Carbon emission accurate accounting system based on big data analysis

Abstract

The invention discloses a carbon emission accurate accounting system based on big data analysis, and relates to the technical field of carbon emission management, and the system comprises a data obtaining module which obtains carbon emission data in real time, and carries out the processing and fusion of the data based on a distributed big data storage technology; the carbon emission accounting engine module is used for processing the fused data based on a big data analysis algorithm, establishing a carbon emission calculation model and calculating the carbon emission; the intelligent prediction module is used for predicting future carbon emission by adopting a deep learning model; the carbon credit automatic evaluation and transaction module establishes a carbon credit scoring algorithm, dynamically evaluates the carbon emission reduction performance of the user based on the real-time carbon emission data and the carbon emission prediction data, and automatically generates the carbon credit. According to the invention, by integrating multi-source large-scale data and combining adaptive data weighting, a deep learning prediction model and a carbon credit automatic evaluation algorithm, accurate accounting, dynamic prediction and intelligent carbon credit evaluation of carbon emission can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon emission management, and in particular to a carbon emission precise accounting system based on big data analysis. Background Art

[0002] As the global climate change problem becomes increasingly serious, carbon emission management has become a global focus. In order to cope with climate change, many countries and regions have formulated strict carbon emission targets and implemented carbon emission trading mechanisms. In this context, accurate, real-time and comprehensive accounting of carbon emissions has become the key to implementing carbon emission management and evaluating carbon credits.

[0003] Existing carbon emission accounting systems usually rely on traditional carbon emission factors and fixed statistical data. This approach often has problems such as data lag and insufficient precision, and it is difficult to cope with large-scale, real-time, and dynamic data needs. Especially in the assessment and trading of carbon credits, how to accurately assess the carbon emission reduction effects of users and generate reasonable carbon credits has always been a technical problem in the industry. Therefore, a carbon emission accurate accounting system based on big data analysis has emerged. It aims to achieve real-time and accurate carbon emission accounting and prediction by integrating multi-source, large-scale, and high-precision carbon emission data, using advanced algorithms and deep learning technology, and providing intelligent decision-making support for carbon emission management. Summary of the invention

[0004] In order to solve the above technical problems, a precise carbon emissions accounting system based on big data analysis is provided. This technical solution solves the above problems.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] The carbon emission accurate accounting system based on big data analysis includes:

[0007] Data acquisition module: acquires carbon emission data from different sources in real time based on API interface, and processes and integrates data based on distributed big data storage technology;

[0008] Carbon emission accounting engine module: the carbon emission accounting engine module is electrically connected to the carbon emission accounting engine module, and the carbon emission accounting engine module is used to process the fused data based on the big data analysis algorithm, establish a carbon emission calculation model, and calculate the carbon emissions according to the regional carbon emission accounting standards;

[0009] Intelligent prediction module: The intelligent prediction module is electrically connected to the carbon emission accounting engine module, and is used to predict future carbon emissions based on real-time carbon emission data using a deep learning model;

[0010] Carbon credit automatic assessment and trading module: The carbon credit automatic assessment and trading module is electrically connected to the intelligent prediction module and the carbon emission accounting engine module. The carbon credit automatic assessment and trading module is used to establish a carbon credit scoring algorithm, dynamically evaluate the user's carbon emission reduction performance based on real-time carbon emission data and carbon emission prediction data, and automatically generate carbon credits.

[0011] Preferably, the data acquisition module specifically includes:

[0012] Data source identification and classification unit: Obtain carbon emission data from multiple sources, including energy consumption data, traffic flow data, industrial production data, climate data, sensor monitoring data, and industry carbon emission statistics, and classify data sources into different categories, including enterprise level, regional level, real-time sensor level, and climate model level;

[0013] Data capture and access unit: capture data from the data source in real time based on the API interface, and use the JSON unified data format for data transmission;

[0014] Data preprocessing unit: cleans the captured raw data, including missing value filling, outlier detection and correction, and duplicate data removal, normalizes data from different sources according to unified standards, and converts different units of energy consumption uniformly;

[0015] Data fusion and integration unit: Based on distributed big data storage technology, data is processed and integrated in parallel, and data from different sources are combined using data fusion algorithms to form a complete carbon emission data set. The data fusion algorithm formula is:

[0016]

[0017] In the formula, D is the fused data result, D i is the observation value of the ith data source, w i For w i The weight of a data source, n is the total number of data sources;

[0018] Adaptive data weighting unit: Based on the adaptive data weighting algorithm, different data sources are given different weights according to the relevance of the data source to carbon emissions, data accuracy and data credibility.

[0019] Preferably, the adaptive data weighting algorithm is based on which different weights are assigned to different data sources according to the relevance of the data source to carbon emissions, data accuracy and data credibility, and specifically includes:

[0020] The correlation between each data source and carbon emissions is measured based on the Pearson correlation coefficient, and the accuracy of each data source and data credibility assessment are obtained based on historical data;

[0021] Calculate the weight of each data source, and sum the relevance, data source accuracy, and data credibility by weighted average. The weight calculation formula is:

[0022]

[0023] In the formula, w i For w i The weight of the data source, α is the weight factor of the correlation term, β is the weight factor of the data source accuracy term, δ is the weight factor of the data credibility term, n is the total number of data sources, g j is the jth correlation value, c j is the precision value of the jth data source, h j is the jth data credibility value.

[0024] Preferably, the carbon emission accounting engine module specifically includes:

[0025] Carbon emission factor and accounting standard selection unit: Carry out carbon emission accounting based on regional standards and obtain carbon emission factors for energy and industrial activities;

[0026] Carbon emission calculation model building unit: Based on the collected data and the selected carbon emission factors, a carbon emission calculation model is built, and the carbon emissions consumed are calculated based on the weighted average formula and the weight calculation results. The carbon emission calculation formula is:

[0027]

[0028] In the formula, C is the total carbon emissions, D x is the xth fused data result, A x The activity volume of the kth type of activity, n is the total number of activities.

[0029] Preferably, the intelligent prediction module specifically includes:

[0030] Prediction model building unit: acquire carbon emission data in real time, build a carbon emission prediction model based on a recurrent neural network algorithm, and predict carbon emissions based on the carbon emission prediction model;

[0031] Model training unit: measures the difference between the model prediction value and the true value based on the loss function;

[0032] Model parameter updating unit: Use the gradient descent method to update the model parameters and repeat multiple iterations until the loss function converges.

[0033] Preferably, the real-time acquisition of carbon emission data, establishing a carbon emission prediction model based on a recurrent neural network algorithm, and predicting carbon emissions based on the carbon emission prediction model specifically include:

[0034] Among them, the carbon emission prediction model formula is:

[0035]

[0036] In the formula, h t is the hidden state at time t, x t is the input feature at time t, w h is the weight matrix at the input moment, b h is the bias term, σ is the activation function, is the predicted value of the model at time t, s y is the weight matrix from the hidden layer to the output layer, b y is the bias of the output layer;

[0037] Based on the predicted value of the model at time t, the predicted carbon emission value is output to obtain the carbon emission reduction, and the difference between the model predicted value and the true value is measured based on the loss function.

[0038] Preferably, the method of measuring the difference between the model prediction value and the true value based on the loss function specifically includes:

[0039]

[0040] Where MSE is the mean square error, T is the total number of samples, t is the current time step, and y t The true value in the tth sample, The predicted value in the tth sample;

[0041] Based on the calculated mean square error, the model parameters are updated using the gradient descent method, and repeated for multiple iterations until the loss function converges.

[0042] Preferably, the method of updating the model parameters using the gradient descent method and repeating multiple iterations until the loss function converges specifically includes:

[0043] Get the gradient of the predicted value of the loss function and backpropagate the gradient through each layer of the network. The gradient of each layer is backpropagated recursively.

[0044] Based on the gradient of the weights and biases with respect to the loss function, the gradient descent algorithm is used to update the parameters, where the update rule is:

[0045]

[0046] In the formula, is the parameter of the model, is the learning rate, is the gradient of the loss function with respect to the parameters;

[0047] The parameters are updated by back-propagating the gradients obtained, and repeated for multiple iterations until the loss function converges.

[0048] Preferably, the carbon credit automatic assessment and trading module specifically includes:

[0049] Carbon credit scoring algorithm design unit: Based on real-time collected data and budget results, design a carbon credit scoring algorithm to evaluate the user's carbon emission reduction performance;

[0050] Carbon credit generation and allocation unit: Based on the calculated carbon credit score, dynamically allocate carbon credits according to the user's carbon reduction performance, and record the carbon credit balance of each user.

[0051] Preferably, the designing of a carbon credit scoring algorithm to evaluate the user's carbon emission reduction performance based on the collected data specifically includes:

[0052] Set the baseline emission value for each user, obtain the actual emission reduction, predicted emission reduction and additionality assessment, weight different types of emission reductions, and establish a scoring model. The scoring model formula is:

[0053]

[0054] In the formula, S(t) is the carbon credit score, p is the weight coefficient, and C o is the baseline carbon emission, R(t) is the carbon emission reduction, R F (t) is the predicted carbon emission reduction, C a (t) is the actual carbon emission, k is the weight coefficient, and j is the weight coefficient.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] The present invention proposes to integrate large-scale data from multiple sources, combine adaptive data weighting, deep learning prediction models and carbon credit automatic assessment algorithms, so as to achieve accurate accounting, dynamic prediction and intelligent carbon credit assessment of carbon emissions. The implementation of this system will provide more accurate, real-time and operational data support for carbon emission management and carbon market transactions, promote the realization of carbon emission reduction targets, and promote the sustainable development of global climate change response. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a system framework diagram of the present invention;

[0058] Figure 2 It is a system framework diagram of the data acquisition module in the present invention;

[0059] Figure 3 This is a system framework diagram of the carbon emission accounting engine module in the present invention;

[0060] Figure 4 This is a system framework diagram of the intelligent prediction module in the present invention;

[0061] Figure 5 This is a system framework diagram of the carbon credit automatic assessment and trading module in the present invention. DETAILED DESCRIPTION

[0062] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art may think of other obvious variations.

[0063] Reference Figure 1 As shown in the figure, the carbon emission accurate accounting system based on big data analysis includes:

[0064] Data acquisition module: acquires carbon emission data from different sources in real time based on API interface, and processes and integrates data based on distributed big data storage technology;

[0065] Carbon emission accounting engine module: the carbon emission accounting engine module is electrically connected to the carbon emission accounting engine module, and the carbon emission accounting engine module is used to process the fused data based on the big data analysis algorithm, establish a carbon emission calculation model, and calculate the carbon emissions according to the regional carbon emission accounting standards;

[0066] Intelligent prediction module: The intelligent prediction module is electrically connected to the carbon emission accounting engine module, and is used to predict future carbon emissions based on real-time carbon emission data using a deep learning model;

[0067] Carbon credit automatic assessment and trading module: The carbon credit automatic assessment and trading module is electrically connected to the intelligent prediction module and the carbon emission accounting engine module. The carbon credit automatic assessment and trading module is used to establish a carbon credit scoring algorithm, dynamically evaluate the user's carbon emission reduction performance based on real-time carbon emission data and carbon emission prediction data, and automatically generate carbon credits.

[0068] Reference Figure 2 As shown, the data acquisition module specifically includes:

[0069] Data source identification and classification unit: Obtain carbon emission data from multiple sources, including energy consumption data, traffic flow data, industrial production data, climate data, sensor monitoring data, and industry carbon emission statistics, and classify data sources into different categories, including enterprise level, regional level, real-time sensor level, and climate model level;

[0070] Data capture and access unit: capture data from the data source in real time based on the API interface, and use the JSON unified data format for data transmission;

[0071] Data preprocessing unit: cleans the captured raw data, including missing value filling, outlier detection and correction, and duplicate data removal, normalizes data from different sources according to unified standards, and converts different units of energy consumption uniformly;

[0072] Data fusion and integration unit: Based on distributed big data storage technology, data is processed and integrated in parallel, and data from different sources are combined using data fusion algorithms to form a complete carbon emission data set. The data fusion algorithm formula is:

[0073]

[0074] In the formula, D is the fused data result, D i is the observation value of the ith data source, w i For w i The weight of a data source, n is the total number of data sources;

[0075] Adaptive data weighting unit: Based on the adaptive data weighting algorithm, different data sources are assigned different weights according to the relevance of the data source to carbon emissions, data accuracy, and data credibility;

[0076] The adaptive data weighting algorithm assigns different weights to different data sources according to the relevance of the data source to carbon emissions, data accuracy, and data credibility, specifically including:

[0077] The correlation between each data source and carbon emissions is measured based on the Pearson correlation coefficient, and the accuracy of each data source and data credibility assessment are obtained based on historical data;

[0078] Calculate the weight of each data source, and sum the relevance, data source accuracy, and data credibility by weighted average. The weight calculation formula is:

[0079]

[0080] In the formula, w i For w i The weight of the data source, α is the weight factor of the correlation term, β is the weight factor of the data source accuracy term, δ is the weight factor of the data credibility term, n is the total number of data sources, g j is the jth correlation value, c j is the precision value of the jth data source, h j is the jth data credibility value.

[0081] Reference Figure 3 As shown, the carbon emission accounting engine module specifically includes:

[0082] Carbon emission factor and accounting standard selection unit: Carry out carbon emission accounting based on regional standards and obtain carbon emission factors for energy and industrial activities;

[0083] Carbon emission calculation model building unit: Based on the collected data and the selected carbon emission factors, a carbon emission calculation model is built, and the carbon emissions consumed are calculated based on the weighted average formula and the weight calculation results. The carbon emission calculation formula is:

[0084]

[0085] In the formula, C is the total carbon emissions, D x is the xth fused data result, A x The activity volume of the kth type of activity, n is the total number of activities.

[0086] Reference Figure 4 As shown, the intelligent prediction module specifically includes:

[0087] Prediction model building unit: acquire carbon emission data in real time, build a carbon emission prediction model based on a recurrent neural network algorithm, and predict carbon emissions based on the carbon emission prediction model;

[0088] Model training unit: measures the difference between the model prediction value and the true value based on the loss function;

[0089] Model parameter updating unit: Use the gradient descent method to update the model parameters, repeating multiple iterations until the loss function converges;

[0090] The real-time acquisition of carbon emission data, establishing a carbon emission prediction model based on a recurrent neural network algorithm, and predicting carbon emissions based on the carbon emission prediction model specifically include:

[0091] Among them, the carbon emission prediction model formula is:

[0092]

[0093] In the formula, h t is the hidden state at time t, x t is the input feature at time t, w h is the weight matrix at the input moment, b h is the bias term, σ is the activation function, is the predicted value of the model at time t, s y is the weight matrix from the hidden layer to the output layer, b y is the bias of the output layer;

[0094] Based on the model's predicted value at time t, the predicted carbon emission value is output to obtain the carbon emission reduction, and the difference between the model's predicted value and the true value is measured based on the loss function;

[0095] The method of measuring the difference between the model prediction value and the true value based on the loss function specifically includes:

[0096]

[0097] Where MSE is the mean square error, T is the total number of samples, t is the current time step, and y t The true value in the tth sample, The predicted value in the tth sample;

[0098] Based on the calculated mean square error, the model parameters are updated using the gradient descent method, and repeated for multiple iterations until the loss function converges.

[0099] The method of updating the model parameters by using the gradient descent method and repeating multiple iterations until the loss function converges specifically includes:

[0100] Get the gradient of the predicted value of the loss function and backpropagate the gradient through each layer of the network. The gradient of each layer is backpropagated recursively.

[0101] Based on the gradient of the weights and biases with respect to the loss function, the gradient descent algorithm is used to update the parameters, where the update rule is:

[0102]

[0103] In the formula, is the parameter of the model, is the learning rate, is the gradient of the loss function with respect to the parameters;

[0104] The parameters are updated by back-propagating the gradients obtained, and repeated for multiple iterations until the loss function converges.

[0105] Reference Figure 5 As shown, the carbon credit automatic assessment and trading module specifically includes:

[0106] Carbon credit scoring algorithm design unit: Based on real-time collected data and budget results, design a carbon credit scoring algorithm to evaluate the user's carbon emission reduction performance;

[0107] Carbon credit generation and allocation unit: Based on the calculated carbon credit score, dynamically allocate carbon credits according to the user's carbon reduction performance, and record each user's carbon credit balance;

[0108] The carbon credit scoring algorithm designed based on the collected data to evaluate the user's carbon emission reduction performance specifically includes:

[0109] Set the baseline emission value for each user, obtain the actual emission reduction, predicted emission reduction and additionality assessment, weight different types of emission reductions, and establish a scoring model. The scoring model formula is:

[0110]

[0111] In the formula, S(t) is the carbon credit score, p is the weight coefficient, and C o is the baseline carbon emission, R(t) is the carbon emission reduction, R F (t) is the predicted carbon emission reduction, C a (t) is the actual carbon emission, k is the weight coefficient, and j is the weight coefficient.

[0112] The use process of the present invention is:

[0113] Step 1: Connect data sources from different sources through API interfaces, including energy consumption data, traffic flow and industrial production data;

[0114] Step 2: Obtain carbon emission data from different sources and classify them according to type;

[0115] Step 3: Use the API interface to capture carbon emission data in real time, ensuring that data transmission uses a unified JSON format;

[0116] Step 4: Clean the original data, fill in missing values, detect and correct outliers, remove duplicate data, and perform unit conversion and normalization;

[0117] Step 5: Process data from different sources in parallel based on distributed big data storage technology and merge them into a complete carbon emissions dataset;

[0118] Step 6: Use an adaptive weighting algorithm to assign weights based on the relevance, accuracy and credibility of the data source to carbon emissions;

[0119] Step 7: Calculate the correlation between each data source and carbon emissions through the Pearson correlation coefficient, evaluate the accuracy and credibility based on historical data, and finally calculate the weight;

[0120] Step 8: Determine the standards and factors for carbon emission accounting to provide a basis for subsequent calculations;

[0121] Step 9: Obtain carbon emission factors for energy and industrial activities according to regional standards;

[0122] Step 10: Based on the collected data and selected factors, a carbon emission calculation model is constructed and the carbon emissions are calculated using a weighted average formula;

[0123] Step 11: Calculate the carbon emissions generated by various activities through the carbon emissions calculation formula to generate carbon emissions results;

[0124] Step 12: Use the recurrent neural network algorithm to establish a carbon emission prediction model to predict future carbon emissions;

[0125] Step 13: Use the loss function to measure the difference between the model prediction value and the true value, and optimize the model through training;

[0126] Step 14: Use the gradient descent method to back-propagate the error and update the parameters of the prediction model until the loss function converges;

[0127] Step 15: Based on the trained prediction model, predict future carbon emissions in real time and calculate carbon emission reductions based on the prediction results;

[0128] Step 16: Evaluate the difference between the predicted value and the true value through the mean square error to guide further optimization of the model;

[0129] Step 17: Based on real-time carbon emission data and forecast results, dynamically calculate the user's carbon emission reduction performance and generate a carbon credit score;

[0130] Step 18: Based on the carbon credit scoring algorithm, calculate each user's carbon credit, allocate corresponding carbon credits according to their carbon emission reduction performance, and record the user's carbon credit balance.

[0131] In summary, the advantages of the present invention are:

[0132] The system can obtain carbon emission data from multiple sources in real time, and process and integrate it through big data storage technology to form a comprehensive and accurate carbon emission data set. This fusion of multi-source data helps to improve the comprehensiveness and accuracy of carbon emission data;

[0133] Through the adaptive data weighting algorithm, the system dynamically adjusts the weight of each data source according to its relevance, accuracy and credibility, further improving the accuracy of data fusion. This dynamic adjustment ensures that the system can maintain high accuracy and credibility when facing different data quality and sources.

[0134] The recurrent neural network model is used for carbon emission forecasting, which can predict future carbon emission trends based on real-time data. Through continuous model training and optimization, the accuracy of the forecast results is continuously improved, providing users with more accurate emission reduction targets and early warning information;

[0135] The system automatically evaluates the user's carbon emission reduction performance based on real-time collected carbon emission data and forecast data, and generates carbon credits for the user through the designed carbon credit scoring algorithm. This process is automated, ensuring the fairness, transparency and timeliness of carbon credits, and providing strong support for the carbon emission trading platform;

[0136] Through the carbon emission factor and accounting standard selection unit, combined with the weighted average calculation model, the system can accurately calculate the carbon emissions of each activity category and provide accurate carbon emission accounting results for governments, enterprises and other organizations. This provides strong data support for policymakers, enterprises and the public, and helps to formulate more scientific and reasonable carbon emission reduction measures;

[0137] The carbon credit generation and allocation unit designed in the system can dynamically allocate carbon credits based on the user's carbon emission reduction performance and record the carbon credit balance of each user. This intelligent management method not only improves the liquidity of carbon credits, but also enhances the transparency and efficiency of carbon emission trading, and helps to realize a market-driven carbon emission management mechanism.

[0138] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions only describe the principles of the present invention. The present invention may be subject to various changes and improvements without departing from the spirit and scope of the present invention. These changes and improvements fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the attached claims and their equivalents.

Claims

1. The carbon emission accurate accounting system based on big data analysis is characterized by: include: Data acquisition module: acquires carbon emission data from different sources in real time based on API interface, and processes and integrates data based on distributed big data storage technology; Carbon emission accounting engine module: the carbon emission accounting engine module is electrically connected to the carbon emission accounting engine module, and the carbon emission accounting engine module is used to process the fused data based on the big data analysis algorithm, establish a carbon emission calculation model, and calculate the carbon emissions according to the regional carbon emission accounting standards; Intelligent prediction module: The intelligent prediction module is electrically connected to the carbon emission accounting engine module, and is used to predict future carbon emissions based on real-time carbon emission data using a deep learning model; Carbon credit automatic assessment and trading module: The carbon credit automatic assessment and trading module is electrically connected to the intelligent prediction module and the carbon emission accounting engine module. The carbon credit automatic assessment and trading module is used to establish a carbon credit scoring algorithm, dynamically evaluate the user's carbon emission reduction performance based on real-time carbon emission data and carbon emission prediction data, and automatically generate carbon credits.

2. The carbon emission accurate accounting system based on big data analysis according to claim 1 is characterized in that: The data acquisition module specifically includes: Data source identification and classification unit: Obtain carbon emission data from multiple sources, including energy consumption data, traffic flow data, industrial production data, climate data, sensor monitoring data, and industry carbon emission statistics, and classify data sources into different categories, including enterprise level, regional level, real-time sensor level, and climate model level; Data capture and access unit: capture data from the data source in real time based on the API interface, and use the JSON unified data format for data transmission; Data preprocessing unit: cleans the captured raw data, including missing value filling, outlier detection and correction, and duplicate data removal, normalizes data from different sources according to unified standards, and converts different units of energy consumption uniformly; Data fusion and integration unit: Based on distributed big data storage technology, data is processed and integrated in parallel, and data from different sources are combined using data fusion algorithms to form a complete carbon emission data set. The data fusion algorithm formula is: In the formula, D is the fused data result, D i is the observation value of the ith data source, w i For w i The weight of the data source, n is the total number of data sources; Adaptive data weighting unit: Based on the adaptive data weighting algorithm, different data sources are given different weights according to the relevance of the data source to carbon emissions, data accuracy and data credibility.

3. The carbon emission accurate accounting system based on big data analysis according to claim 2 is characterized in that: The adaptive data weighting algorithm assigns different weights to different data sources according to the relevance of the data source to carbon emissions, data accuracy, and data credibility, specifically including: The correlation between each data source and carbon emissions is measured based on the Pearson correlation coefficient, and the accuracy of each data source and data credibility assessment are obtained based on historical data; Calculate the weight of each data source, and sum the relevance, data source accuracy, and data credibility by weighted average. The weight calculation formula is: In the formula, w i For w i The weight of the data source, α is the weight factor of the correlation term, β is the weight factor of the data source accuracy term, δ is the weight factor of the data credibility term, n is the total number of data sources, g j is the jth correlation value, c j is the precision value of the jth data source, h j is the jth data credibility value.

4. The carbon emission accurate accounting system based on big data analysis according to claim 3 is characterized in that: The carbon emission accounting engine module specifically includes: Carbon emission factor and accounting standard selection unit: Carry out carbon emission accounting based on regional standards and obtain carbon emission factors for energy and industrial activities; Carbon emission calculation model building unit: Based on the collected data and the selected carbon emission factors, a carbon emission calculation model is built, and the carbon emissions consumed are calculated based on the weighted average formula and the weight calculation results. The carbon emission calculation formula is: In the formula, C is the total carbon emissions, D x is the xth fused data result, A x The activity volume of the kth type of activity, n is the total number of activities.

5. The carbon emission accurate accounting system based on big data analysis according to claim 4 is characterized in that: The intelligent prediction module specifically includes: Prediction model building unit: acquire carbon emission data in real time, build a carbon emission prediction model based on a recurrent neural network algorithm, and predict carbon emissions based on the carbon emission prediction model; Model training unit: measures the difference between the model prediction value and the true value based on the loss function; Model parameter updating unit: Use the gradient descent method to update the model parameters and repeat multiple iterations until the loss function converges.

6. The carbon emission accurate accounting system based on big data analysis according to claim 5 is characterized in that: The real-time acquisition of carbon emission data is based on a recurrent neural network algorithm to establish a carbon emission prediction model, based on which the carbon emission prediction model is used to predict the carbon emission. include: Among them, the carbon emission prediction model formula is: In the formula, h t is the hidden state at time t, x t is the input feature at time t, w h is the weight matrix at the input moment, b h is the bias term, σ is the activation function, is the predicted value of the model at time t, s y is the weight matrix from the hidden layer to the output layer, b y is the bias of the output layer; Based on the predicted value of the model at time t, the predicted carbon emission value is output to obtain the carbon emission reduction, and the difference between the model predicted value and the true value is measured based on the loss function.

7. The carbon emission accurate accounting system based on big data analysis according to claim 6 is characterized in that: The method of measuring the difference between the model prediction value and the true value based on the loss function specifically includes: Where MSE is the mean square error, T is the total number of samples, t is the current time step, and y t The true value in the tth sample, The predicted value in the tth sample; Based on the calculated mean square error, the model parameters are updated using the gradient descent method, and repeated for multiple iterations until the loss function converges.

8. The carbon emission accurate accounting system based on big data analysis according to claim 7 is characterized in that: The method of updating the model parameters by using the gradient descent method and repeating multiple rounds of iterations until the loss function converges specifically includes: Get the gradient of the predicted value of the loss function and backpropagate the gradient through each layer of the network. The gradient of each layer is backpropagated recursively. Based on the gradient of the weights and biases with respect to the loss function, the gradient descent algorithm is used to update the parameters, where the update rule is: Where is the model parameter, is the learning rate, is the gradient of the loss function with respect to the parameters; The parameters are updated by back-propagating the gradients obtained, and repeated for multiple iterations until the loss function converges.

9. The carbon emission accurate accounting system based on big data analysis according to claim 8 is characterized in that: The carbon credit automatic assessment and trading module specifically includes: Carbon credit scoring algorithm design unit: Based on real-time collected data and budget results, design a carbon credit scoring algorithm to evaluate the user's carbon emission reduction performance; Carbon credit generation and allocation unit: Based on the calculated carbon credit score, dynamically allocate carbon credits according to the user's carbon reduction performance, and record the carbon credit balance of each user.

10. The carbon emission accurate accounting system based on big data analysis according to claim 9 is characterized in that: The carbon credit scoring algorithm designed based on the collected data to evaluate the user's carbon emission reduction performance specifically includes: Set the baseline emission value for each user, obtain the actual emission reduction, predicted emission reduction and additionality assessment, weight different types of emission reductions, and establish a scoring model. The scoring model formula is: In the formula, S(t) is the carbon credit score, p is the weight coefficient, and C o is the baseline carbon emission, R(t) is the carbon emission reduction, R F (t) is the predicted carbon emission reduction, C a (t) is the actual carbon emission, k is the weight coefficient, and j is the weight coefficient.

Images