] This was obtained through extensive experiments by staff, demonstrating that as long as the correlation parameter of the prediction curve falls within the range of the correlation parameter of the prediction curve, the target fitted carbon emission future curve is highly similar to the highly similar fitted carbon emission future curve, proving that the development trend of the fitted carbon emission future curve is the same as the trend of similar buildings, and is a reliable carbon emission future curve.

[0182] Step 5416: If the predicted correlation parameter does not fall within the range of the predicted curve correlation parameter, then issue a numerical anomaly warning signal.

[0183] When the predicted correlation parameter does not fall within the preset range of the predicted curve correlation parameter, it indicates that the correlation between the future carbon emission trends of the target building and highly similar buildings has deviated significantly. At this time, the system determines that the historical correlation pattern between the two cannot effectively support future predictions and needs to trigger an early warning to indicate that there are potential risks in the real-time data and predicted curve of the target building and that there is a need for data correction.

[0184] A method for issuing a numerical anomaly warning signal even when the predicted correlation parameter does not fall within the range of the predicted curve correlation parameter, the method includes:

[0185] Step 54160: Obtain the target carbon emission types and target carbon emission percentages of the target building, and obtain the highly similar carbon emission types and highly similar carbon emission percentages of highly similar buildings.

[0186] The target carbon emission categories refer to the carbon emission categories of the target buildings, including direct carbon emissions and indirect carbon emissions. Direct carbon emissions are further subdivided into fuel combustion carbon emissions and industrial process carbon emissions, while indirect carbon emissions are further subdivided into electricity consumption carbon emissions and food waste carbon emissions.

[0187] The target carbon emission percentage is the proportion of various types of carbon emissions in the total emissions of the target building, used to quantify the contribution of different emission sources.

[0188] The definitions of highly similar carbon emission types and highly similar carbon emission percentages are similar to those of the target carbon emission types and target carbon emission percentages mentioned above, and will not be repeated here.

[0189] Step 54161: Perform a significance test on the target carbon emission percentage and the highly similar carbon emission percentage to obtain the significant difference value.

[0190] Significance testing is a verification method that uses statistical methods, such as chi-square test and t-test, to determine whether there are systematic differences in the types and proportions of carbon emissions between the target building and highly similar buildings.

[0191] The significance difference value is a quantitative indicator output by the test, such as the p-value or effect size, used to measure the degree of difference between the target building and similar buildings in terms of carbon emission structure: if the value is lower than the preset threshold, it indicates that the distribution or proportion of carbon emission types of the two buildings have statistical significance, which may be due to differences in function, equipment efficiency or different operating modes; otherwise, it indicates that the difference is mainly caused by accidental factors, and the carbon emission structure of the two buildings can be considered to be basically the same.

[0192] Step 54162: If the significant difference value is greater than the preset difference value, an abnormal value warning signal is issued.

[0193] The difference value refers to a preset threshold used to determine whether a significant difference exceeds the normal range, and it is obtained by staff through extensive analysis. Staff determine the minimum difference between the carbon emission proportions of two buildings by analyzing the carbon emission proportions of different buildings, and define this value as the difference value.

[0194] When the significant difference exceeds a preset threshold, it indicates that the difference between the target building and highly similar buildings in terms of carbon emission types or proportions has exceeded the statistically significant range of random fluctuations, suggesting a possible systematic deviation in their carbon emission structures, such as functional changes, equipment malfunctions, or data anomalies. This significant deviation may lead to errors in the prediction of the target building, thus requiring an alert to prompt a re-verification of the data source.

[0195] When the correlation parameter is higher than the preset correlation threshold, the target fitted carbon emission future curve is corrected based on the correlation-fitted carbon emission future curve. The methods for obtaining the corrected carbon emission future curve include:

[0196] Step 610: Construct a network of relationships between related buildings based on the relationships between buildings.

[0197] The interconnected network is a mesh structure built with buildings as nodes and relationships as edges. It represents the linkage pattern of carbon emission behavior of building groups through the weights and directions between nodes, providing a structured model for analyzing group synergy effects, locating key influencing nodes, and optimizing overall carbon management strategies.

[0198] By identifying relationships between buildings, such as similar emission trends, opposite emission trends, and similar fluctuation cycles, different buildings are linked by line segments. Each line segment has a corresponding correlation parameter. A correlation parameter close to 1 indicates a positive correlation in carbon emission changes among different similar building groups, while a correlation parameter close to -1 indicates a negative correlation. The specific method for constructing these relationships is described in step 07 and will not be repeated here.

[0199] Step 611: Perform network analysis to obtain the association attributes and association strength between related buildings.

[0200] Association attributes refer to the qualitative or quantitative relationship categories formed between buildings based on characteristics such as functional type, carbon emission category and carbon emission type, and are used to describe the specific dimensions of the association.

[0201] The correlation strength is a numerical value quantified by methods such as trend correlation coefficient, cycle consistency index or complementarity index, reflecting the degree of synergy or difference between buildings in a specific attribute. The higher the absolute value, the stronger the linkage between the two in that dimension.

[0202] Step 612: Based on the correlation attributes, correlation strength and correlation, fit the future carbon emission curve and obtain the correction method for the fitted future carbon emission curve from the preset building data correction library.

[0203] The Building Data Correction Library is a structured database built on carbon emission data of historical building complexes and the results of correlation analysis. It stores a set of fitting curve correction strategies corresponding to different correlation attributes and correlation strengths. Its data comes from the summarization of historical correction examples and the training of machine learning models.

[0204] The method of fitting the future carbon emission curve refers to the specific correction operation performed by matching from the correction library based on the correlation attributes and intensity between the target building and related buildings. For example, weighted averaging, trend offset compensation, or periodic parameter calibration of the prediction curve can be performed to improve the prediction accuracy.

[0205] Step 613: Based on the method of correcting the fitted carbon emission future curve, correct the target fitted carbon emission future curve of the target building to obtain the corrected carbon emission future curve.

[0206] The modified carbon emission future curve refers to the curve obtained by dynamically adjusting and optimizing the target fitted carbon emission future curve through a fitting carbon emission future curve correction method.

[0207] Step 614: Based on the corrected future carbon emission curve, obtain the target carbon emission intensity and the target carbon emission growth trend intensity within the future time interval.

[0208] The target carbon emission intensity for the future time interval refers to the unit baseline carbon emission predicted within a specified time period based on the revised future carbon emission curve.

[0209] The target carbon emission growth trend intensity refers to the rate of change of carbon emissions within the same time period, reflecting the speed at which emissions increase or decrease.

[0210] Step 6140: If the target carbon emission intensity is less than the carbon emission threshold and the target carbon emission growth trend intensity is less than the growth trend threshold, then no carbon emission exceedance warning signal or abnormal carbon emission growth warning signal will be issued.

[0211] If the target carbon emission intensity is less than the carbon emission threshold and the target carbon emission growth trend intensity is less than the growth trend threshold, it means that the current carbon emissions have not exceeded the set upper limit, the emission growth rate is within a safe range, and the overall carbon status is normal.

[0212] Step 6141: If the target carbon emission intensity is greater than the carbon emission threshold, a carbon emission exceedance warning signal is issued.

[0213] If the target carbon emission intensity is greater than the carbon emission threshold, it means that the current carbon emission level has exceeded the allowable value, and emission reduction measures need to be addressed immediately.

[0214] Step 6142: If the intensity of the target carbon emission growth trend is greater than the growth trend threshold, then issue an early warning signal for abnormal carbon emission growth.

[0215] If the intensity of the target carbon emission growth trend is greater than the growth trend threshold, it means that although the emissions have not exceeded the standard, the growth rate is too fast and may exceed the threshold in the future, requiring early intervention.

[0216] The method for issuing a carbon emission exceedance warning signal even when the carbon emission intensity is less than the carbon emission threshold and the carbon emission growth trend intensity is less than the growth trend threshold, includes:

[0217] Step 61400: Based on the inter-building relationships between the target building and related buildings, classify related buildings into buildings with the same trend and buildings with different trends.

[0218] Buildings exhibiting similar carbon emission trends to the target building are those whose carbon emission trends are consistent with or positively correlated with the target building. When the carbon emissions of such buildings increase, due to inter-building relationships, such as shared energy systems, collaborative operation models, and physical structural linkages, the carbon emissions of the target building will increase synchronously, either directly or indirectly.

[0219] For example, if multiple buildings share the same central air conditioning or heating network, and one building increases the operating power of its central air conditioning, the buildings sharing the network will also increase their energy consumption and carbon emissions.

[0220] Non-trend buildings refer to related buildings whose carbon emission changes are not significantly correlated with or are inversely correlated with those of the target building. Their carbon emission increases do not have a significant positive effect on the carbon emissions of the target building, or they have a negative effect.

[0221] For example, in the park's work area and accommodation area, during working hours, as employees move from the dormitory to the work area, carbon emissions in the dormitory area decrease, while equipment in the work area begins to operate, and energy consumption and carbon emissions begin to increase.

[0222] Step 61401: Obtain the future carbon emission curves of buildings with similar trends.

[0223] The trend-fitted future carbon emission curve refers to a curve obtained by analyzing real-time and historical carbon emission data of buildings exhibiting the same trend. The trend of this curve should be the same as the target-fitted future carbon emission curve, and its acquisition method is the same as that used in step 4.

[0224] Step 61402: Perform correlation analysis between the corrected future carbon emission curve and the fitted future carbon emission curve with the same trend to obtain the trend correlation value.

[0225] Trend correlation value is a quantitative indicator obtained by calculating the correlation between the modified future carbon emission curve of the target building and the fitted future carbon emission curve of the same trend. It is used to measure the degree of synchronization between the two trends. Its value range is usually [0, 1]. The larger the value, the closer the shape, slope and change pattern of the two curves are, that is, the stronger the correlation between the carbon emission growth patterns of the target building and buildings with the same trend.

[0226] Step 61403: When the trend correlation value is greater than the preset trend threshold, and a carbon emission exceeding warning signal is received from a building with the same trend, a carbon emission abnormal growth warning signal is also issued to the target building.

[0227] When the trend correlation value exceeds a preset threshold, and a building exhibiting the same trend triggers an abnormal carbon emission growth warning signal, it indicates that the carbon emission growth patterns of the target building and the building exhibiting the same trend have been confirmed to be highly consistent through correlation analysis. The fact that a building exhibiting the same trend triggers an abnormal growth warning means that its carbon emission growth rate has exceeded the normal range. Since the two buildings exhibit the same trend, the target building is highly likely to be affected by the same risk factors. Even if the current parameters of the target building do not exceed the limits, its emission growth trend already implies an abnormal risk. Therefore, a carbon emission exceeding warning signal should still be issued.

[0228] A method for not issuing an abnormal carbon emission growth warning signal even when the target carbon emission growth trend intensity is greater than a growth trend threshold, the method including:

[0229] Step 614030: Obtain the trend growth reasons and growth trend lines corresponding to the target carbon emission growth trend intensity from the preset trend database.

[0230] The causes of the trend growth refer to the historical triggers that cause abnormal growth in target carbon emissions, such as changes in production plans or the collective operation of air conditioners due to abnormal weather. The growth trend line is a visualized trajectory generated through historical data analysis, reflecting the change in carbon emission intensity over time under this cause. The two form a "cause-curve" correspondence in the trend database.

[0231] A trend database is a pre-established collection of historical carbon emission data and corresponding causes of increase. It stores potential triggers for different intensities of increase trends, such as equipment failure and production adjustments, along with their corresponding typical growth trend lines. By analyzing a large number of historical abnormal increases in carbon emissions, staff identify common causes and corresponding abnormal increase curves, storing this information in the trend growth cause module and the corresponding growth trend line module. This ensures that the trend database includes various conditions for abnormal increases in carbon emissions, their trend growth causes, and future growth trends.

[0232] When the system detects that the target carbon emission growth rate is too high, it will retrieve historical causes and corresponding trend curves from the database that match the intensity of the current growth trend.

[0233] Step 614031: Conduct a credibility analysis on the revised future carbon emission curve and growth trend line to obtain credibility.

[0234] Credibility analysis is a technical means of assessing whether the current growth pattern conforms to typical historical scenarios by quantifying the correlation between the intensity of the target carbon emission growth trend and the growth trend line of the corresponding causes of the growth trend. Credibility is defined as the degree of matching between two growth patterns in terms of form, trend and change law, and its value ranges from [0, 1]. The higher the value, the better the consistency between the intensity of the current carbon emission growth trend and the historical attribution scenario.

[0235] Step 6140310: If the credibility is lower than the preset credibility threshold, an early warning signal for abnormal carbon emission growth is issued.

[0236] If the credibility is lower than the preset credibility threshold, it means that the analysis found that the current carbon emission growth trend intensity is not highly correlated with the growth trend line of the corresponding trend cause. This indicates that the trend cause cannot reflect the actual reason for the abnormal carbon emission growth rate of the target building. Therefore, it is still necessary to issue an early warning signal for abnormal carbon emission growth.

[0237] Step 6140311: If the credibility is higher than the credibility threshold and the target carbon emission intensity is higher than the carbon emission threshold, then issue a carbon emission exceeding warning signal.

[0238] If the credibility level is higher than the preset credibility threshold, and the target carbon emission intensity is higher than the carbon emission threshold, it indicates that the current carbon emission growth pattern has been matched with historically known causes, ruling out the possibility of numerical anomalies. However, even if the causes conform to historical patterns, quantitative predictions indicate that emissions will still exceed the limit, requiring human intervention. Therefore, it is necessary to issue a carbon emission exceedance warning signal.

[0239] Step 6140312: If the credibility is higher than the credibility threshold and the target carbon emission intensity is lower than the carbon emission threshold, then no carbon emission exceeding warning signal will be issued.

[0240] The condition that the credibility is higher than the preset credibility threshold and the target carbon emission intensity is less than the carbon emission threshold indicates that although the current target carbon emission growth trend intensity exceeds the growth trend threshold, the future carbon emission intensity will not exceed the carbon emission threshold because the carbon emission growth trend is highly consistent with a typical scenario in the historical database. The growth pattern trend attribution is credible and the risk is controllable. Therefore, it is not necessary to issue a carbon emission exceedance warning signal.

[0241] A method for issuing a numerical anomaly warning signal when the target carbon emission intensity is less than the carbon emission threshold and the target carbon emission growth trend intensity is less than the growth trend threshold, the method comprising:

[0242] Step 61404: Identify carbon emission pathways based on inter-building relationships.

[0243] A carbon emission pathway refers to the path of energy flow and carbon emission impact based on the relationships between buildings, and its path is determined by the direction of energy input. Researchers construct carbon emission impact pathways by analyzing the correlation between the target building and related buildings in terms of energy supply and demand and carbon emission intensity, ensuring that buildings along the pathway have the same energy source and similar carbon emissions.

[0244] Step 61405: Determine nodal buildings based on carbon emission pathways.

[0245] Nodal buildings are a collective term for buildings along carbon emission pathways.

[0246] Step 61406: Obtain the total energy level input for each carbon emission path.

[0247] Total energy volume refers to the total amount of various energy inputs transmitted to the target building and its associated nodes along a certain carbon emission path within a specific time period, quantified using a unified unit.

[0248] Step 61407: Obtain real-time carbon emission data for the node building.

[0249] The method for obtaining real-time carbon emission data of nodes is the same as the method for obtaining real-time carbon emission data of targets in step 3, and will not be repeated here.

[0250] Step 61408: Determine the total carbon emission level based on the real-time carbon emission data of all node buildings on the carbon emission path to which the target building belongs.

[0251] Total carbon emissions refer to the cumulative value obtained by aggregating real-time carbon emission data from all buildings along the carbon emission path of the target building and converting it to a uniform energy unit using an energy conversion factor. This value is calculated by weighting the relationship between the carbon emission intensity of each node and its energy input, and is used to characterize the correspondence between the overall carbon emissions and energy consumption of the path.

[0252] Step 61409: When the total carbon emission level does not match the total energy level, an abnormal value warning signal is issued for all buildings along the carbon emission path.

[0253] When the total carbon emissions level deviates from the total energy level, it indicates that the efficiency of energy input and carbon emission output within the path deviates significantly from historical benchmarks or theoretical values. This suggests a potential for inaccurate carbon emission figures along the path, triggering a full-path warning to check the accuracy of carbon emission intensity for all buildings along the path.

[0254] Optional, also includes:

[0255] Step 8: When historical carbon emission data for the target building is missing, obtain historical carbon emission data for highly similar buildings.

[0256] Step 80: Replace the target building's historical target carbon emission data with highly similar historical carbon emission data.

Claims

1. A method for early warning of carbon emissions of a smart building based on big data analysis, characterized in that, include: Step 1: In response to the received target building, find the corresponding related buildings and the relationship between buildings according to the preset building carbon emission database. The relationship between buildings refers to the relationship between carbon emissions between buildings obtained based on the correlation analysis of multi-dimensional data in the building carbon emission database. Step 2: Based on the relationships between buildings, classify the associated buildings into similar buildings and related buildings. Similar buildings refer to building groups whose carbon emission trends are synchronized with the target building in real time and whose carbon emission magnitude, carbon emission type and building functional attributes are highly matched. Related buildings refer to building groups whose carbon emissions are inversely linked, delayed in growth or complementary to the target building. The determination of similar buildings and related buildings is based on the relationships between buildings. If the relationships between buildings obtained from the building carbon emission database are the same in emission trends or the same in fluctuation cycle, then the two buildings are identified as similar buildings. If the relationship between buildings obtained from the building carbon emission database is that the emission trends are opposite or complementary, then the two buildings are identified as related buildings. Step 3: Obtain the target real-time carbon emission data and target historical carbon emission data of the target building, obtain the similar real-time carbon emission data and similar historical carbon emission data of the similar buildings, and obtain the relevant real-time carbon emission data and relevant historical carbon emission data of the related buildings; Step 4: Obtain the target fitted carbon emission future curve based on the target real-time carbon emission data and the target historical carbon emission data; obtain the similar fitted carbon emission future curve based on the similar real-time carbon emission data and the similar historical carbon emission data; obtain the relevant fitted carbon emission future curve based on the relevant real-time carbon emission data and the relevant historical carbon emission data. Step 5: Determine the correlation parameters based on the target fitted future carbon emission curve and the similar fitted future carbon emission curve; Step 60: When the correlation parameter is lower than the preset correlation threshold, a preset numerical anomaly warning signal is output. Step 61: When the correlation parameter is higher than the correlation threshold, the target fitted carbon emission future curve is corrected based on the correlation fitted carbon emission future curve to obtain the corrected carbon emission future curve; Step 62: Based on the modified carbon emission future curve, issue a preset carbon emission exceedance warning signal and a preset carbon emission abnormal growth warning signal. The carbon emission exceedance warning signal is a warning signal triggered when the modified carbon emission future curve shows that its carbon emission intensity within a set time interval will exceed a preset threshold, indicating that the building is at risk of exceeding the allowable carbon emission limit. The carbon emission abnormal growth warning signal is a warning signal triggered when the carbon emission growth trend intensity of the target building exceeds a preset rate threshold, indicating that although the current carbon emission intensity of the target building has not yet exceeded the limit, the predicted value shows a sharp upward trend, prompting early intervention to avoid the risk of exceeding the limit.

2. The method of claim 1, wherein the method is a smart building carbon emission early warning method for big data analysis. The method for establishing the building carbon emissions database includes: Step 7: In response to the received sample building and sample historical carbon emission data, input them into the building carbon emission database; Step 71: Perform correlation analysis based on all the sample buildings and their historical carbon emission data to obtain building correlation parameters; Step 72: If the absolute value of the building association parameter between the sample buildings is greater than the preset association parameter, then an association mapping relationship is constructed for the corresponding sample buildings; Step 73: Based on the historical carbon emission data of the sample buildings, classify the sample buildings into different preset building carbon emission ranges; Step 74: Identify the building type and carbon emission type of all the sample buildings; Step 75: Based on the building type and the carbon emission type, classify the sample buildings in each building carbon emission range into similar building groups, wherein the building type, the carbon emission type and the building carbon emission range to which all the sample buildings in the similar building groups belong are the same; Step 76: Extract the common features of the similar building groups; Step 77: Construct the inter-building relationships among the sample buildings based on the common features of the similar building groups; Step 78: Establish the building carbon emission database based on the association mapping relationship and the relationship between buildings. 3.The smart building carbon emission early warning method of big data analysis of claim 1, wherein, Also includes: Step 50: Assemble the similar buildings into a verification building group; Step 51: Perform a correlation test on the fitted future carbon emission curves of the similar buildings to obtain the similarity correlation parameters between the similar buildings and all other similar buildings; Step 52: When the similarity correlation parameter is less than a preset correlation threshold, the similar building corresponding to the similarity correlation parameter is removed from the verification building group to form a corrected verification building group; Step 53: Obtain the group-fitted carbon emission future curve based on the modified verification building complex, wherein the group-fitted carbon emission future curve is the average curve of the similar fitted carbon emission future curves of all similar buildings in the modified verification building complex. Step 54: Perform a correlation test on the target fitted future carbon emission curve and the group fitted future carbon emission curve to obtain the target correlation parameter; Step 540: When the target correlation parameter is less than the preset target correlation threshold, the numerical anomaly warning signal is issued; Step 541: When the target correlation parameter is greater than the target correlation threshold, the numerical abnormality warning signal will not be issued.

4. The method of claim 3, wherein the method further comprises: It also includes a method for issuing a numerical anomaly warning signal when the target correlation parameter is greater than the target correlation threshold, the method comprising: Step 5410: Sort the corrected verification building groups from high to low based on the correlation parameters to obtain the ordered corrected verification building groups; Step 5411: Select a preset number of similar buildings from the building group in descending order of the ordered correction verification, define them as highly similar buildings, define the similar real-time carbon emission data of highly similar buildings as highly similar real-time carbon emission data, and define the similar historical carbon emission data of highly similar buildings as highly similar historical carbon emission data. Step 5412: Perform historical data correlation analysis on the target historical carbon emission data and the highly similar historical carbon emission data to obtain historical data correlation parameters; Step 5413: Obtain a highly similar fitted future carbon emission curve based on the highly similar real-time carbon emission data and the highly similar historical carbon emission data; Step 5414: Perform correlation analysis on the target fitted future carbon emission curve and the highly similar fitted future carbon emission curve to obtain the correlation parameters of the predicted curve; Step 5415: Based on the predicted curve correlation parameters, find the corresponding predicted curve correlation parameter range from the preset correlation database; Step 5416: If the correlation parameter of the predicted curve does not fall within the range of the correlation parameter of the predicted curve, then issue the numerical anomaly warning signal.

5. The method of claim 4, wherein the method further comprises: It also includes a method for issuing the numerical anomaly warning signal even when the correlation parameter of the prediction curve does not fall within the range of the correlation parameter of the prediction curve, the method comprising: Step 54160: Obtain the target carbon emission types and target carbon emission percentages of the target building, and obtain the highly similar carbon emission types and highly similar carbon emission percentages of the highly similar buildings; Step 54161: Perform a significance test on the target carbon emission percentage and the highly similar carbon emission percentage to obtain the significant difference value; Step 54162: If the significant difference value is greater than the preset difference value, then issue the abnormal value warning signal.

6. The method of claim 1, wherein the method further comprises: When the correlation parameter is higher than the correlation threshold, the target fitted carbon emission future curve is corrected based on the correlation fitted carbon emission future curve. The method for obtaining the corrected carbon emission future curve includes: Step 610: Construct a network of relationships between the related buildings based on the relationships between the buildings; Step 611: Analyze the network to obtain the association attributes and association strength between the associated buildings; Step 612: Based on the correlation attributes, the correlation strength, and the correlation fitted future carbon emission curve, obtain the fitting carbon emission future curve correction method from the preset building data correction library; Step 613: Based on the fitted carbon emission future curve correction method, correct the target fitted carbon emission future curve of the target building to obtain the corrected carbon emission future curve; Step 614: Based on the modified carbon emission future curve, obtain the target carbon emission intensity and the target carbon emission growth trend intensity within the future time interval; Step 6140: If the target carbon emission intensity is less than the preset carbon emission threshold and the target carbon emission growth trend intensity is less than the preset growth trend threshold, then the carbon emission exceeding warning signal and the carbon emission abnormal growth warning signal will not be issued. Step 6141: If the target carbon emission intensity is greater than the carbon emission threshold, then issue the carbon emission exceedance warning signal; Step 6142: If the intensity of the target carbon emission growth trend is greater than the growth trend threshold, then issue the abnormal carbon emission growth warning signal.

7. The method of claim 6, wherein the method further comprises: The method also includes issuing an abnormal carbon emission growth warning signal even when the carbon emission intensity is less than the carbon emission threshold and the carbon emission growth trend intensity is less than the growth trend threshold. Step 61400: Based on the inter-building relationship between the target building and the related buildings, classify the related buildings into buildings with the same trend and buildings with different trends; Step 61401: Obtain the future carbon emission curves of the same-trend buildings; Step 61402: Perform correlation analysis between the modified future carbon emission curve and the fitted future carbon emission curve with the same trend to obtain the trend correlation value; Step 61403: When the trend correlation value is greater than the preset trend threshold, and the abnormal carbon emission growth warning signal is received from the building with the same trend, the abnormal carbon emission growth warning signal is also issued to the target building. 8.The smart building carbon emission early warning method of big data analysis of claim 7, wherein, A method for not issuing a carbon emission abnormal growth warning signal when the trend correlation value is greater than the trend threshold and a carbon emission abnormal growth warning signal is received from a building exhibiting the same trend may be issued. The method includes: Step 614030: Obtain the trend growth reasons and growth trend lines corresponding to the target carbon emission growth trend intensity from the preset trend database; Step 614031: Perform a credibility analysis on the modified future carbon emission curve and the growth trend line to obtain the credibility. Step 6140310: If the credibility is lower than a preset credibility threshold, then issue the abnormal carbon emission growth warning signal; Step 6140311: If the credibility is higher than the credibility threshold and the target carbon emission intensity is higher than the carbon emission threshold, then issue a carbon emission exceeding warning signal; Step 6140312: If the credibility is higher than the credibility threshold and the target carbon emission intensity is less than the carbon emission threshold, then the carbon emission exceeding warning signal will not be issued. 9.The smart building carbon emission early warning method of big data analysis of claim 6, wherein, It also includes a method for issuing the numerical anomaly warning signal when the target carbon emission intensity is less than the carbon emission threshold and the target carbon emission growth trend intensity is less than the growth trend threshold, the method comprising: Step 61404: Confirm carbon emission pathways based on the relationships between the buildings; Step 61405: Determine the nodal buildings based on the carbon emission pathway; Step 61406: Obtain the total energy level of each carbon emission path input; Step 61407: Obtain the real-time carbon emission data of the node building; Step 61408: Determine the total carbon emission level based on the real-time carbon emission data of all node buildings on the carbon emission path to which the target building belongs; Step 61409: When the total carbon emission level does not match the total energy level, an abnormal value warning signal is issued to all buildings along the carbon emission path. 10.The smart building carbon emission early warning method of big data analysis of claim 4, wherein, Also includes: Step 8: When the target building lacks the target historical carbon emission data, obtain the highly similar historical carbon emission data of the highly similar building; Step 80: Replace the target historical carbon emission data of the target building with the highly similar historical carbon emission data.

Images