Real-time scheduling method for ultra-high voltage power transmission based on carbon emission constraint

Abstract

This invention provides a real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints, belonging to the field of clean energy dispatching technology. This invention predicts the electricity consumption of a target area based on historical electricity data, predicts the power generation based on the historical status of wind and solar power plants, and combines the actual electricity consumption of the target area with the power generation of wind and solar power plants to analyze deviations in real time and adjust the predicted electricity consumption and power generation. Thermal power is used as a supplement to wind and solar power. Carbon emissions are acquired in real time, and the "future" carbon emission gap is determined in real time based on the actual electricity consumption and power generation, providing data support for the dispatching center's dispatching work under carbon emission constraints.

Description

Technical Field

[0001] This invention relates to the field of clean energy dispatching technology, specifically to a real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints. Background Technology

[0002] With the development of ecological construction in my country, policies targeting carbon emissions have been implemented. All units need to operate within the specified carbon emission standards, especially in the power sector. Since a large portion of my country's electricity is generated by thermal power, its carbon emissions are enormous. Therefore, it is now necessary to adjust the power structure by incorporating clean energy to reduce carbon emissions. Clean energy includes wind, solar, hydropower, and nuclear power. Hydropower and nuclear power have high construction costs, so wind and solar power, as easily accessible power resources, actively participate in the electricity market. However, wind and solar power are constrained by instability and limited power generation. Therefore, thermal power is still needed to make up for the shortcomings in their utilization. So, it is currently necessary to control the production of thermal power while making full use of wind and solar power to keep carbon emissions within the allowable range. More importantly, it is necessary to realize the utilization of clean energy. Only by utilizing clean energy can we truly reduce carbon emissions.

[0003] In the prior art, CN118761551A discloses a method, system, equipment, and storage medium for real-time scheduling of ultra-high voltage power transmission based on carbon emission constraints. This method acquires raw time-series information about wind, solar, and energy storage; calculates characteristic data values ​​for wind, solar, and energy storage using the raw time-series information as a time period; after determining the correlation curve corresponding to the characteristic data values, calculates and displays boundary data for wind, solar, and energy storage using the correlation curve and the characteristic data values; and after acquiring user-inputted production chain information, calculates carbon emission values ​​based on the production chain information, where the production chain information is energy storage unit information input by the user after viewing the boundary data and using the boundary data as a condition.

[0004] According to publicly available technical documents, although it has achieved the goal of reducing the prediction bias of wind, solar and carbon emissions, it ignores the actual conditions at the electricity consumption end, cannot determine the clean energy being used, and cannot provide guidance for real-time scheduling and carbon emission target allocation.

[0005] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The purpose of this invention is to provide a real-time scheduling method for ultra-high voltage power transmission based on carbon emission constraints, so as to solve the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: The real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints includes the following steps: Step 1: Obtain the historical power consumption of the target area, once per hour and add a timestamp, summarize into a power consumption dataset, and train it in a long short-term memory model to form a power consumption prediction model. Use the power consumption prediction model to obtain the predicted power consumption curve of the target area for the next day. Step 2: Obtain historical meteorological and power generation data from wind and solar power plants to form a meteorological dataset, and train it in a convolutional neural network model to form a power generation prediction model. Obtain the forecast meteorological data for the next day from the meteorological station and input it into the power generation prediction model to obtain the predicted power generation curve for the next day. Step 3: The next day, generate electricity and transmit it to the target area by combining wind, solar and thermal power. Obtain the actual power consumption curve and the actual power generation curve of wind and solar power in the target area. Compare the predicted power consumption curve and the actual power consumption curve to obtain the power consumption instability. Compare the predicted power generation curve and the actual power generation curve to obtain the power generation instability. Step 4: Correct the predicted power consumption curve and the predicted power generation curve by using the unstable power consumption and unstable power generation respectively, obtain the corrected predicted power consumption curve and the corrected predicted power generation curve, and obtain the thermal power generation increase and the corresponding carbon emission increase based on the difference between the two. Step 5: Obtain the increase in carbon emissions and the actual carbon emissions that have already been generated in real time, form the total carbon emissions, obtain carbon emission indicators, and determine the gap between them and the total carbon emissions.

[0008] Furthermore, the historical power consumption of the target area is obtained hourly, with timestamps added. Based on the holiday characteristics of the timestamps, holidays are marked as 1, and non-holidays as 0. The timestamps are compared with meteorological data from weather stations to obtain temperature characteristics, specifically the temperature of the target area at each timestamp. Power consumption is then divided into daily units, and the hourly power consumption is obtained for each day, forming a daily power consumption dataset. The formula used is as follows: in, This represents a dataset of electricity consumption data. Indicates the time of day. Indicates the characteristics of holidays, This represents the temperature characteristics at this specific time of day. This represents the power consumption at this specific time on this day.

[0009] Furthermore, the electricity consumption data is sorted by date, and a sliding window time length is set to select subsequences in the electricity consumption dataset during the sorting process. The subsequences are used as samples to input into the Long Short-Term Memory Network model, and the electricity consumption power in the electricity consumption dataset of the first day after each sample is used as the label to obtain the electricity consumption prediction model. By acquiring historical power consumption data and corresponding timestamps within a sliding window period in the target area, a daily power consumption dataset is formed. This dataset is then input into the power consumption prediction model to obtain the power consumption at each time point on the second day. A fitting curve is generated according to the time series and labeled as the predicted power consumption curve.

[0010] Furthermore, historical meteorological data and historical power generation data of the wind and solar power plants are acquired. The historical meteorological data includes temperature, wind speed, and solar irradiance, along with corresponding timestamps. The historical power generation data includes power output and timestamps. The historical meteorological data and historical power generation data are correlated based on the timestamps to form a meteorological dataset, using the following formula: in, Represents the power generation dataset, For a point in time, This is the temperature corresponding to this time point. This represents the light intensity at this time point. This represents the wind speed at this point in time.

[0011] Furthermore, the power generation dataset is used to train a convolutional neural network model, with the power generation power corresponding to each meteorological dataset as the label, to obtain a power generation prediction model; The forecast meteorological data for the next day is obtained from the meteorological station. The forecast meteorological data is matched with the timestamps of historical meteorological data to obtain the predicted power generation for the next day. The predicted power generation curve is then formed by fitting the time series.

[0012] Furthermore, on the second day, the dispatch center transmitted electricity to the target area through a combination of wind and solar power plants and thermal power plants. Priority was given to using wind and solar power plants to transmit electricity to the target area, and any shortfall was made up by thermal power plants. During the process of wind and solar power plants supplying power to the target area, the dispatch center acquires the actual power consumption curve of the target area in real time, and simultaneously acquires the actual power generation curve of the wind and solar power plants in real time. Starting from the beginning of the day and ending at the current time, the predicted power consumption curve segment within this time period is compared with the actual power consumption curve to obtain the power instability. The formula used is as follows: in, This indicates unstable electricity consumption. The total duration up to the current time. This is the actual power consumption curve. To predict segments of the power consumption curve, For time-based retrieval variables, ; By comparing the predicted power generation curve segment with the actual power generation curve during this time period, the power generation instability is obtained, based on the following formula: in, This indicates the unstable amount of power generation. The total duration up to the current time. This is the actual power generation curve. To predict segments of the power generation curve, For time-based retrieval variables, .

[0013] Furthermore, the unstable quantities of electricity consumption and generation are acquired in real time, and the remaining predicted power consumption curve and the remaining predicted power generation curve are corrected based on these quantities, respectively, to obtain the corrected predicted power consumption curve and the corrected predicted power generation curve. The formulas used are as follows: in, To correct the predicted power consumption curve, For the remaining predicted power consumption curve, This indicates unstable electricity consumption. To correct the predicted power generation curve, For the remaining predicted power generation curve, This indicates the unstable amount of power generation. For time-based retrieval variables, , This represents the total duration of the day.

[0014] Furthermore, the increase in thermal power generation is obtained based on the real-time acquired modified predicted power consumption curve and modified predicted power generation curve, using the following formula: in, To increase thermal power generation, To correct the predicted power consumption curve, To correct the predicted power generation curve, For time-based retrieval variables, , This is the total duration of today; The correlation between power generation and carbon emissions from thermal power plants is obtained, and the carbon emissions from increased power generation from thermal power plants are calculated based on this correlation and denoted as the increase in carbon emissions. The formula used is as follows: in, For the increase in carbon emissions, To increase thermal power generation, This indicates the correlation between electricity generation and carbon emissions.

[0015] Furthermore, the actual carbon emissions and carbon emission increases generated by thermal power up to the current moment are obtained in real time, and their sum is calculated and labeled as the total carbon emissions. The total carbon emissions are obtained in real time and plotted as a curve. The carbon emission target for the day is obtained through the dispatch center, and the carbon emission target is compared with the total carbon emissions to obtain the target gap. The target gap is the total carbon emissions minus the carbon emission target. A gap threshold is set. When the target gap is greater than the gap threshold, an alarm is triggered to remind the dispatch center to purchase quotas. When the target gap is not greater than the gap threshold, no alarm is triggered.

[0016] Compared with the prior art, the beneficial effects of the present invention are: This invention predicts the electricity consumption of a target area based on historical electricity data, predicts the power generation of wind and solar power plants based on historical data, and combines the actual electricity consumption of the target area with the power generation of wind and solar power plants to analyze deviations in real time and adjust the predicted electricity consumption and power generation. It uses thermal power as a supplement to wind and solar power, acquires carbon emissions in real time, and determines the "future" carbon emission gap for the day based on the actual electricity consumption and power generation, providing data support for the dispatch center's dispatching work under carbon emission constraints. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the overall method flow of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0019] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0020] Example: Please see Figure 1 The present invention provides a technical solution: The real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints includes the following steps: Step 1: Obtain the historical power consumption of the target area, once per hour and add a timestamp, summarize into a power consumption dataset, and train it in a long short-term memory model to form a power consumption prediction model. Use the power consumption prediction model to obtain the predicted power consumption curve of the target area for the next day. Step 1 includes the following: Step 101: Obtain the historical power consumption of the target area. The power consumption is obtained hourly and a timestamp is added. Based on the holiday characteristics of the timestamps, holidays are marked as 1 and non-holidays as 0. The timestamps are compared with meteorological data from weather stations to obtain temperature characteristics. The temperature characteristics are the temperature of the target area at each timestamp node. The power consumption is divided into daily units, and the corresponding hourly power consumption is obtained to form a daily power consumption dataset. The formula is as follows: in, This represents a dataset of electricity consumption data. Indicates the time of day. Indicates the characteristics of holidays, This represents the temperature characteristics at this specific time of day. This represents the power consumption at this specific time on this day.

[0021] Through multi-dimensional data collection and structured processing, comprehensive, detailed, and high-quality input data was prepared for training the electricity consumption forecasting model. Hourly power consumption data collection combined with timestamps ensured the continuity and accuracy of the data over time, providing a high-resolution time series foundation for subsequent model analysis. The labeling of holiday characteristics distinguished between holiday and non-holiday electricity consumption patterns, capturing the impact of social activity patterns on electricity consumption behavior. The introduction of temperature characteristics further integrated the significant moderating effect of meteorological conditions on electricity demand, such as peak electricity consumption caused by summer heat or winter heating. Dividing the data into daily units not only facilitated the analysis of intraday periodic changes in electricity consumption but also provided a standardized sample format for model training, enhancing the data's operability. This data preparation significantly enhanced the model's understanding and adaptability to complex electricity consumption scenarios. Its practical significance lies in providing a solid guarantee for the reliability of the prediction results, avoiding prediction biases caused by single data dimensions or incomplete data collection, and ensuring that the prediction results accurately reflect the electricity consumption characteristics of the target area. In the preceding and following steps, it directly provides rich data materials for the training of the long short-term memory model in step 102, and at the same time, it sets a high starting point and scientific basis for the entire electricity consumption prediction process through the integration of multiple features.

[0022] Step 102: Sort the electricity consumption data by date and set the sliding window time length. Select subsequences in the electricity consumption dataset during sorting. Use the subsequences as samples to input the long short-term memory network model. Use the electricity consumption power in the electricity consumption dataset of the first day after each sample as the label to obtain the electricity consumption prediction model. Set up 2 LSTM layers, each with 128 units, using the tanh function as the activation function, a learning rate of 0.001, and 100 training epochs; By acquiring historical power consumption data and corresponding timestamps within a sliding window period in the target area, a daily power consumption dataset is formed. This dataset is then input into the power consumption prediction model to obtain the power consumption at each time point on the second day. A fitting curve is generated according to the time series and labeled as the predicted power consumption curve.

[0023] Data preprocessing and advanced model training efficiently transform historical electricity consumption data into forecasts for the next day, providing the dispatch center with accurate and practical decision-making tools. Sort by date and combine it with sliding window technology to organize massive amounts of data systematically and extract diverse time segments, enabling the Long Short-Term Memory (LSTM) network model to comprehensively learn short-term fluctuations, long-term trends, and potential seasonal variations in electricity consumption during training. The design using next day's electricity consumption as a label directly focuses on the prediction objective, ensuring the model's output highly aligns with actual application needs, while also reflecting the rigor and forward-looking nature of the data processing logic. The generation of the predicted electricity consumption curve is a crucial step in supply-demand matching, providing an intuitive, accurate, and actionable reference for next day's power supply planning. Furthermore, this prediction process provides technical support for the intelligent upgrade of demand-side management. It builds upon the carefully prepared electricity dataset from step 101, and the output prediction curve provides a scientific benchmark for comparing actual electricity consumption and calculating instability in step 3, fully demonstrating the close logical progression from data acquisition to model application and result output.

[0024] Step 2: Obtain historical meteorological and power generation data from wind and solar power plants to form a meteorological dataset, and train it in a convolutional neural network model to form a power generation prediction model. Obtain the forecast meteorological data for the next day from the meteorological station and input it into the power generation prediction model to obtain the predicted power generation curve for the next day. Step 2 includes the following: Step 201: Obtain historical meteorological data and historical power generation data of the wind and solar power plant. The historical meteorological data includes temperature, wind speed, and solar irradiance, along with corresponding timestamps. The historical power generation data includes power generation and timestamps. The historical meteorological data and historical power generation data are correlated based on the timestamps to form a meteorological dataset, using the following formula: in, Represents the power generation dataset, For a point in time, This is the temperature corresponding to this time point. This represents the light intensity at this time point. This represents the wind speed at this point in time.

[0025] By comprehensively collecting historical data from wind and solar power plants and performing time-series matching, a detailed and multi-dimensional data foundation was laid for training the power generation prediction model. The collection of temperature, wind speed, and solar irradiance data covered all key meteorological variables affecting wind and solar power generation, fully demonstrating the comprehensiveness and scientific rigor of the data collection. The corresponding timestamps ensured strict synchronization between meteorological conditions and power generation over time, providing high-quality causal relationship data for subsequent model analysis. The meteorological dataset provided rich learning material for the convolutional neural network, enabling the model to deeply understand the power generation patterns under different meteorological conditions. Its practical significance lies in significantly improving the adaptability and accuracy of prediction results across various meteorological scenarios, avoiding prediction distortion caused by incomplete data. Furthermore, this process also provided data support for the refined management of new energy power generation, helping to optimize the operation strategies of wind and solar power plants. It provided indispensable raw data for the subsequent training of the convolutional neural network, directly supporting the construction of the power generation prediction model and demonstrating the crucial role of data preparation in the prediction process.

[0026] Step 202: Train the power generation dataset into the input convolutional neural network model, and use the power generation power corresponding to each meteorological dataset as the label to obtain the power generation prediction model; The number of convolutional layers is set to 4, the kernel size is 5x5, the max pooling window is 2x2, the ReLU function is used as the activation function, the learning rate is 0.001, and the training epochs are 100.

[0027] The forecast meteorological data for the next day is obtained from the meteorological station. The forecast meteorological data is matched with the timestamps of historical meteorological data to obtain the predicted power generation for the next day. The predicted power generation curve is then formed by fitting the time series.

[0028] Through efficient model training and predictive applications, historical meteorological and power generation data are transformed into expected power generation capacity for the following day, providing the dispatch center with specific and reliable supply-side references. Training a convolutional neural network using power generation as a label accurately captures the complex mapping relationship between meteorological conditions and power generation. The introduction of forecast meteorological data extends the model's predictive capabilities into the future, achieving a leap from historical analysis to real-time application. The generation of predicted power generation curves is crucial for optimizing the energy structure and improving the efficiency of new energy utilization. Its practical significance lies in providing scientific guidance for the output planning of wind and solar power plants, helping the dispatch center achieve stable power supply even with significant fluctuations in new energy power generation, and providing technical support for reducing thermal power usage and carbon emissions. Furthermore, this prediction result also ensures the dynamic balance of the power grid and the reliability of new energy grid connection. This step generates prediction results based on the carefully prepared data in step 201 and provides a fundamental reference for the actual power generation comparison and instability calculation in step 3, demonstrating a complete logical progression from data integration to model output and application implementation.

[0029] Step 3: The next day, generate electricity and transmit it to the target area by combining wind, solar and thermal power. Obtain the actual power consumption curve and the actual power generation curve of wind and solar power in the target area. Compare the predicted power consumption curve and the actual power consumption curve to obtain the power consumption instability. Compare the predicted power generation curve and the actual power generation curve to obtain the power generation instability. Step 3 includes the following: On the second day, the dispatch center transmitted electricity to the target area through a combination of wind and solar power plants and thermal power plants. Priority was given to using wind and solar power plants to transmit electricity to the target area, and any shortfall was made up by thermal power plants. During the process of wind and solar power plants supplying power to the target area, the dispatch center acquires the actual power consumption curve of the target area in real time, and simultaneously acquires the actual power generation curve of the wind and solar power plants in real time. Starting from the beginning of the day and ending at the current time, the predicted power consumption curve segment within this time period is compared with the actual power consumption curve to obtain the power instability. The formula used is as follows: in, This indicates unstable electricity consumption. The total duration up to the current time. This is the actual power consumption curve. To predict segments of the power consumption curve, For time-based retrieval variables, ; This represents "electricity consumption instability," which is the average deviation between actual and predicted electricity consumption in a target area over a specific time period. Specifically, it reflects the accuracy of the prediction model in the real-world environment, quantifying the gap between actual and predicted electricity demand. In the real-world environment, this corresponds to the volatility of electricity demand, such as when residential electricity consumption deviates from expectations due to weather, holidays, or unforeseen events. The independent variable includes time. (From 0 to ),in It represents the total duration from the beginning of the day to the present moment, signifying the time frame of the assessment; It is the actual power consumption curve, representing the real-time recorded power demand of the target area; This is a predicted power consumption curve, representing the expected electricity demand calculated based on historical data and models. These variables collectively reflect the challenges of electricity consumption forecasting in the real world: actual electricity consumption is affected by a variety of uncontrollable factors, and the predicted values ​​may not fully capture these changes, leading to biases.

[0030] The data represents actual measured data, reflecting the real electricity consumption in the target area, while The representative forecast data represents ideal values ​​calculated based on historical patterns and models. The reliability of the prediction is assessed by comparing these two. If actual electricity consumption... consistently higher than predicted values ,but A positive value indicates that the forecast underestimates demand; conversely, if... Below ,but A negative value indicates that the forecast overestimated demand. The independent variable is time. and time range right The impact is reflected in the processes of integration and averaging: as The increase, the accumulation of the difference over a longer period of time, and the division by The total deviation is then standardized to the average. This relationship indicates that... It depends not only on the deviation at a certain moment, but also on the distribution of deviation over the entire time period, reflecting the overall trend of prediction error.

[0031] By comparing the predicted power generation curve segment with the actual power generation curve during this time period, the power generation instability is obtained, based on the following formula: in, This indicates the unstable amount of power generation. The total duration up to the current time. This is the actual power generation curve. To predict segments of the power generation curve, For time-based retrieval variables, .

[0032] This refers to "unstable power generation," specifically reflecting the power generation instability of wind and solar power plants from the beginning of the day to the present moment (time range 0 to 1). Within a given timeframe, this represents the average deviation between actual and predicted power generation. It quantifies the accuracy of the power generation prediction model in real-world environments, reflecting the gap between predicted and actual values. In real-world environments, this corresponds to the volatility of renewable energy generation, such as real-time changes in meteorological conditions like wind speed and solar radiation, or uncertainties in equipment operation, leading to deviations in actual power generation from expectations. The independent variable includes time. (From 0 to ), It is the actual power generation curve, representing the real-time power generation data recorded by the wind and solar power plant, which is directly affected by meteorological conditions and equipment status. This is a predicted power generation curve, calculated based on historical data and weather forecast models. However, due to forecast errors, it may not perfectly match real-time conditions. These variables collectively reflect the challenges faced by new energy power generation forecasting in the real world: the uncontrollability of the external environment and the biases caused by the limitations of the forecasting models.

[0033] This represents the actual output of wind and solar power plants under current conditions, reflecting their real-time power generation capacity. It is an ideal output calculated based on historical patterns and forecast data. The error of the prediction is assessed by comparing these two. When actual power generation... Higher than predicted hour, A positive value indicates that the forecast underestimates the power generation capacity; conversely, a negative value indicates that the forecast underestimates the power generation capacity. Below , A negative value indicates that the forecast overestimates the power generation capacity. Independent variable: Time and time range right The impact is reflected in the processes of integration and averaging: integration The total deviation over the entire time period was calculated and divided by This is converted into the average deviation per unit time. This design allows... It not only reflects the deviation at a certain moment, but also integrates the deviation trend over a period of time, thus having higher representativeness and stability.

[0034] By comparing actual power supply with data, the deviation between prediction and reality was quantified, providing a basis for subsequent dynamic adjustments. The combined power supply method of wind, solar, and thermal power prioritizes renewable energy and supplements it with thermal power, reflecting both the priority principle of green energy and ensuring the stability of power supply. In the overall scheme, the prediction results from steps 1 and 2 are applied to actual operation. Its practical significance lies in revealing the shortcomings of the prediction model through the calculation of unstable electricity consumption and generation, providing real-time feedback for system optimization. In the preceding and following steps, it receives the predicted data, and the output of unstable quantities provides key input for curve correction in step 4, demonstrating a closed-loop connection between prediction and reality.

[0035] By comparing actual power supply operation with real-time data, the deviation between predicted and actual values ​​was scientifically quantified, providing a crucial basis for subsequent dynamic adjustments and optimizations. The combined power supply of wind, solar, and thermal power prioritizes renewable energy and supplements it with thermal power, fully embodying the principle of prioritizing green energy while ensuring the stability and continuity of power supply through the flexibility of thermal power. This step applies the prediction results from steps 1 and 2 to the actual operating scenario. Through the calculation of power consumption instability and power generation instability, the shortcomings of the prediction model in the real environment are clearly revealed, providing real-time feedback data for continuous system improvement and accuracy enhancement. This process not only achieves closed-loop verification between prediction and reality but also provides a scientific basis for the dispatch center to dynamically adjust power supply strategies, helping to maintain stable grid operation under the dual challenges of new energy fluctuations and changes in electricity demand, while also providing a practical foundation for carbon emission control. This step inherits the prediction data from the previous two steps, and the output instability provides direct input for the prediction curve correction in step 4, demonstrating the complete logical chain of the solution from prediction to application to optimization.

[0036] Step 4: Correct the predicted power consumption curve and the predicted power generation curve by using the unstable power consumption and unstable power generation respectively, obtain the corrected predicted power consumption curve and the corrected predicted power generation curve, and obtain the thermal power generation increase and the corresponding carbon emission increase based on the difference between the two. Step 4 includes the following: Step 401: Obtain the unstable power consumption and unstable power generation quantities in real time, and then correct the remaining predicted power consumption curve and the remaining predicted power generation curve using the unstable power consumption and unstable power generation quantities respectively, and obtain the corrected predicted power consumption curve and the corrected predicted power generation curve respectively, based on the following formula: in, To correct the predicted power consumption curve, For the remaining predicted power consumption curve, This indicates unstable electricity consumption. To correct the predicted power generation curve, For the remaining predicted power generation curve, This indicates the unstable amount of power generation. For time-based retrieval variables, , This represents the total duration of the day.

[0037] By correcting the prediction curve in real time, the prediction results for the remaining time of the next day are ensured to better reflect the actual operating conditions, providing more accurate guidance for subsequent dispatch decisions. Based on adjustments to unstable electricity consumption and generation, deviations in the prediction can be captured and compensated for in a timely manner, ensuring the prediction model maintains high applicability and accuracy in dynamic environments. This correction process is the core of dynamic optimization, significantly improving the reliability and real-time performance of the power supply plan, providing more accurate reference data for supplementary thermal power needs, and avoiding resource waste or power shortages caused by prediction errors. Furthermore, this process provides technical support for the system's adaptive capabilities, enabling the dispatch center to respond quickly to fluctuations in electricity consumption and generation. In both steps, it directly utilizes the unstable data calculated in step 3, and the output corrected predicted electricity consumption and generation curves lay a solid foundation for the thermal power generation increase calculation in step 402, demonstrating the seamless integration of deviation correction and prediction optimization.

[0038] Step 402: Obtain the thermal power generation increase based on the real-time acquired modified predicted power consumption curve and modified predicted power generation curve, using the following formula: in, To increase thermal power generation, To correct the predicted power consumption curve, To correct the predicted power generation curve, For time-based retrieval variables, , This is the total duration of today; The term "increased thermal power generation" specifically reflects the increase in thermal power generation from the current moment. By the end of the day During the remaining time period, additional electricity needs to be provided to compensate for the shortfall between the predicted power generation from wind and solar power plants and the predicted electricity consumption in the target area. This means quantifying the total additional power generation that thermal power plants must generate in the future to maintain a balance between power supply and demand. In a real-world environment, this corresponds to the uncertainty of new energy power generation and the dynamic fluctuations in electricity demand: because wind and solar energy are intermittent due to weather conditions, the predicted power generation may not fully meet the predicted electricity demand, and thermal power plants need to flexibly adjust their output according to this gap. Independent variables It is a time variable, ranging from arrive This indicates the remaining time period of the day, reflecting the scheduling time window; It is the revised predicted power consumption curve, representing the expected power demand of the target area at future points in time, driven by actual power consumption behaviors such as residential life and industrial production. This is the revised projected power generation curve, representing the expected power generation capacity of wind and solar power plants at future points in time, affected by weather forecasts and equipment performance. These variables collectively reflect the dynamic supply-demand balance challenges faced by power dispatch in a real-world environment: the volatility of new energy power generation and the real-time changes in electricity demand necessitate thermal power as a supplement to ensure grid stability.

[0039] With internal variables and It is closely related because it is obtained by integrating the difference between the two, directly reflecting the total gap between the predicted electricity demand and the predicted power generation capacity of wind and solar power in the future time period. Specifically, This represents the future power demand of the target area, while This indicates the power generation capacity of a wind and solar power plant at the same point in time. The integral of the difference between these two values ​​quantifies the electricity shortfall that thermal power plants need to fill. > This indicates that wind and solar power cannot meet the electricity demand. If the value is positive, thermal power generation needs to increase to make up the difference; conversely, if... ≤ If the difference is zero or negative, then only positive gaps are typically considered when calculating the increase in issuance. Independent variable Time range (from) arrive This determines the range of points that can be accumulated, and has an impact. The value: The longer the time period, the more supply and demand differences are covered by the points. The larger the potential, the more likely it is to be. This relationship indicates that... It not only depends on the supply and demand gap at a certain moment, but is also affected by the duration and magnitude of the supply and demand imbalance throughout the remaining time period, providing a basis for the dynamic nature of thermal power dispatch.

[0040] The correlation between power generation and carbon emissions from thermal power plants is obtained, and the carbon emissions from increased power generation from thermal power plants are calculated based on this correlation and denoted as the increase in carbon emissions. The formula used is as follows: in, For the increase in carbon emissions, To increase thermal power generation, This indicates the correlation between electricity generation and carbon emissions.

[0041] By calculating the increase in thermal power generation and carbon emission generation, the supplementary demand for thermal power in power supply and its specific environmental impact are clarified, providing a clear and quantitative basis for energy dispatch and carbon emission management. The difference between the revised predicted power consumption curve and the revised predicted power generation curve accurately reflects the power gap that wind and solar power cannot meet. Its practical significance lies in providing scientific guidance for the output planning of thermal power plants, ensuring the continuity and stability of power supply. Simultaneously, by calculating the increase in carbon emission generation, environmental constraints are integrated into dispatch decisions, achieving the dual goals of supply-demand balance and green development. It serves as a crucial bridge between supply-demand matching and carbon emission control, not only optimizing thermal power dispatch strategies but also providing data support for dynamic carbon emission monitoring, helping dispatch centers achieve synergistic improvement in economic and environmental benefits in complex operating environments. In the preceding and following steps, it inherits the correction results from step 401, and the output of increased thermal power generation and carbon emission generation provides key inputs for the total carbon emission assessment and management in step 5, demonstrating the logical integrity from prediction correction to practical application.

[0042] Step 5: Obtain the increase in carbon emissions and the actual carbon emissions that have already been generated in real time, form the total carbon emissions, obtain carbon emission indicators, and determine the gap between them and the total carbon emissions.

[0043] Step 5 includes the following: The system acquires in real-time the actual carbon emissions and carbon emission increases generated by thermal power plants up to the current moment, calculates their sum, and defines it as the total carbon emissions. The total carbon emissions are acquired in real-time and plotted as a curve. The system obtains the daily carbon emission target from the dispatch center, compares the target with the total carbon emissions, and identifies the target gap. The target gap is the total carbon emissions minus the target. A gap threshold is set. When the target gap exceeds the threshold, an alarm is triggered, reminding the dispatch center to purchase quotas. When the target gap is not greater than the threshold, no alarm is triggered.

[0044] By calculating total carbon emissions in real time and combining this with visualized curves and alarm mechanisms, the system significantly enhances its refined management and dynamic control capabilities, ensuring environmental compliance during operation. Real-time updates and curve plotting of total carbon emissions allow the dispatch center to intuitively grasp emission trends and quickly identify potential risks. The assessment of emission gaps and threshold alarm mechanisms provide a scientific basis and operational guidance for timely countermeasures (such as purchasing carbon emission allowances). This final step in carbon emission control is significant because dynamic monitoring and intelligent early warning help the dispatch center strictly control carbon emissions while ensuring stable power supply, avoiding economic costs and policy pressures caused by exceeding limits. It also provides data accumulation and practical experience for subsequent carbon management strategy optimization. Furthermore, this process demonstrates the deep responsiveness of the technology to carbon constraints, injecting new momentum into the sustainable development of the energy system. Throughout these steps, it fully utilizes the carbon emission increase data from step 4, completing a full closed loop from prediction to operation and environmental target management, showcasing a high degree of integration between technology and environmental goals.

[0045] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0046] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0047] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0048] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints, characterized in that, The specific steps include: Step 1: Obtain the historical power consumption of the target area, once per hour and add a timestamp, summarize into a power consumption dataset, and train it in a long short-term memory model to form a power consumption prediction model. Use the power consumption prediction model to obtain the predicted power consumption curve of the target area for the next day. Step 2: Obtain historical meteorological and power generation data from wind and solar power plants to form a meteorological dataset, and train it in a convolutional neural network model to form a power generation prediction model. Obtain the forecast meteorological data for the next day from the meteorological station and input it into the power generation prediction model to obtain the predicted power generation curve for the next day. Step 3: The next day, generate electricity and transmit it to the target area by combining wind, solar and thermal power. Obtain the actual power consumption curve and the actual power generation curve of wind and solar power in the target area. Compare the predicted power consumption curve and the actual power consumption curve to obtain the power consumption instability. Compare the predicted power generation curve and the actual power generation curve to obtain the power generation instability. Step 4: Correct the predicted power consumption curve and the predicted power generation curve by using the unstable power consumption and unstable power generation respectively, obtain the corrected predicted power consumption curve and the corrected predicted power generation curve, and obtain the thermal power generation increase and the corresponding carbon emission increase based on the difference between the two. Step 5: Obtain the increase in carbon emissions and the actual carbon emissions that have already been generated in real time, form the total carbon emissions, obtain carbon emission indicators, and determine the gap between them and the total carbon emissions.

2. The real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints according to claim 1, characterized in that: Historical power consumption data for the target area is obtained hourly, with timestamps added. Holiday data is marked as 1, and non-holiday data as 0, based on the holiday characteristics of the timestamps. The timestamps are compared with meteorological data from weather stations to obtain temperature characteristics, specifically the temperature of the target area at each timestamp. Power consumption is then divided into daily units, and hourly power consumption data is obtained for each day, forming a daily power consumption dataset. The formula used is as follows: in, This represents a dataset of electricity consumption data. Indicates the time of day. Indicates the characteristics of holidays, This represents the temperature characteristics at this specific time of day. This represents the power consumption at this specific time on this day.

3. The real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints according to claim 2, characterized in that: The electricity consumption data is sorted by date, and a sliding window time length is set. Subsequences are selected from the electricity consumption dataset during the sorting process. The subsequences are used as samples to input into the Long Short-Term Memory Network model. The electricity consumption power in the electricity consumption dataset of the first day after each sample is used as the label to obtain the electricity consumption prediction model. By acquiring historical power consumption data and corresponding timestamps within a sliding window period in the target area, a daily power consumption dataset is formed. This dataset is then input into the power consumption prediction model to obtain the power consumption at each time point on the second day. A fitting curve is generated according to the time series and labeled as the predicted power consumption curve.

4. The real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints according to claim 3, characterized in that: Historical meteorological and power generation data of wind and solar power plants are acquired. The historical meteorological data includes temperature, wind speed, and solar irradiance, along with corresponding timestamps. The historical power generation data includes power output and timestamps. The historical meteorological and power generation data are correlated based on timestamps to form a meteorological dataset, using the following formula: in, Represents the power generation dataset, For a point in time, This is the temperature corresponding to this time point. This represents the light intensity at this time point. This represents the wind speed at this point in time.

5. The real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints according to claim 4, characterized in that: The power generation dataset is input into a convolutional neural network model for training, and the power generation power corresponding to each meteorological dataset is used as a label to obtain a power generation prediction model. The forecast meteorological data for the next day is obtained from the meteorological station. The forecast meteorological data is matched with the timestamps of historical meteorological data to obtain the predicted power generation for the next day. The predicted power generation curve is then formed by fitting the time series.

6. The real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints according to claim 5, characterized in that: On the second day, the dispatch center transmitted electricity to the target area through a combination of wind and solar power plants and thermal power plants. Priority was given to using wind and solar power plants to transmit electricity to the target area, and any shortfall was made up by thermal power plants. During the process of wind and solar power plants supplying power to the target area, the dispatch center acquires the actual power consumption curve of the target area in real time, and simultaneously acquires the actual power generation curve of the wind and solar power plants in real time. Starting from the beginning of the day and ending at the current time, the predicted power consumption curve segment within this time period is compared with the actual power consumption curve to obtain the power instability. The formula used is as follows: in, This indicates unstable electricity consumption. The total duration up to the current time. This is the actual power consumption curve. To predict segments of the power consumption curve, For time-based retrieval variables, ; By comparing the predicted power generation curve segment with the actual power generation curve during this time period, the power generation instability is obtained, based on the following formula: in, This indicates the unstable amount of power generation. The total duration up to the current time. This is the actual power generation curve. To predict segments of the power generation curve, For time-based retrieval variables, .

7. The real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints according to claim 6, characterized in that: Real-time acquisition of unstable electricity consumption and generation data is used to correct the remaining predicted power consumption curve and the remaining predicted power generation curve, respectively. The corrected predicted power consumption curve and the corrected predicted power generation curve are then obtained, based on the following formulas: in, To correct the predicted power consumption curve, For the remaining predicted power consumption curve, This indicates unstable electricity consumption. To correct the predicted power generation curve, For the remaining predicted power generation curve, This indicates the unstable amount of power generation. For time-based retrieval variables, , This represents the total duration of the day.

8. The real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints according to claim 7, characterized in that: The increase in thermal power generation is obtained based on the real-time acquired modified predicted power consumption curve and modified predicted power generation curve, using the following formula: in, To increase thermal power generation, To correct the predicted power consumption curve, To correct the predicted power generation curve, For time-based retrieval variables, , This is the total duration of today; The correlation between power generation and carbon emissions from thermal power plants is obtained, and the carbon emissions from increased power generation from thermal power plants are calculated based on this correlation and denoted as the increase in carbon emissions. The formula used is as follows: in, For the increase in carbon emissions, To increase thermal power generation, This indicates the correlation between electricity generation and carbon emissions.

9. The real-time dispatching method for ultra-high voltage power transmission based on carbon emission constraints according to claim 8, characterized in that: The system acquires in real-time the actual carbon emissions and carbon emission increases generated by thermal power plants up to the current moment, calculates their sum, and defines it as the total carbon emissions. The total carbon emissions are acquired in real-time and plotted as a curve. The system obtains the daily carbon emission target from the dispatch center, compares the target with the total carbon emissions, and identifies the target gap. The target gap is the total carbon emissions minus the target. A gap threshold is set. When the target gap exceeds the threshold, an alarm is triggered, reminding the dispatch center to purchase quotas. When the target gap is not greater than the threshold, no alarm is triggered.

Images