Power consumer carbon emission monitoring method, computer equipment and readable storage medium

Abstract

The invention belongs to the technical field of carbon emission monitoring, and particularly relates to a power consumer carbon emission monitoring method, computer equipment and a readable storage medium. The method comprises the following steps: firstly, acquiring data required for predicting carbon emission of a power consumer at a prediction moment; processing the obtained data by adopting PCA (Principal Component Analysis) to obtain feature values and feature vectors of k principal components with relatively large feature values; for a principal component, calculating a result obtained by multiplying a feature value of the principal component by a transpose of a feature vector and then multiplying by a matrix corresponding to the obtained data, and adding calculation results corresponding to the principal components to obtain a comprehensive feature vector; and inputting the comprehensive feature vector into a comprehensive carbon emission prediction model to obtain a carbon emission prediction value of the power consumer at a prediction moment. According to the method, the comprehensive feature vector reflects the sum of the expressions of all the data on all the principal components, the mutual relation and importance of the data needed for predicting the carbon emission of the power consumer can be deeply observed, and the accuracy of carbon emission prediction is effectively improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of carbon emission monitoring, and in particular relates to a carbon emission monitoring method for power users, a computer device and a readable storage medium. Background Art

[0002] As the global climate change problem becomes increasingly serious, how to effectively control carbon emissions has become a key issue of common concern for governments and enterprises. In this context, the power industry, as one of the main sources of energy consumption and carbon emissions, needs to introduce new technical means to achieve accurate monitoring and control of carbon emissions. In recent years, with the rapid development of the Internet of Things, big data analysis and artificial intelligence technology, real-time collection and analysis of power data has become possible. By monitoring the operating data of power user terminal equipment, a reliable data basis can be provided for the accurate assessment of carbon emissions. However, traditional carbon emission monitoring methods mainly rely on the macro-statistics of power data, which cannot reflect the carbon emissions of individual power users in real time and in detail, resulting in a lack of pertinence in the formulation and implementation of energy-saving and emission reduction measures; in addition, although some existing carbon emission monitoring can collect basic power data through smart meter equipment, they often lack the ability to integrate and analyze multivariate data, making it difficult to provide accurate carbon emission predictions and personalized energy-saving suggestions; the construction of some carbon emission prediction models usually relies on simple linear regression and statistical analysis methods, which are difficult to extract complex nonlinear relationships from massive data and cannot fully reflect the multi-dimensional impact of various factors on carbon emissions.

[0003] The Chinese invention patent with application publication number CN116957850A and application publication date 2023.10.27 discloses an enterprise carbon emission prediction algorithm driven by multidimensional feature data. The method first collects multidimensional data of carbon emission influencing factors and performs preprocessing, and then uses principal component analysis and kernel function algorithm to extract multidimensional data features, and then transmits the extracted multidimensional data features to the fruit fly-differential evolution optimization RBF neural network model to obtain the best enterprise carbon emission prediction result. This method directly uses the multidimensional feature data obtained by principal component analysis as the input of the neural network model. The multidimensional data features here usually contain a large amount of redundant information and noise data, which will increase the complexity of the model and cause overfitting, and reduce the generalization ability of the model. In addition, high-dimensional data may mask the key variables that are truly related to carbon emissions, affecting the prediction accuracy of the model; moreover, it will also cause the computational cost of model training and prediction to be high, the training speed of the model to be low, and the complex feature space may cause the time complexity and storage requirements of the algorithm to grow rapidly, affecting the operability of practical applications. In summary, this approach has the problem of low training speed of the model used to predict carbon emissions and low carbon emission prediction accuracy. Summary of the invention

[0004] The object of the present invention is to provide a method for monitoring carbon emissions of power users, a computer device and a readable storage medium, so as to solve the problems in the prior art of low training speed of models for predicting carbon emissions and low carbon emissions prediction accuracy.

[0005] In order to solve the above technical problems, the present invention provides a method for monitoring carbon emissions of power users, which comprises the following steps:

[0006] 1) Obtain the data required for predicting carbon emissions of power users at the prediction time; process the acquired data using the principal component analysis method to obtain the eigenvalues ​​and eigenvectors of k principal components with relatively large eigenvalues, where k>1; for a principal component, calculate its eigenvalue multiplied by the transpose of the eigenvector and then multiplied by the matrix corresponding to the acquired data to obtain the result, and add the calculation results corresponding to each principal component to obtain a comprehensive eigenvector;

[0007] 2) Input the comprehensive feature vector into the comprehensive carbon emission prediction model to obtain the carbon emission prediction value of the power user at the prediction time; the comprehensive carbon emission prediction model is obtained by training the machine learning model using the data set, and the data set contains the comprehensive feature vector and carbon emission data corresponding to the historical data required for the carbon emission of the power user.

[0008] Furthermore, the machine learning model is a support vector machine.

[0009] Furthermore, the kernel function in the support vector machine is RBF.

[0010] Furthermore, the method further comprises the following steps: 3) generating corresponding energy conservation and emission reduction suggestions according to the obtained carbon emission prediction value: when T 1 ≤C(t) <T 2 When T 2 ≤C(t), generate energy-saving and emission-reduction suggestions for users to turn off some of the electrical appliances they use; where C(t) represents the obtained carbon emission prediction value, T 1 and T 2 are the first carbon emission threshold and the second carbon emission threshold set respectively, and T 1 <T 2 .

[0011] Furthermore, T 1 and T 2 All are dynamically adjusted, and the dynamic adjustment method is:

[0012]

[0013] Where, T 1 (t) and T 2(t) represent the first carbon emission threshold and the second carbon emission threshold at time t, respectively. and Respectively represent T 1 (t) and T 2 The initial value of (t), α and β represent the 1 (t) and T 2 (t) The corresponding adjustment coefficient, C mean Represents the average carbon emissions over a period of time, C base Indicates the baseline value of carbon emissions.

[0014] Furthermore, the data required for predicting carbon emissions of power users include power data of power users, and the power data includes at least one of voltage, current and power.

[0015] Furthermore, the data required for predicting carbon emissions from electricity users also include other multivariate data, which include at least one of weather data, electricity price information, and regional energy consumption statistics, and all data required for predicting carbon emissions from electricity users are standardized data.

[0016] Furthermore, the power data is power data obtained from the data lake, and the power data stored in the data lake is obtained in the following manner: receiving data that is first compressed and then encoded from the originally obtained power data, and then verifying the received data, decoding and then decompressing the verified data, and storing it in the data lake after processing.

[0017] Furthermore, wavelet transform is used for the compression, and an inverse wavelet transform corresponding to the wavelet transform is used for the decompression, and the wavelet basis function used is a Daubechies wavelet basis function.

[0018] Furthermore, a Huffman encoder is used for the encoding, and a Huffman decoder corresponding to the Huffman encoder is used for the decoding.

[0019] Furthermore, the originally acquired power data is power data obtained by preprocessing the originally collected power data, and the preprocessing method includes removing high-frequency noise and / or removing abnormal values.

[0020] Furthermore, the other multivariate data are data after missing value completion and / or outlier elimination.

[0021] In order to solve the above technical problems, the present invention also provides a computer device, including a processor, wherein the processor executes a computer program to implement the steps of the above-mentioned method for monitoring carbon emissions of electricity users.

[0022] In order to solve the above technical problems, the present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned method for monitoring carbon emissions of electricity users are implemented.

[0023] The present invention is an improved invention creation, and its beneficial effects are as follows: the present invention applies the principal component analysis method to carbon emission monitoring, firstly uses the principal component analysis method to analyze the acquired data required for predicting carbon emissions of power users, and obtains the eigenvalue and eigenvector of the principal component with relatively large eigenvalue. Then, for a principal component, the eigenvalue is calculated, multiplied by the transposition of the eigenvector, and then multiplied by the matrix corresponding to the acquired data to obtain the result, and the calculation result reflects the performance of all data on the principal component, and then the calculation results corresponding to each principal component are added to obtain a comprehensive eigenvector. The comprehensive feature vector reflects the sum of the performance of all data on each principal component. It extracts the main information from multidimensional data through principal component analysis, reduces high-dimensional data to low-dimensional data, and can significantly reduce the dimension of model input and the training cost of the model. The parameter can provide a deeper insight into the relationship and importance between the data required to predict carbon emissions of power users, condense key information from complex data, retain key information, remove redundancy and noise, improve the training efficiency and calculation speed of the model, and reduce the risk of overfitting. The result presented by the comprehensive feature vector is that only the most relevant features of carbon emissions are retained. Therefore, the information input to the model is more refined, which helps the model to more accurately capture the nonlinear relationship in the data and improve the prediction accuracy of the model. Finally, using the comprehensive feature vector and the comprehensive carbon emission prediction model, the accurate carbon emission prediction value of the power user at the prediction time can be obtained. On the basis of the predicted carbon emission prediction value, it can not only help users optimize their electricity consumption behavior and reduce carbon emissions, but also provide effective decision-making support for power management departments and promote the energy conservation and carbon reduction goals of the whole society. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 is a flow chart of the carbon emission monitoring method for power users of the present invention;

[0025] Figure 2 It is a structural diagram of the carbon emission monitoring device for power users of the present invention. DETAILED DESCRIPTION

[0026] The present invention applies principal component analysis to carbon emission monitoring. The specific process is to first obtain the data required for predicting carbon emissions of power users at the prediction time; use principal component analysis to process the obtained data to obtain the eigenvalues ​​and eigenvectors of k principal components with relatively large eigenvalues; for a principal component, calculate its eigenvalue multiplied by the transposition of the eigenvector and then multiplied by the matrix corresponding to the obtained data to obtain the result, add the calculation results corresponding to each principal component to obtain a comprehensive eigenvector; then input the comprehensive eigenvector into the comprehensive carbon emission prediction model to obtain the carbon emission prediction value of the power user at the prediction time; the comprehensive carbon emission prediction model is obtained by training the machine learning model. The comprehensive eigenvector reflects the sum of the performance of all data on each principal component. This parameter can provide a deeper insight into the relationship and importance between the data required to predict the carbon emissions of power users, and can condense key information from complex data; and then use the comprehensive eigenvector and the comprehensive carbon emission prediction model to obtain an accurate carbon emission prediction value of the power user at the prediction time.

[0027] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is further described in detail below with reference to the accompanying drawings and embodiments.

[0028] Example of carbon emission monitoring method for power users:

[0029] An embodiment of a method for monitoring carbon emissions of power users of the present invention has a process as follows: Figure 1 As shown, the details are as follows:

[0030] Step 1: Use sensors deployed on power users' terminal devices to collect power data from power users in real time.

[0031] In this embodiment, the sensors installed on the terminal device include a voltage sensor, a current sensor and a power meter, which collect real-time power data of voltage V, current I and power P. As other implementations, only one or two of the voltage V, current I and power P data may be selected, or other data that can be used for carbon emission prediction of power users may be added.

[0032] In order to improve the quality of the collected data, the collected data can be preprocessed. The specific preprocessing methods can be: removing high-frequency noise and removing outliers to solve the noise and data deviation problems that may exist in the power data, as well as data standardization processing; of course, if the quality of the collected data is acceptable, only high-frequency noise or only outliers can be removed. Among them, a low-pass filter can be used to remove high-frequency noise, and the Z-score method can be used to detect outliers and remove them. The data set of preprocessed power data is x={x 1 、x 2 、x 3}, where x1 Represents voltage V, x 2 Represents the current I, x 3 Represents power P. The data preprocessing here is particularly critical in the implementation of the solution of the present invention, because power data often has high-frequency fluctuations and abnormal conditions, and this method can greatly improve the accuracy of the data; compared with the general power data collection method, this step ensures that the basic data for subsequent carbon emission analysis is more reliable.

[0033] Step 2: compress and encode the power data obtained in step 1, upload the compressed and encoded power data to the central server, and then perform verification, decompression, decoding and storage in the data lake.

[0034] In this embodiment, wavelet transform is used for data compression. The wavelet basis function used is Daubechies wavelet. The compressed data set after wavelet transform is x′={x′ 1 , x′ 2 , x′ 3 This processing method solves the problem of large amount of power data and limited transmission bandwidth. In particular, the use of wavelet transform to compress power data and the selection of Daubechies wavelet basis function can effectively retain the key information of power data while significantly reducing the amount of data.

[0035] In this embodiment, the compressed data is encoded using a Huffman encoder to generate an encoded data set x″={x 1 ″、x 2 ″、x 3 ″}, upload the encoded data set x″ to the central server through the MQTT protocol.

[0036] The received data is first hashed using the SHA-256 algorithm to generate a checksum H(x″)=SHA-256(x″); the Huffman decoder corresponding to the Huffman encoder is then used to decode the data set x″ to restore the compressed power data set x′={x′ 1 , x′ 2 , x′ 3}; Then, the decoded data x′ is reconstructed using the inverse wavelet transform corresponding to the wavelet transform, and the power data set x={x 1 、x 2 、x 3}; Using Hadoop HDFS as the data lake storage platform, the restored power data set x = {x 1 、x 2 、x 3}Stored in the data lake and indexed by timestamp.

[0037] Compared with conventional data transmission in the prior art, the present invention is more flexible in the selection of compression and encoding algorithms, ensuring the efficiency and security of data transmission, and further enhancing the reliability of data integrity verification through the verification of the SHA-256 algorithm.

[0038] Step three, collect multivariate data from multivariate data sources through the Application Programming Interface (API), integrate it with the power data to obtain a multivariate data set, and perform time alignment and standardization processing.

[0039] In this embodiment, weather data, electricity price information, and regional energy consumption statistics are collected from multiple data sources through an API interface. Of course, as other implementations, in order to reduce the complexity of data processing and improve processing efficiency, only one or two of the weather data, electricity price information, and regional energy consumption statistics can be obtained, and other data can also be added to improve the accuracy of the prediction by using multiple information.

[0040] For multivariate data obtained from different APIs, outlier removal and missing value filling are first performed. Of course, when the data quality is relatively good, only outlier removal or missing value filling can be performed, and the Z-score method can be used to detect and remove outliers; then linear interpolation is used for time alignment; and then the ETL method (a data processing technology that represents the process of extracting, transforming and loading data from the source end to the destination end) is used to integrate the power data with the multivariate data obtained from the API, and Pandas (a tool based on NumPy) is used to convert and clean the data format to generate a standardized multivariate data set. Among them, the use of linear interpolation for time alignment ensures the synchronization of data from different sources, which is crucial in carbon emission forecasting. Compared with the forecasting model based only on power data, the present invention can further improve the accuracy and scientificity of the forecast through the integration of multivariate data. The integrated multivariate data set is D=(τ 1 、x 1 、x 2 、x 3 ,y 1 ,y 2 ,y 3 ),...,(τ c 、x 1 、x 2 、x 3 ,y 1 ,y 2 ,y 3 ), where τ c Indicates the cth timestamp, y1 Represents weather data, y 2 Indicates electricity price information, y 3 Represents regional energy consumption statistics.

[0041] This process supplements the single nature of electricity data with other multivariate data, solving the problem of insufficient accuracy in existing carbon emission monitoring systems due to the lack of external factors.

[0042] In step 4, principal component analysis (PCA) is used to extract comprehensive feature vectors from the multivariate dataset after time alignment and standardization.

[0043] Calculate the covariance matrix S of the multivariate data set, expressed as:

[0044]

[0045] In the formula, h i represents the i-th sample data, μ represents the sample mean, and n represents the number of samples.

[0046] Perform eigenvalue decomposition on the covariance matrix S to obtain the eigenvalue λ 1 ,λ 2 ,…,λ m and the corresponding eigenvector v 1 ,v 2 ,…,v m , where m represents the quantity index.

[0047] According to the size of the eigenvalue, the first k (k>1) principal components with larger eigenvalues ​​are selected by sorting from large to small, and the comprehensive eigenvector F(t) is extracted based on the selected principal components. The expression is:

[0048]

[0049] In the formula, F(t) represents the comprehensive feature vector at time t, λ j represents the eigenvalue of the jth principal component, The eigenvector v representing the jth principal component j , D(t) represents the matrix corresponding to the standardized multivariate data at time t.

[0050] The principal component analysis method is used to extract comprehensive feature vectors from multivariate data sets. Through dimensionality reduction and feature extraction, data redundancy is reduced while retaining the key information that has the greatest impact on carbon emission prediction. Moreover, this step extracts key feature vectors through eigenvalue decomposition and principal component selection, solving the impact of high-dimensional data on the training efficiency of machine learning models. Compared with traditional feature engineering methods, the present invention uses PCA to simplify the processing of complex data and effectively improves the training speed and prediction accuracy of machine learning models.

[0051] Step 5: Collect carbon emission data and integrate it with the summed feature vector obtained in step 4 to obtain a data set, and divide it into training set and test set according to a certain ratio.

[0052] Among them, carbon emission data can be collected by monitoring the CO2 concentration and spectral characteristics in the air.

[0053] Step six, build a machine learning model, use the training set and test set obtained in step five to train and test the constructed machine learning model accordingly, and finally obtain a comprehensive carbon emission prediction model.

[0054] In this embodiment, the selected machine learning model is the support vector machine SVM, and the kernel function used is the RBF kernel function K(F a ,F b ), whose expression is:

[0055]

[0056] In the formula, F a and F b represents the comprehensive feature vector at time a and time b, γ represents the adjustment parameter of the kernel function, and exp represents the natural exponential function. Then the expression of the carbon emission prediction value C(t) obtained based on SVM prediction is:

[0057]

[0058] In the formula, the range of C(t) is [0, +∞], α l represents the Lagrange multiplier, and c represents the bias term.

[0059] Step 7: Obtain the data required to predict the carbon emissions of power users at the prediction time, use the principal component analysis method to obtain a comprehensive feature vector, input the obtained comprehensive feature vector into the comprehensive carbon emission prediction model, and obtain the predicted carbon emissions value of the power user at the prediction time. The specific process is as follows:

[0060] 1) Obtain the data required to predict carbon emissions of power users at the prediction time, which includes power data (including voltage, current and power), weather data, electricity price information and regional energy consumption statistics of power users. The power data is obtained from the data lake mentioned in step 2, and other multivariate data (including weather data, electricity price information and regional energy consumption statistics) is obtained from the API interface, and outliers are removed and missing values ​​are filled. After processing, all data needs to be standardized.

[0061] 2) Use PCA to process the data obtained in step 1) to obtain the eigenvalues ​​and eigenvectors of k principal components with relatively large eigenvalues, and combine the calculation formula of the comprehensive eigenvector F(t) in step 4 to obtain the comprehensive eigenvector.

[0062] 3) Input the comprehensive feature vector into the comprehensive carbon emission prediction model obtained in step 6 to obtain the carbon emission prediction value of the power user at the predicted time.

[0063] Step eight, based on the set carbon emission threshold and the carbon emission forecast value obtained in step seven, generate corresponding processing suggestions.

[0064] The set carbon emission threshold includes the first carbon emission threshold T 1 and the second carbon emission threshold T 2 When C(t) <T 1 , indicating that the carbon emissions are very small and no energy-saving recommendations are generated; when T 1 ≤C(t) <T 2 When T 2 When ≤C(t), it means that the carbon emissions are very large, and users are advised to turn off some of the appliances they use to reduce carbon emissions. That is, recommendations are only generated when the carbon emissions exceed a certain threshold, avoiding excessive interference to users.

[0065] In addition, T can be adjusted dynamically according to the user's electricity consumption behavior and seasonal factors. 1 and T 2 , to ensure that the threshold setting is more reasonable, the expression is:

[0066]

[0067] Where, T 1 (t) and T 2 (t) represent the first carbon emission threshold and the second carbon emission threshold at time t, respectively. and Respectively represent T 1 (t) and T 2 The initial value of (t), α and β represent the 1 (t) and T 2 (t) The corresponding adjustment coefficient, C mean Represents the average carbon emissions over a period of time; C base Indicates the benchmark value of carbon emissions. mean To adjust T 1 and T 2 , the average carbon emission C mean The user's electricity consumption behavior and seasonal factors can be taken into account. This dynamic adjustment of the carbon emission threshold T 1 and T2 This overcomes the limitation that the static threshold cannot adapt to different power usage environments.

[0068] The above method can be Figure 2 The carbon emission monitoring device for power users shown in the figure is implemented. The carbon emission monitoring device for power users includes:

[0069] Data acquisition module: Deploy sensors on power user terminal devices to collect power data in real time and perform pre-processing to provide high-quality input data for subsequent processing;

[0070] Data transmission module: compresses and encodes the pre-processed power data to reduce the data transmission bandwidth, and uploads the compressed data to the central server;

[0071] Data storage module: Verify, decompress and decode data on the central server, store valid data in the data lake, and provide a data basis for subsequent analysis;

[0072] Multivariate data integration module: collects external multivariate data through the API interface, integrates it with power data to form a multivariate data set, and performs time alignment and format conversion;

[0073] Feature extraction module: extracts comprehensive feature vectors from time-aligned and format-converted multivariate datasets to provide high-dimensional input features for the carbon emission prediction model;

[0074] Carbon emission prediction model building module: Based on the comprehensive feature vector, a comprehensive carbon emission prediction model is built using machine learning algorithms to predict the carbon emissions of power users;

[0075] Emission reduction suggestion module: Based on the carbon emission prediction results, combined with user historical data and external factors, it provides personalized energy-saving and emission reduction suggestions to help users optimize their electricity usage behavior.

[0076] The above-mentioned device is actually a computer solution based on the corresponding method flow of the present invention, that is, a software framework, and the above-mentioned various modules are the processing processes or programs corresponding to the method flow.

[0077] Computer equipment example:

[0078] A computer device embodiment of the present invention is applicable to a method for monitoring carbon emissions of power users. The computer device includes a memory, a processor, an internal bus, and a computer program stored in the memory. The processor and the memory communicate and exchange data with each other through the internal bus. The processor executes the computer program to implement the steps of a method for monitoring carbon emissions of power users described in an embodiment of the method for monitoring carbon emissions of power users of the present invention.

[0079] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0080] Computer readable storage medium embodiment:

[0081] The present embodiment provides a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the method for realizing carbon emission monitoring of power users proposed in the above embodiment is implemented. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0082] In summary, the present invention has the following characteristics: by real-time collection of power data and preprocessing, the accuracy of the data is ensured, and by compression and encoding, efficient data transmission and storage are achieved; by multivariate data integration and PCA feature extraction, the accuracy and efficiency of the carbon emission prediction model are improved, and through machine learning algorithms, an accurate carbon emission prediction model is constructed, and personalized energy-saving suggestions can be provided to users based on the prediction results, which can not only help users optimize electricity consumption behavior and reduce carbon emissions, but also provide effective decision-making support for power management departments, and promote the energy conservation and carbon reduction goals of the entire society.

[0083] Specific implementation methods are given above, but the present invention is not limited to the described implementation methods. The basic idea of ​​the present invention lies in the above basic scheme. For ordinary technicians in this field, it does not take creative work to design various deformed models, formulas, and parameters according to the teachings of the present invention. Changes, modifications, substitutions, and variations of the implementation methods without departing from the principles and spirit of the present invention still fall within the scope of protection of the present invention.

Claims

1. A method for monitoring carbon emissions of power users, characterized in that: The method includes the following steps: 1) Obtain the data required for predicting the carbon emissions of power users at the prediction moment; process the obtained data using the principal component analysis method to obtain the eigenvalues and eigenvectors of the k principal components with relatively large eigenvalues, where k > 1; for a principal component, calculate the result obtained by multiplying its eigenvalue by the transpose of the eigenvector and then multiplying by the matrix corresponding to the obtained data, and add the calculation results corresponding to each principal component to obtain the comprehensive eigenvector; 2) Input the comprehensive eigenvector into the comprehensive carbon emission prediction model to obtain the carbon emission prediction value of the power user at the prediction moment; the comprehensive carbon emission prediction model therein is obtained by training a machine learning model using a data set, and the data set includes the comprehensive eigenvector corresponding to the historical data required for the carbon emissions of power users and the carbon emission data.

2. The method for monitoring carbon emissions of power users according to claim 1, characterized in that: The method further includes the following steps: 3) Generate corresponding energy conservation and emission reduction suggestions based on the obtained carbon emission prediction value: when T1 ≤ C(t) < T2, generate energy conservation and emission reduction suggestions for suggesting users to reduce the use of high-power equipment; when T2 ≤ C(t), generate energy conservation and emission reduction suggestions for suggesting users to turn off some of the used electrical appliances; where C(t) represents the obtained carbon emission prediction value, and T1 and T2 are the set first carbon emission threshold and second carbon emission threshold respectively, and T1 < T2.

3. The method for monitoring carbon emissions of power users according to claim 2, characterized in that: Both T1 and T2 are dynamically adjusted, and the dynamic adjustment method is: Where T1(t) and T2(t) represent the first carbon emission threshold and the second carbon emission threshold at time t, respectively. and They represent the initial values ​​of T1(t) and T2(t), α and β represent the adjustment coefficients corresponding to T1(t) and T2(t), respectively. mean Represents the average carbon emissions over a period of time, C base Indicates the baseline value of carbon emissions.

4. The method for monitoring carbon emissions of power users according to claim 1, characterized in that: The data required for predicting the carbon emissions of power users includes the power data of power users and also includes other multivariate data. The other multivariate data includes at least one of weather data, electricity price information, and regional energy consumption statistical data, and all the data required for predicting the carbon emissions of power users is data after standardized processing.

5. The method for monitoring carbon emissions of power users according to claim 4, characterized in that: The power data is the power data obtained from the data lake, and the power data stored in the data lake is obtained in the following manner: receive the data obtained by first compressing and then encoding the originally obtained power data, and then check the received data, and perform first decoding and then decompression processing on the data that passes the check, and store the processed data in the data lake.

6. The method for monitoring carbon emissions of power users according to claim 5, characterized in that: The wavelet transform is used for the compression, and the inverse wavelet transform corresponding to the wavelet transform is used for the decompression, and the wavelet basis function used is the Daubechies wavelet basis function.

7. The method for monitoring carbon emissions of power users according to claim 5, characterized in that: The Huffman encoder is used for the encoding, and the Huffman decoder corresponding to the Huffman encoder is used for the decoding.

8. The method for monitoring carbon emissions of power users according to claim 5, characterized in that: The originally obtained power data is the power data after preprocessing the originally collected power data, and the preprocessing method includes removing high-frequency noise and / or removing outliers.

9. A computer device comprising a processor, characterized in that: The processor executes a computer program to implement the steps of the method according to any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the method according to any one of claims 1 to 8.

Images