Operation scheduling optimization method and device for new energy access power grid, and medium

Abstract

The invention discloses an operation scheduling optimization method and device for a new energy access power grid, and a medium, and belongs to the technical field of power grid scheduling, and the method comprises the steps: obtaining power grid data, carrying out the feature extraction of the power grid data, and obtaining a comprehensive power grid data feature; acquiring historical power grid data, constructing and training a medium and long term prediction model by using the historical power grid data to obtain a first medium and long term prediction model, and inputting the comprehensive power grid data features into the first medium and long term prediction model to predict power grid operation data; performing iterative optimization on the power grid operation data by using an optimization algorithm to generate an optimal power grid short-term scheduling demand; and scheduling adjustment is performed on the power grid according to the optimal short-term scheduling demand of the power grid, so that operation scheduling optimization of new energy access to the power grid is realized, and optimization and accurate control of power grid scheduling are realized.

Description

Technical Field

[0001] This invention relates to the field of power grid dispatching technology, specifically to an operation and dispatching optimization method, equipment, and medium for new energy grid access. Background Technology

[0002] With the transformation and development of the global energy structure, the proportion of new energy sources (such as wind and solar power) in the global power system is gradually increasing. The integration of new energy sources has brought many technical challenges to the power system, especially in grid dispatch and operation management. Traditional power systems are mainly dispatched based on traditional energy sources such as coal, gas, and water, which are relatively stable and easy to predict. However, new energy power generation is affected by natural factors such as sunlight intensity and wind speed, exhibiting strong volatility and uncertainty. This poses a greater challenge to the stability and security of grid operation, especially when the proportion of new energy is high. How to accurately predict and optimize grid operation and dispatch has become one of the current research hotspots. Although there has been some research progress on grid dispatch optimization methods after the integration of new energy, existing technologies still have shortcomings. Traditional grid dispatch optimization methods are mostly based on load forecasting and the dispatch optimization of traditional generating units, often ignoring the volatility and uncertainty of new energy power generation. They are unable to cope with the complex impact of new energy power generation on grid load balance and stability, and lack dynamic adjustment and optimization of grid operation modes.

[0003] This invention proposes an operation scheduling optimization method, equipment, and medium for new energy grid access. By collecting grid data and extracting grid data features, the invention predicts the medium- and long-term operation mode of the grid based on the grid data features, thereby improving the prediction accuracy. Furthermore, by removing the impact of new energy from the medium- and long-term operation mode of the grid, an optimization objective function is constructed to optimize the short-term scheduling needs of the grid, effectively improving the grid's adaptability and scheduling capability to new energy fluctuations, and realizing the optimization and precise control of grid scheduling. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by this invention is that traditional power grid dispatch optimization methods are mostly based on load forecasting and dispatch optimization of traditional generator sets, which ignore the volatility and uncertainty of new energy power generation, make it difficult to cope with the complex impact of new energy power generation on the load balance and stability of the power grid, and lack dynamic adjustment and optimization of the power grid operation mode.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for optimizing the operation and scheduling of new energy sources connected to the power grid, comprising:

[0007] Acquire power grid data, extract features from the power grid data, and obtain comprehensive power grid data features;

[0008] Historical power grid data is acquired, and a medium- and long-term prediction model is constructed and trained using the historical power grid data to obtain a first medium- and long-term prediction model. The comprehensive power grid data features are then input into the first medium- and long-term prediction model to predict power grid operation data.

[0009] The power grid operation data is iteratively optimized using optimization algorithms to generate the optimal short-term power grid dispatch requirements;

[0010] The power grid is adjusted according to the optimal short-term grid dispatch requirements to optimize the operation and dispatch of new energy grid access.

[0011] As a preferred embodiment of the operation and scheduling optimization method for new energy grid access described in this invention, the feature extraction of the grid data includes:

[0012] Calculate the mean and variance of the power grid data respectively;

[0013] Fourier transform is used to extract the frequency domain representation and spectral density of power grid data, and autocorrelation analysis is performed on the power grid data to calculate the autocorrelation of the power grid data.

[0014] By combining the mean, variance, frequency domain representation, spectral density, and autocorrelation of power grid data, comprehensive power grid data characteristics are obtained.

[0015] As a preferred embodiment of the operation and scheduling optimization method for new energy grid access described in this invention, the method for predicting grid operation data includes:

[0016] Power grid data includes load data and generation data;

[0017] For load data, the first medium- to long-term forecasting model includes a long short-term memory network model;

[0018] The comprehensive power grid data characteristics are input into the long short-term memory network model to obtain the predicted power grid load and phase angle;

[0019] The daily periodic fluctuations of the power grid are obtained by Fourier transform analysis based on the grid load and phase angle.

[0020] The beneficial effects of this preferred technical solution are that it can improve the accuracy of power grid load forecasting, enhance the power grid's adaptability to daily periodic fluctuations, and optimize power grid dispatching.

[0021] As a preferred embodiment of the operation and scheduling optimization method for new energy grid access described in this invention, it further includes:

[0022] For power generation data, the first medium- to long-term forecasting model includes the autoregressive moving average model;

[0023] Extract new energy power generation data from power generation data, input the new energy power generation data into the autoregressive moving average model, and obtain the predicted new energy power generation data;

[0024] Calculate the mean and standard deviation of new energy power generation data, and assess the grid reserve capacity demand based on the mean and standard deviation of new energy power generation data.

[0025] The beneficial effects of this preferred technical solution are that it can effectively predict new energy power generation, accurately assess reserve capacity demand, and improve the grid's adaptability to new energy fluctuations and dispatch flexibility.

[0026] As a preferred embodiment of the operation and scheduling optimization method for new energy grid access described in this invention, generating the optimal short-term grid scheduling demand includes:

[0027] The difference between the predicted grid load and the predicted new energy power generation data is used to obtain the grid load regulation demand.

[0028] The grid load dispatch cost and the reserve power cost are obtained by multiplying the grid load regulation demand and the unit load dispatch cost, and by multiplying the grid reserve capacity demand and the unit reserve power capacity cost.

[0029] As a preferred embodiment of the operation and scheduling optimization method for new energy grid access described in this invention, it further includes:

[0030] The predicted load change and reserve capacity demand change are calculated by predicting the grid load and the grid reserve capacity demand, respectively.

[0031] The dynamic contribution of the backup power supply is obtained by adding the square of the predicted load change, the coupling coefficient, the predicted load change, the reserve capacity demand change, and the load adjustment coefficient.

[0032] As a preferred embodiment of the operation and scheduling optimization method for new energy grid access described in this invention, it further includes:

[0033] The objective function for short-term dispatch demand is defined as minimizing the sum of grid load dispatch cost, standby power cost, and standby power dynamic contribution.

[0034] The predicted load change and reserve capacity demand change are used as the initial optimization group, and a reference optimization group is randomly generated.

[0035] The iterative optimization using the simulated annealing algorithm includes setting an initial temperature by calculating the difference between the objective function values ​​of the initial optimization group and the randomly generated reference optimization group, gradually reducing the temperature during the iteration process, and calculating the acceptance probability of the new solution.

[0036] When the acceptance probability is greater than the first threshold, the iteration stops, and the reference optimization group that maximizes the difference in the objective function value is selected as the optimization result. The predicted load change in the optimization group is taken as the optimal short-term power grid dispatch demand.

[0037] The beneficial effects of this preferred technical solution are: optimizing grid dispatch costs, improving grid operation efficiency, and enhancing the grid's adaptability and stability to new energy fluctuations.

[0038] As a preferred embodiment of the operation and scheduling optimization method for new energy grid access described in this invention, the method includes: adjusting the grid scheduling based on the optimal short-term grid scheduling demand, which includes:

[0039] The grid load is adjusted according to the optimal short-term grid dispatch requirements, and grid data is collected in real time during the adjustment process. The collected real-time grid data is used to generate short-term grid dispatch records and overwrite the timestamps.

[0040] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the operation scheduling optimization method for new energy grid access.

[0041] The present invention provides a computer-readable storage medium having a computer program stored thereon, characterized in that the computer program, when executed by a processor, implements the steps of the operation scheduling optimization method for new energy grid access.

[0042] The beneficial effects of this invention are as follows: This invention effectively solves the shortcomings of traditional power grid dispatch optimization methods after the integration of new energy power generation. It uses the Long Short-Term Memory (LSTM) network model and the Autoregressive Moving Average (ARIMA) model to accurately predict the power grid load and new energy power generation, thereby enhancing the power grid's adaptability to daily periodic fluctuations and new energy fluctuations. After removing the influence of new energy through the medium- and long-term operation mode of the power grid, an optimization objective function is constructed to optimize the short-term dispatch requirements of the power grid, effectively improving the power grid's adaptability and dispatch capability to new energy fluctuations, realizing the optimization and precise control of power grid dispatch. Iterative optimization using the simulated annealing algorithm finds the optimal short-term dispatch requirements of the power grid, further improving the flexibility and stability of power grid operation. Attached Figure Description

[0043] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 The above is a flowchart of an operation scheduling optimization method for new energy grid access provided in one embodiment of the present invention. Detailed Implementation

[0045] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0046] Example 1, referring to Figure 1 This is one embodiment of the present invention, which provides an operation scheduling optimization method for new energy grid access, comprising:

[0047] S100: Acquire power grid data, extract features from the power grid data, and obtain comprehensive power grid data features;

[0048] S200: Acquire historical power grid data, use the historical power grid data to build and train a medium- and long-term prediction model to obtain the first medium- and long-term prediction model, and input the comprehensive power grid data features into the first medium- and long-term prediction model to predict the power grid operation data;

[0049] S300: Iteratively optimizes power grid operation data using optimization algorithms to generate optimal short-term power grid dispatch requirements;

[0050] S400: Adjusts the power grid according to the optimal short-term grid dispatch requirements to optimize the operation and dispatch of new energy sources connected to the grid.

[0051] It should be noted that, through precise feature extraction and historical data analysis, an efficient medium- and long-term prediction model was constructed, significantly improving the prediction accuracy of power grid operation data. Utilizing optimization algorithms to iteratively optimize power grid operation data, the generated optimal short-term dispatch demand effectively reduced power grid load dispatch costs, enhanced the power grid's adaptability to new energy fluctuations, and achieved dynamic optimization of power grid operation through real-time dispatch adjustments. This ensured the stability and economic viability of new energy sources after their integration into the grid, providing strong support for intelligent power grid management.

[0052] In this embodiment of the invention, step S100 includes the following sub-steps A1-A3;

[0053] In A1: Calculate the mean and variance of the power grid data respectively;

[0054] In A2: Fourier transform is used to extract the frequency domain representation and spectral density of the power grid data, autocorrelation analysis is performed on the power grid data, and the autocorrelation of the power grid data is calculated.

[0055] In A3: The mean, variance, frequency domain representation, spectral density, and autocorrelation of the power grid data are combined to obtain the comprehensive power grid data characteristics.

[0056] Specifically, acquiring power grid data includes collecting power grid data, including load data and generation data, by deploying multimodal sensors at power grid nodes. The multimodal sensors are connected through a network to form a sensor network and connected to the power dispatch center. The multimodal sensors transmit the collected power grid data to the power dispatch center for preprocessing, including data cleaning, filtering and noise reduction, missing value imputation, and outlier removal. At the same time, historical power grid data is acquired based on the power dispatch center and used as training data for subsequent models.

[0057] It should be noted that by deploying multimodal sensors at key nodes of the power grid to collect power grid data, the problem of data diversity in power grid monitoring is solved. Traditional power grid monitoring often relies on a single type of sensor, which easily overlooks other important factors in the power grid, such as meteorological data and power quality. The introduction of multimodal sensors can simultaneously collect data on load, generation, voltage, frequency, and other aspects, forming more comprehensive and detailed power grid status data. This not only improves the accuracy of power grid monitoring but also helps the power dispatch center to fully understand the power grid operation and promptly detect potential system problems or faults. Connecting multimodal sensors through a network to form a sensor network and connecting it to the power dispatch center allows the power dispatch center to seamlessly connect to various power grid nodes, obtain various types of information in real time, and improve the timeliness and accuracy of data. Since power grid data is often subject to noise interference, preprocessing can effectively improve the accuracy and reliability of the data. Through data fusion and analysis, the power dispatch center can integrate data from different sensors to form more representative and accurate power grid operation characteristics. The beneficial effect of this process is that it improves the quality of data analysis, enabling power dispatch decisions to be based on a more realistic and reliable power grid operation status, thereby conducting more precise dispatch.

[0058] Specifically, feature extraction of power grid data includes calculating the mean and variance of each type of preprocessed power grid data, expressed as follows: ; ;

[0059] in, and The first Mean and variance of power grid-like data For the first The total number of grid-like data, For the first Class-A power grid data One data point;

[0060] Fourier transform is used to extract the frequency domain representation and spectral density of each type of power grid data. Autocorrelation analysis is performed on each type of power grid data to calculate the autocorrelation of each type of power grid data. The mean, variance, frequency domain representation, spectral density and autocorrelation of each type of power grid data are used to form the features of each type of power grid data. The features of all types of power grid data are combined to form the comprehensive power grid data features.

[0061] Fourier transform is used to extract the frequency domain representation and spectral density of each type of power grid data. The amplitude spectrum is represented as follows: ;

[0062] in, For frequency indexing, by sampling frequency Acquire, sampling frequency Corresponding to the power grid data sequence, The imaginary unit, For amplitude spectrum, For the first The total number of grid-like data, For the first Class-A power grid data One data point;

[0063] Phase extraction based on the imaginary and real parts of the amplitude spectrum Represented as: ;

[0064] in, and They are represented by the imaginary part and the real part, respectively. For phase, For amplitude spectrum, It is the arctangent function;

[0065] The amplitude spectrum and phase are used as the frequency domain representation of the power grid data, and the spectral density is further calculated based on the amplitude spectrum. Represented as: ;

[0066] in, For power grid data sampling frequency, For spectral density, For amplitude spectrum, For the first The total number of data related to power grids.

[0067] The autocorrelation of each type of power grid data is expressed as follows: ;

[0068] in, For the first Autocorrelation of power grid-like data The lag coefficient, For the first The total number of grid-like data, For the first Class-A power grid data One data point, For the first Class-A power grid data Data points.

[0069] It should be noted that by performing mean and variance calculations and frequency domain analysis on each type of power grid data, key characteristics of power grid operation can be accurately captured. For example, the mean of the power grid can provide information about the stability of overall load or power generation, while the variance reveals the amplitude of fluctuations. Frequency domain representation and spectral density can help identify periodic fluctuations in the power grid. For example, when renewable energy generation is integrated, the fluctuations of wind or solar power generation often exhibit periodic characteristics. Autocorrelation analysis can identify the continuous changes in the power grid within a specific time period, providing in-depth time-series characteristic analysis for power grid operation. Combining these characteristics helps to comprehensively improve the accuracy and efficiency of power grid operation monitoring. Feature extraction of power grid data not only helps in the analysis of the current power grid status but also helps in predicting future power grid operation trends. For example, the frequency domain representation obtained through Fourier transform can reveal the energy distribution of the power grid at different frequencies, thereby helping to predict possible load fluctuations and renewable energy generation fluctuations. Autocorrelation analysis, by identifying the time-series dependencies of power grid data, can provide support for short-term load forecasting and power generation dispatch. Combining the extraction of power grid data characteristics helps to improve the power grid dispatching system's ability to predict the future power grid operation status, thereby enabling better power grid dispatching and optimization.

[0070] In this embodiment of the invention, step S200 includes the following sub-steps B1-B4;

[0071] In B1: Grid data includes load data and generation data;

[0072] In B2: For load data, the first medium- to long-term forecasting model includes a long short-term memory network model;

[0073] In B3: The integrated power grid data characteristics are input into the long short-term memory network model to obtain the predicted power grid load and phase angle;

[0074] In B4: The daily periodic fluctuations of the power grid are obtained by Fourier transform analysis based on the grid load and phase angle.

[0075] Specifically, the first medium- to long-term prediction model is a trained medium- to long-term prediction model, including a Long Short-Term Memory Network (LSTM) model and an Autoregressive Moving Average (ARIMA) model.

[0076] Constructing a medium- to long-term forecasting model includes building and training an LSTM model. The LSTM model is defined with inputs of integrated power grid data features and outputs of predicted power grid load and phase angle. The integrated power grid data features are input into the trained LSTM model to obtain the predicted power grid load and phase angle.

[0077] Based on the grid load and phase angle, Fourier transform analysis yields the following daily periodic fluctuations in the power grid: ;

[0078] in, For the daily periodic fluctuations of the power grid, To analyze the number of cycles, For the analysis period number, For time, the unit is days. To predict grid load, The phase angle, It is a cosine function;

[0079] It should be noted that using LSTM models for power grid load forecasting has significant advantages, especially when power grid load data exhibits strong time-series characteristics. LSTM models can accurately capture the long-term fluctuation trend of power grid load by learning the time dependencies in historical data, thereby achieving accurate prediction of future power grid load. In addition, LSTM models can not only predict load but also output phase angle information, providing key data for power flow calculation in power grid dispatch. By accurately predicting power grid load and phase angle, dispatch centers can better allocate and optimize power grid resources, ensuring the stable operation of the power grid.

[0080] In this embodiment of the invention, after completing steps B1-B4, step S200 also includes steps B5-B7.

[0081] In B5: For power generation data, the first medium- to long-term forecasting model includes an autoregressive moving average model;

[0082] In B6: Extract new energy power generation data from the power generation data, input the new energy power generation data into the autoregressive moving average model, and obtain the predicted new energy power generation data;

[0083] In B7: Calculate the mean and standard deviation of new energy power generation data, and assess the grid reserve capacity demand based on the mean and standard deviation of new energy power generation data.

[0084] Specifically, new energy power generation data is extracted from power generation data, an ARIMA model is constructed and trained, the input of the ARIMA model is defined as new energy power generation data, and the output is predicted new energy power generation data. The collected new energy power grid power generation data is input into the trained ARIMA model to obtain predicted new energy power generation data.

[0085] The mean and standard deviation of new energy power generation data are calculated, and the grid reserve capacity demand is assessed as follows: ;

[0086] in, To meet the power grid's reserve capacity requirements, To predict grid load, To predict new energy power generation data, The standard deviation of new energy power generation data. This refers to the daily periodic fluctuations in the power grid.

[0087] It should be noted that using the ARIMA model to predict renewable energy generation data can help the power grid dispatch center predict the volatility of wind or solar power generation. Since renewable energy generation is usually greatly affected by meteorological conditions and is highly volatile, accurate prediction of its power generation changes is of great significance. The ARIMA model can identify seasonal and trend changes in renewable energy generation data, thereby generating short-term power generation forecast data to provide decision support for power dispatch. Based on the renewable energy generation data predicted by the ARIMA model, the dispatch center can make advance arrangements for power grid resource allocation to ensure that the power grid can fully absorb renewable energy power and balance load demand. By calculating the standard deviation of renewable energy generation data, the power dispatch center can understand the degree of uncertainty in power generation and estimate the required reserve capacity of the power grid accordingly. The larger the standard deviation, the higher the uncertainty in power generation, and therefore the higher the reserve capacity is required to cope with possible fluctuations. By accurately assessing the reserve capacity requirements, the dispatch center can better allocate flexible resources such as energy storage systems and peak-shaving power sources, thereby improving the stability and emergency response capabilities of the power grid.

[0088] It should also be noted that by using advanced prediction and analysis methods such as LSTM models, Fourier transforms, and ARIMA models, combined with historical and real-time data of the power grid, the level of intelligence in power grid dispatch can be significantly improved. Based on accurate load forecasts and new energy generation forecasts, the dispatch center can make adjustments and optimizations to power grid resources in advance, improving the power grid's response to load fluctuations and new energy fluctuations. The intelligent dispatch system can not only improve the operating efficiency of the power grid, but also significantly reduce manual intervention and errors, making the power grid operation more stable and reliable.

[0089] In this embodiment of the invention, step S300 includes the following sub-steps C1-C4;

[0090] In C1: The grid load regulation demand is obtained based on the difference between the predicted grid load and the predicted new energy generation data;

[0091] In C2: The grid load dispatch cost and the backup power cost are obtained by multiplying the grid load regulation demand and the unit load dispatch cost, and the grid reserve capacity demand and the unit backup power capacity cost.

[0092] In C3: the predicted load change and the change in reserve capacity demand are calculated by predicting the grid load and the grid reserve capacity demand, respectively;

[0093] In C4: The dynamic contribution of the backup power supply is obtained by adding the square of the predicted load change, the coupling coefficient, the predicted load change, the reserve capacity demand change, and the load regulation coefficient.

[0094] Specifically, calculating grid load regulation demand based on the difference between predicted grid load and predicted renewable energy generation data includes: ;

[0095] in, To predict grid load, To predict new energy power generation data, For time The target load of the power grid is usually determined by the power grid dispatch plan, reflecting the load required by the power grid in the absence of any new energy fluctuations. To meet the needs of power grid load dispatching.

[0096] The difference between the predicted grid load and the predicted renewable energy generation data is used to remove the impact of renewable energy generation on the grid load. The grid load regulation demand represents the difference between the load demand and the target load, ensuring the real-time and accuracy of the load regulation demand, and helping grid dispatchers to formulate more refined dispatch plans.

[0097] The costs of power grid load dispatching and backup power supply are expressed as follows: ; ;

[0098] in, For unit load dispatching cost, Cost per unit of backup power capacity For the cost of power grid load dispatching, For backup power cost, To meet the needs of power grid load dispatching, This is to meet the power grid's reserve capacity requirements.

[0099] The predicted load change is expressed as follows: ;

[0100] in, To predict load changes, To predict grid load, For time Forecasted grid load;

[0101] The change in reserve capacity demand, calculated based on the grid reserve capacity demand, is expressed as follows: ;

[0102] in, This represents the change in reserve capacity demand. To meet the power grid's reserve capacity requirements, For time The grid reserve capacity requirement.

[0103] The dynamic contribution of backup power is calculated based on the predicted load changes and changes in backup capacity demand, and is expressed as follows: ;

[0104] in, This is the load adjustment coefficient. The coupling coefficient is set empirically. For time analysis, the timeframe is determined based on the short-term scheduling. To predict load changes, This represents the change in reserve capacity demand. Dynamic contribution to backup power.

[0105] It should be noted that by calculating grid load regulation demand in real time, grid dispatching can quickly respond to fluctuations in grid load and renewable energy generation. By accurately calculating the difference between the target load and the actual load, grid dispatchers can monitor the grid's operating status in real time and make timely adjustments. The load regulation demand calculation takes into account the volatility of renewable energy generation, effectively eliminating the impact of renewable energy fluctuations on grid load. By combining the calculation of the difference between grid load and renewable energy generation data, grid load can be adjusted in real time to ensure that the grid can balance load demand and generation capacity. Especially when renewable energy generation is unstable, grid dispatching can dynamically adjust load demand, thereby avoiding frequent grid fluctuations and improving the stability of the power system. The predicted load changes and reserve capacity demand changes can help the power dispatching center plan the configuration of backup power sources in advance, ensuring that backup power sources can be put into use in a timely manner when grid load fluctuations or renewable energy generation fluctuations are large, thereby improving the grid's emergency response capability. Optimized backup power source configuration can reduce grid dispatching costs and reduce the waste of over-configuration of backup power sources.

[0106] In this embodiment of the invention, after completing steps C1-C4, step S300 further includes steps C5-C8;

[0107] In C5: the objective function for short-term dispatch demand is defined as minimizing the sum of grid load dispatch cost, standby power cost, and standby power dynamic contribution;

[0108] In C6: the predicted load change and reserve capacity demand change are used as the initial optimization group, and a reference optimization group is randomly generated.

[0109] In C7: Iterative optimization using simulated annealing algorithm includes setting the initial temperature by calculating the difference between the objective function values ​​of the initial optimization group and the randomly generated reference optimization group, gradually reducing the temperature during the iteration process, and calculating the acceptance probability of the new solution;

[0110] In C8: When the acceptance probability is greater than the first threshold, the iteration stops, and the reference optimization group that maximizes the difference in the objective function value is selected as the optimization result. The predicted load change in the optimization group is taken as the optimal short-term power grid dispatch demand.

[0111] In this embodiment of the invention, the optimization algorithm includes a simulated annealing algorithm;

[0112] Specifically, the objective function for short-term scheduling requirements is defined as follows: ;

[0113] in, Let be the objective function. For the cost of power grid load dispatching, For backup power cost, Dynamic contribution to backup power supply For analysis time;

[0114] Using the predicted load changes and reserve capacity demand changes as the initial optimization group, and randomly generating a reference optimization group, the objective function values ​​of the initial optimization group are calculated. The differences between the objective function values ​​of the initial optimization group and the reference optimization group are then calculated. The initial temperature is calculated based on these differences and is expressed as follows: ;

[0115] in, To initialize the difference between the objective function values ​​of the optimization group and the reference optimization group, The initial temperature;

[0116] The reference optimization set is iteratively adjusted using the simulated annealing algorithm, and the temperature is gradually reduced. The acceptance probability of the new solution is calculated after each iteration and expressed as follows: ;

[0117] in, The difference between the values ​​of the objective function during iteration. To reduce the temperature for iteration, To accept probability, It is an exponential function.

[0118] After obtaining the acceptance probability, it is compared with the first threshold. If the acceptance probability is greater than the first threshold, the iteration stops, all reference optimization groups of the last iteration are output, and the reference optimization group that maximizes the difference in the objective function value is extracted as the final optimization group. The predicted load change in the final optimization group is used as the final short-term grid dispatch demand. The first threshold can be obtained through historical data analysis or set through expert experience.

[0119] It should be noted that by simulating the physical annealing process, continuously adjusting the reference optimization set and reducing the temperature, the optimal ratio between grid load regulation and backup power demand is found. This algorithm can find the global optimal solution through multiple iterations, avoiding the trap of local optima, and ensuring the accuracy and efficiency of grid dispatch. The optimization of unit load dispatch cost and unit backup power capacity cost can reduce the additional costs required by the grid to cope with load fluctuations and new energy power generation fluctuations. This not only improves the operating efficiency of the grid, but also provides a guarantee for the economics of the power system. By combining advanced technologies such as real-time data, predictive models, and simulated annealing algorithms, it provides a brand-new intelligent decision support for grid dispatch.

[0120] In this embodiment of the invention, step S400 includes the following sub-step D1;

[0121] In D1: The grid load is adjusted according to the optimal short-term grid dispatch demand, and grid data is collected in real time during the adjustment process. The collected real-time grid data is used to generate short-term grid dispatch records and overwrite the timestamp.

[0122] Specifically, adjusting the power grid according to short-term dispatch needs and recording it in real time means adjusting the power grid load according to the generated short-term dispatch needs, collecting power grid data in real time during the adjustment process, generating short-term power grid dispatch records from the collected real-time power grid data and overwriting the timestamp.

[0123] Storing power grid data and short-term dispatch requirements in a database refers to storing collected power grid data, short-term dispatch requirements, and short-term dispatch records in a database. The database regularly performs integrity and security checks on the stored data and uploads the stored data to the cloud for backup.

[0124] It should be noted that real-time monitoring and dispatching enable timely responses to sudden load changes and fluctuations in new energy power generation, thereby ensuring more stable and flexible grid operation and reducing system fluctuations caused by emergencies. Load dispatching is a crucial task in grid operation, especially when multiple power generation methods (including wind and solar power) are integrated, making load regulation more complex. By adjusting the grid load according to short-term dispatching needs, flexible grid dispatching can be effectively achieved, ensuring scientific and rational load allocation. At the same time, the collection and recording of real-time data can provide timely feedback to power dispatchers, helping them make more accurate decisions. During grid dispatching, timestamp coverage ensures that each piece of grid data has a clear recording time, ensuring data timeliness. The application of timestamps allows grid dispatchers to track the specific time of each dispatching adjustment and compare historical data with current operating data to quickly identify potential problems.

[0125] The above is an illustrative scheme of an operation scheduling optimization method for new energy grid access according to this embodiment. It should be noted that the technical solution of this new energy grid access operation scheduling optimization system belongs to the same concept as the technical solution of the above-described new energy grid access operation scheduling optimization method. Details not described in detail in the technical solution of the new energy grid access operation scheduling optimization system in this embodiment can be found in the description of the above-described new energy grid access operation scheduling optimization method.

[0126] This embodiment also provides a computer device applicable to a method for optimizing the operation and scheduling of new energy grid access, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for optimizing the operation and scheduling of new energy grid access as proposed in the above embodiment.

[0127] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements an operation scheduling optimization method for new energy grid access as proposed in the above embodiments.

[0128] The storage medium proposed in this embodiment belongs to the same inventive concept as the operation scheduling optimization method for new energy grid access proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0129] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0130] Example 2, referring to Table 1, differs from the first example. It provides a verification test of the operation scheduling optimization method for new energy grid access, and verifies and explains the technical effects adopted in this method.

[0131] The existing technology and the method of the present invention are compared with the parameters including load forecasting error, new energy forecasting, reserve capacity redundancy rate, short-term scheduling cost and system frequency stability rate. The comparison experiment is shown in Table 1.

[0132] Table 1. Comparison of the method of the present invention with existing technologies:

[0133] As shown in Table 1, the method of this invention represents a significant improvement over existing technologies. Under conditions of high renewable energy integration, it can significantly reduce prediction errors and reserve redundancy, thereby improving the economy and stability of power grid operation. For example, load prediction error decreased from 6.8% to 2.9%, renewable energy prediction error decreased from 95MW to 38MW, reserve capacity redundancy rate decreased from 22.5% to 9.3%, short-term dispatch cost decreased from RMB 1.58 million / day to RMB 1.12 million / day, and system frequency stability increased from 92.4% to 98.6%. This demonstrates that the present invention can more accurately predict grid load and renewable energy generation, more effectively optimize grid dispatch, reduce reserve capacity redundancy, lower dispatch costs, and improve system frequency stability, thus achieving optimized and precise control of grid dispatch under conditions of high renewable energy integration.

[0134] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for optimizing the operation and scheduling of new energy sources connected to the power grid, characterized in that, include: Acquire power grid data, extract features from the power grid data, and obtain comprehensive power grid data features; Historical power grid data is acquired, and a medium- and long-term prediction model is constructed and trained using the historical power grid data to obtain a first medium- and long-term prediction model. The comprehensive power grid data features are then input into the first medium- and long-term prediction model to predict power grid operation data. The power grid operation data is iteratively optimized using optimization algorithms to generate the optimal short-term power grid dispatch requirements; The power grid is adjusted according to the optimal short-term grid dispatch requirements to optimize the operation and dispatch of new energy grid access.

2. The operation and scheduling optimization method for new energy grid access as described in claim 1, characterized in that, Feature extraction of the power grid data includes: Calculate the mean and variance of the power grid data respectively; Fourier transform is used to extract the frequency domain representation and spectral density of power grid data, and autocorrelation analysis is performed on the power grid data to calculate the autocorrelation of the power grid data. By combining the mean, variance, frequency domain representation, spectral density, and autocorrelation of power grid data, comprehensive power grid data characteristics are obtained.

3. The operation and scheduling optimization method for new energy grid access as described in claim 2, characterized in that, Forecasting power grid operation data includes: Power grid data includes load data and generation data; For load data, the first medium- to long-term forecasting model includes a long short-term memory network model; The comprehensive power grid data characteristics are input into the long short-term memory network model to obtain the predicted power grid load and phase angle; The daily periodic fluctuations of the power grid are obtained by Fourier transform analysis based on the grid load and phase angle.

4. The operation and scheduling optimization method for new energy grid access as described in claim 3, characterized in that, Also includes: For power generation data, the first medium- to long-term forecasting model includes the autoregressive moving average model; Extract new energy power generation data from power generation data, input the new energy power generation data into the autoregressive moving average model, and obtain the predicted new energy power generation data; Calculate the mean and standard deviation of new energy power generation data, and assess the grid reserve capacity demand based on the mean and standard deviation of new energy power generation data.

5. The operation and scheduling optimization method for new energy grid access as described in claim 4, characterized in that, Generating optimal short-term power grid dispatch requirements includes: The difference between the predicted grid load and the predicted new energy power generation data is used to obtain the grid load regulation demand. The grid load dispatch cost and the reserve power cost are obtained by multiplying the grid load regulation demand and the unit load dispatch cost, and by multiplying the grid reserve capacity demand and the unit reserve power capacity cost.

6. The operation scheduling optimization method for new energy grid access as described in claim 5, characterized in that, Also includes: The predicted load change and reserve capacity demand change are calculated by predicting the grid load and the grid reserve capacity demand, respectively. The dynamic contribution of the backup power supply is obtained by adding the square of the predicted load change, the coupling coefficient, the predicted load change, the reserve capacity demand change, and the load adjustment coefficient.

7. The operation and scheduling optimization method for new energy grid access as described in claim 6, characterized in that, Also includes: The objective function for short-term dispatch demand is defined as minimizing the sum of grid load dispatch cost, standby power cost, and standby power dynamic contribution. The predicted load change and reserve capacity demand change are used as the initial optimization group, and a reference optimization group is randomly generated. The iterative optimization using the simulated annealing algorithm includes setting an initial temperature by calculating the difference between the objective function values ​​of the initial optimization group and the randomly generated reference optimization group, gradually reducing the temperature during the iteration process, and calculating the acceptance probability of the new solution. When the acceptance probability is greater than the first threshold, the iteration stops, and the reference optimization group that maximizes the difference in the objective function value is selected as the optimization result. The predicted load change in the optimization group is taken as the optimal short-term power grid dispatch demand.

8. The operation and scheduling optimization method for new energy grid access as described in claim 7, characterized in that, Adjusting the power grid according to the optimal short-term power grid dispatch requirements includes: The grid load is adjusted according to the optimal short-term grid dispatch requirements, and grid data is collected in real time during the adjustment process. The collected real-time grid data is used to generate short-term grid dispatch records and overwrite the timestamps.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the operation scheduling optimization method for new energy grid access as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the operation scheduling optimization method for new energy grid access as described in any one of claims 1 to 8.

Images