Smart algorithm-based method and apparatus for constructing carbon emission data graph, and device

Abstract

Disclosed in the embodiments of the present invention are a smart algorithm-based method and apparatus for constructing a carbon emission data graph, and a device. A specific embodiment of the method comprises: from hierarchical time series carbon emission data corresponding to a target carbon emission node, extracting time series carbon emission data corresponding to a last layer as last-layer time series carbon emission data; determining each layer of time series carbon emission data in the hierarchical time series carbon emission data that differs from the last-layer time series carbon emission data to be a prior-layer time series carbon emission data set; acquiring a carbon emission amount prediction result set corresponding to the prior-layer time series carbon emission data set; on the basis of the last-layer time series carbon emission data and a carbon emission amount prediction model, generating a carbon emission amount prediction result; on the basis of carbon emission amount prediction constraint information, performing correction processing on the carbon emission prediction result to obtain a carbon emission amount prediction output result; and, on the basis of the carbon emission amount prediction output result, constructing a carbon emission amount trend data graph. The present embodiments increase the accuracy of predicted carbon emission amount hierarchical time series.

Description

Method, device and equipment for constructing carbon emission data graph based on intelligent algorithm Technical Field

[0001] The embodiments of the present disclosure relate to the field of computer technology, and in particular to a method, apparatus, and device for constructing a carbon emission data graph based on an intelligent algorithm. Background Art

[0002] Hierarchical time series forecasting is a technique in the field of time series forecasting. Hierarchical time series refers to multiple time series that can be linked through certain hierarchical relationships. Time series at different levels can be decomposed or aggregated based on certain attributes (such as carbon emission nodes and carbon emission levels) to achieve mutual transformation. Currently, when conducting hierarchical time series forecasting of carbon emissions, the commonly used approach is to perform single-series forecasts on each level of time series, regardless of the hierarchical relationships.

[0003] However, the above approach usually has the following technical problems:

[0004] First, the prediction results at different levels cannot automatically meet consistency and are not applicable to the collaborative decision-making process at all levels. This leads to poor accuracy of the layered time series of predicted carbon emissions, making it difficult to shut down some carbon emission points in a timely manner, resulting in excessive carbon emissions.

[0005] Second, the anomalies in the carbon emission data were not analyzed in advance, resulting in a long time for carbon data analysis;

[0006] Third, manual detection of carbon emission data is inefficient and difficult to detect abnormal data in a timely manner.

[0007] The above information disclosed in this Background section is only for enhancement of understanding of the background of the inventive concept and therefore it may contain information that does not form the prior art that is already known in this country to a person of ordinary skill in the art.

[0008] Summary of the Invention

[0009] The content of this disclosure is used to briefly introduce concepts that will be described in detail in the detailed description section below. The content of this disclosure is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0010] Some embodiments of the present disclosure propose methods, devices, electronic devices, and computer-readable media for constructing carbon emission data graphs based on intelligent algorithms to solve one or more of the technical problems mentioned in the above background technology section.

[0011] In a first aspect, some embodiments of the present disclosure provide a method for constructing a carbon emission data graph based on an intelligent algorithm, the method comprising: extracting the time series carbon emission data corresponding to the last layer from the layered time series carbon emission data corresponding to the target carbon emission node as the last layer time series carbon emission data; determining the time series carbon emission data of each layer in the above-mentioned layered time series carbon emission data that is different from the above-mentioned last layer time series carbon emission data as a front-layer time series carbon emission data set; obtaining the carbon emission prediction result set corresponding to the preset future time period of the above-mentioned front-layer time series carbon emission data set as carbon emission prediction constraint information, wherein, Each carbon emission prediction result in the above-mentioned carbon emission prediction result set meets the consistency condition; based on the above-mentioned last-layer time series carbon emission data and the carbon emission prediction model, the carbon emission prediction result corresponding to the above-mentioned preset future time period is generated; based on the above-mentioned carbon emission prediction constraint information, the above-mentioned carbon emission prediction result is corrected to obtain the corrected carbon emission prediction result as the carbon emission prediction output result of the above-mentioned carbon emission prediction model; based on the above-mentioned carbon emission prediction output result, a carbon emission trend data graph is constructed, and the above-mentioned carbon emission trend data graph is sent to the associated display terminal for display.

[0012] In the second aspect, some embodiments of the present disclosure provide a carbon emission data graph construction device based on an intelligent algorithm, the device comprising: an extraction unit, configured to extract the time series carbon emission data corresponding to the last layer from the layered time series carbon emission data corresponding to the target carbon emission node as the last layer time series carbon emission data; a determination unit, configured to determine the time series carbon emission data of each layer in the above-mentioned layered time series carbon emission data that is different from the above-mentioned last layer time series carbon emission data as a front-layer time series carbon emission data set; an acquisition unit, configured to acquire the carbon emission prediction result set corresponding to the preset future time period of the above-mentioned front-layer time series carbon emission data set as carbon emission prediction constraint information, wherein, Each carbon emission prediction result in the above-mentioned carbon emission prediction result set meets the consistency condition; the generation unit is configured to generate the carbon emission prediction result corresponding to the above-mentioned preset future time period based on the above-mentioned last-layer time series carbon emission data and the carbon emission prediction model; the correction unit is configured to correct the above-mentioned carbon emission prediction result according to the above-mentioned carbon emission prediction constraint information, and obtain the corrected carbon emission prediction result as the carbon emission prediction output result of the above-mentioned carbon emission prediction model; the construction unit is configured to construct a carbon emission trend data graph based on the above-mentioned carbon emission prediction output result, and send the above-mentioned carbon emission trend data graph to the associated display terminal for display.

[0013] In a third aspect, some embodiments of the present disclosure provide an electronic device comprising: one or more processors; a storage device on which one or more programs are stored, and when the one or more programs are executed by one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0014] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method described in any implementation of the first aspect is implemented.

[0015] The various embodiments of the present disclosure described above have the following beneficial effects: Through the intelligent algorithm-based carbon emission data graph construction method of some embodiments of the present disclosure, the accuracy of the layered time series of predicted carbon emissions is improved, thereby enabling the timely closure of some carbon emission points with excessive carbon emissions to avoid excessive carbon emissions. Specifically, the cause of excessive carbon emissions is that the prediction results at different levels cannot automatically meet consistency and are not applicable to the collaborative decision-making process at each level, resulting in poor accuracy of the layered time series of predicted carbon emissions and difficulty in timely closure of some carbon emission points. Based on this, the intelligent algorithm-based carbon emission data graph construction method of some embodiments of the present disclosure first extracts the time series carbon emission data corresponding to the last layer from the layered time series carbon emission data corresponding to the target carbon emission node as the last layer time series carbon emission data. Thus, the extracted last layer time series data can represent the last layer time series data in the original layered time series data. Secondly, the time series carbon emission data of each layer in the layered time series carbon emission data that differs from the last layer time series carbon emission data is determined as the previous layer time series carbon emission data set. In this way, time series data for each layer prior to the final layer can be extracted. Next, a set of carbon emission forecast results for the preceding time series carbon emission dataset corresponding to a preset future time period is obtained as carbon emission forecast constraint information. Each carbon emission forecast result in the set of carbon emission forecast results satisfies consistency conditions. Thus, the obtained set of carbon emission forecast results can represent the externally input forecast results for each layer prior to the final layer, thereby using the externally input forecast results of higher-level time series as consistency constraints for lower-level time series. Subsequently, based on the final time series carbon emission data and the carbon emission forecast model, a carbon emission forecast result for the preset future time period is generated. Thus, the carbon emission forecast model can generate the original forecast results for the final time series data. Then, based on the carbon emission forecast constraint information, the carbon emission forecast results are corrected, resulting in the corrected carbon emission forecast results as the carbon emission forecast output of the carbon emission forecast model. Thus, the original prediction results of the last layer of time series data can be corrected using the externally input carbon emission prediction result set as a constraint, so that the correction result can be used as the output result of the carbon emission generation model. Finally, based on the above carbon emission prediction output results, a carbon emission trend data graph is constructed, and the above carbon emission trend data graph is sent to the associated display terminal for display. Thus, it is possible to use the prediction results of the externally input high-level time series as the consistency constraint of the low-level time series. In addition, the time series data of carbon emissions at different levels can meet the consistency, so that it can be applied to the collaborative decision-making process at each level, thereby improving the accuracy of the layered time series data of the predicted carbon emissions.In this way, some carbon emission points that exceed the carbon emission standards can be shut down in time to avoid exceeding the carbon emission standards. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The above and other features, advantages, and aspects of the various embodiments of the present disclosure will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and that components and elements are not necessarily drawn to scale.

[0017] FIG1 is a flowchart of some embodiments of a method for constructing a carbon emission data graph based on an intelligent algorithm according to the present disclosure;

[0018] FIG2 is a schematic diagram of hierarchical time series carbon emission data including two levels according to some embodiments of the carbon emission data graph construction method based on the intelligent algorithm of the present disclosure;

[0019] FIG3 is a flow chart of some embodiments of a carbon emission data graph construction device based on an intelligent algorithm according to the present disclosure;

[0020] FIG4 is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0021] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein. On the contrary, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are for illustrative purposes only and are not intended to limit the scope of protection of the present disclosure.

[0022] It should also be noted that, for ease of description, only the parts related to the invention are shown in the drawings. In the absence of conflict, the embodiments and features in the embodiments of the present disclosure may be combined with each other.

[0023] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0024] It should be noted that the modifications of "one" and "multiple" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, they should be understood as "one or more".

[0025] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0026] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0027] FIG1 is a flow chart of some embodiments of the method for constructing a carbon emission data graph based on an intelligent algorithm according to the present disclosure. It shows a process 100 of some embodiments of the method for constructing a carbon emission data graph based on an intelligent algorithm according to the present disclosure. The method for constructing a carbon emission data graph based on an intelligent algorithm includes the following steps:

[0028] Step 101 : extracting the time series carbon emission data corresponding to the last layer from the layered time series carbon emission data corresponding to the target carbon emission node as the last layer time series carbon emission data.

[0029] In some embodiments, the execution subject (for example, a computing device) of the carbon emission data graph construction method based on an intelligent algorithm can extract the time series carbon emission data corresponding to the last layer from the layered time series carbon emission data corresponding to the target carbon emission node as the last layer time series carbon emission data. The target carbon emission node can be a monitoring device node in the carbon emission area. The layered time series carbon emission data corresponding to the target carbon emission node can be the carbon emission data of the above-mentioned target carbon emission node organized in the form of a layered time series. Here, the carbon emissions in the layered time series carbon emission data can be the carbon emissions of the above-mentioned target carbon emission node in the historical time period. The above-mentioned last layer can be the last layer of the layered time series carbon emission data. The last layer time series carbon emission data can include the carbon emission time series of each last layer. The carbon emission time series can include each carbon emission in the historical time period. Each carbon emission time series can correspond to a node in the last layer.

[0030] As an example, the structure of the two-layered time series carbon emission data can refer to Figure 2. In Figure 2, the first layer of time series carbon emission data can include y1. The second layer of time series carbon emission data can include y2 and y3.

[0031] Step 102: determine the time series carbon emission data of each layer in the layered time series carbon emission data that is different from the last layer time series carbon emission data as a previous layer time series carbon emission data set.

[0032] In some embodiments, the execution entity may determine the time series carbon emission data of each layer in the layered time series carbon emission data that is different from the last layer of time series carbon emission data as the previous layer time series carbon emission data set. The time series carbon emission data of each layer in the layered time series carbon emission data that is before the last layer of time series carbon emission data may be determined as the previous layer time series carbon emission data set.

[0033] Step 103: Obtain a carbon emission prediction result set corresponding to a preset future time period of the above-mentioned front-layer time series carbon emission dataset as carbon emission prediction constraint information.

[0034] In some embodiments, the execution entity may obtain the carbon emission prediction result set corresponding to the preset future time period of the previous layer time series carbon emission data set as carbon emission prediction constraint information. The individual carbon emission prediction results in the above carbon emission prediction result set meet the consistency condition. The above carbon emission prediction result set may be an externally input prediction result for the time series carbon emission data of each layer before the last layer. Here, the carbon emission prediction result of each layer may be the individual carbon emissions of each node of the layer within the preset future time period predicted based on expert experience knowledge. The individual carbon emission prediction results in the above carbon emission prediction result set meet the consistency condition. The consistency condition may be that the sum of the individual carbon emissions corresponding to the same time point and the same branch in the carbon emission prediction result of the current layer is equal to the sum of the individual carbon emissions of the nodes corresponding to the branch and the time point in the previous layer.

[0035] Step 104 : Generate a carbon emission prediction result corresponding to the preset future time period based on the last layer time series carbon emission data and the carbon emission prediction model.

[0036] In some embodiments, the execution entity may generate a carbon emissions forecast result corresponding to the preset future time period based on the last-layer time series carbon emissions data and the carbon emissions forecast model. The carbon emissions forecast model may be a model for forecasting carbon emissions for each future time period. The carbon emissions forecast model may be an algorithmic model or a neural network model. For example, the carbon emissions forecast model may be, but is not limited to, ARIMA, ETS, xgboost, lightgbm, or deepar.

[0037] In practice, the above-mentioned execution entity can generate the carbon emission forecast results corresponding to the above-mentioned preset future time period through the following steps:

[0038] In the first step, in response to determining that there are outliers in the last layer of the time series carbon emission data, the outliers in the last layer of the time series carbon emission data are removed. The outliers in the last layer of the time series carbon emission data can be removed.

[0039] The second step is to perform missing value filling on the final-layer time series carbon emission data after the removal process to update the final-layer time series carbon emission data. First, in response to determining that missing values ​​exist in the final-layer time series data, a carbon emission time series corresponding to the missing values ​​can be extracted from the layered time series carbon emission data. Then, the carbon emission time series can be used to generate fill-in values ​​corresponding to the missing values. For example, interpolation can be used to generate fill-in values ​​corresponding to the missing values ​​based on the carbon emission time series.

[0040] The third step is to generate a carbon emissions forecast for the preset future time period based on the updated last-layer time series carbon emissions data and the carbon emissions forecast model. The updated last-layer time series carbon emissions data can be input into the carbon emissions forecast model to obtain a carbon emissions forecast for the preset future time period. Here, the carbon emissions forecast can be the predicted carbon emissions for each node in the last layer within the preset future time period.

[0041] Step 105 , correcting the carbon emission prediction result according to the carbon emission prediction constraint information, and obtaining the corrected carbon emission prediction result as the carbon emission prediction output result of the carbon emission prediction model.

[0042] In some embodiments, the execution entity may modify the carbon emission prediction result based on the carbon emission prediction constraint information, and obtain the modified carbon emission prediction result as the carbon emission prediction output result of the carbon emission prediction model. For example, a preset optimization objective may be solved based on the carbon emission constraint information to obtain the modified carbon emission prediction result. Here, the preset optimization objective may be to minimize the absolute value of the difference between the carbon emission prediction result and the modified carbon emission prediction result.

[0043] In practice, the above-mentioned execution entities can revise the above-mentioned carbon emission forecast results through the following steps:

[0044] In the first step, residual information is generated based on the historical fitting values ​​and actual values ​​of the carbon emissions prediction model corresponding to the historical time period and the final-layer time series carbon emissions data. The residual between each actual value and the corresponding historical fitting value can be determined to obtain a residual set as the residual information. The historical time period can be a period prior to the preset future time period.

[0045] The second step is to generate a covariance matrix based on the residual information. The variance-covariance matrix of the residual information can be determined as the covariance matrix.

[0046] The third step is to determine a constraint matrix corresponding to the layered time series carbon emission data. The constraint matrix can be determined by a pre-set summation matrix that aggregates the lowest level time series of the layered time series carbon emission data to the higher level time series.

[0047] The fourth step is to generate carbon emission constraint information based on the constraint matrix and the carbon emission prediction constraint information. The constraint formula can be determined as the carbon emission constraint information. The constraint formula can be expressed as Among them, b represents the carbon emission prediction result after correction. Represents the above carbon emission prediction constraint information. Here, b and All are in matrix form.

[0048] The constraint matrix C can be:

[0049] The fifth step is to construct the carbon emission prediction optimization target based on the above carbon emission prediction results and the above covariance matrix.

[0050] The following formula can be determined as the carbon emission prediction optimization objective:

[0051] in, Indicates the carbon emission prediction results after correction. W represents the above carbon emission prediction result. b represents the above covariance matrix. It can represent the adjustment term for the corrected bottom-level prediction result. The above formula can be used to minimize the sum of squares of the adjustment term weighted by the residual covariance matrix. It can be understood that for time series with larger residual standard deviations (less accurate predictions), larger adjustment terms can be tolerated under the same loss function value.

[0052] Step 6: Generate a revised carbon emissions forecast based on the carbon emissions constraint information and the carbon emissions forecast optimization objective. The carbon emissions constraint information can be used as a constraint to solve the carbon emissions forecast optimization objective to obtain a revised carbon emissions forecast. For example, the Lagrangian method can be used to solve the problem. Specifically, the following formula can be used:

[0053] Where T represents the matrix transpose.

[0054] Step 106 : constructing a carbon emission trend data graph based on the carbon emission prediction output result, and sending the carbon emission trend data graph to a related display terminal for display.

[0055] In some embodiments, the above-mentioned execution entity can construct a carbon emission trend data graph based on the above-mentioned carbon emission prediction output result, and send the above-mentioned carbon emission trend data graph to an associated display terminal for display. For example, the above-mentioned execution entity can construct a data trend graph of each carbon emission prediction amount corresponding to the above-mentioned carbon emission prediction output result as a carbon emission trend data graph through an Excel spreadsheet (for example, each carbon emission prediction amount can be input into an Excel spreadsheet, and then a carbon emission trend data graph can be fitted by linear regression). The display terminal can be a computing terminal with a display function that is communicatively connected to the above-mentioned execution entity. For example, the display terminal can be a display.

[0056] Optionally, the above-mentioned carbon emission prediction output results are input into a pre-trained abnormal data detection model to obtain the node position of each carbon emission prediction amount in each binary tree in the above-mentioned carbon emission prediction output results.

[0057] In some embodiments, the execution entity may input the carbon emission prediction output results into a pre-trained anomaly data detection model to obtain the node position of each carbon emission prediction value in each binary tree. The anomaly data detection model may include at least two binary trees, each containing different randomly selected attributes. The anomaly data detection model may be a pre-trained isolation forest model.

[0058] The carbon emission prediction output results can be input into a pre-created isolation forest (iForest, Isolation Forest) model. Among them, the isolation forest model usually contains at least two binary trees. And the randomly selected attributes between these binary trees in the isolation forest model are different. In this way, through the isolation forest model, the node position of each carbon emission prediction amount in each binary tree in the carbon emission prediction output results can be output. The node position here can be the position data of the node to which the carbon emission prediction amount belongs in the binary tree, such as node identifiers such as node number or node height (or number of layers), or node path length, etc.

[0059] It should be noted that iForest is generally a non-parametric and unsupervised method, that is, it does not require the definition of a mathematical model, nor does it require labeled training. iForest uses a very efficient strategy to find out which points are easily isolated. Suppose we use a random hyperplane to split the data space, and one cut can generate two subspaces. Then we continue to use a random hyperplane to split each subspace. The cycle continues until there is only one data point in each subspace. Intuitively speaking, we can find that clusters with high density will not stop being cut until they are split many times, but those points with low density can easily stop in a subspace very early.

[0060] The iForest algorithm draws on the concept of random forests. Just as random forests are composed of a large number of decision trees, iForest forests are also composed of a large number of binary trees. The trees in iForest are called isolation trees, or iTrees for short, which are binary trees. iTrees are different from decision trees and their construction process is simpler, using a completely random process.

[0061] Assuming that the data set has N data, when building an iTree, n samples can be uniformly sampled from the N data as training samples for the tree. Generally, sampling is done without replacement. In the sample, a feature is randomly selected, and a value is randomly selected from the range of all values ​​of this feature (between the minimum and maximum values), and the sample is binary divided. The samples with values ​​less than the value are divided to the left of the node, and the samples with values ​​greater than or equal to the value are divided to the right of the node. In this way, a splitting condition and the data sets on the left and right sides can be obtained. Then repeat the above process on the data sets on the left and right sides respectively until the data set has only one record or the limited height of the tree is reached.

[0062] Optionally, the target node position of each carbon emission prediction amount is determined according to the node position of each carbon emission prediction amount in each binary tree included in the above-mentioned carbon emission prediction output result.

[0063] In some embodiments, the execution entity may determine the target node position of each carbon emission prediction amount based on the node positions of each carbon emission prediction amount in each binary tree included in the carbon emission prediction output result. For each carbon emission prediction amount in the carbon emission prediction output result, the average value of the node positions of the carbon emission prediction amount in each binary tree in the isolation forest model may be determined as the target node position of the carbon emission prediction amount. The execution entity may determine the average value of the node positions of the carbon emission prediction amount in each binary tree in the isolation forest model, i.e., the height average value (PathLength), as the target node position of the carbon emission prediction amount. The weighted average value of these node positions may also be determined as the target node position of the carbon emission prediction amount.

[0064] Optionally, at least one abnormal carbon emission prediction amount in the above carbon emission prediction output result is determined based on the target node position of each carbon emission prediction amount.

[0065] In some embodiments, the execution entity may determine at least one abnormal carbon emission prediction amount in the carbon emission prediction output result based on the target node position of each carbon emission prediction amount.

[0066] In practice, the execution entity can determine the carbon emission predictions for target node positions that are less than a preset position threshold as abnormal carbon emission predictions, in descending order of target node positions, to obtain at least one abnormal carbon emission prediction. The first few target node positions with the lowest position can be selected as candidate node positions. It can then be determined whether these candidate node positions are less than the position threshold. Thus, the carbon emission information corresponding to the candidate node positions that are less than the position threshold, i.e., the carbon emission predictions close to the root node, are determined as abnormal carbon emission predictions.

[0067] Optionally, multiple indicator parameters related to the at least one abnormal carbon emission prediction amount are determined.

[0068] In some embodiments, the execution entity may determine multiple indicator parameters related to the at least one abnormal carbon emission prediction. Multiple indicator parameters related to the at least one abnormal carbon emission information are determined. The multiple indicator parameters are typically at least two. The related indicator parameters may be directly or indirectly related. For example, carbon emissions are typically directly related to indicator parameters such as carbon emission rate and carbon emission time.

[0069] Optionally, the impact indicator parameters of the above-mentioned multiple indicator parameters on the above-mentioned at least one abnormal carbon emission prediction amount are analyzed.

[0070] In some embodiments, the execution entity may analyze the impact of the multiple indicator parameters on the at least one abnormal carbon emission prediction amount.

[0071] Attribution theory (LMDI, Logarithmic Mean Index Method) can be used to analyze the impact of multiple indicator parameters on at least one abnormal carbon emissions forecast. For a certain abnormal carbon emissions forecast, attribution theory analysis can be used to break down the data changes to a specific indicator parameter factor, thereby further identifying the main cause.

[0072] As an example, if we assume the indicator: Y = X + Z. At this time, when comparing the indicators Y1 and Y0 of two different periods, the change in the indicator is expressed as: Y1-Y0. Since the change in the indicator is caused by the changes in two factors X and Z, the change in the indicator can be further divided into the changes in two factors X and Z: Y1-Y0 = (X1+Z1)-(X0+Z0) = (X1-X0)+(Z1-Z0). Therefore, the absolute value of the contribution of factor X to the indicator is: X1-X0; the relative value of the contribution is: (X1-X0) / (Y1-Y0). The absolute value of the contribution of factor Z to the indicator is: Z1-Z0; the relative value of the contribution is: (Z1-Z0) / (Y1-Y0). It can be seen from this that if the predicted amount of abnormal carbon emissions is in an additive relationship with multiple indicator parameters, then for each of the multiple indicator parameters, the difference between the current value of the indicator parameter and the comparison value can be determined as the influencing indicator parameter affecting the predicted amount of abnormal carbon emissions.

[0073] In practice, the execution entity may analyze the impact of the multiple indicator parameters on the at least one abnormal carbon emission prediction amount through the following steps:

[0074] The first step is to determine that the at least one abnormal carbon emission prediction amount and the multiple indicator parameters are in an additive relationship, and for each of the multiple indicator parameters, determine the difference between the current indicator value and the comparison indicator value of the indicator parameter as the contribution value affecting the at least one abnormal carbon emission prediction amount.

[0075] In the second step, in response to determining that the above-mentioned at least one abnormal carbon emission prediction amount is in a multiplication relationship with the above-mentioned multiple indicator parameters, for each indicator parameter among the above-mentioned multiple indicator parameters, the natural logarithm of the ratio of the current indicator value of the above-mentioned indicator parameter to the comparative indicator value is determined, and based on the above-mentioned natural logarithm and the difference between the current indicator value of the above-mentioned indicator parameter and the comparative indicator value, the contribution value of the above-mentioned indicator parameter to the above-mentioned at least one abnormal carbon emission prediction amount is generated.

[0076] The third step is to generate an influencing index parameter that has the greatest impact on the at least one abnormal carbon emission prediction amount based on the contribution values ​​corresponding to the above multiple index parameters.

[0077] Optionally, the at least one abnormal carbon emission prediction amount and the impact index parameter are sent to the display terminal.

[0078] In some embodiments, the execution entity may send the at least one abnormal carbon emission prediction amount and the impact indicator parameter to the display terminal.

[0079] The above-mentioned related content, as an inventive feature of the present disclosure, addresses the second technical problem mentioned in the background technology: "long carbon data parsing time." Factors that contribute to long carbon data parsing times are often as follows: Failure to pre-parse outliers in each carbon emission data set results in a long carbon data parsing time. Addressing these factors can shorten carbon data parsing time. To achieve this, first, the carbon emission forecast output results are input into a pre-trained outlier data detection model to obtain the node position of each carbon emission forecast value in each binary tree. The outlier data detection model includes at least two binary trees, each containing different randomly selected attributes. Secondly, based on the node positions of each carbon emission forecast value in each binary tree, a target node position for each carbon emission forecast value is determined. Next, based on the target node position of each carbon emission forecast value, at least one outlier carbon emission forecast value in the carbon emission forecast output results is determined. Finally, multiple indicator parameters related to the at least one outlier carbon emission forecast value are determined, and the impact of these multiple indicator parameters on the at least one outlier carbon emission forecast value is analyzed. Finally, the at least one abnormal carbon emission forecast and the influencing indicator parameters are sent to the display terminal. This allows analysis of each carbon emission forecast and the identification of anomalies based on the temporal trend characteristics of the data. This reduces the likelihood that a change will be identified as an anomaly, improving the accuracy of anomaly analysis results and, in turn, shortening carbon data analysis time.

[0080] Optionally, before inputting the carbon emission prediction output result into a pre-trained abnormal data detection model to obtain the node position of each carbon emission prediction amount in the carbon emission prediction output result in each binary tree, the method further includes:

[0081] The first step is to obtain a training sample dataset. Here, the training sample dataset can refer to sample data used to train the abnormal data detection model. The training sample data can be a sample carbon emission sequence.

[0082] In the second step, for each binary tree in the abnormal data detection model, a preset number of training sample data are selected from the training sample data set and placed into the root node of the binary tree.

[0083] The third step is to select an attribute dimension that is different from other binary trees from the multiple attribute dimensions of the above-mentioned preset number of training sample data.

[0084] In practice, the above execution entity can select an attribute dimension that is different from other binary trees by following the steps below:

[0085] In the first sub-step, at least two candidate attribute dimensions are selected from the plurality of attribute dimensions of the preset number of sample data using the kurtosis coefficient.

[0086] In the second sub-step, an attribute dimension that is different from other binary trees is selected from the at least two candidate attribute dimensions.

[0087] The fourth step is to determine the division value of the above attribute dimension based on the value range of the above preset number of training sample data in the selected attribute dimension, divide the training sample data of the current node into two child nodes, and record the division value of the above attribute dimension at the above current node until it is divided into a leaf node or the height of the above binary tree reaches the height threshold.

[0088] It should be noted that for the two child nodes obtained by division, the value range of the training sample data on the child node in the attribute dimension can be determined, so as to determine the division value of the attribute dimension again to divide the training sample data on the child node into two child nodes.

[0089] As an example, the construction of an isolation forest first requires the construction of an iTree. An iTree is a random binary tree where each node either has two daughters or is a leaf node with no children. Given a dataset D, where all attributes of D are continuous variables, the construction process of an iTree is as follows: 1. Randomly select an attribute Attr; 2. Randomly select a value Value for the attribute; 3. Classify each record according to Attr, placing records with Attr less than Value in the left daughter and records with Attr greater than or equal to Value in the right child; 4. Then recursively repeat steps 2 and 3 to continue constructing the left and right daughters until the following conditions are met: the incoming training sample dataset has only one or more identical records; or the tree reaches a limited height.

[0090] It is understandable that the maximum height is set for each iTree, generally because at least one abnormal carbon emission prediction record is relatively small and its path length is relatively low. And we only need to distinguish normal records from abnormal records. Therefore, we only need to care about the part below the average height. This can make the analysis efficiency of the isolation forest model more efficient. As an example, the maximum height can be determined according to the number of samples in each tree (sub-sampling size), such as log2ψ. The number of samples in each tree can be the same or different. In addition, the negative impact of sampling on the performance of the isolation forest model algorithm is usually small. And through experiments, it can be found that when the sampling size exceeds 256, the effect of the isolation forest model is often not greatly improved.

[0091] Once t iTrees are obtained, iForest training is generally complete. The resulting iForest can then be used to analyze test data. It should be noted that to ensure the accuracy of the Isolation Forest model's predictions, each iTree can be tested on the test data after it is obtained. Alternatively, the entire Isolation Forest model can be tested on the test data.

[0092] Here, the testing process typically involves running the test data through each constructed iTree to determine which leaf node the test data is located in. Understandably, the assumption underlying iTree's ability to effectively detect anomalies is that outliers are generally rare and are quickly classified as leaf nodes in the iTree. In this case, the accuracy of the iTree analysis results can be determined based on the anomaly index of the test data. This index is typically determined based on the natural logarithm of the current analysis data (i.e., the data input into the iTree) and the amount of data in the time series data (i.e., the dataset).

[0093] The above-mentioned related content, as an inventive feature of the present disclosure, addresses the third technical problem mentioned in the background art: "difficulty in timely detecting abnormal data." Factors that often make it difficult to detect abnormal data in a timely manner include the following: manual detection of individual carbon emission data is inefficient and difficult to detect in a timely manner. If these factors are addressed, the effect of timely detecting abnormal data can be achieved. To achieve this effect, first, a training sample dataset is obtained. Second, for each binary tree in the abnormal data detection model, a preset number of training sample data are selected from the training sample dataset and placed into the root node of the binary tree. Next, from the multiple attribute dimensions of the preset number of training sample data, an attribute dimension that is different from the other binary trees is selected. Then, based on the value range of the preset number of training sample data in the selected attribute dimension, a partition value for the attribute dimension is determined, and the training sample data of the current node is partitioned into two child nodes. The partition value of the attribute dimension is recorded at the current node until the node is partitioned into a leaf node or the height of the binary tree reaches a height threshold. Thus, abnormal data can be detected using the trained abnormal data detection model. Therefore, the model can be used to quickly detect abnormal carbon emission predictions in the carbon emission prediction output results, thereby improving detection efficiency.

[0094] Further referring to Figure 3, as an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a carbon emission data graph construction device based on an intelligent algorithm. These embodiments of the carbon emission data graph construction device based on an intelligent algorithm correspond to those method embodiments shown in Figure 1. The carbon emission data graph construction device based on an intelligent algorithm can be specifically applied to various electronic devices.

[0095] As shown in FIG3 , some embodiments of the carbon emission data graph construction device 300 based on the intelligent algorithm include: an extraction unit 301, a determination unit 302, an acquisition unit 303, a generation unit 304, a correction unit 305 and a construction unit 306. The extraction unit 301 is configured to extract the time series carbon emission data corresponding to the last layer from the layered time series carbon emission data corresponding to the target carbon emission node as the last layer time series carbon emission data; the determination unit 302 is configured to determine the time series carbon emission data of each layer in the above-mentioned layered time series carbon emission data that is different from the above-mentioned last layer time series carbon emission data as the previous layer time series carbon emission data set; the acquisition unit 303 is configured to obtain the carbon emission prediction result set of the above-mentioned previous layer time series carbon emission data set corresponding to the preset future time period as the carbon emission prediction constraint information, wherein each carbon emission prediction result set in the above-mentioned carbon emission prediction result set is The measurement result meets the consistency condition; the generation unit 304 is configured to generate the carbon emission prediction result corresponding to the above-mentioned preset future time period according to the above-mentioned last-layer time series carbon emission data and the carbon emission prediction model; the correction unit 305 is configured to correct the above-mentioned carbon emission prediction result according to the above-mentioned carbon emission prediction constraint information, and obtain the corrected carbon emission prediction result as the carbon emission prediction output result of the above-mentioned carbon emission prediction model; the construction unit 306 is configured to construct a carbon emission trend data graph according to the above-mentioned carbon emission prediction output result, and send the above-mentioned carbon emission trend data graph to the associated display terminal for display.

[0096] It is understood that the various units described in the intelligent algorithm-based carbon emission data map construction device 300 correspond to the various steps in the method described with reference to FIG1 . Therefore, the operations, features, and beneficial effects described above with respect to the method are also applicable to the intelligent algorithm-based carbon emission data map construction device 300 and the units contained therein, and will not be further described here.

[0097] In another implementation example, the above-mentioned carbon emission data graph construction device 300 based on intelligent algorithm includes: a processor, wherein the processor is used to execute the above-mentioned program module stored in the memory, including: an extraction unit 301, a determination unit 302, an acquisition unit 303, a generation unit 304, a correction unit 305 and a construction unit 306.

[0098] Reference is now made to FIG4 , which illustrates a schematic diagram of the structure of an electronic device (e.g., a computing device) 400 suitable for implementing some embodiments of the present disclosure. The electronic devices in some embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. The electronic device illustrated in FIG4 is merely an example and should not limit the functionality or scope of use of the embodiments of the present disclosure.

[0099] As shown in Figure 4, electronic device 400 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 401, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 402 or a program loaded from a storage device 408 into a random access memory (RAM) 403. Various programs and data required for the operation of electronic device 400 are also stored in RAM 403. Processing device 401, ROM 402, and RAM 403 are connected to each other via a bus 404. An input / output (I / O) interface 405 is also connected to bus 404.

[0100] Typically, the following devices can be connected to the I / O interface 405: an input device 406 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 407 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 408 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 409. The communication device 409 can allow the electronic device 400 to communicate with other devices wirelessly or by wire to exchange data. Although FIG4 shows an electronic device 400 with various devices, it should be understood that it is not required to implement or have all of the devices shown. More or fewer devices may be implemented or provided alternatively. Each box shown in FIG4 can represent one device, or multiple devices as needed.

[0101] In particular, according to some embodiments of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes program code for executing the method shown in the flowchart. In some such embodiments, the computer program can be downloaded and installed from the network via the communication device 409, or installed from the storage device 408, or installed from the ROM 402. When the computer program is executed by the processing device 401, the above-mentioned functions defined in the method of some embodiments of the present disclosure are performed.

[0102] It should be noted that the computer-readable medium described in some embodiments of the present disclosure may be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, the computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, device, or device. In some embodiments of the present disclosure, the computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium that can transmit, propagate, or transport a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0103] In some embodiments, the client and server can communicate using any currently known or later developed network protocol, such as HTTP (HyperText Transfer Protocol), and can be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or later developed network.

[0104] The computer-readable medium may be included in the electronic device; or it may exist independently without being installed in the electronic device. The computer-readable medium carries one or more programs. When the one or more programs are executed by the electronic device, the electronic device: extracts the time series carbon emission data corresponding to the last layer from the layered time series carbon emission data corresponding to the target carbon emission node as the last layer time series carbon emission data; determines the time series carbon emission data of each layer in the layered time series carbon emission data that is different from the last layer time series carbon emission data as the previous layer time series carbon emission data set; obtains the carbon emission prediction result set corresponding to the preset future time period of the previous layer time series carbon emission data set as the carbon emission prediction constraint information, Among them, each carbon emission prediction result in the above-mentioned carbon emission prediction result set meets the consistency condition; based on the above-mentioned last-layer time series carbon emission data and the carbon emission prediction model, the carbon emission prediction result corresponding to the above-mentioned preset future time period is generated; based on the above-mentioned carbon emission prediction constraint information, the above-mentioned carbon emission prediction result is corrected to obtain the corrected carbon emission prediction result as the carbon emission prediction output result of the above-mentioned carbon emission prediction model; based on the above-mentioned carbon emission prediction output result, a carbon emission trend data graph is constructed, and the above-mentioned carbon emission trend data graph is sent to the associated display terminal for display.

[0105] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0106] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the module, program segment, or a part of code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order than that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of the boxes in the block diagram and / or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0107] The units described in some embodiments of the present disclosure may be implemented by software or by hardware. The described units may also be set in a processor, for example, it may be described as: a processor including: an extraction unit, a determination unit, an acquisition unit, a generation unit, a correction unit and a construction unit. Among them, the names of these units do not constitute a limitation on the unit itself under certain circumstances. For example, the extraction unit may also be described as "a unit that extracts the time series carbon emission data corresponding to the last layer from the layered time series carbon emission data corresponding to the target carbon emission node as the last layer time series carbon emission data."

[0108] The functions described above herein may be performed, at least in part, by one or more hardware logic components. For example, and without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chip (SOCs), complex programmable logic devices (CPLDs), and the like.

[0109] The above description is only an illustration of some preferred embodiments of the present disclosure and the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combination of the above-mentioned technical features, but should also cover other technical solutions formed by any combination of the above-mentioned technical features or their equivalent features without departing from the above-mentioned inventive concept. For example, the above-mentioned features are replaced with (but not limited to) technical features with similar functions disclosed in the embodiments of the present disclosure.

Claims

1. A method for constructing a carbon emission data graph based on an intelligent algorithm, comprising: From the layered time series carbon emission data corresponding to the target carbon emission node, extract the time series carbon emission data corresponding to the last layer as the last layer time series carbon emission data; Determine the time series carbon emission data of each layer in the layered time series carbon emission data that is different from the last layer time series carbon emission data as a previous layer time series carbon emission data set; Obtaining a carbon emission prediction result set corresponding to a preset future time period of the front-layer time series carbon emission data set as carbon emission prediction constraint information, wherein each carbon emission prediction result in the carbon emission prediction result set satisfies a consistency condition; Generate a carbon emission prediction result corresponding to the preset future time period according to the last layer time series carbon emission data and the carbon emission prediction model; According to the carbon emission prediction constraint information, the carbon emission prediction result is corrected to obtain the corrected carbon emission prediction result as the carbon emission prediction output result of the carbon emission prediction model; A carbon emission trend data graph is constructed according to the carbon emission prediction output result, and the carbon emission trend data graph is sent to an associated display terminal for display.

2. The method according to claim 1, wherein: The generating of the carbon emission prediction result corresponding to the preset future time period according to the last layer time series carbon emission data and the carbon emission prediction model includes: In response to determining that there are abnormal values ​​in the last layer of time series carbon emission data, removing the abnormal values ​​in the last layer of time series carbon emission data; Perform missing value filling processing on the last layer of time series carbon emission data after elimination to update the last layer of time series carbon emission data; According to the updated last-layer time series carbon emission data and the carbon emission prediction model, a carbon emission prediction result corresponding to the preset future time period is generated.

3. The method according to claim 1, wherein: The method further comprises: Inputting the carbon emission prediction output result into a pre-trained abnormal data detection model to obtain the node position of each carbon emission prediction amount in the carbon emission prediction output result in each binary tree, wherein the abnormal data detection model includes at least two binary trees, and the randomly selected attributes in the at least two binary trees are different; Determine the target node position of each carbon emission prediction amount according to the node position of each carbon emission prediction amount in each binary tree included in the carbon emission prediction output result; Determine at least one of the carbon emission prediction output results according to the target node position of each carbon emission prediction amount. The predicted amount of abnormal carbon emissions; Determining a plurality of indicator parameters related to the at least one predicted abnormal carbon emission amount; Analyze the impact of the multiple indicator parameters on the at least one abnormal carbon emission prediction amount; The at least one abnormal carbon emission prediction amount and the influencing indicator parameter are sent to the display terminal.

4. The method according to claim 3, wherein: The step of determining at least one abnormal carbon emission prediction amount in the carbon emission prediction output result according to the target node position of each carbon emission prediction amount includes: According to the order of the target node positions from low to high, the carbon emission prediction amount of the target node position less than the preset position threshold is determined as the abnormal carbon emission prediction amount, and at least one abnormal carbon emission prediction amount is obtained.

5. A carbon emission data graph construction device based on an intelligent algorithm, comprising: The extraction unit is configured to extract the time series carbon emission data corresponding to the last layer from the layered time series carbon emission data corresponding to the target carbon emission node as the last layer time series carbon emission data; a determination unit configured to determine the time series carbon emission data of each layer different from the last layer time series carbon emission data in the layered time series carbon emission data as a previous layer time series carbon emission data set; an acquisition unit configured to acquire a carbon emission prediction result set corresponding to a preset future time period of the front-layer time series carbon emission data set as carbon emission prediction constraint information, wherein each carbon emission prediction result in the carbon emission prediction result set satisfies a consistency condition; A generating unit configured to generate a carbon emission prediction result corresponding to the preset future time period according to the last layer time series carbon emission data and the carbon emission prediction model; a correction unit configured to correct the carbon emission prediction result according to the carbon emission prediction constraint information, and obtain the corrected carbon emission prediction result as the carbon emission prediction output result of the carbon emission prediction model; The construction unit is configured to construct a carbon emission trend data graph according to the carbon emission prediction output result, and send the carbon emission trend data graph to an associated display terminal for display.

6. An electronic device comprising: one or more processors; a storage device having one or more programs stored thereon; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 4.

7. A computer readable medium having a computer program stored thereon, wherein: When the program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images