Electric power carbon emission influence factor analysis method and system

Abstract

The invention discloses an electric power carbon emission influence factor analysis method and system. The system comprises a multi-source data acquisition module, a data preprocessing module, an influence factor analysis model construction module, a real-time analysis and dynamic display module and a decision support module. Through the multi-source data acquisition module, the data preprocessing module, the influence factor analysis model construction module, the real-time analysis and dynamic display module and the decision support module, the comprehensiveness of data acquisition, the accuracy of a deep learning algorithm and advanced data processing and the real-time performance of real-time data acquisition and analysis are realized; and the decision support module can provide support for a plurality of optimization decision schemes according to the analysis result.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon emission monitoring and analysis in the power industry, and specifically to a method and system for analyzing factors affecting power carbon emissions. The method can be used to assist power companies in optimizing their production structures, formulating emission reduction strategies, and providing data support for energy management and policy formulation by government departments. Background Art

[0002] With growing global concern about climate change, reducing carbon emissions has become a critical task for all countries. As a major source of carbon emissions, the power industry necessitates an accurate analysis of the factors influencing its emissions. Currently, traditional methods for analyzing factors influencing power carbon emissions suffer from the following major issues: First, data sources are limited, often relying on limited operational parameters of the power production process, such as power generation and coal consumption. This fails to comprehensively capture the full range of factors influencing power carbon emissions. Factors such as changes in energy mix, grid efficiency, and demand-side management measures are often overlooked, leading to biased analysis results. Second, analytical methods are relatively simplistic, often employing traditional statistical methods or empirical models. These methods cannot effectively handle complex nonlinear relationships, making it difficult to deeply explore the inherent connections between various factors and carbon emissions, and unable to accurately quantify the impact of each factor on carbon emissions. Third, existing analysis systems lack dynamic and real-time capabilities, failing to promptly reflect the impact of changes in various factors during power system operation on carbon emissions, and failing to meet the power industry's growing demand for refined management and real-time decision-making. Therefore, a method and system for comprehensive, accurate, and real-time analysis of factors influencing power carbon emissions is urgently needed. Summary of the Invention

[0003] The present invention aims to provide a method and system for analyzing factors affecting carbon emissions from electricity. By integrating multi-source data and applying advanced artificial intelligence algorithms, a comprehensive and accurate analysis of factors affecting carbon emissions from electricity can be achieved, the degree of influence of each factor on carbon emissions can be quantified, and changes in each factor can be tracked dynamically in real time, providing a scientific and effective decision-making basis for energy conservation and emission reduction in the power industry.

[0004] The technical solution adopted by the present invention to solve its technical problems is: a multi-source data acquisition module: This module is responsible for collecting multi-source data related to electricity carbon emissions, including but not limited to power production data (such as generator type, power generation, fuel consumption, boiler efficiency, etc.), energy structure data (the proportion of various power generation energy sources, such as coal, natural gas, wind power, solar power, etc.), power grid operation data (transmission line losses, transformer efficiency, grid load rate, etc.), power demand data (power demand in different time periods and different regions), and macroeconomic and policy data (GDP growth rate, energy price fluctuations, environmental protection policy changes, etc.). By connecting with the power company's production management system, energy monitoring platform, power grid dispatching system, and data interfaces of relevant government departments, real-time and automatic data collection is achieved.

[0005] The preferred technical solution, the data preprocessing module, cleans, converts, and normalizes collected multi-source data. Data cleaning algorithms are used to remove noise, outliers, and duplicate data. Data format conversion and encoding are performed based on the data type and analysis requirements. Normalization methods are used to map data of different dimensions to a unified interval, improving data usability and the accuracy of analytical models. Missing data is filled in using machine learning-based interpolation algorithms, such as using regression or neural network models to predict missing values.

[0006] As the preferred technical solution, the influencing factor analysis model construction module: constructs an analysis model of factors affecting electricity carbon emissions based on deep learning algorithms. The selection of model structures such as multi-layer perceptron (MLP) or long short-term memory network (LSTM), combined with the attention mechanism, can effectively capture the complex nonlinear relationship and time series characteristics between various factors and carbon emissions. The preprocessed data is divided into training set, validation set and test set. The model parameters are optimized through the training set, the model hyperparameters are adjusted using the validation set, and the generalization ability of the model is evaluated with the test set to ensure the accuracy and reliability of the model. During the model training process, feature importance analysis algorithms are introduced, such as SHAP (SHapley Additive exPlanations) value calculation, to quantify the degree of influence of each factor on carbon emissions.

[0007] As a preferred technical solution, the real-time analysis and dynamic display module inputs real-time collected data into a trained influencing factor analysis model for real-time analysis, rapidly outputting the real-time impact of each factor on electricity carbon emissions and carbon emission forecasts. Through a visual interface, the analysis results are presented in intuitive charts and graphs, including trends in the influence weights of various factors, comparisons of factors influencing carbon emissions over different time periods, and future carbon emission forecast curves. Furthermore, an early warning mechanism is implemented to issue timely warnings when changes in key influencing factors may cause carbon emissions to exceed preset thresholds, prompting relevant personnel to take action.

[0008] As a preferred technical solution decision support module, this module provides users with a variety of optimized decision-making options based on the results of influencing factor analysis and carbon emission forecasts, combined with the power company's production goals and energy conservation and emission reduction requirements. For example, based on the analysis of the impact of energy structure adjustments on carbon emissions, it proposes recommendations for optimizing energy allocation; based on the relationship between grid operating efficiency and carbon emissions, it formulates grid optimization scheduling strategies; and for power demand-side management, it provides measures to guide users in rational electricity use. This module also supports users in conducting simulation evaluations of decision-making options. By varying relevant parameters, it predicts the changes in carbon emissions under different decision-making options, helping users select the optimal solution.

[0009] Compared with the prior art, the present invention has the following advantages:

[0010] 1. Comprehensiveness: By collecting multi-source data, it covers the influencing factors of power production, energy structure, grid operation, power demand, macroeconomics and policies, etc., and can fully reflect the influencing factor system of power carbon emissions, avoiding one-sided analysis.

[0011] 2. Accuracy: The use of deep learning algorithms and advanced data processing technology can effectively handle complex nonlinear relationships, accurately quantify the impact of various factors on carbon emissions, and improve the accuracy and reliability of analysis results.

[0012] 3. Real-time: It realizes real-time data collection and analysis, which can promptly reflect the impact of changes in various factors during the operation of the power system on carbon emissions, providing strong support for real-time decision-making.

[0013] 4. Decision support: The decision support module can provide a variety of optimized decision-making plans based on the analysis results, and support simulation evaluation to help users formulate scientific and reasonable energy-saving and emission reduction strategies. It has strong practicality and decision-making support value. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is an architecture diagram of the power carbon emission influencing factor analysis system of the present invention;

[0015] Figure 2 This is a schematic diagram of the influencing factor analysis model structure;

[0016] Figure 3 It is a data processing flow chart; DETAILED DESCRIPTION

[0017] In order to make the purpose, technical solutions and advantages of the present invention clearer, the present invention is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In addition, in the description of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0018] Example

[0019] refer to Figure 1-Figure 3The multi-source data collection system shown includes power production data collection: This system connects with the power company's generator control system and fuel management system to collect real-time operating parameters such as power generation, fuel consumption (coal, natural gas, etc.), boiler efficiency, and unit start and stop times. For example, at a thermal power plant, sensors and data acquisition devices are installed on the generator equipment to collect data every 15 minutes and transmit it to a data acquisition server. Energy structure data collection: This system obtains data on the proportion of various power generation energy sources from energy management departments and relevant statistical agencies, and updates it regularly (e.g., monthly). Furthermore, this system collects construction and operation data on new energy power generation projects, including installed capacity and power generation of wind farms and panel area and power generation efficiency of solar photovoltaic power plants, to keep abreast of changes in the energy structure. Grid operation data collection: This system connects with the grid dispatching system to collect data such as transmission line losses, transformer efficiency, grid load factor, and voltage compliance rate. Smart meters and grid monitoring equipment are used to obtain real-time grid operation parameters, providing data support for analyzing the impact of grid operation on carbon emissions. Electricity demand data collection: Utilizing the power marketing system and load forecasting system, electricity demand data for different time periods and regions is collected, including data for various types of users, including industrial, residential, and commercial electricity users. Meteorological data (such as temperature and humidity) and holiday information are also combined to analyze factors influencing electricity demand and improve the accuracy of demand forecasts. Macroeconomic and policy data collection: Data such as GDP growth rate, energy price index, and environmental protection policy documents are obtained from government public data platforms. Web crawler technology is used to track the release and adjustment of relevant policies in real time, analyzing the impact of macroeconomic and policy changes on electricity carbon emissions. Data preprocessing and cleaning: Statistical methods are used to identify outliers in the collected power production data. For example, data is set within upper and lower limits, and data outside the range is considered outliers and removed. Fields with few missing values are imputed using methods such as the mean and median. Fields with many missing values are predicted using random forest algorithms or neural networks. For example, when processing power generation data for a certain period of time, by setting a reasonable power range, five outliers were removed, and two missing values were imputed using the random forest algorithm. Data conversion and normalization: Convert different types of data into different formats, such as converting time data into a unified timestamp format and encoding categorical data (e.g., using one-hot encoding). Then, use the Min-Max normalization method to map the data to the [0,1] interval. The formula is: Where X is the original data, X min and X max are the minimum and maximum values of the data, respectively, normThe data is normalized. Data feature engineering: Extract and combine features from the raw data based on analysis requirements. For example, derived features such as fuel consumption per unit of power generation and transmission line loss rate are calculated to increase the dimensionality and information content of the data and improve the model's analytical capabilities. Influencing factor analysis model construction, model selection and architecture design: A multi-layer perceptron (MLP) is selected as the basic model structure, and an attention mechanism is introduced. The MLP contains three hidden layers with 128, 64, and 32 neurons, respectively. The input layer receives preprocessed multi-source data, and the output layer outputs the impact weight of each influencing factor on carbon emissions and the predicted carbon emissions value. The attention mechanism is used to enhance the model's focus on key influencing factors and improve the model's accuracy. Model training: The preprocessed data is divided into training, validation, and test sets in a ratio of 7:1:2. The model is trained using the training set, using the Adam optimizer, setting the learning rate to 0.001, and the mean squared error (MSE) as the loss function. During training, the model performance was evaluated on the validation set every 10 training batches. Model hyperparameters, such as the number of hidden layer neurons and the learning rate, were adjusted based on the validation set loss. After 500 training epochs, the model achieved a mean squared error of 0.05 on the test set, demonstrating good generalization. Feature Importance Analysis: The trained model was analyzed using the SHAP value calculation method. By calculating the SHAP value for each feature, the impact of each factor on carbon emissions was quantified and ranked from greatest to least significant. For example, the analysis results showed that fuel consumption, power generation energy mix, and grid load rate were the top three factors influencing electricity carbon emissions. Real-time Analysis and Dynamic Display: Real-time analysis: After preprocessing, the collected data is fed into the trained influencing factor analysis model. The model then outputs the impact of each factor on current electricity carbon emissions and a carbon emissions forecast within one minute. For example, if a sudden increase in grid load rate is detected during a certain period, the model can quickly analyze its impact on carbon emissions and predict future carbon emissions trends. Dynamic display: Through the web interface or mobile application, the analysis results are presented in a variety of chart formats, such as line charts, bar charts, and pie charts. Line charts are used to show the changing trends of the influence weights of various factors, bar charts compare the impact of various factors on carbon emissions in different time periods, and pie charts show the proportion of energy structure. At the same time, an early warning prompt area is set up on the interface. When changes in key influencing factors cause carbon emissions to approach or exceed the preset threshold, a warning message is issued in the form of a red warning light and a pop-up window. Decision support and decision plan generation: Based on the analysis results of influencing factors and carbon emission forecasts, the decision support module automatically generates a variety of optimized decision plans.For example, if the analysis finds that the excessively high proportion of coal-fired power generation leads to increased carbon emissions, the system will propose suggestions such as increasing the proportion of clean energy (such as wind and solar power) in power generation and optimizing the combustion efficiency of coal-fired power generation units. If large grid losses affect carbon emissions, the system will provide scheduling strategies such as optimizing the grid topology and improving transformer operating efficiency. Simulation evaluation: Users can perform simulation evaluations on different decision-making plans on the decision support interface. By inputting different parameters (such as the adjustment range of the proportion of clean energy power generation, the implementation time of grid optimization measures, etc.), the system uses the trained model to predict the changes in carbon emissions in the future under different decision-making plans, and displays the evaluation results in the form of charts and data to help users choose the best energy-saving and emission reduction plan.

[0020] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method and system for analyzing factors affecting electricity carbon emissions, including: The multi-source data acquisition module, data preprocessing module, influencing factor analysis model construction module, real-time analysis and dynamic display module, and decision support module include: The multi-source data acquisition module is responsible for collecting multi-source data related to electricity carbon emissions, including but not limited to power production data (such as generator type, power generation, fuel consumption, boiler efficiency, etc.), energy structure data (the proportion of various power generation energy sources, such as coal, natural gas, wind power, solar power, etc.), power grid operation data (transmission line losses, transformer efficiency, power grid load rate, etc.), power demand data (power demand in different time periods and regions), and macroeconomic and policy data (GDP growth rate, energy price fluctuations, environmental protection policy changes, etc.). By connecting with the power company's production management system, energy monitoring platform, power grid dispatching system, and data interfaces of relevant government departments, real-time and automatic data collection is achieved.

2. The method and system for analyzing factors affecting carbon emissions of electricity according to claim 1 are characterized by The data preprocessing module cleans, converts, and normalizes the collected multi-source data. It uses data cleaning algorithms to remove noise, outliers, and duplicate data; it converts and encodes data based on data type and analysis requirements; and it uses normalization to map data of different dimensions to a unified range, improving data usability and the accuracy of analytical models. At the same time, for missing data, interpolation algorithms based on machine learning are used to fill in the missing data, such as using regression models or neural network models to predict missing values.

3. The method and system for analyzing factors affecting carbon emissions of electricity according to claim 1 are characterized by The influencing factor analysis model construction module constructs an analysis model of influencing factors of electricity carbon emissions based on a deep learning algorithm. The selection of model structures such as multi-layer perceptron (MLP) or long short-term memory network (LSTM), combined with the attention mechanism, can effectively capture the complex nonlinear relationship and time series characteristics between various factors and carbon emissions. The preprocessed data is divided into training set, validation set and test set. The model parameters are optimized through the training set, the model hyperparameters are adjusted using the validation set, and the generalization ability of the model is evaluated with the test set to ensure the accuracy and reliability of the model. During the model training process, feature importance analysis algorithms are introduced, such as SHAP (SHapley Additive exPlanations) value calculation, to quantify the degree of influence of each factor on carbon emissions.

4. The method and system for analyzing factors affecting carbon emissions of electricity according to claim 1 are characterized by The real-time analysis and dynamic display module feeds real-time collected data into a trained influencing factor analysis model for real-time analysis, rapidly outputting the real-time impact of each factor on electricity carbon emissions and carbon emission forecasts. Through a visual interface, the analysis results are presented in intuitive charts and graphs, including trends in the weighting of each factor, comparisons of factors influencing carbon emissions over different time periods, and future carbon emission forecasts. Furthermore, an early warning mechanism is implemented to issue timely alerts when changes in key influencing factors could cause carbon emissions to exceed preset thresholds, prompting relevant personnel to take appropriate action.

5. The method and system for analyzing factors affecting carbon emissions of electricity according to claim 1 are characterized by The decision support module provides users with a variety of optimization decision-making solutions based on the analysis results of influencing factors and carbon emission forecasts, combined with the production goals and energy conservation and emission reduction requirements of power companies. For example, based on the analysis of the impact of energy structure adjustment on carbon emissions, it provides suggestions for optimizing energy allocation; Based on the relationship between grid efficiency and carbon emissions, an optimized grid dispatch strategy is developed. For power demand-side management, measures are provided to guide users in optimizing electricity usage. This module also supports simulation evaluation of decision-making scenarios. By varying relevant parameters, it predicts changes in carbon emissions under different decision-making scenarios, helping users select the optimal option.

Images