A system and method for dynamically evaluating supply capacity under new energy load conditions

Abstract

The present application provides a system and method for dynamically evaluating the supply guarantee capability under the condition of new energy load, which relates to the field of smart grid technology. The power load prediction module predicts the power load and generates a load prediction sequence; the power generation demand analysis module analyzes the power generation demand based on the load prediction sequence and combines frequent power distribution losses to generate a preset power generation sequence; the prediction model establishment module establishes a hybrid power generation prediction model for the new energy power grid; the energy scheduling fitting module performs energy scheduling fitting based on the hybrid power generation prediction model in the preset evaluation time zone and takes the preset power generation sequence as the target, and generates an energy scheduling sequence; the supply guarantee capability evaluation module compares the energy scheduling sequence with the preset power generation sequence to generate a supply guarantee capability evaluation result. The present application solves the technical problem that the traditional method lacks dynamic analysis and accurate prediction of real-time data, resulting in inaccurate power supply capability evaluation, and improves the power supply stability and quality of the power grid.

Description

Technical Field

[0001] The present application relates to the field of smart grid technology, and specifically to a system and method for dynamically evaluating power supply capability under new energy load conditions. Background Art

[0002] With the continuous growth of global energy demand and the enhancement of environmental protection awareness, renewable energy power generation has gradually become an important part of power supply. New energy such as solar energy and wind energy have attracted widespread attention and application due to their renewable and environmentally friendly characteristics. However, renewable energy power generation has obvious volatility and uncertainty. For example, solar power generation is significantly affected by weather conditions, and wind power generation depends on changes in wind speed and direction. These unstable factors have brought great challenges to the load forecasting and power generation scheduling of the power grid.

[0003] Traditional methods of power dispatching and power supply capacity assessment mainly rely on historical data and fixed models, lacking the ability to dynamically analyze and predict real-time data. This method often fails to adjust the power supply strategy in a timely manner when faced with the volatility of renewable energy generation, resulting in the power system being unable to guarantee stable power supply during peak demand periods or wasting power during low demand periods. In addition, traditional methods do not adequately consider the loss of power during transmission, further affecting the efficiency and reliability of the power supply system. Summary of the invention

[0004] The present application provides a system and method for dynamically evaluating power supply capacity under renewable energy load conditions, which solves the technical problems that traditional power dispatching and power supply capacity evaluation methods lack the ability to dynamically analyze and predict real-time data, and are difficult to adapt to the volatility and uncertainty of renewable energy power generation, thus leading to delayed power dispatching decisions and inaccurate power supply capacity evaluation, and achieves the technical effect of improving the power supply stability and power supply quality of renewable energy power systems under various load conditions.

[0005] In view of the above problems, on the one hand, the present application provides a system for dynamically evaluating power supply capacity under new energy load conditions, the system comprising: an electricity load prediction module, the electricity load prediction module is used to predict the electricity load in the power supply area of ​​the new energy power grid in a preset evaluation time zone, and generate a load prediction time series; a power generation demand analysis module, the power generation demand analysis module is used to perform power generation demand analysis based on the load prediction time series and the frequent power distribution losses of the new energy power grid, and generate a preset power generation time series; a prediction model establishment module, the prediction model establishment module is used to establish a hybrid power generation prediction model of the new energy power grid; an energy scheduling fitting module, the energy scheduling fitting module is used to perform energy scheduling fitting based on the hybrid power generation prediction model in the preset evaluation time zone, with the preset power generation time series as the target, to generate an energy scheduling time series with the minimum fitting deviation index; a power supply capacity evaluation module, the power supply capacity evaluation module is used to compare the energy scheduling time series with the preset power generation time series, and generate a power supply capacity evaluation result.

[0006] On the other hand, the present application also provides a method for dynamically evaluating power supply capability under new energy load conditions, the method comprising: in a preset evaluation time zone, predicting the power load in the power supply area of ​​the new energy power grid to generate a load prediction time series; based on the load prediction time series, combined with the frequent power distribution losses of the new energy power grid, analyzing the power generation demand to generate a preset power generation time series; establishing a hybrid power generation prediction model for the new energy power grid; in the preset evaluation time zone, taking the preset power generation time series as the target, performing energy scheduling fitting based on the hybrid power generation prediction model to generate an energy scheduling time series with the smallest fitting deviation index; comparing the energy scheduling time series with the preset power generation time series to generate a power supply capability evaluation result.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The power load forecasting module predicts the power load of the power supply area of ​​the new energy power grid in the preset evaluation time zone and generates a load forecast time series. This module dynamically adjusts the forecast by analyzing historical data and real-time data to accurately reflect future load demand. The power generation demand analysis module analyzes the power generation demand based on the load forecast time series and the frequent power distribution losses of the new energy power grid to generate a preset power generation time series. This module comprehensively considers the load forecast and the actual power grid loss, accurately calculates the power generation that meets the power demand in different time periods, and ensures that the power supply can meet the actual demand. The prediction model establishment module establishes a hybrid power generation prediction model for the new energy power grid. This module integrates the power generation characteristics of multiple energy sources to establish a comprehensive power generation prediction model, improves the accuracy and reliability of the prediction of new energy power generation, and provides more accurate prediction data for energy scheduling. The energy scheduling fitting module performs energy scheduling fitting based on the hybrid power generation prediction model in the preset evaluation time zone with the preset power generation time series as the target, and generates an energy scheduling time series with the smallest fitting deviation index. This module uses an optimization algorithm to adjust the energy dispatch strategy in real time, so that the actual power generation is closest to the demand forecast, reducing the gap between supply and demand, ensuring that power generation resources can be effectively dispatched under different load conditions, and improving energy utilization efficiency. The supply guarantee capacity evaluation module compares the energy dispatch timing and the preset power generation timing to generate a supply guarantee capacity evaluation result. This module can dynamically evaluate the power supply capacity of the power grid, identify the deficiencies and optimization space of the dispatch strategy, and provide a decision-making basis for improving the stability of the power grid and the reliability of power supply.

[0009] To sum up, through the collaborative work of the above modules, this application realizes the dynamic evaluation and optimal scheduling of the supply guarantee capacity under the new energy load, effectively responds to the volatility and uncertainty of new energy power generation, and greatly improves the power supply reliability and quality of the new energy power system, ensuring that the power grid can supply stable power under various load conditions, improves the energy utilization rate of new energy power generation, and ensures the stable operation of the power grid.

[0010] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A schematic diagram of the structure of a system for dynamically evaluating supply capacity under new energy load conditions provided in an embodiment of the present application;

[0012] Figure 2 A schematic diagram of a flow chart of generating a load forecast time series in a system for dynamically evaluating supply capacity under new energy load conditions provided in an embodiment of the present application;

[0013] Figure 3 A schematic diagram of a flow chart of generating an energy dispatching sequence in a dynamic evaluation system for supply guarantee capability under new energy load conditions provided in an embodiment of the present application;

[0014] Figure 4 A flowchart of a method for dynamically evaluating power supply capability under renewable energy load conditions provided in an embodiment of the present application.

[0015] Explanation of the reference numerals: power load prediction module 10 , power generation demand analysis module 20 , prediction model establishment module 30 , energy scheduling fitting module 40 , power supply capability assessment module 50 . DETAILED DESCRIPTION

[0016] The embodiments of the present application provide a system and method for dynamically evaluating power supply capabilities under renewable energy load conditions, thereby solving the technical problems that traditional power dispatching and power supply capacity evaluation methods lack the ability to dynamically analyze and predict real-time data, and are difficult to adapt to the volatility and uncertainty of renewable energy power generation, thereby leading to delayed power dispatching decisions and inaccurate power supply capacity evaluation, thereby achieving the technical effect of improving the power supply stability and power supply quality of renewable energy power systems under various load conditions.

[0017] Embodiment 1, as Figure 1 As shown, an embodiment of the present application provides a system for dynamically evaluating supply capacity under new energy load conditions, the system comprising:

[0018] The power load prediction module 10 is used to predict the power load of the power supply area of ​​the new energy power grid in a preset evaluation time zone and generate a load prediction time sequence.

[0019] Specifically, the preset evaluation time zone refers to a pre-set time range for load forecasting and evaluation, which can be the next 24 hours, a week or a month. The power supply area refers to the geographical area covered by the new energy power grid, including cities, villages, industrial areas, etc. The new energy power grid is an electric power system powered by renewable energy such as wind power generation and photovoltaic power generation. Its power generation is affected by natural conditions such as weather and light, and is volatile and uncertain. The load forecast time series refers to a sequence of power load forecast data arranged in chronological order, which is used to represent the predicted power demand at each time point in the preset evaluation time zone.

[0020] First, collect historical and real-time electricity consumption data. Historical data can include electricity consumption records over the past few months or years, while real-time data includes current electricity usage and environmental parameters such as temperature, humidity, etc. For example, use the power company's database to obtain electricity consumption data for the past year, and use sensors and smart meters to obtain real-time data.

[0021] Next, use data analysis and machine learning tools to preprocess and analyze the collected data. Then use the preprocessed data to train a model to predict the electricity load in the preset evaluation time zone. Common tools used in this process include Python's Pandas and NumPy libraries for data processing, and the Scikit-Learn library for machine learning model training. Time series models can be used, including ARIMA, SARIMA, LSTM, GRU, etc. For example, historical data can be organized into a time series data frame through the Pandas library, and then the linear regression model in the Scikit-Learn library can be used to train and predict the data. The electricity consumption data for the past year can be divided into a training set and a test set to train and verify the model. The prediction accuracy can be optimized by adjusting the model parameters.

[0022] Finally, the trained model is used to predict the power load in the preset evaluation time zone to generate a load forecast time series. The generated load forecast time series is a sequence showing the future changes in power demand. For example, to predict the power demand per hour in the next 24 hours, the model will output a load forecast time series containing 24 data points, each data point representing the power demand forecast for one hour within 24 hours.

[0023] Through the above process, the module can dynamically predict the power load in the future and provide basic data for subsequent power generation demand analysis.

[0024] The power generation demand analysis module 20 is used to perform power generation demand analysis based on the load forecast time sequence and in combination with the frequent power distribution losses of the new energy power grid to generate a preset power generation time sequence.

[0025] Specifically, power distribution loss refers to power loss caused by line loss, transformer loss, etc. during power transmission and distribution. These losses are frequent and unavoidable and need to be considered in the power generation demand analysis. The power generation demand value sequence generated by the power generation demand analysis module 20 for each hour or shorter time interval in the future period of time when the power generation time sequence is preset reflects the power generation required to meet the load forecast demand under the condition of considering frequent power distribution losses.

[0026] The power generation demand analysis module 20 obtains the load forecast time series from the power load forecast module 10 and collects the frequent power distribution losses of the new energy power grid. Power distribution losses are losses caused by various factors in the power system during the power transmission and distribution process, including but not limited to heat losses caused by resistance and other factors on the transmission line, i.e. line losses; copper losses and iron losses generated when power is converted in the transformer, i.e. transformer losses; losses caused by aging equipment, poor lines, etc. in the distribution system, i.e. distribution losses. Usually, power companies provide the average power distribution loss rate of the system.

[0027] Then, the module calculates the power generation demand per hour or shorter time intervals in the future based on the load forecast time series and frequent power distribution losses. This process can use data analysis tools and algorithms, such as Python's Pandas library for data processing and calculation to calculate the total amount of power actually required at each time point. For example, if the load forecast for a certain hour is 500MW, considering the 10% power distribution loss, the required power generation is 550MW / (1-10%)=555.56MW.

[0028] Finally, the calculated power generation demand values ​​are sorted in time series to form a preset power generation time series, that is, a sequence of power generation demand values ​​every hour or shorter time intervals in the future, providing a basis for power system scheduling and power generation planning.

[0029] Through the above process, the power generation demand analysis module 20 effectively combines load forecasting and power distribution loss, performs detailed power generation demand analysis, generates preset power generation timing, and ensures stable operation of the power system and efficient resource allocation.

[0030] The prediction model building module 30 is used to build a hybrid power generation prediction model for the new energy power grid.

[0031] Specifically, the hybrid power generation forecast model is a comprehensive model that combines multiple power generation sources to predict the total power generation of the new energy grid. It provides more accurate forecast results by comprehensively considering the power generation characteristics and volatility of different energy sources.

[0032] First, collect data from multiple sources of power generation. This data includes historical power generation data, real-time power generation data, and environmental parameters that affect power generation. For example, data for solar power generation may include daily power generation in the past few years, current solar radiation intensity, and meteorological data; data for wind power generation may include wind speed, wind direction, and past power generation records. Using data acquisition systems and sensors, this data can be obtained in real time.

[0033] Next, preprocess and analyze the collected data. Use Python's Pandas and NumPy libraries to clean, normalize and extract features from the data. Select a suitable machine learning algorithm to establish a hybrid power generation prediction model. Common algorithms include regression analysis, time series analysis and neural network models. Then, train and verify the prediction model. Divide the preprocessed data into a training set and a test set, train the model using the training set, and verify the accuracy of the model through cross-validation and the test set. Improve the prediction accuracy of the model by adjusting model parameters and optimizing algorithms. Finally, deploy the trained hybrid power generation prediction model to predict the power generation of the new energy grid in real time. For example, when predicting the power generation in the next 24 hours, the model can combine real-time solar radiation and wind speed data to output the predicted power generation per hour.

[0034] Through this process, the prediction model building module 30 effectively improves the prediction accuracy and reliability of the power generation of the new energy power grid.

[0035] The energy scheduling fitting module 40 is used to perform energy scheduling fitting based on the hybrid power generation prediction model in the preset evaluation time zone and with the preset power generation timing as the target, to generate an energy scheduling timing with the minimum fitting deviation index.

[0036] Specifically, the module obtains the preset power generation time series generated by the power generation demand analysis module 20. Then, the hybrid power generation prediction model is used to predict the power generation capacity of each power generation source in the preset evaluation time zone. Assuming that the new energy grid includes solar and wind power generation, the hybrid power generation prediction model will predict the solar and wind power generation per hour in the next 24 hours based on historical data and real-time environmental parameters such as solar radiation, wind speed, etc. Then, the energy scheduling fitting module 40 optimizes the scheduling based on these predicted data. Using optimization algorithms such as linear programming, genetic algorithms or particle swarm optimization algorithms, the module fits the predicted power generation of each power generation source with the preset power generation time series to generate the optimal energy scheduling plan. The optimization goal is to minimize the fitting deviation index, which is an indicator that measures the difference between the energy scheduling plan and the actual demand. The smaller the deviation, the better the scheduling plan. The module finally generates an energy scheduling time series with the smallest fitting deviation index. This energy scheduling time series is a sequence of electricity that should be generated by different energy sources in different time periods. The energy scheduling time series with the smallest fitting deviation index means that the deviation between the actual scheduling result and the preset power generation time series is the smallest, that is, the energy scheduling fitting effect is the best.

[0037] Through this process, the energy scheduling fitting module 40 ensures that the power grid can operate efficiently and stably under various load conditions, minimizes the difference between actual power generation and target power generation, and thus improves the stability and reliability of power supply.

[0038] The power supply capability assessment module 50 is used to compare the energy scheduling sequence with the preset power generation sequence to generate a power supply capability assessment result.

[0039] Specifically, the supply guarantee capacity assessment result is the result of assessing whether the power grid can meet the power demand within the preset assessment time zone. This result is used to determine whether the power grid has sufficient power supply capacity and identify potential power supply shortage risks.

[0040] The supply guarantee capability assessment module 50 obtains the energy scheduling timing and the preset power generation timing generated by the energy scheduling fitting module 40 and the power generation demand analysis module 20 respectively, and then compares the two timing data hour by hour. The purpose of the comparison is to evaluate whether the actual power generation arrangement at each time point meets or exceeds the preset power generation demand. Then, the module calculates the power supply deviation at each time point based on the comparison results, and counts the power supply situation in the entire evaluation time zone. The power supply deviation refers to the difference between the actual power generation and the preset power generation, which can be evaluated using indicators such as deviation rate, absolute deviation, and mean square error. Finally, the supply guarantee capability assessment module 50 generates a supply guarantee capability assessment report based on the statistical results. The report includes a time series chart of the power supply deviation, total power supply deviation statistics, and identified potential power supply risk time periods.

[0041] Through the above process, the power supply capacity assessment module 50 effectively identifies and assesses the power supply capacity of the power grid in different time periods, provides detailed assessment results and improvement suggestions, and ensures that the power grid can operate stably under various new energy load conditions. In addition, the module can also be connected to an early warning device. When there is a power supply risk or abnormal situation, an early warning signal is issued through the early warning device. For example, when the power supply deviation exceeds the deviation threshold predefined in the system, an early warning is issued to relevant personnel through sound and light signals or message push.

[0042] Further, such as Figure 2 As shown, the power load prediction module 10 in the embodiment of the present application is also used to perform the following steps:

[0043] Collect historical electricity consumption records of the power supply area, and build a load forecasting model based on the historical electricity consumption records; perform network heat identification on the power supply area in the preset evaluation time zone, and generate a network heat mapping result, wherein the network heat mapping result includes multiple groups of network heat information types and heat indicators with mapping relationships; perform feedback impact analysis on the network heat mapping through the power load feedback network, and generate a feedback factor; perform load forecasting in the preset evaluation time zone through the load forecasting model, and perform feedback adjustment in combination with the feedback factor to generate the load forecasting time series.

[0044] Specifically, historical electricity consumption records refer to electricity consumption data recorded in the power system over a period of time in the past. These data usually include hourly or daily electricity consumption and are used to analyze and predict future electricity demand. The load forecasting model is a mathematical model or algorithm used to predict future electricity demand based on historical electricity consumption data and other relevant factors. Network heat identification is to identify the activity heat of a specific area in a certain period of time by analyzing Internet data such as search trends and social media activity. The network heat mapping result is generated by network heat identification and analysis, and contains multiple groups of network heat information types and heat index results with mapping relationships, which are used to reflect the potential impact of factors such as social activities and weather changes on electricity load. The power load feedback network is a system component used to analyze the impact of network heat mapping results on power load, and adjust the load forecast through a feedback mechanism. The feedback factor refers to a parameter derived from the feedback network that indicates the degree of impact of network heat information on power load.

[0045] First, collect historical electricity consumption records in the power supply area. These data are usually provided by the power company's database and cover electricity consumption in the past few months or years. The collected historical electricity consumption records are preprocessed to ensure the quality and applicability of the data. The preprocessing steps include: data cleaning, that is, removing outliers, filling or deleting missing values ​​to ensure the integrity of the data; feature engineering, that is, extracting or constructing features from the original data that are helpful for model prediction, including periodic features, such as daily, weekly, and monthly periodic electricity consumption patterns, and trend features, such as the long-term growth trend of electricity consumption; and data standardization or normalization. Based on historical electricity consumption records, use machine learning algorithms such as linear regression, decision trees, or time series models to train historical data and build load forecasting models. These models can capture the periodic and trend characteristics of electricity consumption patterns and make preliminary predictions of future electricity loads.

[0046] For example, a load forecasting system based on the LSTM model is constructed to predict the power load of a city in the next week. The long short-term memory network (LSTM) is a deep learning model that is particularly suitable for processing time series data. It can capture long-term dependencies in data and is very suitable for power load forecasting. The specific construction process is as follows: define the structure of the LSTM model in a deep learning framework such as TensorFlow, Keras, etc., including input layer, LSTM layer, output layer, etc. The LSTM layer can be set to multiple layers to enhance the expressiveness of the model. Collect hourly power consumption data for the past year from the power supply area, as well as weather data related to power load, including temperature, humidity, wind speed, etc. Preprocess the collected data, clean the data, fill in missing values, and perform feature engineering to extract trend features in weather data. Divide the preprocessed data into a training set and a test set. For example, the data of the past year is used as a training set, and the data of the most recent month is used as a test set. For the LSTM model, it is usually necessary to convert the data into a format suitable for time series prediction, such as the sliding window method, to convert continuous time series data into a supervised learning problem. Use the training set data to train the LSTM model, and optimize the model parameters through the back propagation algorithm to minimize the prediction error. This process may require adjusting hyperparameters such as learning rate, batch size, and number of training rounds to obtain the best model performance. Verify the generalization ability of the model on the test set data and evaluate the accuracy of the model prediction. Common evaluation indicators include mean square error, root mean square error, mean absolute error, etc. Use the trained LSTM model to predict the power load in the power supply area, input auxiliary information at future time points, and predict the corresponding power load.

[0047] Next, the network heat of the power supply area is identified within the preset evaluation time zone. By analyzing multiple data sources such as social media, news reports, and weather forecasts, network heat information related to power load is identified, such as the holding of large-scale events, warnings of extreme weather, and approaching holidays. This information is mapped into multiple groups of network heat information types and heat indicators with mapping relationships to form network heat mapping results, providing an additional dimension for power load forecasting. For example, during holidays in a certain area, social media activity and search engine queries increase significantly, and these data can reflect the activity heat in the area.

[0048] Then, the feedback impact analysis of the network heat mapping results is performed through the power load feedback network. This network builds a dynamic and adaptive feedback mechanism by analyzing the correlation between network heat and power load in historical data. For example, a neural network model is built using TensorFlow or Keras libraries to analyze the relationship between network heat data and historical power consumption data. When the network heat mapping results show major events or changes that may occur in the future, the feedback network will generate a feedback factor to adjust the original prediction value of the load forecasting model. This feedback factor reflects the expected impact of network heat on power load, which can help the model more accurately capture power load fluctuations caused by social activities, weather changes and other factors, thereby generating more realistic forecast results.

[0049] Finally, the load forecasting model is used to forecast the load in the preset evaluation time zone, and the feedback factor is combined to make feedback adjustments to generate the load forecast time series. For example, if the network heat feedback factor of a certain area indicates that the electricity demand increases by 10% during holidays, this feedback factor is applied to the load forecast results in the load forecasting model to generate a more accurate forecast time series.

[0050] The power load prediction module 10 realizes accurate prediction of power demand in the power supply area through deep learning models, network heat identification and feedback adjustment mechanism, providing accurate and reliable data support for subsequent analysis and prediction.

[0051] Preferably, the power load prediction module 10 of the embodiment of the present application is also used to perform the following steps:

[0052] The heat-people flow conversion database is called through the electricity load feedback network to identify heat-people flow conversion data samples whose similarity with the network heat mapping result is greater than a preset similarity, and generate traffic surge data; based on the traffic surge data, the traffic-load influence curve is called to identify the load influence coefficient and generate the feedback factor.

[0053] Optionally, the power load prediction module 10 in the embodiment of the present application is further configured to perform the following steps:

[0054] The heat-people flow conversion database includes multiple groups of heat-people flow conversion data samples, wherein any group of heat-people flow conversion data samples includes heat information type samples, heat index samples and traffic surge samples; traverse the heat-people flow conversion database, perform similarity comparison on any group of network heat information types and heat indexes in the network heat mapping results, and generate multiple groups of matching traffic surge samples; perform weighted fusion on the multiple groups of matching traffic surge samples to generate the traffic surge data.

[0055] Specifically, the heat-traffic conversion database is a database that stores multiple sets of heat-traffic conversion data samples, each of which includes a heat information type sample, a heat index sample, and a traffic surge sample, which is used to quantify the relationship between network heat information and traffic changes. Heat information type samples refer to the types of network activities, such as social media activity, search engine query volume, etc. Heat index samples are specific values ​​that describe the heat information type, such as the specific value of social media activity or the specific value of search engine query volume. Traffic surge samples refer to traffic change data corresponding to specific heat information types and heat indicators.

[0056] Traffic surge data refers to data that reflects the impact of a specific event on changes in passenger flow, and is used to predict sudden increases in power load. Similarity is a metric used to measure the degree of similarity between two data sets. The preset similarity is a threshold set by the system, which is used to filter out historical data samples that are similar to the current network heat mapping results. The traffic-load impact curve is a curve model that quantifies the relationship between traffic surges and power load changes. By analyzing historical data, it identifies the degree of impact of different traffic surge levels on the power load and is used to generate a load impact coefficient. The load impact coefficient is a parameter extracted from the traffic-load impact curve, which is used to describe the degree of impact of changes in passenger flow on the power load.

[0057] The module calls the heat-traffic conversion database through the power load feedback network. The database contains historical network heat and corresponding traffic data, such as the relationship between social media activity and shopping mall traffic during certain holidays. Data can be easily extracted from this database using SQL database queries or Python's Pandas library.

[0058] Then, traverse the heat-people flow conversion database, perform a similarity comparison on any group of network heat information types and heat indicators in the network heat mapping results, and identify heat-people flow conversion data samples whose similarity to the current network heat mapping results is greater than the preset similarity by calculating the similarity, and generate multiple groups of matching traffic surge samples. This can use algorithms such as cosine similarity or Euclidean distance to calculate data similarity. The matching traffic surge samples are traffic surge data samples that are selected from the heat-people flow conversion database through similarity comparison and match the current network heat mapping results. Through weighted fusion algorithms, such as weighted average, weighted sum, etc., multiple groups of matching traffic surge samples are weighted fused to generate traffic surge data that reflects the impact of specific events on changes in traffic.

[0059] Next, based on the identified traffic surge data, the traffic-load impact curve is called to identify the load impact coefficient. The traffic-load impact curve describes the impact of changes in pedestrian flow on the power load. For example, high pedestrian flow in a shopping mall will increase the power consumption of air conditioning and lighting. By analyzing the traffic surge data, a load impact coefficient is generated that reflects the degree of impact of a specific traffic surge level on the power load. Regression analysis or neural network models, such as using the TensorFlow or Keras library, can be used to establish a mapping relationship model between pedestrian flow and power load.

[0060] Finally, based on the load impact coefficient generated by the flow-load impact curve, the power load forecasting module 10 generates a feedback factor for dynamically adjusting the load forecast results to reflect the expected impact of network heat information on the power load. The load impact coefficient is applied to the load forecasting model as a feedback factor to adjust the forecast results and generate a more accurate load forecast time series. For example, assuming that the traffic surge data shows that the flow of people in a shopping mall increases by 50% during holidays, according to the flow-load impact curve, it is calculated that the increase in flow will cause the power load to increase by 20%. The 20% load impact coefficient is the feedback factor used to adjust the load forecasting model.

[0061] Through the above process, the module combines historical data and current network heat to generate feedback factors, further improving the prediction accuracy of the load forecasting model and generating more accurate power load forecasting results.

[0062] Furthermore, the prediction model building module 30 of the embodiment of the present application is also used to perform the following steps:

[0063] Acquire multiple power generation sources of the new energy power grid and establish multiple sets of unstable influencing factors of the multiple power generation sources; collect historical power generation data based on the multiple sets of unstable influencing factors, construct multiple power generation prediction branches, and generate the hybrid power generation prediction model.

[0064] Specifically, the multiple power generation sources of the new energy grid refer to the multiple renewable energy sources used in the grid, such as solar energy, wind energy, geothermal energy, hydropower, etc. These power generation sources are volatile and uncertain. The set of unstable influencing factors refers to the set of unstable factors that affect the power generation of new energy, such as weather conditions, seasonal changes, geographical location, etc. The power generation prediction branch is an independent prediction model branch constructed based on different power generation sources and corresponding unstable influencing factors, and each branch predicts for one power generation source.

[0065] First, classify the various power sources in the new energy grid, identify the instability influencing factors of each power source, and obtain the instability influencing factor set corresponding to each power source. For example, photovoltaic power generation is affected by light intensity and cloud cover, and wind power generation is affected by wind speed and temperature. This process can use meteorological knowledge and data analysis techniques, such as cluster analysis and principal component analysis (PCA), to quantify and normalize the influencing factors.

[0066] Next, collect historical power generation data for each power source under different non-stable influencing factors over a period of time, including power generation records at hourly or shorter time intervals, and corresponding non-stable influencing factor data. For example, collect solar power generation and corresponding weather data for a certain place over the past three years.

[0067] Then, multiple power generation prediction branches are constructed. According to different power generation sources and unstable influencing factors, independent prediction model branches are constructed for each power generation source, and each branch model separately predicts the future power generation of its corresponding power generation source. Finally, the various power generation prediction branches are integrated to generate a hybrid power generation prediction model. The prediction results of each branch model are combined to generate a total power generation forecast for the entire new energy grid. For example, the prediction results of each prediction model branch are combined using methods such as weighted average and ensemble learning to form the final power generation prediction model.

[0068] Through this process, the prediction model building module 30 performs power generation prediction according to the characteristics of different power generation sources, and effectively integrates the prediction results of multiple power generation sources to adapt to various new energy load conditions, thereby improving the prediction accuracy and reliability of the new energy power grid.

[0069] Further, such as Figure 3 As shown, the energy scheduling fitting module 40 in the embodiment of the present application is also used to perform the following steps:

[0070] In the preset evaluation time zone, real-time data collection is performed on the multiple sets of unstable influencing factors, and the data is input into the multiple power generation prediction branches to generate multiple first predicted power generation time series; the multiple first predicted power generation time series are aligned and added to generate a fused power generation time series; the energy storage module and the fused power generation time series are combined to perform energy scheduling with the preset power generation time series as the target, and a fitting loss minimization analysis is performed through an iterative fitting loss function to generate the energy scheduling time series.

[0071] Preferably, the energy scheduling fitting module 40 of the embodiment of the present application is also used to perform the following steps:

[0072] Acquire the energy storage capacity of the energy storage module; perform timing node mapping on the fused power generation timing and the preset power generation timing; when the fused power generation of the first timing mapping node is greater than the preset power generation, perform electric energy storage fitting in combination with the energy storage capacity to generate the first node stored electric energy; acquire the fused power generation of the second timing mapping node, and add it with the first node stored electric energy; compare the sum with the preset power generation of the second timing mapping node, perform electric energy storage fitting again, and so on, until all timing mapping nodes are traversed to generate the energy scheduling timing.

[0073] Specifically, first, in a preset evaluation time zone, real-time data collection is performed on a set of multiple non-stable influencing factors. These influencing factors include weather conditions, temperature, humidity, wind speed, etc., which are acquired in real time through sensors and data acquisition systems. For example, the data from the meteorological station and the IoT sensor are used to obtain the current temperature, humidity, wind speed and other data. Then, these real-time data are input into multiple power generation prediction branches. Each power generation prediction branch makes predictions for a power generation source. For example, the solar power generation prediction branch uses real-time solar radiation intensity and temperature data, and the wind power generation prediction branch uses real-time wind speed and wind direction data to generate multiple first predicted power generation time series. Among them, the first predicted power generation time series is a sequence of predicted power generation data arranged in chronological order generated by each power generation prediction branch.

[0074] Then, align and sum the multiple first predicted power generation time series. Use data processing tools to time align the predicted data of different power generation sources, and add them up to generate a fused power generation time series. Among them, the fused power generation time series refers to aligning and summing the predicted power generation time series of multiple power generation sources to form a unified power generation time series, reflecting the power generation capacity of the entire new energy grid. For example, the solar power generation prediction time series and the wind power generation prediction time series are the power generation data for each hour. By aligning and summing, the comprehensive power generation forecast for each hour is obtained.

[0075] Next, energy scheduling is performed based on the preset power generation time series in combination with the energy storage module and the fused power generation time series. The energy storage module refers to the facilities used to store electric energy in the new energy grid, such as battery energy storage, pumped storage, etc., which are used to store electric energy when the power generation exceeds the demand and release electric energy when the power generation is insufficient to balance the supply and demand. The specific process includes: obtaining the energy storage capacity of the energy storage module, that is, the maximum storage capacity of the energy storage system. The fused power generation time series and the preset power generation time series are mapped to time series nodes, that is, the fused power generation time series and the preset power generation time series are matched one by one to each time point. The purpose of this step is to match the actual power generation at each time point with the target power generation. Starting from the first time series mapping node, when the fused power generation of the first time series mapping node is greater than the preset power generation, the excess electric energy is combined with the energy storage capacity for electric energy storage fitting to generate the first node storage electric energy, that is, according to the difference between the actual power generation and the demand, the electric energy storage and release strategy is adjusted to achieve the effective use of electric energy. The first time series mapping node refers to the first time node in the time series mapping. The first node stored electric energy refers to the electric energy stored in the energy storage module at the first time-series mapping node with excess power generation. For example, assuming that the fused power generation of the first time-series mapping node is 600MW, the preset power generation is 500MW, and the excess 100MW of electric energy is stored in the energy storage module. Next, obtain the fused power generation of the second time-series mapping node and add it with the first node stored electric energy. Compare the summed result with the preset power generation of the second time-series mapping node to determine whether it is necessary to perform electric energy storage fitting or release again. For example, the fused power generation of the second time-series mapping node is 450MW, the preset power generation is 500MW, the first node stored electric energy is 100MW, and the summed result is 550MW. At this time, the preset power generation can be met and the remaining 50MW can be stored. By analogy, traverse all time-series mapping nodes, gradually adjust the energy storage and power generation strategies, and generate the final energy scheduling time series. The entire process can use Python's Pandas library for time series data processing, and use the NumPy library for mathematical operations and storage fitting calculations.

[0076] In the energy dispatch process, the fitting loss minimization analysis is performed by iteratively fitting the loss function to generate the energy dispatch instruction sequence. The fitting loss function is used to measure the difference between the actual power generation and the preset power generation. The scheduling strategy is iteratively adjusted through optimization algorithms such as gradient descent or genetic algorithms to minimize this difference. The optimization algorithm can be implemented using Python's SciPy library or TensorFlow.

[0077] For example, assume that a city's new energy grid needs to perform energy scheduling for the next 24 hours. The energy storage capacity of the energy storage module is 1000MWh. Get the fused power generation time series and the preset power generation time series. Assume that the fused power generation of the first time series mapping node is 600MW, the preset power generation is 500MW, and the excess 100MW is stored in the energy storage module. Get the fused power generation of the second time series mapping node as 450MW, and the preset power generation is 500MW. The first node stores 100MW of electrical energy, and the sum is 550MW. After meeting the preset power generation demand, the remaining 50MW continues to be stored. The fused power generation of the third time series mapping node is 700MW, and the preset power generation is 600MW. The current energy storage module stores 150MW of electrical energy, and the excess 100MW is stored, and the total stored electrical energy becomes 250MW. The fused power generation of the fourth time series mapping node is 300MW, and the preset power generation is 500MW. 200MW needs to be released from the energy storage module, and the total stored energy becomes 50MW. Through this node-by-node energy storage fitting and release, the energy dispatch fitting module 40 generates an energy dispatch time sequence within 24 hours to ensure stable operation of the power grid under various load conditions.

[0078] Through the above process, the energy scheduling fitting module 40 can achieve accurate matching of power generation and demand, ensuring the stable operation of the power grid while improving energy utilization efficiency.

[0079] In summary, the embodiment of the present application provides a system for dynamically evaluating supply capacity under new energy load conditions, which has the following technical effects:

[0080] The power load forecasting module 10, in the preset evaluation time zone, combines historical data and network heat identification, builds and adjusts the load forecasting model, forecasts the power load in the power supply area of ​​the new energy power grid, and generates a load forecasting time series. This module dynamically adjusts the forecast by analyzing historical data and real-time data to accurately reflect future load demand. The power generation demand analysis module 20, based on the load forecasting time series, combines the frequent power distribution losses of the new energy power grid to analyze the power generation demand and generate a preset power generation time series. This module comprehensively considers the load forecast and the actual power grid loss, accurately calculates the power generation that meets the power demand in different time periods, and ensures that the power supply can meet the actual demand. The prediction model establishment module 30, according to different power generation sources and their unstable influencing factors, constructs multiple power generation prediction branches, and summarizes and establishes a hybrid power generation prediction model for the new energy power grid. This module integrates the power generation characteristics of multiple energy sources to establish a comprehensive power generation prediction model, improves the accuracy and reliability of the prediction of new energy power generation, and provides more accurate prediction data for energy scheduling. The energy scheduling fitting module 40 performs energy scheduling fitting based on the hybrid power generation prediction model in the preset evaluation time zone and takes the preset power generation timing as the target, and generates an energy scheduling timing with the minimum fitting deviation index. This module adjusts the energy scheduling strategy in real time through the optimization algorithm, so that the actual power generation is closest to the demand forecast, reduces the supply and demand gap, ensures that the power generation resources can be effectively scheduled under different load conditions, and improves energy utilization efficiency. The supply guarantee capacity evaluation module 50 compares the energy scheduling timing with the preset power generation timing to generate a supply guarantee capacity evaluation result. This module can dynamically evaluate the power supply capacity of the power grid, identify the deficiencies and optimization space of the scheduling strategy, and provide a decision-making basis for improving the stability of the power grid and the reliability of power supply.

[0081] Overall, the embodiments of the present application realize dynamic evaluation and optimal scheduling of the power supply capacity under new energy loads through the coordinated work of the above modules, effectively cope with the volatility and uncertainty of new energy power generation, greatly improve the power supply reliability and quality of the new energy power system, ensure that the power grid can supply stable power under various load conditions, improve the energy utilization rate of new energy power generation, and ensure the stable operation of the power grid.

[0082] Embodiment 2, as Figure 4 As shown, the embodiment of the present application provides a method for dynamically evaluating the supply guarantee capability under the condition of new energy load, and the method includes:

[0083] In the preset evaluation time zone, the power load of the power supply area of ​​the new energy power grid is predicted to generate a load prediction time series; based on the load prediction time series, the power generation demand analysis is performed in combination with the frequent power distribution losses of the new energy power grid to generate a preset power generation time series; a hybrid power generation prediction model of the new energy power grid is established; in the preset evaluation time zone, with the preset power generation time series as the target, energy scheduling fitting is performed based on the hybrid power generation prediction model to generate an energy scheduling time series with the minimum fitting deviation index; the energy scheduling time series is compared with the preset power generation time series to generate a supply guarantee capacity evaluation result.

[0084] Furthermore, the embodiment of the present application performs power load forecasting in a preset evaluation time zone and generates a load forecast time series, and further includes:

[0085] Collect historical electricity consumption records of the power supply area, and build a load forecasting model based on the historical electricity consumption records; perform network heat identification on the power supply area in the preset evaluation time zone, and generate a network heat mapping result, wherein the network heat mapping result includes multiple groups of network heat information types and heat indicators with mapping relationships; perform feedback impact analysis on the network heat mapping through the power load feedback network, and generate a feedback factor; perform load forecasting in the preset evaluation time zone through the load forecasting model, and perform feedback adjustment in combination with the feedback factor to generate the load forecasting time series.

[0086] Preferably, the embodiment of the present application performs feedback impact analysis on the network heat map through the power load feedback network to generate a feedback factor, and further includes:

[0087] The heat-people flow conversion database is called through the electricity load feedback network to identify heat-people flow conversion data samples whose similarity with the network heat mapping result is greater than a preset similarity, and generate traffic surge data; based on the traffic surge data, the traffic-load influence curve is called to identify the load influence coefficient and generate the feedback factor.

[0088] Optionally, the method of generating traffic surge data in the embodiment of the present application further includes:

[0089] The heat-people flow conversion database includes multiple groups of heat-people flow conversion data samples, wherein any group of heat-people flow conversion data samples includes heat information type samples, heat index samples and traffic surge samples; traverse the heat-people flow conversion database, perform similarity comparison on any group of network heat information types and heat indexes in the network heat mapping results, and generate multiple groups of matching traffic surge samples; perform weighted fusion on the multiple groups of matching traffic surge samples to generate the traffic surge data.

[0090] Furthermore, the embodiment of the present application establishes a hybrid power generation prediction model for a new energy power grid, and further includes:

[0091] Acquire multiple power generation sources of the new energy power grid and establish multiple sets of unstable influencing factors of the multiple power generation sources; collect historical power generation data based on the multiple sets of unstable influencing factors, construct multiple power generation prediction branches, and generate the hybrid power generation prediction model.

[0092] Furthermore, in the preset evaluation time zone, the embodiment of the present application takes the preset power generation time sequence as the target, performs energy scheduling fitting based on the hybrid power generation prediction model, and generates an energy scheduling time sequence with the minimum fitting deviation index, and also includes:

[0093] In the preset evaluation time zone, real-time data collection is performed on the multiple sets of unstable influencing factors, and the data is input into the multiple power generation prediction branches to generate multiple first predicted power generation time series; the multiple first predicted power generation time series are aligned and added to generate a fused power generation time series; the energy storage module and the fused power generation time series are combined to perform energy scheduling with the preset power generation time series as the target, and a fitting loss minimization analysis is performed through an iterative fitting loss function to generate the energy scheduling time series.

[0094] Preferably, the embodiment of the present application combines the energy storage module and the fusion power generation time sequence, performs energy scheduling with the preset power generation time sequence as the target, performs fitting loss minimization analysis by iterative fitting loss function, and generates the energy scheduling time sequence, and also includes:

[0095] Acquire the energy storage capacity of the energy storage module; perform timing node mapping on the fused power generation timing and the preset power generation timing; when the fused power generation of the first timing mapping node is greater than the preset power generation, perform electric energy storage fitting in combination with the energy storage capacity to generate the first node stored electric energy; acquire the fused power generation of the second timing mapping node, and add it with the first node stored electric energy; compare the sum with the preset power generation of the second timing mapping node, perform electric energy storage fitting again, and so on, until all timing mapping nodes are traversed to generate the energy scheduling timing.

[0096] Through the above detailed description of a system for dynamically evaluating power supply capability under renewable energy load conditions, those skilled in the art can clearly understand a method for dynamically evaluating power supply capability under renewable energy load conditions in this embodiment. As for the method disclosed in Embodiment 2, since it corresponds to the system disclosed in Embodiment 1 and has corresponding execution steps and beneficial effects, please refer to the description of the system part for the relevant matters.

[0097] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A dynamic evaluation system for supply capacity under new energy load conditions, characterized in that: include: An electricity load prediction module, which is used to predict the electricity load of the power supply area of ​​the new energy power grid in a preset evaluation time zone and generate a load prediction time sequence; A power generation demand analysis module, the power generation demand analysis module is used to perform power generation demand analysis based on the load forecast time sequence and in combination with the frequent power distribution losses of the new energy power grid, and generate a preset power generation time sequence; A prediction model building module, wherein the prediction model building module is used to build a hybrid power generation prediction model for a new energy power grid; An energy scheduling fitting module, the energy scheduling fitting module is used to perform energy scheduling fitting based on the hybrid power generation prediction model in the preset evaluation time zone and with the preset power generation time sequence as the target, to generate an energy scheduling time sequence with the minimum fitting deviation index; A supply guarantee capability evaluation module, the supply guarantee capability evaluation module is used to compare the energy scheduling sequence with the preset power generation sequence to generate a supply guarantee capability evaluation result; The prediction model building module is also used to perform the following steps: Acquire multiple power generation sources of the new energy power grid, and establish multiple sets of unstable influencing factors of the multiple power generation sources; Based on the plurality of unstable influencing factor sets, historical power generation data are collected, a plurality of power generation prediction branches are constructed, and the hybrid power generation prediction model is generated; The energy scheduling fitting module is also used to perform the following steps: In the preset evaluation time zone, real-time data collection is performed on the plurality of unstable influencing factor sets, and input into the plurality of power generation prediction branches to generate a plurality of first predicted power generation time series; Aligning and adding the multiple first predicted power generation time series to generate a fused power generation time series; Combining the energy storage module and the fused power generation time sequence, performing energy scheduling with the preset power generation time sequence as the target, performing fitting loss minimization analysis through iterative fitting loss function, and generating the energy scheduling time sequence; The energy scheduling fitting module is also used to perform the following steps: Obtaining the energy storage capacity of the energy storage module; Performing time series node mapping on the fused power generation time series and the preset power generation time series, when the fused power generation of the first time series mapping node is greater than the preset power generation, performing electric energy storage fitting in combination with the energy storage capacity to generate first node storage electric energy; The fused power generation of the second timing mapping node is obtained, and added to the stored electric energy of the first node, the sum result is compared with the preset power generation of the second timing mapping node, and the electric energy storage fitting is performed again, and so on, until all timing mapping nodes are traversed to generate the energy scheduling timing.

2. A system for dynamically evaluating supply capacity under new energy load conditions as claimed in claim 1, characterized in that: The power load prediction module is also used to perform the following steps: Collecting historical electricity consumption records of the power supply area, and building a load forecasting model based on the historical electricity consumption records; Performing network heat identification on the power supply area in the preset evaluation time zone to generate a network heat mapping result, wherein the network heat mapping result includes a plurality of groups of network heat information types and heat indicators having a mapping relationship; Performing feedback impact analysis on the network heat map through an electricity load feedback network to generate a feedback factor; The load forecasting model is used to perform load forecasting in the preset evaluation time zone, and feedback adjustment is performed in combination with the feedback factor to generate the load forecasting time series.

3. A dynamic evaluation system for supply capacity under new energy load conditions as claimed in claim 2, characterized in that: The power load prediction module is also used to perform the following steps: Calling the heat-people flow conversion database through the power load feedback network, identifying the heat-people flow conversion data samples whose similarity with the network heat mapping result is greater than a preset similarity, and generating traffic surge data; Based on the traffic surge data, the traffic-load impact curve is called to identify the load impact coefficient and generate the feedback factor.

4. A system for dynamically evaluating supply capacity under new energy load conditions as claimed in claim 3, characterized in that: The heat-people flow conversion database includes multiple groups of heat-people flow conversion data samples, wherein any group of heat-people flow conversion data samples includes heat information type samples, heat index samples and traffic surge samples, and the power load prediction module is further used to perform the following steps: Traversing the heat-traffic conversion database, performing a similarity comparison on any group of network heat information types and heat indicators in the network heat mapping results, and generating multiple groups of matching traffic surge samples; The plurality of groups of matching traffic surge samples are weightedly fused to generate the traffic surge data.

5. A method for dynamically evaluating supply capacity under new energy load conditions, characterized in that: The method is performed by a system for dynamically evaluating supply capacity under new energy load conditions according to any one of claims 1 to 4, comprising: In the preset evaluation time zone, the power load of the power supply area of ​​the new energy power grid is predicted to generate a load prediction time series; Based on the load forecast time series, combined with the frequent power distribution losses of the new energy grid, the power generation demand analysis is performed to generate a preset power generation time series; Establish a hybrid power generation prediction model for new energy grids; In the preset evaluation time zone, taking the preset power generation time sequence as the target, performing energy scheduling fitting based on the hybrid power generation prediction model, and generating an energy scheduling time sequence with the minimum fitting deviation index; The energy dispatching sequence and the preset power generation sequence are compared to generate a power supply capability assessment result.

Images