Flexible resource adaptability configuration method based on multi-scene fusion

Abstract

The invention relates to a flexible resource adaptive configuration method based on multi-scene fusion. The method comprises the steps of obtaining wind power, photovoltaic and load power predicted values of a regional power grid; constructing a scene recognition target function by taking the weighted sum of the flexibility characteristic indexes of the power balance, peak regulation and frequency modulation scenes as a minimum target, and solving the scene recognition target function based on the scene recognition constraint to obtain a flexibility characteristic index vector; based on the flexibility characteristic index vector, a scene identification result is obtained through judgment; if the scene is one or more of power balance, peak regulation and frequency modulation, constructing a two-stage adaptive configuration model with minimum flexible resource capacity configuration and resource control loss of each side of the source network load storage as a target; based on a scene identification result and the first-stage constraint, solving to obtain a flexible resource configuration scale; and solving and obtaining a flexible resource operation plan based on wind power, photovoltaic and load power predicted values, the newly added flexible resource configuration scale and the determinacy and uncertainty constraints of the second stage.

Description

Technical Field

[0001] The present invention relates to the technical field of power systems, and in particular, to a flexible resource adaptive configuration method based on multi-scenario fusion. Background Art

[0002] As the proportion of new energy in the power system increases year by year, the requirements for the stability and flexibility of the power system are also increasing. New energy power generation has significant volatility and uncontrollability, which poses greater challenges to the reliability of power supply. How to improve the flexible regulation ability of the power system to adapt to the derivative characteristics brought by new energy fluctuations is an urgent problem to be solved at present. The medium- and long-term seasonal fluctuations, as well as the short-term randomness and intermittency of new energy power generation, require the power system to perform dynamic optimization on time scales such as across days, months, seasons, and even years. The volatility of new energy at each time scale gives rise to different demand scenarios, and single-dimensional regulation means are difficult to meet the flexible requirements of multi-scenario integration. Therefore, in the face of the demand problem of multi-scenario coupling, it is particularly important to construct a scenario recognition mechanism and formulate a differential configuration strategy.

[0003] Current research on the optimization of flexible resource allocation mostly focuses on a single scale or a single demand scenario, lacking differential configuration methods for different scenario requirements. Under multi-scenario integration, how to coordinate the flexible regulation resources on both sides of the source, grid, load, and storage at different time scales to achieve more comprehensive collaborative optimization still needs in-depth study. In addition, in a complex power grid environment where multiple adjustable resources participate together, there are complex coupling relationships among the resources. Most current research focuses on the independent optimization of single resources, lacking a collaborative coupling optimization mechanism among resources, especially insufficient research on the overall optimization control with the participation of large-scale distributed adjustable resources. Summary of the Invention

[0004] In view of the above analysis, the embodiments of the present invention aim to provide a flexible resource adaptive configuration method based on multi-scenario fusion to solve the technical problems that existing methods mostly focus on the optimization of a single time scale or scenario, lack research on differential configuration methods under multi-scenario integration, and focus on the independent optimization of single resources, without fully considering the collaborative coupling among multiple adjustable resources, and are difficult to meet the flexible regulation requirements of the power system under high-proportion new energy access.

[0005] The object of the present invention is mainly achieved by the following technical solutions: The present invention provides a flexible resource adaptive configuration method based on multi-scenario fusion, including the following steps: Obtain the predicted values of wind power, photovoltaic power, and load power of the regional power grid; Construct a scenario recognition objective function with the goal of minimizing the weighted sum of the flexibility characteristic indicators in the power balance scenario, peak shaving scenario, and frequency modulation scenario. Solve the scenario recognition objective function based on the scenario recognition constraints to obtain a flexibility characteristic indicator vector. Make a judgment based on the flexibility characteristic indicator vector to obtain the corresponding scenario recognition result. If the scenario recognition result is one or more of the power balance scenario, peak shaving scenario, and frequency modulation scenario, construct a two-stage adaptive configuration model with the goal of minimizing the capacity configuration of flexibility resources on each side of the source, grid, load, and storage and the control loss of flexibility resources on each side. Solve based on the scenario recognition result and the first-stage constraints to obtain the flexibility resource configuration scale. Solve based on the wind power, photovoltaic power, and load power prediction values, the newly added flexibility resource configuration scale, and the second-stage deterministic and uncertainty constraints to obtain the flexibility resource operation plan corresponding to the flexibility resource configuration scale in the scenario.

[0006] Furthermore, it also includes constructing a source-load model. Perform data fitting and solution on the source-load model based on the historical data of wind power, photovoltaic power, and the power of various types of users in the regional power grid to obtain the wind power, photovoltaic power, and load power prediction values, including: Obtain the load sample data of wind power, photovoltaic power, and various types of users in the regional power grid during the historical period, and perform preprocessing to obtain the preprocessed load sample data of wind power, photovoltaic power, and various types of users. Use the SVM and cross-validation method to perform data fitting and solution on the source-load model based on the load sample data of wind power, photovoltaic power, and various types of users to obtain the wind power, photovoltaic power, and load power prediction values.

[0007] Furthermore, the source-load model includes a wind power model, a photovoltaic power model, and a load power model. Perform data fitting and solution on the wind power model using the wind power sample data to obtain the wind power prediction value. Perform data fitting and solution on the photovoltaic power model using the photovoltaic power sample data to obtain the photovoltaic power prediction value. Perform data fitting and solution on the load power model using the load sample data of various types of users respectively to obtain the load power prediction values corresponding to various types of users. Integrate the load power prediction values of various types of users with the historical load mean value to obtain the load power prediction value.

[0008] Furthermore, the scenario recognition objective function is expressed as follows: ; Where , , are the flexibility characteristic indicators of the power balance scenario, peak shaving scenario, and frequency modulation scenario respectively. , , are respectively , , weights; , are respectively the maximum and minimum output powers of the flexibility resources required by the scenario; , are respectively the rated values of the maximum and minimum output powers; , are respectively the upward and downward ramp rates of the flexibility resources required by the scenario; , are respectively the maximum values of the upward and downward ramp power rates of the flexibility resources; , are respectively the conventional frequency regulation ability and fast frequency regulation ability indexes of the flexibility resources required by the scenario; are respectively the upper reserve and lower reserve regulation abilities of the conventional frequency regulation ability; The scenario recognition constraints include conventional power balance constraints, conventional non-adjustable unit operation constraints, wind power generation constraints, transmission line constraints, flexibility regulation resource operation constraints, load full-load constraints, spinning reserve capacity constraints, and fast frequency regulation reserve constraints.

[0009] Furthermore, based on the current grid parameters, the gurobi solver is used to solve the scenario recognition objective function based on the scenario recognition constraints, and ; Using to form the flexibility characteristic index vector , as follows: ; Judging based on the flexibility characteristic index vector to obtain the corresponding scenario, including: If or is non-zero, it is determined that a power balance scenario occurs; If or is non-zero, it is determined that a peak shaving scenario occurs; If or is non-zero, it is determined that a frequency regulation scenario occurs.

[0010] Furthermore, the two-stage adaptive configuration model includes a two-stage objective function with the minimum control loss of the flexibility resource capacities on each side of the source, grid, load, and storage and the flexibility resources on each side, a first-stage constraint, and a second-stage constraint; The first-stage constraints include configuration scale constraints, power demand constraints, ramp demand constraints, and frequency regulation demand constraints; The second-stage constraints include second-stage deterministic constraints and second-stage uncertainty constraints; The second-stage deterministic constraints include source-side, network, load, and side flexibility resource constraints.

[0011] Furthermore, the two-stage objective function is as follows: ; Wherein, , , , are the unit capacity configuration parameters of source, network, load, and storage resources respectively; is the configuration scale of source, network, load, and storage resources; , are the time scales of the whole year and the typical day respectively; , , , are the control losses of resources on each side of source, network, load, and storage respectively; is the type of flexibility resources for source, network, load, and storage; is the set of load uncertainties.

[0012] Furthermore, the second-stage uncertainty constraint is as follows: ; Wherein, represents an uncertain variable; , , are the output powers of photovoltaic and wind power and the electrical load power after considering uncertainties respectively; , , are the predicted values of wind power, photovoltaic power, and load power respectively; , , are the maximum allowable fluctuation coefficients of wind power, photovoltaic power output, and electrical load power respectively.

[0013] Furthermore, the flexibility resource configuration scale includes the newly added flexibility resource scales on the source, network, load, and storage sides , , and ; The flexibility resource operation plan corresponding to the flexibility resource configuration scale includes the regulation power of the newly added source-side flexibility resources , the regulation power of the newly added network-side flexibility resources , Transfer out of the transferable load of the newly added load-side resources , Transferable load and interruptible load power of the newly added load-side resources , Power generation power of the newly added energy storage-side resources , Charging power of the newly added energy storage-side resources and transfer in of the transferable load of the newly added load-side resources .

[0014] Further, the power demand constraint is as follows: ; Wherein, , , , are the numbers of conventional non-adjustable units, external transmission lines, flexibility adjustment units and loads respectively; is the output of the th conventional non-adjustable unit at time ; is the output of the photovoltaic unit at time; is the output of the wind turbine unit at time; is the output of the th external transmission line at time; is the output of the flexibility adjustment unit at time; is the adjustment power of the newly added flexibility resources; is the load demand of the load node at time; is the load shedding power at time.

[0015] Compared with the prior art, the present invention can at least achieve one of the following beneficial effects: 1. Multi-scenario fusion optimization and differential configuration ability: By constructing a multi-scenario recognition model (power balance scenario, peak shaving scenario and frequency modulation scenario), the present invention comprehensively considers the flexibility characteristic indexes of the power balance scenario, peak shaving scenario and frequency modulation scenario, solves the limitations of single-scenario optimization in the prior art, realizes the differential resource adaptation under multi-scenario coupling, and significantly improves the dynamic response ability of the power system to new energy fluctuations (such as wind and light intermittency, load mutation); 2. Global collaborative coupling optimization of multiple types of resources in the power source, grid, load, and energy storage: The present invention designs a two-stage adaptive configuration model (capacity configuration + operation regulation), integrates the dynamic coupling relationships of resources on the power source side (thermal power, energy storage), grid side (transmission lines), load side (interruptible / transferable loads), and energy storage side (energy storage), breaks through the bottleneck of independent optimization of a single resource in traditional research, realizes full-cycle collaborative optimization across time scales and resource types, improves the overall regulation efficiency, realizes the collaborative optimal configuration of flexible resources on each side of the power source, grid, load, and energy storage, improves resource utilization efficiency, and overcomes the defect of the existing technology that focuses on independent optimization of a single resource; 3. Dynamic adaptation and uncertainty management: The present invention introduces a two-stage adaptive configuration model. The first stage focuses on resource capacity configuration, and the second stage considers the dynamics and uncertainties of operation regulation, making the resource configuration plan more adaptable and robust, and effectively coping with the uncertainties brought by the volatility of new energy; ensuring the stable operation of the system in complex scenarios such as extreme weather and prediction deviations, and reducing the risk of wind and light abandonment rates and load curtailment; 4. Multi-scenario recognition and differential configuration: Based on the flexible feature index vector (such as ramp rate, frequency modulation ability), dynamically determine the demand scenario, and differentially configure resources according to different scenarios (one or more), avoiding the difficulty of a single configuration method in adapting to the changing and complex grid operation conditions; 5. Data-driven prediction and improvement of configuration accuracy: The present invention uses SVM (Support Vector Machine) and cross-validation method to fit the wind power, photovoltaic, and load power prediction models, combines historical data with meteorological response regression analysis, significantly improves the prediction accuracy, provides highly reliable inputs for multi-scenario recognition and resource configuration, and supports the refined operation of the power system.

[0016] In the present invention, the above technical solutions can also be combined with each other to achieve more preferred combination schemes. Other features and advantages of the present invention will be described in the subsequent specification, and some advantages will be obvious from the specification, or can be understood by implementing the present invention. The objectives and other advantages of the present invention can be realized and obtained through the content specifically pointed out in the specification and the drawings. Description of the Drawings

[0017] The drawings are only for the purpose of showing specific embodiments and are not considered as limiting the present invention. Throughout the drawings, the same reference signs represent the same components; Figure 1 It is a flowchart of a method for adaptively configuring flexible resources based on multi-scenario fusion in an embodiment of the present invention. Detailed Embodiments

[0018] The preferred embodiments of the present invention will be specifically described below in conjunction with the accompanying drawings. The accompanying drawings form a part of this application and are used together with the embodiments of the present invention to explain the principles of the present invention, rather than to limit the scope of the present invention.

[0019] Starting from multiple types of power interaction demand scenarios, the present invention proposes a method for adaptively configuring flexibility resources based on multi-scenario fusion. First, multi-scenario characteristic indicators are introduced to quantitatively describe multi-dimensional demands such as power and electricity balance and new energy consumption, a scenario demand identification model is constructed, and the coupling relationship of multi-scenario demands is established; secondly, the scenario identification results are aggregated, and the most severe demand scenario is screened. Finally, based on the flexibility requirement characteristics of the most severe scenario, considering the uncertainties at both the source and load ends, a two-stage adaptive configuration model for flexibility resources is constructed; through this method, the uncertainty problem brought by the volatility of new energy can be solved, and the coordinated configuration and differential configuration of more efficient flexibility resources for multi-scenario fusion can be realized by dynamically adapting to the derivative scenario demands to cope with complex power interaction scenario demands.

[0020] A specific embodiment of the present invention discloses a method for adaptively configuring flexibility resources based on multi-scenario fusion, as Figure 1 shown, including the following steps: Step S1: Obtain the predicted values of wind power, photovoltaic power, and load power of the regional power grid; Step S2: Construct a scenario identification objective function with the minimum weighted sum of the flexibility characteristic indicators of the power balance scenario, peak shaving scenario, and frequency modulation scenario as the objective, solve the scenario identification objective function based on the scenario identification constraints to obtain a flexibility characteristic indicator vector; make a judgment based on the flexibility characteristic indicator vector to obtain the corresponding scenario identification result; Step S3: If the scenario identification result is one or more of the power balance scenario, peak shaving scenario, and frequency modulation scenario, construct a two-stage adaptive configuration model with the minimum capacity configuration of flexibility resources on each side of the source, grid, load, and storage and the minimum control loss of flexibility resources on each side as the objective; solve based on the scenario identification result and the first-stage constraints to obtain the configuration scale of flexibility resources; Step S4: Solve based on the wind power, photovoltaic power, and load power predicted values, the newly added flexibility resource configuration scale, and the second-stage deterministic and uncertainty constraints to obtain the flexibility resource operation plan corresponding to the flexibility resource configuration scale in the scenario.

[0021] Step S1 includes steps S11 - S13.

[0022] Step S11: Obtain the historical data of wind power, photovoltaic power, and load power within the regional power grid.

[0023] Since the output of wind and solar power highly depends on meteorological conditions (such as wind speed, solar irradiance, temperature, etc.), it has significant nonlinearity, volatility, and uncertainty. On the load side, there are obvious "daily" and "weekly" cycle patterns in the time dimension, and it is also affected by various external factors such as temperature, humidity, and holidays. Therefore, building a perception module for wind and solar resources and load characteristics can achieve the prior characterization of available power sources and power demands in future time periods, providing data support for resource allocation, dispatching strategy generation, reserve capacity arrangement, etc.

[0024] Calculate the correlation coefficients between the output power of wind and solar and meteorological factors based on historical wind and solar output data and meteorological data. The data shows that the output power of wind power is related to wind speed, wind direction, and temperature; the output power of photovoltaic power is related to total solar irradiance and temperature; the change rules of load power consumption are also positively correlated with temperature, humidity, and precipitation. According to the above relevant rules, construct a wind-solar and load sample set to obtain the digital characteristics of the source and load.

[0025] Use the historical extreme weather load data and corresponding meteorological data in the region to obtain the sample data of wind power, photovoltaic power, and various types of user loads.

[0026] Construction of wind power sample data is as follows: Formula (1) Where, represents the wind power sample data; 、 、 are respectively the wind speed, wind direction, and air temperature at time is the historical wind power output at time is the total number of observation periods.

[0027] Construction of photovoltaic sample data is as follows: Formula (2) Where, represents the photovoltaic sample data; 、 are respectively the total irradiance and temperature at time is the historical photovoltaic output at time

[0028] Construction of load sample data is as follows: Formula (3) Where, represents the load sample data; 、 、 are respectively The change in temperature, humidity, and precipitation at a moment relative to a moment; is the change in historical load at a moment relative to a moment; and is the user type.

[0029] Step S12 further includes constructing a source-load model; fitting and solving the source-load model based on the historical data of wind power, photovoltaic power, and the power of various types of users in the regional power grid to obtain the predicted values of wind power, photovoltaic power, and load power.

[0030] On the basis of constructing the sample data between the wind power, photovoltaic power, and load data of the regional power grid and the key meteorological variables, mining the response laws and correlations of the sample data in the time series dimension, and establishing the mapping relationship between the output of wind power, photovoltaic power, and load power and environmental factors.

[0031] Wind power mainly fluctuates with the change of wind speed and direction; photovoltaic power is jointly affected by irradiance and temperature; load power shows the superposition response characteristics to the changes of temperature, humidity, and precipitation.

[0032] By fitting these source-load sample data, the accurate prediction of the availability of wind and light and the change trend of load demand can be realized, providing a basic support for the subsequent operation scenario recognition, regulation demand characterization, and configuration optimization of flexible resources.

[0033] Constructing a source-load model, fitting and solving the source-load model based on the historical data of wind power, photovoltaic power, and the power of various types of users in the regional power grid to obtain the predicted values of wind power, photovoltaic power, and load power, includes: Obtaining the load sample data of wind power, photovoltaic power, and various types of users in the historical cycle of the regional power grid, and performing preprocessing to obtain the preprocessed load sample data of wind power, photovoltaic power, and various types of users; Using the SVM and cross-validation method, fitting and solving the source-load model based on the load sample data of wind power, photovoltaic power, and various types of users to obtain the predicted values of wind power, photovoltaic power, and load power.

[0034] The source-load model includes a wind power model, a photovoltaic power model, and a load power model; Using the wind power sample data to fit and solve the wind power model to obtain the predicted value of wind power; Using the photovoltaic power sample data to fit and solve the photovoltaic power model to obtain the predicted value of photovoltaic power; Using the load sample data of each type of user to perform data fitting and solution on the load power model respectively, the load power prediction values corresponding to each type of user are obtained; the load power prediction values of each type of user are integrated with the historical load mean value to obtain the load power prediction value.

[0035] Wind power model , as follows: Formula (4) Wherein, , , are respectively the proportionality coefficients of the wind power increment caused by the changes in meteorological factors such as wind speed, wind direction, and temperature at time represents the wind power fitting error term.

[0036] Exemplarily, takes values from 0.4 to 0.8; takes values from -0.1 to 0.1; is from 0.01 to 0.05; takes values from 0.1 to 0.3; it can be changed according to specific requirements.

[0037] Photovoltaic power model , as follows: Formula (5) Wherein, , respectively represent the proportionality coefficients of the photovoltaic power increment caused by the changes in meteorological factors such as total irradiance and temperature at time represents the photovoltaic power fitting error term.

[0038] Exemplarily, takes values from 0.6 to 0.9; takes values from -0.002 to -0.005; takes values from 0.05 to 0.2; it can be changed according to specific requirements.

[0039] Load power model , as follows: Formula (6) Wherein, is the total load of the th type of user at time under extreme weather; is the load prediction value of the th type of user at time assuming no extreme weather occurs; is the load increment at the th type of user at a certain moment, which is related to various meteorological factors such as temperature, humidity, and precipitation; 、 and are respectively the changes in actual temperature, humidity, and precipitation under extreme weather conditions at a certain moment; 、 and respectively represent the proportionality coefficients of the electricity consumption increment generated by the changes in the meteorological factors of temperature, humidity, and precipitation for the th type of user at a certain moment; ε is the error term, representing the consumption difference that cannot be reflected by the changes in the meteorological factors of temperature, humidity, and precipitation.

[0040] Exemplarily, (1) Temperature change coefficient : Represents the impact of a unit temperature change (such as 1 °C) on the load increment.

[0041] Residential users: 0.05 - 0.15 (when the temperature rises by 1 °C, the air-conditioning load increases by 5% - 15%); Commercial users: 0.03 - 0.10 (the temperature sensitivity is slightly lower than that of residential users); Industrial users: 0.01 - 0.05 (production equipment is not sensitive to temperature changes); For special scenarios: In the case of extreme high temperature (> 35 °C), it may increase to 0.2 - 0.3; In cold regions (< 0 °C), the heating demand is significant, and it may be negative (such as -0.1 - -0.2).

[0042] (2) Humidity change coefficient : Represents the impact of a unit humidity change (such as 1%) on the load increment.

[0043] General range: 0.005 - 0.02 (when the humidity increases by 1%, the load slightly increases by 0.5% - 2%); In high-humidity regions: 0.01 - 0.03 (such as during the rainy season, dehumidification equipment is required and the load increases significantly).

[0044] (3) Precipitation change coefficient : Represents the impact of a unit precipitation change (such as 1 mm) on the load increment.

[0045] General range: 0.001 - 0.01 (when the precipitation increases by 1 mm, the load slightly increases by 0.1% - 1%); Rainstorm scenario: It may increase to 0.02 - 0.05 (e.g., a sharp increase in the demand for drainage pumps and emergency power). Arid regions: It may be negative (a decrease in precipitation leads to a decrease in the agricultural irrigation load).

[0046] In the present invention, extreme weather is exemplarily defined as follows. When any one of the following occurs, extreme weather has occurred: (1) Temperature: Beyond the statistical range of local historical temperature data for the same period, such as high temperature above 35°C or low temperature below -10°C; (2) Humidity: The relative humidity is higher than 80% or lower than 20% for a long time; (3) Precipitation: The precipitation within 24 hours exceeds 50 mm (e.g., rainstorm), or extreme precipitation events such as heavy snow occur.

[0047] (4) Wind speed: When the wind speed reaches the storm range of 17 - 25 m / s, the wind turbine frequently triggers overspeed protection or derating operation.

[0048] (5) Irradiance: Under the influence of rainstorm, thunderstorm, sandstorm, etc., the irradiance can drop sharply from 800 - 1000 W / m² on sunny days to 100 - 300 W / m², or even lower than 100 W / m², accompanied by a minute - level fluctuation amplitude exceeding 500 W / m², resulting in a "saw - tooth - shaped" rapid drop in the photovoltaic output.

[0049] For the user load types, exemplarily, they include industrial, commercial, residential, agricultural, and public service users.

[0050] Due to the differences in user load types (industrial, commercial, and residential), as well as various factors such as geographical location, climate environment, economy, industry, and population, there are differences in the coefficient and parameter values in the formula.

[0051] Obtain the digital characteristics of the power source and load, and construct the sample data of wind power, photovoltaic power, and load 、 、 ; Establish the mapping relationship between the power output and environmental factors.

[0052] Step S13: Train the power source - load model and conduct wind power, photovoltaic power, and various types of load forecasting.

[0053] After constructing the power source - load model, use SVM (Support Vector Machine) for training. During the training process, the cross - validation method is adopted to select the optimal parameters. Then, use the trained parameters to predict the power source - load, and finally obtain the wind power prediction value 、photovoltaic power prediction value 、load power prediction value .

[0054] (1) Preprocess the wind power, photovoltaic, and load sample data, including missing value filling and data normalization; Missing value filling: Use linear interpolation or the mean value of adjacent time periods to fill in the missing values; Data normalization: Normalize the features with different dimensions such as wind speed, irradiance, and temperature; Normalize the load data separately according to user types to avoid the influence of magnitude differences on the model performance.

[0055] (2) Train the source-load model.

[0056] Configure the SVM, set the task type to a regression task; select the radial basis function (RBF) as the kernel function, which is suitable for non-linearity; Hyperparameters, including: (a) Penalty coefficient , with a value range of [0.1, 100], which controls the model's tolerance for errors; (b) Kernel parameter , which controls the width of the RBF kernel, with a value range of [0.001, 0.1]; (c) Insensitive loss function , set to 5% of the predicted target dimension (e.g., in load forecasting, = 0.05 × maximum load value).

[0057] Adopt k-fold cross-validation. Exemplarily, k is equal to 10; Divide the sample data set into 1 subset as the validation set, and the remaining 9 as the training set; For each combination of hyperparameters ( , , ), calculate the average validation error (such as mean squared error, MSE); select the combination of hyperparameters with the minimum validation error as the final model parameters.

[0058] The training process includes: Input the sample data as follows: Wind power sample data: wind speed, wind direction, air temperature , ,

[0059] Photovoltaic sample data: irradiance, temperature ,

[0060] Load sample data: changes in temperature, humidity, and precipitation , , , historical load change , and the corresponding user type .

[0061] Train the SVMs for wind power, photovoltaic power, and various types of users respectively: (a) Train independent SVM models for wind power and photovoltaic power respectively; (b) Train independent SVM models for the load according to user types respectively, and then output the total load prediction through weighted integration.

[0062] (3) Call the trained SVM models to obtain the predicted values of wind power output, photovoltaic power output, and load increment predictions for various types of users; integrate the load increment prediction results for various types of users into the load prediction value as follows: Load prediction value = Base load + Load increment prediction values for various types of users Formula (7) Among them, the base load is the mean value of historical load data.

[0063] Use the support vector machine method to train each influencing feature factor. During the training process, the cross-validation method is used to select the optimal parameters. Then use the trained parameters to predict the source load, and finally obtain the predicted value of wind power , the predicted value of photovoltaic power and the predicted value of load power .

[0064] The function of step S1 is to obtain the historical data of the regional power grid and construct a source-load model, use methods such as SVM for data fitting and training, and finally obtain the predicted values of wind power, photovoltaic power, and load power, providing data support for subsequent scenario recognition and flexible resource allocation.

[0065] Step S2 includes steps S21 - S22.

[0066] Step S21: Construct a scenario recognition model.

[0067] The present invention proposes a power grid multi-scenario configuration recognition mechanism. To quantify the requirements of each scenario, by introducing multiple scenario feature indicators as slack variables, identify scenarios such as power balance, peak shaving, and frequency modulation existing in the power grid operation.

[0068] Under the large-scale access of new energy, affected by the volatility and intermittency of wind and light, the requirements for power balance, peak shaving, and frequency modulation scenarios are derived. And different resources have different capabilities to provide power balance, peak shaving, and frequency modulation for the power grid. Under the current existing resource settings, more flexible resources need to be invested to balance the requirements of different scenarios.

[0069] To identify the most prominent demand scenarios in the current power grid, the scenario recognition model aims to minimize the weighted sum of the new resource flexibility characteristic indicators. By introducing six new resource flexibility characteristic indicators, namely the maximum output power, minimum output power, upward ramp rate, downward ramp rate, conventional frequency regulation ability, and fast frequency regulation ability, a multi-scenario demand recognition model is constructed. Then, under the condition of minimizing the investment in new resources, the most severe demand scenarios are identified.

[0070] The objective function of scenario recognition is to minimize the weighted average of the new resource flexibility indicators, that is, to minimize the increase in the resource flexibility adjustment parameters required to balance the scenario demands.

[0071] Construct a scenario recognition model, which includes a scenario recognition objective function and scenario recognition constraints.

[0072] The scenario recognition objective function is expressed as follows: Equation (8) Where 、 、 are the flexibility characteristic indicators of the power balance scenario, peak shaving scenario, and frequency modulation scenario respectively; 、 、 are respectively 、 、 weights; 、 are the maximum and minimum output powers of the flexibility resources required by the scenario respectively; 、 are the rated values of the maximum and minimum output powers respectively; 、 are the upward and downward ramp rates of the flexibility resources required by the scenario respectively; 、 are the maximum values of the upward and downward ramp power rates of the flexibility resources respectively; 、 are the conventional frequency regulation ability and fast frequency regulation ability indicators of the flexibility resources required by the scenario respectively; are the upper and lower reserve regulation abilities of the conventional frequency regulation ability respectively; The scenario recognition constraints include conventional power balance constraints, conventional non-adjustable unit operation constraints, wind power generation constraints, transmission line constraints, flexibility regulation resource operation constraints, load full-load constraints, spinning reserve capacity constraints, and fast frequency regulation reserve constraints.

[0073] 、 、 weights ; Exemplarily, , , The values are as follows and can be revised according to specific requirements.

[0074] Power balance scenario dominant,

[0075] Peak shaving pressure scenario Prominent scenario:

[0076] Frequency modulation scenario Priority: .

[0077] Scenario recognition constraints are as follows: (1) Conventional system power balance constraint Formula (9) Where, is the output of the th conventional non-adjustable unit at time ; is the output of the photovoltaic unit at time; is the output of the wind turbine unit at time; is the output of the nd external transmission line at time; is the output of the flexibility regulation unit at time; , , , are the numbers of conventional non-adjustable units, external transmission lines, flexibility regulation units, and loads respectively; is the load demand of the load node at time; is the load shedding power at time.

[0078] Formula (9) is an indirect constraint condition of the objective function formula (8). Power balance affects the output of units, and the output of each unit corresponds to various flexibility regulation indicators. The minimum weighted sum of flexibility regulation indicators is the scenario recognition objective function.

[0079] Conventional non-adjustable units refer to generator sets with fixed output power that are difficult to quickly adjust according to the real-time needs of the power system. These units are processed according to a pre-set plan and rarely participate in the real-time power regulation of the system; for example, traditional thermal power generator sets, nuclear power generator sets, and hydraulic generator sets. A flexible unit refers to a generator set that can quickly adjust its output power according to the grid demand, aiming to help the grid better cope with load fluctuations and the uncertainty of new energy power generation, and improve the stability and flexibility of the power system; it has fast regulation ability, flexible start-stop, and can quickly respond to the changes in grid load, including short-term regulation from minute level to hour level.

[0080] Flexible resources on the source side include: gas turbine generator sets, pumped storage units; Flexible regulation resources on the grid side refer to the power transmission between regional power grids; Flexible regulation resources on the load side refer to electric vehicles, air-conditioning loads, big data centers, etc.; Flexible regulation resources on the storage side refer to new energy storage devices such as electrochemical energy storage and flywheel energy storage.

[0081] Load shedding power at a certain moment It refers to the load power reduced at a certain moment due to the operation needs or emergencies of the power system, which is used to reduce part of the load through manual or automatic control means to maintain the stable operation of the power system when the power system cannot meet all load demands.

[0082] (2) Operation constraints of conventional non-adjustable units: Formula (10) Among them, , are respectively the minimum and maximum power generation of the th conventional non-adjustable unit.

[0083] (3) Wind power generation constraints: Formula (11) Among them, , are respectively the minimum and maximum power generation of the wind turbine unit.

[0084] (4) Photovoltaic power generation constraints: Formula (12) Among them, , are respectively the minimum and maximum power generation of the photovoltaic unit.

[0085] (5) Transmission line constraints: Formula (13) Among them, , are respectively the minimum and maximum transmission power of the th line; Indicates the safety margin coefficient of the th line.

[0086] Exemplarily, the safety margin coefficient takes values between 0 and 1.

[0087] Indicates the maximum transmission power limit. At any moment, the actual transmission power of the transmission line cannot exceed times its maximum transmission power; the line cannot operate at full load and must leave a certain margin; , indicates the minimum transmission power limit. At any moment, the actual transmission power of the transmission line cannot be lower than times its maximum transmission power, that is, the transmission power of the line cannot reverse and exceed this limit.

[0088] (6) Operational constraints of flexible regulation resources: Equation (14) Wherein, is the start-stop flag of the th type of flexible regulation resource at the moment. The values of 1 and 0 indicate that the unit is in the enabled and disabled states respectively; is the minimum output of the th type of flexible regulation unit; is the maximum output of the th type of flexible regulation unit; , are respectively the minimum output power and the maximum output power of the flexible regulation unit to be newly added.

[0089] (7) Load full-load constraint: Equation (15) Wherein, is the maximum load full-load ratio. Exemplarily, takes values between 0.8 and 0.95. The specific value is determined according to the actual operation requirements and load characteristics of the power grid. This ratio is used to limit the upper limit of the load, prevent system overload, and leave a certain margin to cope with emergencies and ensure the safe and stable operation of the system.

[0090] (8) Ramping constraint: Equation (16) Wherein, is the output of the flexible regulation unit at the moment; , are the maximum uphill climbing and maximum downhill climbing rates respectively; and are respectively time, the start-stop status of the th type of flexibility resource at time; and are respectively the regulation capabilities for the required resource to climb uphill and downhill; and are respectively the durations of uphill climbing and downhill climbing.

[0091] (9) Spinning reserve capacity constraint Equation (17) wherein, is the proportion of spinning reserve capacity in the total load; is the proportion of spinning reserve capacity in the installed capacity of renewable energy; and are respectively the upward and downward spinning regulation capabilities of the flexibility regulation unit to be newly added; is the quantity of renewable energy; is the output of renewable energy at time.

[0092] Exemplarily, takes values between 0.05 and 0.1; takes values between 0.1 and 0.2; adjusted according to specific grid operation requirements and the access scale of renewable energy.

[0093] (10) Fast frequency regulation reserve constraint: Equation (18) wherein, is the output of the th type of flexibility regulation resource at time; is the climbing rate of the th type of flexibility regulation resource; represents the frequency regulation duration; is the fast spinning reserve capacity of the flexibility regulation unit to be newly added; and are respectively the frequency regulation capacity ratios of the load and renewable energy.

[0094] Exemplarily, and take values between 0.05 and 0.15; specific values need to be adjusted according to the frequency regulation requirements of the grid, load characteristics, and the access scale of renewable energy to ensure the stability and reliability of the system during the frequency regulation process.

[0095] Step S22: Solve the scenario recognition objective function based on the scenario recognition constraints to obtain a flexibility feature index vector; make a judgment based on the flexibility feature index vector to obtain the corresponding scenario.

[0096] According to the scenario recognition objective function and scenario recognition constraints set in the scenario recognition model, solve the defined flexibility feature index. Solve to obtain the maximum output power , the minimum output power , the upward ramp rate , the downward ramp rate , the conventional frequency regulation capacity , the fast frequency regulation capacity .

[0097] To further optimize the allocation of flexibility resources for the most prominent scenarios in different regions of the power grid at present, on the basis of solving each flexibility feature index, construct a flexibility feature index vector, as shown in formula (18). If the flexibility feature index vector is a non-zero vector, it means that there is one or more scenarios. Further judge which flexibility feature index is non-zero, which represents the existence of the scenario corresponding to this index.

[0098] Based on the current power grid parameters, use the gurobi solver to solve the scenario recognition objective function based on the scenario recognition constraints to obtain ; Take to form the flexibility feature index vector , as follows: Formula (19) Make a judgment based on the flexibility feature index vector to obtain the corresponding scenario, including: If or is non-zero, it is determined that a power balance scenario occurs; If or is non-zero, it is determined that a peak shaving scenario occurs; If or is non-zero, it is determined that a frequency regulation scenario occurs.

[0099] The current power grid parameters include the installed capacity data of current wind turbines, photovoltaic units, non-adjustable units, adjustable resources and load data; input these data into the scenario recognition model and use the gurobi solver in matlab for solution.

[0100] If all variables in the flexibility feature index vector are zero, then steps S3 - S4 do not need to be executed.

[0101] The function of step S2 is to identify the most prominent demand scenarios (power balance scenario, peak shaving scenario, and frequency modulation scenario) of the power grid by constructing a scenario recognition model and solving the flexibility feature index vector, providing a basis for subsequent differential resource collaborative configuration.

[0102] Step S3 includes steps S31 - S32.

[0103] Step S31: Construct a two - stage adaptive configuration model.

[0104] To achieve systematic and scientific flexibility resource allocation based on scenario recognition, a two - stage adaptive configuration model integrating capacity configuration and operation regulation is constructed. The two - stage adaptive configuration model aims to evaluate the capacity deployment pressure and dynamic control cost of various types of source - grid - load - storage flexibility resources in the power system over the full cycle, ensuring that the configuration plan can take into account physical feasibility and operation regulation performance.

[0105] The two - stage adaptive configuration model includes a two - stage objective function with the minimum capacity configuration of flexibility resources on each side of the source - grid - load - storage and the minimum control loss of flexibility resources on each side, the first - stage constraints, and the second - stage constraints; The first - stage constraints include configuration scale constraints, power demand constraints, ramp - up demand constraints, and frequency regulation demand constraints; The second - stage constraints include second - stage deterministic constraints and second - stage uncertainty constraints; The second - stage deterministic constraints include flexibility resource constraints on the source side, grid side, load side.

[0106] In the two - stage adaptive configuration model, in the first stage, through capacity configuration parameters, the capacity deployment pressure and investment cost of various types of flexibility resources under multiple nodes and multiple types are measured; In the second stage, according to the time scale of typical periods, the dynamic regulation load and operation loss of various resources are introduced to quantify the real - time loss and regulation intensity of resources during the operation stage. The two - stage objective function realizes the efficient configuration of flexibility resources and the guarantee of operation adaptability through the comprehensive optimization of capacity configuration and operation regulation.

[0107] The two - stage objective function is as follows: Formula (20) Among them, 、 、 、 are the unit capacity configuration parameters of source, grid, load, and storage resources respectively; is the configuration scale of source, grid, load, and storage resources; 、 The time scales are the whole year and typical day, respectively; , , , They are the control losses of resources on the source, network, load and storage sides respectively; Provide various flexible resource categories for source, network, load and storage; is the load uncertainty set.

[0108] , The time scales are the whole year and the typical day respectively. The selection of typical days in the present invention is based on the historical data of the whole year's load and wind power and photovoltaic light output. The daily curves are classified by the K-means clustering analysis method, and the date with the largest peak-to-valley difference is selected as the typical day for research.

[0109] (1) Source-side flexibility resource control loss ,as follows: Formula (21) in, , , They are the constant term, first-order term, and second-order term control loss coefficients of the power supply side flexibility resources (coal power, gas power), respectively; For the Source-side flexibility resources The power generated during the period; , Indicates fuel consumption and emissions the cost; , Respectively represent Fuel consumption rate of flexible resources on the power supply side and Emission rate; Indicates the total number of flexibility resources.

[0110] Taking coal-fired power as an example, , , The values are as follows: Constant term : Usually in the range of 100~500 yuan / MW, it is used to characterize the fixed costs of coal-fired power flexibility regulation, such as equipment maintenance, personnel on duty, etc.; One-time item : Generally in the range of 5~15 yuan / MW², reflecting the loss cost proportional to the power generation during coal-fired power regulation, mainly related to factors such as fuel consumption; Quadratic term :In the range of approximately 0.1 - 0.5 yuan / MW³, it reflects the loss cost related to the square of the power generation, which is caused by the reduction in equipment operation efficiency and fuel utilization efficiency during the regulation process of coal-fired power generation.

[0111] (2) Grid-side flexibility resource control loss , as follows: Formula (22) Wherein, represents the transmission power of the th inter-network line during the , , respectively represent the voltage deviation cost, the active and reactive operation and maintenance costs of the line; , , respectively represent the boundaries of the voltage, active power and reactive power of each node and line; represents the total number of inter-network lines. represents the reference voltage level; , respectively represent the starting and ending nodes connected at both ends of the line.

[0112] (3) Load-side flexibility resource control loss , as follows: Formula (23) Wherein, , are respectively the unit compensation costs of the time-shiftable and interruptible load response modes of the th type of load-side flexibility resource; , are respectively the response load quantities of the time-shiftable and interruptible modes of the th type of load-side flexibility resource; is the total number of load-side flexibility resources.

[0113] (4) Energy storage-side flexibility resource control loss , as follows: Formula (24) Wherein, , , are respectively the daily maintenance cost, the unit cost in the charging state and the unit cost in the discharging state of the th type of energy storage resource; , are respectively the th type of energy storage resource Charging and discharging power during a period; is the total number of flexibility regulation resources on the energy storage side.

[0114] Exemplarily, The types of energy storage resources include lithium battery energy storage, flow battery energy storage, sodium-sulfur battery energy storage, lead-acid battery energy storage, supercapacitor energy storage, flywheel energy storage, and compressed air energy storage categories.

[0115] The first-stage constraints include configuration scale constraints, power demand constraints, ramp demand constraints, and frequency regulation demand constraints, as follows: (1) Configuration scale constraints, as follows: Equation (25) Among them, is the total scale limit of configurable flexibility resources; is the configuration scale of the newly added flexibility resources of the is the total configuration scale of the newly added flexibility resources; is the unit configuration parameter of the flexibility resources on the energy storage side; is the unit configuration parameter of the flexibility resources on the source side; is the unit configuration parameter of the flexibility resources on the load side; is the unit configuration parameter of the flexibility resources on the grid side.

[0116] (2) Power demand constraints: The power demand constraints are as follows: Equation (26) Among them, , , , are the numbers of conventional non-adjustable units, external transmission lines, flexibility regulation units, and loads, respectively; is the output of the th conventional non-adjustable unit at time is the output of the photovoltaic unit at time; is the output of the wind turbine at time; is the output of the th external transmission line at time is the output of the flexibility regulation unit at time; is the regulation power of the newly added flexibility resources; is the load node at Load demand at a moment; is Load shedding power at a moment.

[0117] 、 are respectively the adjustable power of flexibility resources on the newly added source side and network side at moment; 、 are respectively the power generation and charging power of the newly added storage side resources at moment; 、 、 are respectively the transfer-in power, transfer-out power and interruptible load power of the transferable load of the newly added load side resources at moment; 、 、 、 are respectively the quantities of newly added flexibility resources on the source, network, load and storage sides.

[0118] (3)Ramp demand constraint, as follows: Formula (27) Among them, 、 、 、 、 、 、 、 , as follows: Formula (28) Among them, 、 、 、 are respectively the upward ramp ability coefficients that can be provided by the resources on each unit source, network, load and storage side; 、 、 、 are respectively the downward ramp ability coefficients that can be provided by the resources on each unit source, network, load and storage side; is the adjustable power of flexibility resources of the newly added source side at moment; is the adjustable power of flexibility resources of the newly added network side at moment; is the transfer-out power of the transferable load of the newly added load side resources at moment; is the transfer-in power of the transferable load of the newly added load side resources at moment; is the newly added storage side resources at Power generation at a certain moment; For the newly added energy storage side resources at Charging power at a certain moment.

[0119] (4) Frequency regulation demand constraint, as follows: Equation (29) Wherein, 、 、 、 、 、 , as follows: Equation (30) Wherein, 、 、 、 Are the configuration scales of resources on each side of the source, grid, load, and energy storage respectively; 、 、 Are the conventional frequency regulation capacity coefficients that can be provided by unit resources on each side of the source, grid, load, and energy storage respectively; 、 、 Are the fast frequency regulation capacity coefficients that can be provided by unit resources on each side of the source, grid, load, and energy storage respectively; Is the enabled status of grid-side flexibility resources, 1 means enabled, 0 means disabled; Is the maximum power generation of various types of units on the source side; Is the maximum transmissible power of grid-side flexibility resources; Is the maximum chargeable power of energy storage-side flexibility resources; 、 、 Are the upper limits of the ramping rates of flexibility regulation resources on the source, grid, and energy storage sides respectively; 、 、 Are the continuously adjustable frequency regulation times of flexibility regulation resources on the source, grid, and energy storage sides.

[0120] The second-stage constraints include second-stage deterministic constraints and second-stage uncertainty constraints; the second-stage deterministic constraints include source-side, grid, load, and side flexibility resource constraints.

[0121] Second-stage deterministic constraints: (1) Source-side flexibility regulation resource constraint, as follows: Equation (31) Equation (32) Wherein, 、 are the flexibility resources on the source side respectively , are the power generation powers at , are the minimum and maximum power generation powers of various types of units on the source side respectively; , are the maximum upward and downward ramping capabilities of various types of units on the source side respectively.

[0122] (2) Constraints on flexibility resources on the grid side are as follows: Equation (33) where , are the minimum and maximum values of the power transmission of the external transmission line respectively.

[0123] (3) Constraints on flexibility resources on the load side include interruptible load constraints and shiftable load constraints, as follows: a) Interruptible load constraints are as follows: Equation (34) Equation (35) b) Shiftable load constraints are as follows: Equation (36) Equation (37) where , , are respectively at the transfer-in power of the shiftable load, the transfer-out power of the shiftable load, and the interruptible load power of the -th newly added flexibility resources on the load side; is at the transfer-out power of the shiftable load of the -th newly added flexibility resources on the load side; , , are the maximum interruptible load ratio, the maximum continuous time interruptible load ratio, and the maximum shiftable load ratio respectively; , are the 0-1 state variables of the transferred-out load and the transferred-in load respectively. The variable being 1 indicates enabling this state, and the variable being 0 indicates disabling this state.

[0124] (4) Constraints on flexibility regulation resources on the storage side are as follows: Equation (38) Formula (39) where and are the minimum and maximum discharge powers of the energy storage side flexibility resources, respectively; and are the minimum and maximum charging powers of the energy storage side flexibility resources, respectively; is the 0-1 enabling flag for discharging of the energy storage side flexibility resources, 1 indicates enabling the discharging condition of the energy storage side flexibility resources, and 0 indicates disabling the discharging condition of the energy storage side flexibility resources; is the 0-1 enabling flag for charging of the energy storage side flexibility resources, 1 indicates enabling the charging condition of the energy storage side flexibility resources, and 0 indicates disabling the charging condition of the energy storage side flexibility resources; and are the energy of the energy storage side flexibility resources in time periods and respectively; and are the working efficiencies of charging and discharging, respectively; and are the charging and discharging powers in time period respectively; is the scheduling time step; represents the rated capacity of the energy storage side flexibility resources; is the SOC (State of Charge) of the energy storage side flexibility resources in time period ; is the available SOC of the energy storage side flexibility resources; represents the minimum available SOC of the energy storage side flexibility resources; is the ratio of the maximum energy difference between the start and end of a day of the energy storage side flexibility resources to the rated capacity; and are the initial and terminal energy of the energy storage side flexibility resources, respectively.

[0125] The second-stage uncertainty constraint is as follows: Formula (40) where represents the uncertain variable; and and are the photovoltaic, wind power output and electrical load power after considering uncertainty, respectively; and and are the predicted values of wind power, photovoltaic and load power, respectively; and and They are the maximum allowable fluctuation coefficients for wind power, photovoltaic power output, and electrical load power, respectively.

[0126] Among them, represents an uncertain variable that can take any value in the uncertainty set ; , and are the photovoltaic power output, wind power output, and electrical load power after considering uncertainty, respectively; , , are the predicted wind power output, photovoltaic power output, and electrical load power, respectively; , , represent the maximum allowable fluctuation coefficients for wind power, photovoltaic power output, and electrical load power, respectively.

[0127] Step S32: Solve based on the scenario recognition result and the first-stage constraints to obtain the scale of flexible resource allocation.

[0128] The two-stage adaptive configuration model is transformed as follows: Formula (41) Among them, represents the configuration parameters of each flexible resource; represents the set of decision variables for the new capacity of various flexible resources ( - the scale of each type of resource), and the number of new flexible resources on the source, grid, load, and storage sides , , , ; represents the set of uncertain variables, including the wind and light power output and load at each node ( , , ), represents the set of decision variables, including the wind and solar power generation power, the power generation power of conventional non-adjustable units, the power generation power of original flexible regulation units, the inter-network transmission power, and the regulation power of new flexible resources on the source, grid, load, and storage sides; represents the first-stage constraints; represents the second-stage deterministic constraints; represents the second-stage uncertainty constraints.

[0129] Input various grid parameters of the source-grid-load-storage in the current scenario, including historical meteorological data, historical wind and solar power generation data, historical load data, inter-network transmission power data, the proportion of controllable load, conventional non-adjustable units, the installed capacity of various flexible adjustment units, and the technical parameters of flexible resources on each side of the source-grid-load-storage (including the constant term, first-order term, and second-order coefficient of the control loss function on each side; coal consumption rate; carbon dioxide emission rate, etc.).

[0130] The configured scale of the flexible resources includes the newly added flexible resource scale on the source, grid, load, and storage sides 、 、 and ; The function of step S3 is to construct a two-stage adaptive configuration model integrating capacity configuration and operation regulation based on the identified scenario, and solve to obtain the configured scale of flexible resources on each side of the source-grid-load-storage based on the constraints in the first stage 、 、 、 。

[0131] Step S4, specifically

[0132] Based on the configured scale of various resources of the source-grid-load-storage 、 、 、 , based on the control loss parameters of various resources, wind and solar power output, and load power uncertainty data (including based on the predicted values of wind power, photovoltaic power, and load power), and the deterministic and uncertainty constraints in the second stage, optimize and solve to obtain the operation plans of various newly added flexible resources of the source-grid-load-storage 、 、 、 、 、 、 。

[0133] The operation plan of the flexible resources corresponding to the configured scale of the flexible resources includes the regulation power of the newly added flexible resources on the source side 、the regulation power of the newly added flexible resources on the grid side 、the transfer-out of the transferable load of the newly added load-side resources 、the interruptible load power of the transferable load of the newly added load-side resources 、the power generation power of the newly added storage-side resources 、the charging power of the newly added storage-side resources and the transfer-in of the transferable load of the newly added load-side resources 。

[0134] The function of step S4 is to formulate the specific operation plan of the newly added flexibility resources on each side of the source, network, load, and storage through optimization based on the determined flexibility resource allocation scale, various resource control loss parameters, and the uncertainty data of wind-solar power output and load power, in combination with the deterministic and uncertainty constraints in the second stage.

[0135] In summary, a flexibility resource adaptive allocation method based on multi-scenario fusion in an embodiment of the present invention has the following beneficial effects: 1. Multi-scenario fusion optimization and differential allocation ability: By constructing a multi-scenario recognition model (power balance scenario, peak regulation scenario, and frequency modulation scenario), the present invention comprehensively considers the flexibility characteristic indexes of the power balance scenario, peak regulation scenario, and frequency modulation scenario, solves the limitation of single-scenario optimization in the prior art, realizes differential resource adaptation under multi-scenario coupling, and significantly improves the dynamic response ability of the power system to new energy fluctuations (such as wind-solar intermittency and load mutation); 2. Global collaborative coupling optimization of multiple types of resources on the source, network, load, and storage: The present invention designs a two-stage adaptive allocation model (capacity allocation + operation regulation), integrates the dynamic coupling relationships of resources on the source side (thermal power, energy storage), network side (transmission line), load side (interruptible / transferable load), and storage side (energy storage), breaks through the bottleneck of single-resource independent optimization in traditional research, realizes full-cycle collaborative optimization across time scales and resource types, improves the overall regulation efficiency, realizes the collaborative optimization allocation of flexibility resources on each side of the source, network, load, and storage, improves resource utilization efficiency, and overcomes the defect of focusing on single-resource independent optimization in the prior art; 3. Dynamic adaptation and uncertainty management: The present invention introduces a two-stage adaptive allocation model. The first stage focuses on resource capacity allocation, and the second stage considers the dynamics and uncertainty of operation regulation, making the resource allocation scheme more adaptable and robust, and effectively coping with the uncertainty brought by new energy volatility; ensuring the stable operation of the system in complex scenarios such as extreme weather and prediction deviation, and reducing the risk of wind and light abandonment rate and load curtailment; 4. Multi-scenario recognition and differential allocation: Dynamically determine the demand scenario based on the flexibility characteristic index vector (such as ramp rate, frequency modulation ability), and allocate resources differentially according to different scenarios (one or more), avoiding the difficulty of a single allocation method in adapting to the changing and complex grid operation conditions; 5. Data-driven prediction and improvement of configuration accuracy: The present invention uses SVM (support vector machine) and cross-validation method to fit the wind power, photovoltaic, and load power prediction models, and combines historical data with meteorological response regression analysis to significantly improve the prediction accuracy, provide highly reliable input for multi-scenario recognition and resource allocation, and support the refined operation of the power system.

[0136] Those skilled in the art can understand that all or part of the processes of implementing the methods of the above embodiments can be completed by instructing relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. Among them, the computer-readable storage medium is a disk, an optical disc, a read-only memory or a random access memory, etc.

[0137] As described above, only the specific and preferred embodiments of the present invention are provided, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention.

Claims

1. A flexible resource adaptive configuration method based on multi-scenario fusion, characterized in that Including: Obtaining the predicted values of wind power, photovoltaic power, and load power of the regional power grid; Constructing a scenario recognition objective function with the goal of minimizing the weighted sum of the flexibility characteristic indicators of the power balance scenario, peak shaving scenario, and frequency modulation scenario, and solving the scenario recognition objective function based on the scenario recognition constraints to obtain a flexibility characteristic indicator vector; judging based on the flexibility characteristic indicator vector to obtain the corresponding scenario recognition result; If the scenario recognition result is one or more of the power balance scenario, peak shaving scenario, and frequency modulation scenario, constructing a two-stage adaptive configuration model with the goal of minimizing the capacity configuration of flexibility resources on each side of the source-network-load-storage and the control loss of flexibility resources on each side; solving based on the scenario recognition result and the first-stage constraints to obtain the scale of flexibility resource configuration; Solving based on the predicted values of wind power, photovoltaic power, and load power, the newly added flexibility resource configuration scale, and the second-stage deterministic and uncertainty constraints to obtain the flexibility resource operation plan corresponding to the flexibility resource configuration scale in the scenario.

2. The flexible resource adaptive configuration method based on multi-scenario fusion according to claim 1, characterized in that It also includes constructing a source-load model; fitting and solving the source-load model based on the historical data of wind power, photovoltaic power, and the power of various types of users in the regional power grid to obtain the predicted values of wind power, photovoltaic power, and load power, including: Obtaining the load sample data of wind power, photovoltaic power, and various types of users in the historical cycle of the regional power grid, and performing preprocessing to obtain the preprocessed load sample data of wind power, photovoltaic power, and various types of users; Using the SVM and cross-validation method, fitting and solving the source-load model based on the load sample data of wind power, photovoltaic power, and various types of users to obtain the predicted values of wind power, photovoltaic power, and load power.

3. The flexible resource adaptive configuration method based on multi-scenario fusion according to claim 2, wherein The source-load model includes a wind power model, a photovoltaic power model, and a load power model; Fitting and solving the wind power model using the wind power sample data to obtain the predicted value of wind power; Fitting and solving the photovoltaic power model using the photovoltaic sample data to obtain the predicted value of photovoltaic power; Fitting and solving the load power model using the load sample data of various types of users respectively to obtain the predicted load power values corresponding to various types of users; integrating the predicted load power values of various types of users with the historical load average value to obtain the predicted load power value.

4. The flexible resource adaptive configuration method based on multi-scenario fusion according to claim 1, characterized in that The scenario recognition objective function is expressed as follows: ； Among them, , , are the flexibility characteristic indicators of the power balance scenario, the peak shaving scenario, and the frequency modulation scenario, respectively; , , are respectively , , 's weights; , are respectively the maximum and minimum output powers of the flexibility resources required by the scenario; , are respectively the rated values of the maximum and minimum output powers; , are respectively the upward and downward ramp rates of the flexibility resources required by the scenario; , are respectively the maximum values of the upward and downward ramp power rates of the flexibility resources; , are respectively the conventional frequency modulation ability and fast frequency modulation ability indicators of the flexibility resources required by the scenario; are respectively the upper standby and lower standby regulation capabilities of the conventional frequency modulation ability; The scenario recognition constraints include conventional power balance constraints, operation constraints of conventional non-adjustable units, wind power generation constraints, transmission line constraints, operation constraints of flexibility regulation resources, load full-load constraints, spinning reserve capacity constraints, and fast frequency modulation reserve constraints.

5. The flexible resource adaptive configuration method based on multi-scenario fusion according to claim 4, wherein, Based on the current power grid parameters, the gurobi solver is used to solve the scenario recognition objective function based on the scenario recognition constraints, and obtain ; To form the flexibility characteristic index vector , as follows: ； Judging based on the flexibility characteristic indicator vector to obtain the corresponding scenario, including: If or is non-zero, it is determined that a power balance scenario occurs; If or is non-zero, it is determined that a peak shaving scenario occurs; If or is non-zero, it is determined that a frequency modulation scenario occurs.

6. The flexible resource adaptability configuration method based on multi-scenario fusion according to claim 1, wherein The two-stage adaptive configuration model includes a two-stage objective function with the goal of minimizing the capacity configuration of flexibility resources on each side of the source-network-load-storage and the control loss of flexibility resources on each side, first-stage constraints, and second-stage constraints; The first-stage constraints include configuration scale constraints, power demand constraints, ramping demand constraints, and frequency regulation demand constraints; The second-stage constraints include second-stage deterministic constraints and second-stage uncertainty constraints; The second-stage deterministic constraints include source-side, grid, load, and side flexibility resource constraints.

7. The flexible resource adaptability configuration method based on multi-scenario fusion according to claim 6, characterized in that The two-stage objective function is as follows: ； Among them, , , , are the unit capacity configuration parameters of the source, grid, load, and storage resources respectively; is the configuration scale of the source, grid, load, and storage resources; , are the time scales of the whole year and the typical day respectively; , , , are the control losses of the resources on each side of the source, grid, load, and storage respectively; is the type of flexible resources for the source, grid, load, and storage; is the set of load uncertainties.

8. The flexible resource adaptability configuration method based on multi-scenario integration according to claim 6, characterized in that, The second-stage uncertainty constraint , is as follows: ； Among them, represents an uncertain variable; , , are the photovoltaic power output, wind power output and electrical load power after considering uncertainty, respectively; , , are the predicted values of wind power, photovoltaic power and load power, respectively; , , are the maximum allowable fluctuation coefficients of wind power output, photovoltaic power output and electrical load power, respectively.

9. The flexible resource adaptability configuration method based on multi-scenario fusion according to claim 8, characterized in that The scale of flexible resource allocation includes the newly added scale of flexible resources on the source, grid, load, and storage sides , , and ; The flexibility resource operation plan corresponding to the scale of flexibility resource allocation includes the regulated power of newly added flexibility resources on the source side , the regulated power of newly added flexibility resources on the network side , the transferred-out of the transferable load of newly added load-side resources , the interruptible load power of the transferable load of newly added load-side resources , the generated power of newly added energy storage-side resources , the charging power of newly added energy storage-side resources and the transferred-in of the transferable load of newly added load-side resources .

10. The flexible resource adaptive configuration method based on multi-scenario fusion according to any one of claims 6-9, characterized in that The power demand constraint is as follows: ； Among them, , , , are the quantities of conventional non-adjustable units, external transmission lines, flexibility adjustment units, and loads, respectively; is the output of the th conventional non-adjustable unit at time ; is the output of the photovoltaic unit at time; is the output of the wind turbine unit at time; is the output of the th external transmission line at time; is the output of the flexibility adjustment unit at time; is the adjustment power of the newly added flexibility resource; is the load demand of the load node at time; is the load shedding power at time.

Images