Zero-carbon park carbon emission assessment method based on multi-source data fusion

Abstract

The invention relates to the technical field of carbon emission evaluation, and discloses a zero-carbon park carbon emission evaluation method based on multi-source data fusion, which comprises the steps of dividing a zero-carbon park into a plurality of to-be-evaluated areas, obtaining related data of each to-be-evaluated area, and generating a dominant related data set and a hidden related data set of each to-be-evaluated area; respectively constructing a first data fusion model and a second data fusion model according to the dominant correlation data set and the implicit correlation data set, and obtaining a predicted carbon emission evaluation value of the corresponding to-be-evaluated region in the current monitoring period by combining actual correlation data of the current monitoring period; according to the method, the predicted carbon emission evaluation value is accurately evaluated, whether the correction instruction of the predicted carbon emission evaluation value is generated or not is judged according to the evaluation result, the real-time dynamic carbon emission evaluation requirement is met, the quantitative accuracy of the carbon emission intensity is improved, and the technical problem of hysteresis of carbon accounting is solved.

Description

Technical Field

[0001] This application relates to the technical field of carbon emission assessment, and particularly to a zero-carbon park carbon emission assessment method based on multi-source data fusion. Background Art

[0002] The accurate assessment of carbon emissions is one of the important goals of zero-carbon parks. In the existing technology, the complex and diverse carbon emission data and traditional static algorithms cannot meet the real-time and dynamic carbon emission assessment requirements, reducing the quantitative accuracy of carbon emission intensity and bringing significant economic losses to zero-carbon parks. Therefore, there is an urgent need for a zero-carbon park carbon emission assessment method based on multi-source data fusion to improve the accuracy of carbon emission assessment and solve the technical problem of lag in carbon accounting. Summary of the Invention

[0003] To solve the above technical problems, this application provides a zero-carbon park carbon emission assessment method based on multi-source data fusion. By dividing the zero-carbon park into multiple areas to be evaluated, constructing a first data fusion model and a second data fusion model according to the explicit relevant data set and implicit relevant data set of each area to be evaluated, and generating a predicted carbon emission assessment value, it meets the real-time and dynamic carbon emission assessment requirements, improves the quantitative accuracy of carbon emission intensity, and solves the technical problem of lag in carbon accounting.

[0004] In some embodiments of this application, a zero-carbon park carbon emission assessment method based on multi-source data fusion is provided, including:

[0005] Dividing the zero-carbon park into multiple areas to be evaluated, obtaining the relevant data of each area to be evaluated, and processing the relevant data to generate an explicit relevant data set and an implicit relevant data set for each area to be evaluated;

[0006] Constructing a first data fusion model and a second data fusion model for the corresponding area to be evaluated according to the explicit relevant data set and implicit relevant data set of the same area to be evaluated, and combining the actual relevant data of the current monitoring period to obtain the predicted carbon emission assessment value of the corresponding area to be evaluated in the current monitoring period;

[0007] Evaluating the accuracy of the predicted carbon emission assessment value, and judging whether to generate a correction instruction for the predicted carbon emission assessment value according to the evaluation result.

[0008] In some embodiments of this application, obtaining the relevant data of each area to be evaluated includes:

[0009] Presetting multiple carbon emission assessment indicators;

[0010] Obtaining multiple historical monitoring logs of each area to be evaluated, and extracting the historical monitoring data from each historical monitoring log;

[0011] Taking the historical monitoring period of each historical monitoring log as the time reference line, with the historical evaluation value of each carbon emission assessment index as the ordinate and the historical monitoring data as the abscissa, multiple change relationship curves for each historical monitoring period are constructed;

[0012] Determine whether there is an association relationship between each historical monitoring data and the historical evaluation value of each carbon emission assessment index among the multiple change relationship curves in the same historical monitoring period;

[0013] If within the first preset period, the variable value of the historical monitoring data is greater than the preset first change value threshold and the change value of the historical evaluation value of the corresponding carbon emission assessment index is greater than the preset second change value threshold, there is a dominant association relationship, and the first association degree is calculated;

[0014] If within the second preset period, the variable value of the historical monitoring data is greater than the preset first change value threshold and the change value of the historical evaluation value of the corresponding carbon emission assessment index is greater than the preset second change value threshold, there is a recessive association relationship, and the recessive time of the historical monitoring data and the corresponding carbon emission assessment index and the corresponding second association degree are calculated;

[0015] If the same historical monitoring data in different historical monitoring logs has an association relationship with at least one carbon emission assessment index, set the corresponding historical monitoring data as the relevant data for the corresponding area to be evaluated, and the relevant data includes first relevant data and second relevant data.

[0016] In some embodiments of the present application, the relevant data is processed to generate a dominant relevant data set and a recessive relevant data set for each area to be evaluated, including:

[0017] Process the relevant data, and the processing includes preprocessing and standardization processing;

[0018] Preprocess the relevant data, and the preprocessing includes unifying the time unit, removing outliers, filling in missing values, and data cleaning;

[0019] Perform standardization processing on the preprocessed relevant data. The standardization processing includes classifying the preprocessed first relevant data to obtain multiple types of first relevant data and standardizing the first relevant data of the same type, and also includes classifying the preprocessed second relevant data to obtain multiple types of second relevant data and standardizing the second relevant data of the same type;

[0020] Construct a dominant relevant data set based on the standardized multiple types of first relevant data, and construct a recessive relevant data set based on the standardized multiple types of first relevant data.

[0021] In some embodiments of the present application, classifying the preprocessed first related data includes classifying the first related data of the same carbon emission assessment index with an explicit association relationship, and reclassifying according to the data sources of the first related data of the same carbon emission assessment index with an explicit association relationship to obtain an explicit related data set;

[0022] Classifying the preprocessed second related data includes classifying the second related data of the same carbon emission assessment index with an implicit association relationship, and reclassifying according to the implicit time corresponding to the second related data of the same carbon emission assessment index with an implicit association relationship to obtain an implicit related data set.

[0023] In some embodiments of the present application, constructing a first data fusion model corresponding to the area to be evaluated according to the explicit related data set of the same area to be evaluated includes

[0024] Inputting different types of first related data in the explicit related data set into the first preset carbon emission assessment model of the corresponding carbon emission assessment index respectively, obtaining the predicted carbon emission assessment sub-values of different types of first related data and the corresponding carbon emission assessment index, comparing with the standard carbon emission assessment sub-value, and generating the first accuracy of different types of first related data according to the comparison result;

[0025] Setting the first weight coefficient of different types of first related data according to the first accuracy;

[0026] Setting the second weight coefficient of the corresponding first related data according to the first association degree between the same type of first related data in the explicit related data set and the corresponding carbon emission assessment index;

[0027] Inputting each first related data in the explicit related data set into the first preset carbon emission assessment model of the corresponding carbon emission assessment index respectively, obtaining the predicted carbon emission assessment sub-value corresponding to each first related data, comparing with the standard carbon emission assessment sub-value, and generating the second accuracy of each first related data according to the comparison result;

[0028] Setting the compensation coefficient of the corresponding first related data according to the second accuracy, correcting the second weight coefficient of the corresponding first related data according to the compensation coefficient, and generating the standard weight coefficient of the corresponding first related data according to the corrected second weight coefficient and the first weight coefficient;

[0029] Generating a first data fusion model corresponding to the area to be evaluated according to each first related data in the explicit related data set, the standard weight coefficient of the corresponding first related data, and the weight coefficient of the carbon emission assessment index.

[0030] In some embodiments of the present application, constructing a second data fusion model for a corresponding area to be evaluated according to the implicit correlation dataset of the same area to be evaluated includes:

[0031] Inputting different types of second correlation data in the implicit correlation dataset into the second preset carbon emission assessment model of the corresponding carbon emission assessment index, obtaining the predicted carbon emission assessment sub-values of different types of second correlation data and the corresponding carbon emission assessment index after the corresponding implicit time, comparing with the standard carbon emission assessment sub-values, and generating the third accuracy of different types of second correlation data according to the comparison results;

[0032] Setting the third weight coefficient of different types of second correlation data in the implicit correlation dataset according to the third accuracy;

[0033] Setting the fourth weight coefficient of the corresponding second correlation data according to the second correlation degree between the same type of second correlation data in the implicit correlation dataset and the corresponding carbon emission assessment index;

[0034] Obtaining the historical mutation values of multiple second correlation data of the same type in the implicit correlation dataset;

[0035] Inputting the multiple historical mutation values of the second correlation data of the same type into the second preset carbon emission assessment model of the corresponding carbon emission assessment index, obtaining the predicted carbon emission assessment sub-values of the corresponding second correlation data and the corresponding carbon emission assessment index after the corresponding implicit time, comparing with the corresponding standard carbon emission assessment sub-values, and generating the fourth accuracy of the corresponding second correlation data according to the comparison results;

[0036] Setting the compensation coefficient of the corresponding second correlation data according to the fourth accuracy, correcting the fourth weight coefficient according to the compensation coefficient, and generating the standard weight coefficient of the corresponding second correlation data according to the corrected fourth weight coefficient and the third weight coefficient;

[0037] Generating a second data fusion model for the corresponding area to be evaluated according to each second correlation data in the implicit correlation dataset, the standard weight coefficient of the corresponding second correlation data, and the weight coefficient of the corresponding carbon emission assessment index.

[0038] In some embodiments of the present application, combining the actual correlation data of the current monitoring period to obtain the predicted carbon emission assessment value of the corresponding area to be evaluated in the current monitoring period includes:

[0039] Obtaining the actual correlation data at multiple monitoring time nodes of the current monitoring period, where the actual correlation data includes first actual correlation data and second actual correlation data;

[0040] Input the first actual relevant data at the same monitoring time node into the first data fusion model corresponding to the area to be evaluated, and obtain the first predicted carbon emission assessment value at the corresponding monitoring time node;

[0041] Analyze the second actual relevant data in the current monitoring period, screen out the second actual relevant data with mutation time nodes and the corresponding actual mutation values, and determine the monitoring time nodes when the second actual relevant data affects the corresponding carbon emission assessment indicators;

[0042] Input the second actual relevant data and the corresponding mutation values into the second data fusion model to obtain the predicted lag value of the carbon emission assessment value at the corresponding monitoring time node;

[0043] If there is no lag value at the monitoring time node, set the first predicted carbon emission assessment value as the predicted carbon emission assessment value at the corresponding monitoring time node of the area to be evaluated;

[0044] If there is a lag value at the monitoring time node, correct the first predicted carbon emission assessment value according to the lag value, and set the corrected first predicted carbon emission assessment value as the predicted carbon emission assessment value at the corresponding monitoring time node of the area to be evaluated.

[0045] In some embodiments of the present application, the accuracy evaluation of the predicted carbon emission assessment value includes:

[0046] Extract multiple carbon emission characteristics of the predicted carbon emission assessment value in the current monitoring period;

[0047] Compare the carbon emission characteristics with the corresponding standard carbon emission characteristics to obtain multiple carbon emission characteristic differences;

[0048] Match a corresponding difference coefficient to each carbon emission characteristic difference, and generate a comprehensive characteristic difference coefficient according to the difference coefficients corresponding to multiple carbon emission characteristics and the corresponding weight coefficients;

[0049] Set the accuracy evaluation value of the predicted carbon emission assessment value for the corresponding area to be evaluated in the current monitoring period according to the comprehensive characteristic difference coefficient.

[0050] In some embodiments of the present application, judging whether to generate a correction instruction for the predicted carbon emission assessment value according to the evaluation result includes:

[0051] Preset an accuracy evaluation threshold;

[0052] If the accuracy evaluation value is less than the accuracy evaluation threshold, generate a correction instruction for the predicted carbon emission assessment value;

[0053] If the accuracy evaluation value is not less than the accuracy evaluation threshold, generate a correction instruction for the predicted carbon emission assessment value.

[0054] Compared with the prior art, the carbon emission assessment method for zero-carbon parks based on multi-source data fusion in the embodiments of the present application has the following beneficial effects:

[0055] By dividing the zero-carbon park into multiple areas to be evaluated, constructing a first data fusion model and a second data fusion model according to the explicit relevant data set and the implicit relevant data set of each area to be evaluated, and generating a predicted carbon emission assessment value, it meets the real-time dynamic carbon emission assessment requirements, improves the quantification accuracy of carbon emission intensity, and solves the technical problem of lag in carbon accounting. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a schematic flow chart of the carbon emission assessment method for zero-carbon parks based on multi-source data fusion in the embodiments of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0057] The following further describes the specific embodiments of the present application in detail with reference to the drawings and embodiments. The following embodiments are used to illustrate the present application, but are not intended to limit the scope of the present application.

[0058] In the description of the present application, it should be understood that the orientation or positional relationship indicated by the terms "center", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and thus should not be construed as limiting the present application.

[0059] The terms "first" and "second" are only used for descriptive purposes, and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the present application, unless otherwise specified, the meaning of "a plurality" is two or more.

[0060] In the description of the present application, it should be noted that unless otherwise clearly defined and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and it may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to specific circumstances.

[0061] Such as Figure 1As shown in the figure, the carbon emission assessment method for a zero-carbon park based on multi-source data fusion according to an embodiment of the present application includes:

[0062] Step S101: Divide the zero-carbon park into multiple areas to be evaluated, obtain the relevant data of each area to be evaluated, process the relevant data, and generate an explicit relevant data set and an implicit relevant data set for each area to be evaluated;

[0063] Step S102: Construct a first data fusion model and a second data fusion model for the corresponding area to be evaluated according to the explicit relevant data set and the implicit relevant data set of the same area to be evaluated, and combine the actual relevant data in the current monitoring period to obtain the predicted carbon emission assessment value of the corresponding area to be evaluated in the current monitoring period;

[0064] Step S103: Conduct a precision evaluation on the predicted carbon emission assessment value, and judge whether to generate a correction instruction for the predicted carbon emission assessment value according to the evaluation result.

[0065] In this embodiment, the relevant data refers to the combination of the park's energy production, consumption, and environmental monitoring data, which refers to the characteristic data that extracts the influence of different data sources on the carbon emission assessment value from the historical monitoring data. The multi-source abnormal data is fused to obtain the first data fusion model and the second data fusion model, improving the accurate quantification of the carbon emission assessment value.

[0066] In some embodiments of the present application, obtaining the relevant data of each area to be evaluated includes:

[0067] Preset multiple carbon emission assessment indicators in advance;

[0068] Obtain multiple historical monitoring logs of each area to be evaluated, and extract the historical monitoring data in each historical monitoring log;

[0069] Taking the historical monitoring period of each historical monitoring log as the time reference line, with the historical evaluation value of each carbon emission assessment indicator as the ordinate and the historical monitoring data as the abscissa, construct multiple change relationship curves for each historical monitoring period;

[0070] Judge whether there is an association relationship between each historical monitoring data and the historical evaluation value of each carbon emission assessment indicator among the multiple change relationship curves in the same historical monitoring period;

[0071] If within the first preset period, the variable quantity value of the historical monitoring data is greater than the preset first change quantity value threshold and the change quantity value of the historical evaluation value of the corresponding carbon emission assessment indicator is greater than the preset second change quantity value threshold, there is an explicit association relationship, and calculate the first association degree;

[0072] If within the second preset time period, the variable value of the historical monitoring data is greater than the preset first change value threshold and the change value of the historical evaluation value of the corresponding carbon emission evaluation index is greater than the preset second change value threshold, there is a hidden correlation relationship, and the hidden time between the historical monitoring data and the corresponding carbon emission evaluation index and the corresponding second correlation degree are calculated;

[0073] If the same historical monitoring data in different historical monitoring logs is associated with at least one carbon emission evaluation index, the corresponding historical monitoring data is set as the relevant data for the corresponding area to be evaluated, and the relevant data includes first relevant data and second relevant data.

[0074] In this embodiment, the carbon emission evaluation index refers to an index for accurately evaluating different types of carbon emissions, and the historical evaluation value refers to the accurate evaluation value of the corresponding type of carbon emission.

[0075] In this embodiment, the first preset time period is less than the second preset time period. The explicit correlation relationship means that after the historical monitoring data changes greatly, it immediately affects the historical evaluation value of the carbon emission evaluation index, that is, there is no lag time. The hidden correlation relationship means that after the historical monitoring changes greatly, after a short time interval, it affects the historical evaluation value of the carbon emission evaluation index, that is, there is a lag time, and the current lag time is the hidden time when there is a hidden correlation relationship.

[0076] In this embodiment, the first relevant data refers to the historical monitoring data that has an explicit correlation relationship with at least one carbon emission evaluation index, and the second relevant data refers to the historical monitoring data that has a hidden correlation relationship with at least one carbon emission evaluation index.

[0077] In this embodiment, by collecting the relevant data of different types of carbon emissions, without reducing the accuracy of carbon emissions, the data processing volume and analysis volume are reduced, laying a foundation for the subsequent construction of a multi-source relevant data set and a multi-source data fusion model, improving the accuracy of the multi-source data fusion model, thereby ensuring the accurate value of the carbon emission intensity and solving the problem of lag in carbon accounting.

[0078] In some embodiments of the present application, the relevant data is processed to generate an explicit relevant data set and a hidden relevant data set for each area to be evaluated, including:

[0079] The relevant data is processed, and the processing includes preprocessing and normalization processing;

[0080] The relevant data is preprocessed, and the preprocessing includes unifying the time unit, removing outliers, filling in missing values, and data cleaning processing;

[0081] Perform standardization processing on the preprocessed relevant data. The standardization processing includes classifying the preprocessed first relevant data to obtain multiple types of first relevant data, and standardizing the first relevant data of the same type. It also includes classifying the preprocessed second relevant data to obtain multiple types of second relevant data, and standardizing the second relevant data of the same type;

[0082] Construct an explicit relevant data set based on the standardized multiple types of first relevant data, and construct an implicit relevant data set based on the standardized multiple types of first relevant data.

[0083] In this embodiment, unification is to uniformly process the data to ensure that the data has similar scales and distributions. Missing values refer to abnormal losses and other situations existing in the data. Outliers are data in the data set that are significantly different from other data. Data cleaning processing is a process of identifying and correcting data with errors, inconsistencies, or incompleteness in the data set.

[0084] In this embodiment, standardizing the relevant data of the same type includes:

[0085]

[0086] Among them, x′(i) is the data after standardizing the i-th relevant data in the same type, and n is the number of data in the same type.

[0087] In this embodiment, by performing preprocessing and standardization processing on the relevant data, the quality and credibility of the relevant data are improved, and the relevant data is divided into an explicit relevant data set and an implicit relevant data set, laying a data foundation for subsequent determination of multi-source data fusion and improving the accuracy of carbon emission assessment.

[0088] In some embodiments of the present application, classifying the preprocessed first relevant data includes classifying the first relevant data of the same carbon emission assessment index with an explicit association relationship, and reclassifying according to the data sources corresponding to the first relevant data of the same carbon emission assessment index with an explicit association relationship to obtain an explicit relevant data set;

[0089] Classifying the preprocessed second relevant data includes classifying the second relevant data of the same carbon emission assessment index with an implicit association relationship, and reclassifying according to the implicit time corresponding to the second relevant data of the same carbon emission assessment index with an implicit association relationship to obtain an implicit relevant data set.

[0090] In this embodiment, the second relevant data with the same implicit time is classified according to the preset duration of the implicit time, and the second relevant data with the same preset duration is divided into the same type. The preset duration is set according to the time interval between adjacent monitoring time nodes in the current monitoring period. The first preset duration is the time length of one time interval, the second preset time interval is the time length of two time intervals, and the third preset duration and the fourth preset duration are obtained in sequence.

[0091] In this embodiment, by constructing an explicit relevant data set and an implicit relevant data set, the dynamic relationship between different relevant data and the corresponding carbon emission assessment indicators is determined, and the accurate quantification of the carbon emission assessment value is realized through the real-time dynamic relationship, so as to solve the problem of the lag of static carbon accounting.

[0092] In some embodiments of the present application, a first data fusion model corresponding to the area to be evaluated is constructed according to the explicit relevant data set of the same area to be evaluated, including

[0093] The first relevant data of different types in the explicit relevant data set are respectively input into the first preset carbon emission assessment models of the corresponding carbon emission assessment indicators, and the predicted carbon emission assessment sub-values of the first relevant data of different types and the corresponding carbon emission assessment indicators are obtained, and compared with the standard carbon emission assessment sub-values. According to the comparison results, the first accuracy of the first relevant data of different types is generated;

[0094] The first weight coefficients of the first relevant data of different types are set according to the first accuracy;

[0095] The second weight coefficients corresponding to the first relevant data are set according to the first correlation degree between the first relevant data of the same type in the explicit relevant data set and the corresponding carbon emission assessment indicators;

[0096] Each first relevant data in the explicit relevant data set is respectively input into the first preset carbon emission assessment model of the corresponding carbon emission assessment indicator, and the predicted carbon emission assessment sub-value corresponding to each first relevant data is obtained, and compared with the standard carbon emission assessment sub-value. According to the comparison results, the second accuracy of each first relevant data is generated;

[0097] The compensation coefficient corresponding to the first relevant data is set according to the second accuracy, the second weight coefficient of the corresponding first relevant data is corrected according to the compensation coefficient, and the standard weight coefficient corresponding to the first relevant data is generated according to the corrected second weight coefficient and the first weight coefficient;

[0098] The first data fusion model corresponding to the area to be evaluated is generated according to each first relevant data in the explicit relevant data set, the standard weight coefficient corresponding to the first relevant data, and the weight coefficient of the carbon emission assessment indicator.

[0099] In this embodiment, the first preset carbon emission assessment model for each carbon emission assessment indicator is pre-trained.

[0100] In this embodiment, the first accuracy is used to evaluate the evaluation accuracy of different types of first-related data in the explicit correlation dataset for the carbon emission assessment value of the corresponding carbon emission assessment indicator, that is, the evaluation accuracy of different data sources for the carbon emission assessment value of the same carbon emission assessment indicator. When the first accuracy is greater, the first weight coefficient of the corresponding type of first-related data is greater.

[0101] In this embodiment, the greater the first correlation degree, the greater the corresponding second weight coefficient. The second accuracy refers to the evaluation accuracy of each first-related data in the explicit correlation dataset for the carbon emission assessment value of the corresponding carbon emission assessment indicator. When the second accuracy is greater, the compensation coefficient is greater, and vice versa. The value range of the compensation coefficient is (0.8, 1.2). The second weight coefficient is corrected according to the compensation coefficient to improve the accuracy of each first-related data for evaluating carbon emissions of the corresponding carbon emission assessment indicator.

[0102] In this embodiment, the standard weight coefficient of each first-related data is calculated, and data fusion is performed according to each first-related data, the corresponding standard weight coefficient, and the weight coefficient of the carbon emission assessment indicator corresponding to the first-related data to obtain a first data fusion model. The carbon emission assessment value is predicted according to the first data fusion model to improve the prediction accuracy of the carbon emission assessment value.

[0103] In some embodiments of the present application, a second data fusion model corresponding to the area to be evaluated is constructed according to the implicit correlation dataset of the same area to be evaluated, including:

[0104] Input different types of second-related data in the implicit correlation dataset into the second preset carbon emission assessment model of the corresponding carbon emission assessment indicator to obtain predicted carbon emission assessment sub-values of different types of second-related data and the corresponding carbon emission assessment indicators after the corresponding implicit time, and compare them with the standard carbon emission assessment sub-values. According to the comparison results, the third accuracy of different types of second-related data is generated;

[0105] Set the third weight coefficient of different types of second-related data in the implicit correlation dataset according to the third accuracy;

[0106] Set the fourth weight coefficient of the corresponding second-related data according to the second correlation degree between the same type of second-related data in the implicit correlation dataset and the corresponding carbon emission assessment indicator;

[0107] Obtain the historical mutation values of multiple second-related data of the same type in the implicit correlation dataset;

[0108] Input multiple historical mutation values of the second relevant data of the same type into the second preset carbon emission assessment model of the corresponding carbon emission assessment indicator, obtain the predicted carbon emission assessment sub-value corresponding to the second relevant data and the corresponding carbon emission assessment indicator after the corresponding implicit time, compare it with the corresponding standard carbon emission assessment sub-value, and generate the fourth accuracy corresponding to the second relevant data according to the comparison result;

[0109] Set the compensation coefficient corresponding to the second relevant data according to the fourth accuracy, correct the fourth weight coefficient according to the compensation coefficient, and generate the standard weight coefficient corresponding to the second relevant data according to the corrected fourth weight coefficient and the third weight coefficient;

[0110] Generate the second data fusion model corresponding to the area to be evaluated according to each second relevant data in the implicit relevant data set, the standard weight coefficient corresponding to the second relevant data, and the weight coefficient of the corresponding carbon emission assessment indicator.

[0111] In this embodiment, the second preset carbon emission assessment model of each carbon emission assessment indicator is pre-trained.

[0112] In this embodiment, the second accuracy is used to evaluate the evaluation accuracy of different types of second relevant data of the same carbon emission assessment indicator in the implicit relevant data set for the carbon emission assessment value of the corresponding carbon emission evaluation indicator. When the second accuracy is larger, the third weight coefficient of the corresponding type of second relevant data is larger.

[0113] In this embodiment, the greater the second correlation degree, the greater the corresponding fourth weight coefficient. The fourth accuracy refers to the evaluation accuracy of each second relevant data for the carbon emission assessment value after the implicit time of the corresponding carbon emission evaluation indicator. When the fourth accuracy is larger, the compensation coefficient is larger, and vice versa. The value range of the compensation coefficient is (0.8, 1.2). Correct the fourth weight coefficient according to the compensation coefficient to improve the accuracy of the lag value of the carbon emission assessment value evaluated by each second relevant data for the corresponding carbon emission evaluation indicator.

[0114] In this embodiment, calculate the standard weight coefficient of each second relevant data, perform data fusion according to each second relevant data, the corresponding standard weight coefficient, and the weight coefficient of the carbon emission assessment indicator corresponding to the second relevant data, obtain the second data fusion model, and predict the lag value of the carbon emission assessment value after the corresponding implicit time according to the second data fusion model, improve the prediction accuracy of the carbon emission assessment value, solve the problem of carbon accounting lag, and obtain the accurate carbon emission assessment value in real time.

[0115] In some embodiments of the present application, combining the actual relevant data of the current monitoring period, obtain the predicted carbon emission assessment value of the area to be evaluated in the current monitoring period, including:

[0116] Obtain the actual relevant data at multiple monitoring time nodes in the current monitoring period, where the actual relevant data includes first actual relevant data and second actual relevant data;

[0117] Input the first actual relevant data at the same monitoring time node into the first data fusion model for the corresponding area to be evaluated to obtain the first predicted carbon emission assessment value at the corresponding monitoring time node;

[0118] Analyze the second actual relevant data in the current monitoring period, screen out the second actual relevant data with mutation time nodes and the corresponding actual mutation values, and determine the monitoring time nodes when the second actual relevant data affects the corresponding carbon emission assessment indicators;

[0119] Input the second actual relevant data and the corresponding mutation values into the second data fusion model to obtain the predicted lag value of the carbon emission assessment value at the corresponding monitoring time node;

[0120] If there is no lag value at the monitoring time node, set the first predicted carbon emission assessment value as the predicted carbon emission assessment value at the corresponding monitoring time node of the area to be evaluated;

[0121] If there is a lag value at the monitoring time node, correct the first predicted carbon emission assessment value according to the lag value, and set the corrected first predicted carbon emission assessment value as the predicted carbon emission assessment value at the corresponding monitoring time node of the area to be evaluated.

[0122] In this embodiment, the first predicted carbon emission assessment value at each monitoring time node in the current monitoring period is obtained through the first actual relevant data and the first data fusion model. The second actual relevant data with mutations and the corresponding mutation values are obtained according to the fluctuation degree of the second actual relevant data. The actual lag value after the hidden time is obtained according to the second actual relevant data and the mutation values, improving the accuracy of the predicted carbon emission assessment value at each monitoring time node and solving the problem of lag in carbon accounting.

[0123] In some embodiments of the present application, the accuracy evaluation of the predicted carbon emission assessment value includes:

[0124] Extract multiple carbon emission characteristics of the predicted carbon emission assessment value in the current monitoring period;

[0125] Compare the carbon emission characteristics with the corresponding standard carbon emission characteristics to obtain multiple carbon emission characteristic differences;

[0126] Match a corresponding difference coefficient to each carbon emission characteristic difference, and generate a comprehensive characteristic difference coefficient according to the difference coefficients corresponding to multiple carbon emission characteristics and the corresponding weight coefficients;

[0127] Set the accuracy evaluation value of the predicted carbon emission evaluation value of the corresponding area to be evaluated in the current monitoring period according to the comprehensive feature difference coefficient.

[0128] In this embodiment, the carbon emission characteristics include the change trend of the predicted carbon emission evaluation value in the current monitoring period, the change rate and the change amount value at adjacent monitoring time nodes, as well as the change trend of the predicted carbon emission sub - evaluation value of different carbon emission evaluation indicators in the current monitoring period, the change rate and the change amount value at adjacent monitoring time nodes. The standard carbon emission characteristics are set in advance, including various normal change trends of the normal carbon emission evaluation value, multiple normal change rates and multiple normal change amount values at adjacent monitoring time nodes, as well as various change trends of the normal carbon emission sub - evaluation value of different carbon emission evaluation indicators, multiple normal change rates and multiple normal change amount values at adjacent monitoring time nodes.

[0129] In this embodiment, the greater the difference in carbon emission characteristics, the greater the corresponding difference coefficient, and vice versa. It is set according to the matching relationship between historical feature differences and corresponding historical difference coefficients.

[0130] In this embodiment, a first preset comprehensive feature difference coefficient interval, a second preset comprehensive feature difference coefficient interval, a third preset comprehensive feature difference coefficient interval, and a fourth preset comprehensive feature difference coefficient interval are preset in advance.

[0131] When the comprehensive feature difference coefficient is within the first preset comprehensive feature difference coefficient interval, set the accuracy evaluation value to the first preset accuracy evaluation value.

[0132] When the comprehensive feature difference coefficient is within the second preset comprehensive feature difference coefficient interval, set the accuracy evaluation value to the second preset accuracy evaluation value.

[0133] When the comprehensive feature difference coefficient is within the third preset comprehensive feature difference coefficient interval, set the accuracy evaluation value to the third preset accuracy evaluation value.

[0134] When the comprehensive feature difference coefficient is within the fourth preset comprehensive feature difference coefficient interval, set the accuracy evaluation value to the fourth preset accuracy evaluation value.

[0135] And the first preset comprehensive feature difference coefficient interval < the second preset comprehensive feature difference coefficient interval < the third preset comprehensive feature difference coefficient interval < the fourth preset comprehensive feature difference coefficient interval, and the first preset accuracy evaluation value < the second preset accuracy evaluation value < the third preset accuracy evaluation value < the fourth preset accuracy evaluation value.

[0136] In this embodiment, the larger the preset accuracy evaluation value where the accuracy evaluation value is located, the more accurate the corresponding predicted carbon emission evaluation value is.

[0137] In some embodiments of the present application, judging whether to generate a correction instruction for the predicted carbon emission evaluation value according to the evaluation result includes:

[0138] Presetting an accuracy evaluation threshold in advance;

[0139] If the accuracy evaluation value is less than the accuracy evaluation threshold, generate a correction instruction for the predicted carbon emission evaluation value;

[0140] If the accuracy evaluation value is not less than the accuracy evaluation threshold, generate a correction instruction for the predicted carbon emission evaluation value.

[0141] In this embodiment, if the accuracy evaluation value is less than the accuracy evaluation threshold, screen out the predicted carbon emission evaluation sub-value or the corresponding actual lag value of the carbon emission evaluation index that needs to be corrected according to the accuracy evaluation process, and optimize the first data fusion model and the second data fusion model to improve the accuracy of carbon emission evaluation.

[0142] The above are only the preferred embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the technical principle of the present application, several improvements and replacements can be made, and these improvements and replacements should also be regarded as the protection scope of the present application.

Claims

1. A zero-carbon park carbon emission assessment method based on multi-source data fusion, characterized in that: include: Divide the zero-carbon park into multiple areas to be evaluated, obtain relevant data of each area to be evaluated, process the relevant data, and generate explicit relevant data sets and implicit relevant data sets for each area to be evaluated; According to the explicit relevant data set and the implicit relevant data set of the same area to be evaluated, a first data fusion model and a second data fusion model of the corresponding area to be evaluated are respectively constructed, and combined with the actual relevant data of the current monitoring period, the predicted carbon emission assessment value of the corresponding area to be evaluated in the current monitoring period is obtained; The predicted carbon emission assessment value is evaluated for accuracy, and a determination is made based on the evaluation result whether to generate a correction instruction for the predicted carbon emission assessment value.

2. The zero-carbon park carbon emission assessment method based on multi-source data fusion according to claim 1 is characterized in that: Obtain relevant data for each area to be assessed, including: Pre-set multiple carbon emission assessment indicators; Obtain multiple historical monitoring logs for each area to be evaluated, and extract historical monitoring data from each historical monitoring log; Taking the historical monitoring period of each historical monitoring log as the time reference line, the historical evaluation value of each carbon emission evaluation indicator as the vertical coordinate, and the historical monitoring data as the horizontal coordinate, multiple change relationship curves for each historical monitoring period are constructed; Determine whether there is a correlation between each historical monitoring data in multiple change relationship curves of the same historical monitoring period and the historical evaluation value of each carbon emission evaluation indicator; If, within a first preset period of time, the variable value of the historical monitoring data is greater than a preset first change value threshold and the change value of the historical assessment value of the corresponding carbon emission assessment indicator is greater than a preset second change value threshold, then there is an explicit correlation relationship, and a first correlation degree is calculated; If, within the second preset time period, the variable value of the historical monitoring data is greater than the preset first change value threshold and the change value of the historical assessment value of the corresponding carbon emission assessment index is greater than the preset second change value threshold, then there is an implicit correlation relationship, and the implicit time of the historical monitoring data and the corresponding carbon emission assessment index and the corresponding second correlation degree are calculated; If the same historical monitoring data in different historical monitoring logs is associated with at least one carbon emission assessment indicator, the corresponding historical monitoring data is set as relevant data corresponding to the area to be assessed, and the relevant data includes first relevant data and second relevant data.

3. The zero-carbon park carbon emission assessment method based on multi-source data fusion as claimed in claim 2 is characterized in that: The relevant data are processed to generate an explicit relevant data set and an implicit relevant data set for each area to be evaluated, including: Processing the relevant data, wherein the processing includes preprocessing and standardization; Preprocessing the relevant data, including unifying time units, removing outliers, filling missing values, and data cleaning; The preprocessed related data is subjected to standardization processing, wherein the standardization processing includes classifying the preprocessed first related data to obtain multiple types of first related data, and standardizing the first related data of the same type, and also includes classifying the preprocessed second related data to obtain multiple types of second related data, and standardizing the second related data of the same type; An explicit correlation data set is constructed based on the standardized multiple types of first correlation data, and a hidden correlation data set is constructed based on the standardized multiple types of first correlation data.

4. The zero-carbon park carbon emission assessment method based on multi-source data fusion as claimed in claim 3 is characterized in that: Classifying the preprocessed first relevant data, including classifying the first relevant data of the same carbon emission assessment indicator with an explicit correlation relationship, and reclassifying according to the data source corresponding to the first relevant data of the same carbon emission assessment indicator with an explicit correlation relationship, to obtain an explicit correlation data set; The preprocessed second related data are classified, including classifying the second related data of the same carbon emission assessment indicator with implicit correlation, and reclassifying according to the implicit time corresponding to the second related data of the same carbon emission assessment indicator with implicit correlation, to obtain a cryptobiological related data set.

5. The zero-carbon park carbon emission assessment method based on multi-source data fusion as claimed in claim 4 is characterized in that: A first data fusion model corresponding to the area to be evaluated is constructed according to the explicit related data sets of the same area to be evaluated, including Inputting different types of first relevant data in the explicit relevant data set into the first preset carbon emission assessment model of the corresponding carbon emission assessment index respectively, obtaining different types of first relevant data and predicted carbon emission assessment sub-values ​​of the corresponding carbon emission assessment index, and comparing them with the standard carbon emission assessment sub-values, and generating first accuracies of different types of first relevant data according to the comparison results; Setting first weight coefficients of different types of first correlation data according to the first accuracy; Setting a second weight coefficient corresponding to the first relevant data according to a first correlation degree between the first relevant data of the same type in the explicit relevant data set and the corresponding carbon emission assessment indicator; Input each first relevant data in the explicit relevant data set into the first preset carbon emission assessment model of the corresponding carbon emission assessment index, obtain the predicted carbon emission assessment sub-value corresponding to each first relevant data, and compare it with the standard carbon emission assessment sub-value, and generate the second accuracy of each first relevant data according to the comparison result; Setting a compensation coefficient corresponding to the first correlation data according to the second accuracy, correcting a second weight coefficient corresponding to the first correlation data according to the compensation coefficient, and generating a standard weight coefficient corresponding to the first correlation data according to the corrected second weight coefficient and the first weight coefficient; A first data fusion model corresponding to the area to be evaluated is generated according to each first relevant data in the explicit relevant data set, the standard weight coefficient corresponding to the first relevant data, and the weight coefficient of the carbon emission evaluation index.

6. The zero-carbon park carbon emission assessment method based on multi-source data fusion according to claim 4 is characterized in that: Constructing a second data fusion model corresponding to the area to be evaluated according to the implicit related data set of the same area to be evaluated, including: Input different types of second related data in the implicit related data set into the second preset carbon emission assessment model of the corresponding carbon emission assessment index, obtain the predicted carbon emission assessment sub-values ​​of the different types of second related data and the corresponding carbon emission assessment index after the corresponding implicit time, and compare them with the standard carbon emission assessment sub-values, and generate third accuracies of different types of second related data according to the comparison results; setting third weight coefficients of different types of second related data in the implicit related data set according to a third accuracy; Setting a fourth weight coefficient corresponding to the second relevant data according to a second correlation degree between the same type of second relevant data in the implicit relevant data set and the corresponding carbon emission assessment indicator; Obtaining historical mutation values ​​of multiple second related data of the same type in the implicit related data set; Input multiple historical mutation values ​​of the same type of second relevant data into the second preset carbon emission assessment model of the corresponding carbon emission assessment index, obtain the predicted carbon emission assessment sub-values ​​of the corresponding second relevant data and the corresponding carbon emission assessment index after the corresponding implicit time, and compare them with the corresponding standard carbon emission assessment sub-values, and generate the fourth accuracy of the corresponding second relevant data according to the comparison result; Setting a compensation coefficient corresponding to the second correlation data according to the fourth accuracy, correcting the fourth weight coefficient according to the compensation coefficient, and generating a standard weight coefficient corresponding to the second correlation data according to the corrected fourth weight coefficient and the third weight coefficient; A second data fusion model corresponding to the area to be evaluated is generated according to each second related data in the implicit related data set, the standard weight coefficient corresponding to the second related data, and the weight coefficient corresponding to the carbon emission evaluation index.

7. The zero-carbon park carbon emission assessment method based on multi-source data fusion according to claim 6 is characterized in that: Combined with the actual relevant data of the current monitoring period, the predicted carbon emission assessment value of the corresponding area to be assessed in the current monitoring period is obtained, including: Acquire actual relevant data at multiple monitoring time nodes of the current monitoring cycle, wherein the actual relevant data includes first actual relevant data and second actual relevant data; Inputting the first actual relevant data at the same monitoring time node into the first data fusion model of the corresponding area to be evaluated to obtain the first predicted carbon emission evaluation value at the corresponding monitoring time node; Analyze the second actual relevant data of the current monitoring period, screen out the second actual relevant data with mutation time nodes and the corresponding actual mutation value, and determine the monitoring time nodes at which the second actual relevant data affects the corresponding carbon emission assessment indicators; Inputting the second actual relevant data and the corresponding mutation value into the second data fusion model to obtain the predicted lag value of the carbon emission assessment value at the corresponding monitoring time node; If there is no hysteresis value at the monitoring time node, the first predicted carbon emission assessment value is set to the predicted carbon emission assessment value at the corresponding monitoring time node of the area to be assessed; If there is a lag value at the monitoring time node, the first predicted carbon emission assessment value is corrected according to the lag value, and the corrected first predicted carbon emission assessment value is set as the predicted carbon emission assessment value at the corresponding monitoring time node of the area to be assessed.

8. The zero-carbon park carbon emission assessment method based on multi-source data fusion according to claim 7 is characterized in that: Evaluate the accuracy of the predicted carbon emissions assessment, including: Extract multiple carbon emission characteristics of the predicted carbon emission assessment value in the current monitoring period; Compare the carbon emission characteristics with the corresponding standard carbon emission characteristics to obtain multiple carbon emission characteristic differences; Match the corresponding difference coefficient for each carbon emission characteristic difference, and generate a comprehensive characteristic difference coefficient according to the difference coefficients corresponding to multiple carbon emission characteristics and the corresponding weight coefficients; The accuracy evaluation value of the predicted carbon emission assessment value of the corresponding area to be assessed in the current monitoring period is set according to the comprehensive characteristic difference coefficient.

9. The zero-carbon park carbon emission assessment method based on multi-source data fusion according to claim 8 is characterized in that: Determine whether to generate a correction instruction for the predicted carbon emission assessment value based on the evaluation results, including: Pre-set accuracy evaluation threshold; If the accuracy evaluation value is less than the accuracy evaluation threshold, a correction instruction for the predicted carbon emission evaluation value is generated; If the accuracy evaluation value is not less than the accuracy evaluation threshold, a correction instruction for the predicted carbon emission assessment value is generated.

Images