A virtual power plant scheduling method and system

Abstract

This invention discloses a virtual power plant scheduling method and system in the field of power system dispatching technology. The method includes: acquiring operational data from the distribution network and the virtual power plant; determining the equivalent carbon emission intensity of new energy sources based on the curtailment status of these units, taking a negative value when curtailment exists and zero otherwise; embedding the equivalent carbon emission intensity into the node carbon emission flow calculation to obtain an improved node carbon emission factor; and generating a scheduling scheme based on the factor through a two-layer collaborative scheduling model, where the upper and lower layers of the model interact iteratively by exchanging power and the factor, and introducing dual control constraints on total carbon emissions and intensity. This invention quantifies the marginal emission reduction benefits of absorbing new energy sources during curtailment periods, forming differentiated incentive signals to guide the virtual power plant to actively promote new energy absorption; simultaneously, by rolling correction of the carbon emission factor in two stages (day-ahead and intraday), the carbon emission constraints are dynamically matched with the actual operating status, improving the accuracy and economy of scheduling decisions.

Description

Technical Field

[0001] This invention relates to the field of power system dispatching technology, and more specifically, to a dispatching method and system for a virtual power plant. Background Technology

[0002] With the continuous increase in the installed capacity of new energy sources such as wind power and photovoltaics, the problem of wind and solar curtailment is becoming increasingly prominent. Virtual power plants, as a new type of regulatory resource that aggregates distributed power sources, energy storage, and adjustable loads, play a crucial role in promoting the consumption of new energy and ensuring supply-demand balance. The carbon emission factor is a core parameter connecting dispatch decisions and carbon emission accounting. Existing methods, based on carbon emission flow theory, propagate carbon emissions from the generation side to load nodes along the power flow direction, where the carbon emission intensity of new energy generation is always set to zero, passively reducing the node carbon emission factor only through dilution effects. This approach has good applicability in traditional power systems dominated by thermal power.

[0003] However, in scenarios with a high proportion of renewable energy connected to the grid, the aforementioned existing technologies have the following shortcomings:

[0004] First, the carbon emission reduction contribution achieved by virtual power plants during periods of curtailment through proactive adjustment behaviors such as increasing energy storage charging or shifting loads cannot be quantitatively reflected in carbon emission accounting. This results in the dispatch model lacking effective optimization incentives for virtual power plants to proactively promote the consumption of new energy, thus restricting the effectiveness of low-carbon coordinated dispatch between virtual power plants and the distribution network.

[0005] Second, once the carbon emission factor is determined based on the day-ahead forecast parameters, it is not adjusted during the day. When there is a deviation in the day-ahead forecast, the scheduling scheme based on the fixed carbon emission factor cannot reflect the true carbon emission level, resulting in a disconnect between carbon emission constraints and actual operating conditions.

[0006] Therefore, there is an urgent need to provide a collaborative scheduling method that can quantify the carbon emission reduction benefits of virtual power plant absorption behavior and dynamically update carbon emission factors according to operating status, in order to solve the above problems. Summary of the Invention

[0007] To overcome the shortcomings of existing technologies, this invention provides a scheduling method and system for virtual power plants.

[0008] The technical solution of this invention is as follows:

[0009] A method for scheduling a virtual power plant includes the following steps:

[0010] Acquire operational data of a power system, which includes a distribution network and virtual power plants connected to the distribution network;

[0011] The equivalent carbon emission intensity of new energy is determined based on the curtailment status of new energy units. During periods when curtailment occurs, the equivalent carbon emission intensity of new energy is negative, and during periods when curtailment does not occur, it is zero.

[0012] By embedding the equivalent carbon emission intensity of the new energy source into the calculation of the nodal carbon emission flow of the distribution network, an improved nodal carbon emission factor is obtained.

[0013] Based on the improved node carbon emission factor, a scheduling scheme for the virtual power plant is generated through a two-layer collaborative scheduling model of the virtual power plant and the distribution network. The two-layer collaborative scheduling model includes an upper layer and a lower layer, and the upper layer and the lower layer iteratively interact by exchanging power and the improved node carbon emission factor.

[0014] Preferably, the absolute value of the equivalent carbon emission intensity of the new energy source is positively correlated with the curtailment rate and the carbon emission intensity of marginal thermal power units; the curtailment rate is the ratio of curtailed power to the total output of new energy sources; the marginal thermal power unit refers to the unit with the highest carbon emission intensity among thermal power units that are in operation and whose actual output is greater than their minimum technical output.

[0015] Preferably, the formula for calculating the equivalent carbon emission intensity of the new energy source is:

[0016] ;

[0017] in, For time period The equivalent carbon emission intensity of new energy sources For time period Carbon emission intensity per unit of power generation of marginal thermal power units in the system For time period The power of abandoned electricity, For time period Total output of new energy sources.

[0018] Preferably, the improved formula for calculating the node carbon emission factor is as follows:

[0019] ;

[0020] in, For nodes exist Improved node carbon emission factors over time periods For access nodes A collection of thermal power units, For the unit Carbon emission intensity per unit of electricity generation For the unit exist The amount of effort contributed during a given period; For access nodes A collection of new energy generating units, For the first Taiwan New Energy Equivalent carbon emission intensity over a given period This will contribute to the actual absorption of this new energy source; To the node The set of upstream nodes for output power. upstream node Improved carbon emission factor For the node To the node The active power flow.

[0021] Preferably, the scheduling scheme for generating the virtual power plant includes a day-ahead scheduling phase and an intraday scheduling phase; in the intraday scheduling phase, the improved node carbon emission factor is rolled over based on the actual intraday operating data, and the corrected carbon emission factor is used as the carbon emission constraint input for the intraday scheduling model.

[0022] Preferably, the process of rolling correction of the improved node carbon emission factor includes:

[0023] Distribution network side correction: The day-ahead forecast is replaced by the actual output of new energy sources, the actual amount of abandoned electricity, and the actual output of thermal power units. The improved nodal carbon emission factor is recalculated to obtain the distribution network side correction value.

[0024] Virtual power plant side correction: Calculate the change in node carbon potential caused by the difference between the actual exchange power of the virtual power plant and the day-ahead planned exchange power, and use it as the virtual power plant side correction amount;

[0025] Comprehensive correction: The correction value on the distribution network side is superimposed with the correction amount on the virtual power plant side to obtain the final carbon emission factor for the day.

[0026] Preferably, the rolling correction adopts a dual-mode triggering mechanism: including a regular correction mode executed at a fixed cycle, and an event correction mode triggered when the deviation between the actual output of new energy and the day-ahead forecast exceeds a first predetermined threshold, or the deviation between the actual abandoned power and the day-ahead forecast exceeds a second predetermined threshold.

[0027] Preferably, the two-layer collaborative scheduling model includes a day-ahead two-layer optimization model and an intraday two-layer rolling optimization model;

[0028] The upper layer of the day-ahead two-layer optimization model aims to minimize the total operating cost of the distribution network and includes carbon emission constraints. The lower layer of the day-ahead two-layer optimization model aims to minimize the internal operating cost of the virtual power plant and optimizes the output and regulation of the resources inside the virtual power plant based on the power exchange plan determined by the upper layer.

[0029] The intraday two-layer rolling optimization model is re-solved based on the rolled-corrected carbon emission factor during the intraday phase, and deviation constraints are set with respect to the day-ahead scheduling plan.

[0030] Preferably, the intraday dual-layer rolling optimization model adopts a rolling window optimization architecture and sets energy storage state of charge deviation correction constraints and adjustable load intraday response constraints.

[0031] A second aspect of the present invention provides a scheduling system for a virtual power plant, comprising a data acquisition module, an equivalent carbon emission intensity determination module, a node carbon emission factor calculation module, and a scheduling scheme generation module, wherein:

[0032] The data acquisition module is used to acquire the operating data of the power system, which includes a distribution network and virtual power plants connected to the distribution network.

[0033] The equivalent carbon emission intensity determination module is used to determine the equivalent carbon emission intensity of new energy based on the curtailment status of new energy units. In the period when curtailment occurs, the equivalent carbon emission intensity of new energy is negative, and in the period when curtailment does not occur, it is zero.

[0034] The node carbon emission factor calculation module is used to embed the equivalent carbon emission intensity of the new energy into the calculation of the node carbon emission flow of the distribution network to obtain an improved node carbon emission factor.

[0035] The scheduling scheme generation module is used to generate a scheduling scheme for the virtual power plant based on the improved node carbon emission factor through a two-layer collaborative scheduling model of the virtual power plant and the distribution network. The two-layer collaborative scheduling model includes an upper layer and a lower layer, and the upper layer and the lower layer iteratively interact by exchanging power and the improved node carbon emission factor.

[0036] The scheduling method and system for a virtual power plant based on the above scheme have the following advantages:

[0037] 1. This invention assigns an equivalent carbon emission intensity to new energy sources based on the state of power curtailment: a negative value is taken during curtailment periods and zero during non-curtailment periods; this feature causes the nodal carbon emission factor to decrease due to the negative contribution during curtailment periods, while remaining consistent with traditional methods during non-curtailment periods; when virtual power plants increase energy storage charging or load absorption during curtailment periods, the carbon emission accounting value of the electricity consumed is even lower or even negative, forming a differentiated economic incentive in the scheduling target, and quantifying the carbon emission reduction contribution of absorption behavior;

[0038] 2. The absolute value of the equivalent carbon emission intensity of this invention is positively correlated with the curtailment rate and the carbon intensity of marginal thermal power units: the higher the curtailment rate and the higher the carbon intensity of marginal units, the larger the absolute value of the negative value. After embedding the equivalent carbon emission intensity into the nodal carbon flow formula, the output of new energy sources reduces the carbon potential of the connected nodes and their downstream nodes. The more severe the curtailment, the more significant the decrease in carbon potential. The incentive signal for virtual power plants to actively absorb curtailment during peak curtailment periods is thus strengthened, guiding them to concentrate their regulation resources on the period with the greatest emission reduction benefits.

[0039] 3. In the two-layer collaborative scheduling model of the present invention, the upper layer and the lower layer interact iteratively by exchanging power and improving the node carbon emission factor; the upper layer distribution network incorporates carbon emission cost into the objective function, and the active absorption behavior of the lower layer virtual power plant directly reduces the carbon emission cost term by reducing the node carbon potential, so that the carbon emission reduction contribution is quantitatively fed back in the optimization objective, thereby driving the virtual power plant to spontaneously select the adjustment strategy that is conducive to the absorption of new energy.

[0040] 4. This invention establishes a two-stage architecture of day-ahead and intraday. In the intraday stage, the carbon emission factor is rolled over based on the actual output of new energy sources, the actual amount of abandoned electricity, and the actual output of thermal power. A comprehensive approach is adopted, which combines the correction on the distribution network side and the correction on the virtual power plant side. This ensures that the carbon potential gradually approaches the true value as the operating information is updated, and that the intraday carbon emission constraints are consistent with the actual operating status, avoiding decision-making errors caused by day-ahead forecast deviations. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0042] Figure 1 A flowchart illustrating the scheduling method for a virtual power plant provided by this invention;

[0043] Figure 2 This is a schematic diagram of the overall structure of the virtual power plant scheduling method provided by the present invention. Detailed Implementation

[0044] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0045] It should be noted that when a component is referred to as "fixed," "set," or "connected" to another component, it may be located directly or indirectly on that other component. The terms "upper," "lower," "left," "right," "front," "rear," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or position based on the accompanying drawings, and are for ease of description only, and should not be construed as limiting the technical solution. The terms "first," "second," etc., are used for ease of description only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features. "Many" means two or more, unless otherwise explicitly specified. "Several" means one or more, unless otherwise explicitly specified.

[0046] See Figures 1 to 2 This embodiment provides a scheduling method for a virtual power plant, including the following steps:

[0047] S1. Obtain the operation data of the power system, wherein the power system includes a distribution network and virtual power plants connected to the distribution network;

[0048] S2. Determine the equivalent carbon emission intensity of new energy sources based on the curtailment status of new energy units. In the period when curtailment occurs, the equivalent carbon emission intensity of new energy sources is negative, and in the period when curtailment does not occur, it is zero.

[0049] S3. Embed the equivalent carbon emission intensity of the new energy into the calculation of the node carbon emission flow of the distribution network to obtain the improved node carbon emission factor.

[0050] S4. Based on the improved node carbon emission factor, a scheduling scheme for the virtual power plant is generated through a two-layer collaborative scheduling model of the virtual power plant and the distribution network. The two-layer collaborative scheduling model includes an upper layer and a lower layer, and the upper layer and the lower layer iteratively interact by exchanging power and the improved node carbon emission factor.

[0051] In some embodiments, during step S1, the present invention divides the aggregated multi-source distributed resources of the virtual power plant into three categories, and establishes operating parameter models for each type of resource. Wherein:

[0052] The first category is distributed power generation, including distributed photovoltaic and decentralized wind power. This will be integrated into the virtual power plant. Taiwan distributed power supply The actual output during the time period is expressed as Its output is constrained by natural resource conditions and satisfies:

[0053]

[0054] in, for The power output limit for the time period is determined based on meteorological conditions. When the system's absorption capacity is insufficient, distributed power sources can be subject to power curtailment, with the curtailed power expressed as... ,satisfy .

[0055] The second category is energy storage systems, including electrochemical energy storage and other forms of user-side energy storage. This will be integrated into the virtual power plant. Taiwan Energy Storage The charging and discharging power during a period of time is expressed as ,in Indicates discharge. Indicates charging, and charging power Discharge power , The operational constraints of the energy storage system include charge / discharge power constraints and state of charge constraints:

[0056]

[0057]

[0058] in, and These are the maximum charging power and the maximum discharging power, respectively. and These represent the upper and lower limits of the state of charge. The time-series recursive relationship of the energy storage state of charge (1) is as follows:

[0059] (1)

[0060] in, and These are charging efficiency and discharging efficiency, respectively. This refers to the rated capacity of the energy storage.

[0061] The third category is adjustable loads, including continuously adjustable loads, time-shiftable adjustable loads, and discretely adjustable loads. This will be applied to the virtual power plant. One adjustable load user The actual load power during the time period is expressed as Its baseline load power is expressed as Then the user is The adjustment formula (2) for the time period is:

[0062] (2)

[0063] in, This indicates a load reduction (upward adjustment). This indicates an increase in load (downward adjustment). The adjustment amount is constrained by the adjustment capacity and the duration.

[0064]

[0065]

[0066] in, for Maximum adjustable capacity for a given time period This represents the maximum number of continuous adjustment periods.

[0067] Through the above classification modeling, the operational characteristics of the aggregated resources of the virtual power plant are expressed in a unified parametric function form, forming a set of resource parameters.

[0068] In some embodiments, before performing step S2, it is necessary to establish a propagation calculation model of the carbon emission factor of traditional distribution network nodes based on carbon emission flow theory, as a benchmark model for subsequent improvements to the carbon potential method. The specific establishment process is as follows:

[0069] For containing Taiwan thermal power units and In a distribution network system with n nodes, carbon emissions propagate from the generation side to the load side along the power flow direction. The carbon emission factor of node n in time period t is defined as the carbon flow density flowing into that node, and is calculated according to the following formula (3):

[0070] (3)

[0071] in, For access nodes A collection of thermal power units; For the unit The carbon emission intensity per unit of electricity generation, expressed in tCO2 / MWh; For the unit exist The amount of effort contributed during a given period; To the node The set of upstream nodes for output power; upstream node Carbon emission factors; For the node To the node Active power flow. For inflow nodes The carbon emission intensity per unit of electricity consumption at a given node is obtained by power-weighted averaging of all carbon flows.

[0072] In the traditional model of the aforementioned construction, the carbon emission intensity of new energy units (wind power and photovoltaic) is set to zero, that is... In carbon flow calculations, renewable energy output does not carry carbon emission information; it only generates a dilution effect by increasing the denominator, passively reducing the carbon emission factor of the connected nodes. Therefore, it has the following limitations:

[0073] First, it is impossible to distinguish between the two operating states: full consumption of new energy and the existence of power curtailment.

[0074] Second, it overlooks the marginal emission reduction benefits of absorbing new energy sources during periods of power curtailment;

[0075] Third, the responsibility for carbon emissions rests solely with the power generation side.

[0076] Therefore, this invention connects the virtual power plant as a controllable aggregate to the distribution network, establishing a topological coupling relationship and a carbon flow transfer model between the two. Specifically:

[0077] Representing the distribution network as a topology diagram ,in For a set of nodes, This is a set of branch lines. The virtual power plant connects to nodes. Exchange power with the distribution network. Define a virtual power plant in... The exchange power between the time period and the distribution network is ,in This indicates that the virtual power plant inputs active power into the distribution network. This indicates that the virtual power plant absorbs active power from the distribution network.

[0078] The output and load of various resources within the virtual power plant are jointly determined, and its expression (4) is:

[0079] (4)

[0080] in, For distributed power sources within a virtual power plant, It is a collection of controllable distributed power sources such as gas turbines. A collection of energy storage systems, Number of adjustable load users For controllable distributed power sources Active power output during a given time period. Each item represents the actual power absorbed by the distributed power source, the power output of the controllable power source, the charging and discharging power of energy storage, and the power consumed by the load.

[0081] In the carbon emission flow propagation model, the virtual power plant access node The carbon flow relationship needs to distinguish between power injection and absorption states. When When the virtual power plant injects power into the distribution network, the carbon emission intensity carried by the injected power is determined by the internal power generation structure of the virtual power plant, as expressed in expression (5):

[0082] (5)

[0083] in, Carbon emission intensity per unit of electricity generation for controllable distributed power sources. For the first The equivalent carbon emission intensity of Taiwan's new energy source. The injected carbon emission intensity. As an access node The input for carbon flow calculations participates in the propagation calculation of carbon emission flows in the distribution network.

[0084] when At that time, the virtual power plant absorbs power from the distribution network, and the carbon emission intensity carried by the absorbed power is equal to the carbon emission factor of the access node. It is determined by the carbon flow propagation results on the distribution network side.

[0085] In some embodiments, the process of performing step S2 includes: the present invention defining the equivalent carbon emission intensity of new energy sources. , used to characterize in The equivalent carbon emission signal generated by absorbing one unit of renewable energy electricity during a given time period. The equivalent carbon emission intensity is determined based on whether the system experiences power curtailment, as follows:

[0086] First, obtain Power curtailment during the time period Total output of new energy Curtailed power is the sum of curtailed wind power and curtailed solar power; total output includes both the consumed portion and the curtailed portion.

[0087] Then, determine Determine whether there is any power wastage in the system during a given period, and calculate the equivalent carbon emission intensity based on the determination results.

[0088] when When there is no power curtailment in the system and the output of new energy sources has been fully absorbed, the equivalent carbon emission intensity is taken as zero.

[0089]

[0090] when At this time, the absolute value of the equivalent carbon emission intensity of the new energy is positively correlated with the curtailment rate and the carbon emission intensity of the marginal thermal power units; the curtailment rate is the ratio of curtailed power to the total output of new energy; that is, there is curtailment in the system, and absorbing curtailed power can replace thermal power output and generate emission reduction benefits. At this time, the equivalent carbon emission intensity is calculated according to the following formula (6):

[0091] (6)

[0092] in, For time period The equivalent carbon emission intensity of new energy sources For time period Carbon emission intensity per unit of power generation of marginal thermal power units in the system For time period The power of abandoned electricity, For time period Total output of new energy sources.

[0093] Secondly, the marginal thermal power units are identified. Thermal power units that are in operation during a given period and whose output has not dropped to their minimum technical output are defined as the set of adjustable units. Within the set of adjustable generating units, the carbon emission intensity per unit of power generation for each unit is considered. The units with the highest carbon emission intensity are selected as marginal units, ranked from highest to lowest, and their carbon emission intensity is determined by the following formula (7):

[0094] (7)

[0095] The basis for selecting the operating unit with the highest carbon emission intensity as the marginal unit is that when the system increases the consumption of new energy, the high-carbon unit with the largest adjustment margin will be replaced first, and its carbon emission intensity is the upper limit of the marginal emission reduction benefit generated by consuming new energy.

[0096] Through the above methods, the equivalent carbon emission intensity of new energy sources Zero values ​​are taken during non-curtailment periods, consistent with traditional methods; negative values ​​are taken during curtailment periods, and the higher the curtailment rate and the greater the marginal unit carbon emission intensity, the greater the absolute value of the equivalent carbon emission intensity, and the stronger the incentive signal for renewable energy consumption in the scheduling model.

[0097] When the power curtailment rate approaches 1, Approaching To achieve maximum incentive intensity;

[0098] When the power curtailment rate approaches 0, Approaching 0, it degenerates into the traditional model processing method.

[0099] Furthermore, in step S3, based on the definition of the equivalent carbon emission intensity of new energy, this invention embeds the equivalent carbon emission intensity into the carbon emission flow propagation model of the distribution network and establishes an improved method for calculating the nodal carbon emission factor.

[0100] By adding a carbon flow contribution term representing the equivalent carbon emission intensity carried by new energy output to the numerator and adding a term representing the actual absorption output of new energy to the denominator, an improved node carbon emission factor is formed, and the expression (8) is:

[0101] (8)

[0102] in, For nodes exist Improved node carbon emission factors over time periods For access nodes A collection of thermal power units, For the unit Carbon emission intensity per unit of electricity generation For the unit exist The amount of effort contributed during a given period; For access nodes A collection of new energy generating units, For the first Taiwan New Energy Equivalent carbon emission intensity over a given period This will contribute to the actual absorption of this new energy source; To the node The set of upstream nodes for output power. upstream node Improved carbon emission factor For the node To the node The active power flow.

[0103] The core difference between the improved formula and the traditional carbon emission factor formula lies in the fact that the output of new energy sources no longer participates in the carbon flow calculation with zero carbon intensity, but instead carries a definite... Equivalent carbon emission intensity. Since the equivalent carbon emission intensity is negative during the curtailment period, the output of new energy sources makes a negative contribution to the carbon flow, thus lowering the carbon potential of the connected nodes and their downstream nodes.

[0104] Improve node carbon emission factor It has the following two characteristics:

[0105] First, during periods when there is no power wastage. The contribution of new energy carbon flow in the improved formula is zero, and the improved formula degenerates into the original traditional calculation formula. That is, the improved method and the traditional method are fully compatible in the scenario of no power curtailment.

[0106] Second, during the period of power curtailment, New energy sources contribute negative carbon emission intensity to carbon flow propagation, thus lowering the carbon intensity of the nodes they connect to. carbon emission factors It propagates downstream along the power flow direction. The higher the curtailment rate and the greater the marginal unit carbon emission intensity, the more significant the decrease in carbon potential.

[0107] Based on the above characteristics, when the virtual power plant exchanges power through the above formula (4) during the power curtailment period... Increasing power absorption (such as increasing energy storage charging or transferring adjustable loads to this period) will result in lower or even negative carbon emission accounting values, thereby creating economic incentives in subsequent collaborative scheduling optimization models and guiding virtual power plants to actively participate in the consumption of new energy.

[0108] For virtual power plant access nodes When a virtual power plant injects power into the distribution network, the carbon emission intensity carried by that injected power... The carbon emission intensity of new energy sources is calculated according to the method described in equation (5). The method described in Equation (6) is used to confirm the unified calculation of carbon flow within the virtual power plant and carbon flow in the distribution network.

[0109] In some embodiments, the process of performing step S4 includes:

[0110] S4.1 Generate a scheduling scheme for the virtual power plant;

[0111] S4.2 VPP-distribution network two-layer coordinated dispatch taking into account carbon emission constraints.

[0112] Specifically, during the execution of step S4.1, the node carbon emission factor is improved. The calculation of carbon emission factors relies on input parameters such as renewable energy output, curtailed electricity, and thermal power unit output. In actual dispatching and operation, the accuracy of these parameters differs significantly between the day-ahead and intraday phases: only predicted values ​​are available during the day-ahead phase, while actual operating values ​​are gradually obtained during the intraday phase. If the day-ahead predicted parameters are consistently used to calculate the carbon emission factor, the carbon emission constraints during the intraday operating phase will deviate from reality due to prediction errors. Therefore, this invention establishes a rolling correction method for carbon emission factors in both the day-ahead and intraday phases, enabling the carbon emission factor to gradually approach the true value as operating information is updated. Specifically, it includes the following steps:

[0113] S4.1.1 Generation of day-ahead carbon emission factor baseline values;

[0114] S4.1.2, Intraday rolling correction of carbon emission factors;

[0115] S4.1.3, Modify constraints and triggering mechanisms.

[0116] Specifically, in the execution of step S4.1.1, the present invention uses the day-ahead forecast information as input and, based on the improved carbon potential calculation method, generates the carbon emission factor benchmark value for each node and time period of the next day, which serves as the carbon emission constraint input for the day-ahead collaborative scheduling model and as the benchmark reference for intraday correction.

[0117] First, day-ahead forecast parameters are obtained. During the day-ahead scheduling phase, all forecast parameters related to the calculation of carbon emission factors are collected to form a day-ahead forecast parameter set. This forecast parameter set includes the following four types of data:

[0118] The first category is the current-day renewable energy output forecast. Based on meteorological forecast data (irradiance, wind speed, temperature, etc.) and a new energy power output prediction model, it represents the predicted active power output of each new energy unit in each time period of the next day, with a time granularity of 1 hour, covering 24 time periods of the next day.

[0119] The second category is the prediction of day-ahead power curtailment. Based on the supply and demand balance analysis of day-ahead renewable energy output forecasts and day-ahead load forecasts, this represents the renewable energy output that is not expected to be absorbed by the system in each period. When the forecasted renewable energy output exceeds the system's absorption capacity... ,otherwise .

[0120] The third category is the daytime thermal power unit combination and output plan. Based on the unit combination optimization results obtained earlier, this indicates the planned active power output of each thermal power unit in each time period of the following day, and also determines the set of thermal power units in operation in each time period. .

[0121] The fourth category is day-ahead load forecasting. Based on historical load data, weather forecasts, and calendar features, it represents the predicted load power of each node in the distribution network at each time period on the following day.

[0122] Secondly, the day-ahead equivalent carbon emission intensity of renewable energy is calculated. This is based on the obtained day-ahead curtailment forecast. Compared with recent new energy power output forecasts Calculate the day-ahead equivalent carbon emission intensity of new energy sources for each time period according to the above formula (6). Specifically, for During the period, ;for The time period is primarily based on the current day's thermal power unit assembly. Determine the carbon emission intensity of marginal thermal power units in the near term Then calculate using the following formula (9):

[0123] (9)

[0124] Then, substituting the aforementioned day-ahead forecast parameters into the improved node carbon emission factor formula of equation (8), the carbon emission factors for each node are calculated. At different times The current day's baseline carbon emission factor:

[0125]

[0126] in, The calculation process covers all nodes and 24 time periods, forming a daily carbon emission factor baseline matrix.

[0127] Specifically, during the execution of step S4.1.2, the rolling correction of the improved nodal carbon emission factor includes: distribution network side correction: replacing the day-ahead forecast value with the actual output of new energy sources, the actual amount of power curtailed, and the actual output of thermal power units, and recalculating the improved nodal carbon emission factor to obtain the distribution network side correction value; virtual power plant side correction: calculating the change in nodal carbon potential caused by the difference between the actual exchange power of the virtual power plant and the day-ahead planned exchange power, as the virtual power plant side correction amount; comprehensive correction: superimposing the distribution network side correction value and the virtual power plant side correction amount to obtain the final intraday carbon emission factor.

[0128] In practice, during the intraday scheduling phase, this invention employs a rolling window optimization architecture, using a 15-minute time granularity to perform rolling corrections and comprehensive calculations of carbon emission factors within a 4-hour rolling window, gradually bringing the carbon emission factors closer to the actual operating values. The correction process consists of three stages: distribution network-side correction, virtual power plant-side correction, and comprehensive correction.

[0129] First, the carbon emission factor on the distribution network side is corrected. During intraday operation, as the operation progresses, the actual output of new energy sources... Actual abandoned electricity With the actual output of thermal power units This data can be obtained from the operation monitoring system. Substitute the above actual operating data for the day-ahead forecast parameters in step S4.1.1, and substitute it into the improved carbon potential formula of equation (8) to calculate the corrected carbon emission factor on the distribution network side:

[0130]

[0131] For the time period that has already occurred within the scrolling window ( In the above formula, the operating parameters are taken as actual values; for periods that have not yet occurred ( The operating parameters are estimated forward using ultra-short-term forecast values.

[0132] Secondly, carbon emission factor correction is performed on the virtual power plant side. The adjustment behavior of resources within a virtual power plant (such as adjusting energy storage charging and discharging power, and adjusting load transfer) alters the power injection at its access nodes, thereby affecting the local carbon flow distribution. A virtual power plant-side correction amount is defined. The change in node carbon potential caused by the difference between the actual regulation behavior of the virtual power plant and the day-ahead plan:

[0133]

[0134] in, For virtual power plants Actual switching power during the time period This refers to the planned exchange power for the day-ahead period. When the virtual power plant increases charging or load absorption during periods of curtailment (… When more power is absorbed from the distribution network, This indicates that the absorption behavior of virtual power plants further reduces the node carbon potential, and their emission reduction contribution is quantified through this correction.

[0135] Then, a comprehensive correction is performed. The correction results from the distribution network side are superimposed with the correction amounts from the virtual power plant side to obtain the final intraday carbon emission factor, formula (10):

[0136] (10)

[0137] It comprehensively reflects the dual impact of changes in the actual power generation structure on the distribution network side and the active adjustment behavior on the virtual power plant side on the carbon emission factor, and serves as the carbon emission constraint input for the intraday two-layer rolling optimization model.

[0138] Furthermore, in step S4.1.3, to ensure the rationality and computational efficiency of the intraday carbon emission factor correction, this invention sets constraints on the correction process and designs a dual-mode triggering mechanism.

[0139] The modified constraints include the following two items.

[0140] The first constraint is the correction range. The deviation between the intraday corrected carbon emission factor and the previous day's baseline value must not exceed a set threshold to avoid abnormal fluctuations in the carbon emission factor due to extreme prediction errors, which could affect the stability of the scheduling optimization results.

[0141]

[0142] in, This represents the maximum permissible deviation ratio. The deviation ratio is determined based on the system's renewable energy penetration rate and prediction accuracy level, and is generally taken as... This means that the intraday correction cannot exceed 30% of the previous day's baseline value. If the correction calculation result exceeds the above constraint, Truncate to the constraint boundary value.

[0143] The second constraint is a non-negative lower bound. The improved carbon potential may become negative under extreme curtailment scenarios, representing the net emission reduction effect generated by absorbing new energy sources during that period. However, to avoid physically unreasonable extreme negative values, a lower bound is set for the carbon emission factor:

[0144]

[0145] in, This represents the unit emission value of the thermal power unit with the highest carbon emission intensity in the system. The constraint ensures that the negative value of the carbon potential does not exceed the emission intensity of the unit with the highest carbon emission, and has a clear physical meaning—that is, the maximum marginal emission reduction benefit of absorbing new energy sources does not exceed the emission level of the unit with the highest carbon emission that is being replaced.

[0146] In fact, the rolling correction adopts a dual-mode triggering mechanism, including:

[0147] In the conventional correction mode executed on a fixed cycle, the daily carbon emission factor is rolled over and corrected on a fixed cycle of 1 hour, synchronized with the scheduling cycle of the daily rolling optimization. Within each correction cycle, the latest operating data is recalculated according to the method described in equation (10). The updated results are then input into the intraday two-layer optimization model in subsequent step S4.2.

[0148] The event-triggered correction mode triggers a non-periodic rapid correction when the deviation between the actual output of new energy and the day-ahead forecast exceeds a first predetermined threshold, such as 20% of the installed capacity, or when the deviation between the actual abandoned electricity and the day-ahead forecast exceeds a second predetermined threshold, such as 50% of the forecast value. The rapid correction is not limited by the regular correction cycle. After the triggering conditions are met, the method of formula (10) is immediately executed to correct the calculation, and the updated carbon emission factor is pushed out.

[0149] Through the above constraints and dual-mode triggering mechanism, the intraday carbon emission factor rolling correction method ensures the rationality of the correction results while taking into account both the computational efficiency under normal operation and the rapid response capability under sudden deviations, so that the carbon emission factor can achieve smooth connection and gradual approach on the two time scales of day-ahead and intraday.

[0150] Specifically, during step S4.2, this invention constructs a two-layer collaborative scheduling model of virtual power plants and distribution networks that takes into account carbon emission constraints, based on the improved carbon potential calculation method of equation (8) and the rolling correction mechanism of carbon emission factors in step S4.1. The model adopts a two-stage progressive architecture of day-ahead and intraday, with each stage being a two-layer structure of upper-level distribution network economic scheduling and lower-level virtual power plant resource optimization. The upper and lower layers interact through the exchange of power and carbon emission factors by virtual power plants, realizing the synergy between global optimization of the distribution network and internal optimization of virtual power plants. Specifically, it includes the following steps:

[0151] S4.2.1, Current two-layer optimization model upper layer: economic and low-carbon dispatching of distribution network;

[0152] S4.2.2, Lower layer of the current two-layer optimization model: internal resource optimization of the virtual power plant;

[0153] S4.2.3, Current iterative interaction mechanism between upper and lower layers;

[0154] S4.2.4 Intraday Double-Layer Rolling Optimization Model.

[0155] Specifically, during the execution of step S4.2.1, the day-ahead dual-layer optimization model uses a time granularity of 1 hour and a scheduling cycle of 24 hours. The upper layer of the day-ahead dual-layer optimization model takes minimizing the total operating cost of the distribution network as the optimization objective, incorporates carbon emission costs into the objective function, and introduces dual carbon emission control constraints to achieve synergistic optimization of economic efficiency and low carbon emissions.

[0156] The objective function of the upper-level model consists of four components: thermal power generation cost, curtailment penalty cost, cost of purchasing electricity from the main grid, and carbon emission cost, expressed as equation (11):

[0157] (11)

[0159] in:

[0160] For thermal power units The power generation cost function is a quadratic function. ;

[0161] The penalty price per unit of electricity wasted;

[0162] For new energy units exist The amount of power abandoned during a given period;

[0163] for Time-of-use electricity pricing for purchasing electricity from the main grid during specific time periods;

[0164] To purchase power from the mainnet;

[0165] The price is for carbon emissions trading, expressed in yuan / tCO2.

[0166] for The total carbon emissions of the system during the time period are calculated based on the improved carbon potential and the load power of each node, and are expressed as the sum of the products of the carbon emission factor of each load node and the electricity consumption of the load:

[0167]

[0168] in, For nodes exist The baseline value of the day-ahead carbon emission factor for the period. For nodes exist Load power during a given time period. When When the amount of carbon emissions decreases during the curtailment period due to the negative contribution of equivalent carbon emissions from renewable energy sources, the carbon emission accounting value corresponding to the load consumption at that node decreases accordingly, thereby reducing the carbon emission cost term in the objective function and creating an optimized incentive to increase load absorption during the curtailment period.

[0169] The upper-level model must satisfy the following six types of constraints.

[0170] The first type is the power balance constraint of the distribution network, which requires that the sum of the power output of the generation side, the power purchased by the main grid and the power exchanged by the virtual power plant in the distribution network at each time period equals the sum of the system load and the network loss.

[0171]

[0172] in, Power exchange for virtual power plants.

[0173] The second category is carbon emission total control constraints, which require that the total carbon emissions of the system during the scheduling cycle do not exceed a set upper limit:

[0174]

[0175] The third category is carbon emission intensity control constraints, which require that the carbon emission intensity per unit of electricity supply in each time period not exceed a set upper limit:

[0176]

[0177] The second and third types of constraints together constitute a dual carbon emission control mechanism, which constrains system carbon emissions from both the total amount and intensity dimensions.

[0178] The fourth category is the output and ramping constraints of thermal power units:

[0179]

[0180]

[0181] in, and The units Minimum and maximum technical output, This represents the maximum gradient in a single time period.

[0182] The fifth category is the constraint of curtailment of renewable energy:

[0183]

[0184] in, For new energy units The predicted power output for the day is not more than the predicted power output.

[0185] The sixth category is virtual power plant switching power constraints:

[0186]

[0187] Among them, the upper and lower limits of the switching power and Determined by feedback from the lower-level model, it reflects the actual adjustable range of the virtual power plant in each time period.

[0188] Furthermore, during the execution of step S4.2.2, the lower layer of the daytime dual-layer optimization model takes minimizing the total daily operating cost of the virtual power plant as the optimization objective, and optimizes the scheduling of controllable distributed power sources, energy storage systems, and adjustable loads within the virtual power plant based on the power exchange plan and carbon emission factor constraints issued by the upper-layer model.

[0189] The objective function of the lower-level model consists of three components: controllable distributed power source fuel cost, energy storage cycle loss cost, and adjustable load response compensation cost, expressed as equation (12):

[0190] (12)

[0192] in:

[0193] For controllable distributed power sources (such as gas turbines), the fuel cost function is used.

[0194] Cost of charge-discharge cycle losses for energy storage units;

[0195] For energy storage exist The absolute value of charging and discharging power during a given period;

[0196] For the first Unit adjustment compensation price for each adjustable load user;

[0197] For this user The actual adjustment amount during the time period.

[0198] The lower-level model must satisfy the following five types of constraints.

[0199] The first type is the internal power balance constraint of the virtual power plant, which requires that the net value of the power generation output, energy storage charging and discharging and load consumption within the virtual power plant at each time period equals the exchange power plan issued by the upper-level model:

[0200]

[0201] in, The virtual power plant switching power plan determined for the upper-level model is used as a known boundary condition in the lower-level model.

[0202] The second category involves controllable distributed power generation output and ramping constraints:

[0203]

[0204]

[0205] in, and These represent the minimum and maximum output of the controllable power supply, respectively. This represents the maximum gradient in a single time period.

[0206] The third category consists of constraints on the operation of energy storage systems, including charging and discharging power constraints, state of charge constraints, and consistency constraints between the initial and final state of charge:

[0207]

[0208]

[0209]

[0210] Among them, the state of charge Calculated using a time-series recursive formula. The consistency constraint between the initial and final states of charge ensures that the energy storage returns to its initial state at the end of the scheduling cycle, guaranteeing the continuity of cross-cycle scheduling.

[0211] The fourth category is distributed photovoltaic power output constraints:

[0212]

[0213]

[0214] The fifth category is adjustable load regulation constraints, including regulation capacity constraints and duration constraints:

[0215]

[0216]

[0217] After the lower-level model is solved, the actual adjustable switching power range of the virtual power plant in each time period is determined. With internal operating costs Feedback is sent to the upper-level model to update the virtual power plant exchange power constraints and collaborative optimization objectives in the upper-level model.

[0218] Specifically, during step S4.2.3, the upper and lower layers of the current two-layer optimization model achieve collaborative optimization through the transfer of interaction variables and iterative solution. The specific process is as follows:

[0219] S4.2.3.1, The upper-level model passes two pieces of information to the lower-level model:

[0220] First, the virtual power plant power exchange plan , As boundary conditions for the power balance constraints within the lower-level model;

[0221] Second, recent carbon emission factors The baseline value, calculated in step S4.1.1, is used by the lower-level model to assess the carbon emission impact of the virtual power plant's regulation behavior.

[0222] S4.2.3.2, The lower-level model feeds back two pieces of information to the upper-level model:

[0223] First, the actual adjustable switching power range of the virtual power plant at different times. Based on the various resource constraints in the lower-level model, it is used to update the sixth type of constraint in the upper-level model;

[0224] Second, the internal operating costs of virtual power plants. It is used by upper-level models to evaluate the overall system cost under different switching power schemes.

[0225] The above process is repeated iteratively until the convergence criterion is met:

[0226]

[0227] in, For the number of iterations, and This is a convergence accuracy threshold. When the above conditions are met, the iteration terminates, and a day-ahead scheduling plan is output, which includes the power output of thermal power units for each time period. Wasted electricity from renewable energy sources Virtual power plant switching power And the output and regulation schemes of various resources within the virtual power plant.

[0228] Furthermore, during step S4.2.4, an intraday dual-layer rolling optimization model is constructed based on the day-ahead dual-layer optimization model. This intraday rolling optimization model is used to re-solve the carbon emission factor based on the rolled correction during the intraday phase, and sets deviation constraints from the day-ahead scheduling plan. This allows the day-ahead scheduling plan to be corrected based on the latest operational information during the intraday operation phase, enabling the scheduling scheme to adapt to intraday uncertainties such as new energy output deviations, load fluctuations, and changes in the carbon emission factor. The intraday model uses a 15-minute time granularity and employs a 4-hour rolling window and a 15-minute step rolling optimization architecture.

[0229] Specifically, the intraday upper-level model aims to minimize the intraday operating cost of the distribution network and adopts an intraday corrected carbon emission factor. Replacement day-ahead benchmark As a carbon emission constraint input, the structure of the upper-level objective function is consistent with the previous-day upper-level model, but the optimization time domain is limited to 16 time periods (4 hours / 15 minutes) within the current rolling window, and a deviation tolerance constraint from the previous-day plan is added:

[0230]

[0231]

[0232] in, and These represent the revised thermal power output and the virtual power plant exchange power, respectively, for the day. and This represents the upper limit of the allowable deviation. The deviation constraint ensures that intraday adjustments do not deviate too far from the day-ahead plan, maintaining the continuity and stability of the scheduling scheme. The remaining constraints of the upper-level model are consistent with those of the day-ahead upper-level model, including carbon emissions. Based on intraday carbon emission factors Recalculate.

[0233] Furthermore, the intraday lower-level model aims to minimize the intraday operating cost of the virtual power plant. The objective function structure is consistent with the day-ahead lower-level model, but the following two intraday-specific constraints are added to the constraints.

[0234] The first item is the constraint for correcting the energy storage state of charge deviation. Since actual daily charging and discharging behavior may deviate from the day-ahead plan, causing a deviation in the energy storage state of charge trajectory, a deviation tolerance band constraint is set:

[0235]

[0236] in, This is the planned state of charge trajectory. This is the corrected state of charge for the day. To allow for deviations, the constraints ensure flexible adjustment space for energy storage while preventing excessive shifts in the state of charge that could lead to loss of adjustment capability in subsequent periods.

[0237] The second item is the intraday response constraint for adjustable load. Intraday adjustments to adjustable load must consider the cumulative adjustments already made by the user to avoid exceeding the user's total daily adjustment limit.

[0238]

[0239] in, For the first The daily adjustable power limit for each user.

[0240] The iterative interaction mechanism between the upper and lower layers of the intraday two-layer model is consistent with that of the day-ahead model, with iterative solutions completed independently within each rolling window. The connection between rolling windows is as follows: the state of the last period obtained from the solution of the previous window (thermal power output, energy storage state of charge, cumulative adjustment of adjustable load, etc.) is used as the initial condition for the next window, ensuring the continuity of the scheduling scheme between adjacent windows.

[0241] When the correction mechanism in step S4.1.3 is activated by an event (new energy output deviation exceeds 20% of installed capacity or curtailment deviation exceeds 50% of predicted value), the daily carbon emission factor... When a non-periodic update occurs, the intraday two-layer model is immediately re-solved within the current window, enabling the scheduling scheme to be adjusted promptly in response to sudden changes in carbon emission information.

[0242] A second aspect of the present invention provides a scheduling system for a virtual power plant, comprising a data acquisition module, an equivalent carbon emission intensity determination module, a node carbon emission factor calculation module, and a scheduling scheme generation module, wherein:

[0243] The data acquisition module is used to acquire the operating data of the power system, which includes a distribution network and virtual power plants connected to the distribution network.

[0244] The equivalent carbon emission intensity determination module is used to determine the equivalent carbon emission intensity of new energy based on the curtailment status of new energy units. In the period when curtailment occurs, the equivalent carbon emission intensity of new energy is negative, and in the period when curtailment does not occur, it is zero.

[0245] The node carbon emission factor calculation module is used to embed the equivalent carbon emission intensity of the new energy into the calculation of the node carbon emission flow of the distribution network to obtain an improved node carbon emission factor.

[0246] The scheduling scheme generation module is used to generate a scheduling scheme for the virtual power plant based on the improved node carbon emission factor through a two-layer collaborative scheduling model of the virtual power plant and the distribution network. The two-layer collaborative scheduling model includes an upper layer and a lower layer, and the upper layer and the lower layer iteratively interact by exchanging power and the improved node carbon emission factor.

[0247] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A scheduling method for a virtual power plant, characterized in that, Includes the following steps: Acquire operational data of a power system, which includes a distribution network and virtual power plants connected to the distribution network; The equivalent carbon emission intensity of new energy is determined based on the curtailment status of new energy units. During periods when curtailment occurs, the equivalent carbon emission intensity of new energy is negative, and during periods when curtailment does not occur, it is zero. By embedding the equivalent carbon emission intensity of the new energy source into the calculation of the nodal carbon emission flow of the distribution network, an improved nodal carbon emission factor is obtained. Based on the improved node carbon emission factor, a scheduling scheme for the virtual power plant is generated through a two-layer collaborative scheduling model of the virtual power plant and the distribution network. The two-layer collaborative scheduling model includes an upper layer and a lower layer, and the upper layer and the lower layer iteratively interact by exchanging power and the improved node carbon emission factor.

2. The scheduling method for a virtual power plant according to claim 1, characterized in that, The absolute value of the equivalent carbon emission intensity of the new energy source is positively correlated with the curtailment rate and the carbon emission intensity of marginal thermal power units; the curtailment rate is the ratio of curtailed power to the total output of new energy sources; the marginal thermal power unit refers to the unit with the highest carbon emission intensity among the thermal power units that are in operation and whose actual output is greater than their minimum technical output.

3. The scheduling method for a virtual power plant according to claim 2, characterized in that, The formula for calculating the equivalent carbon emission intensity of the new energy source is as follows: ； in, For time period The equivalent carbon emission intensity of new energy sources For time period Carbon emission intensity per unit of power generation of marginal thermal power units in the system For time period The power of abandoned electricity, For time period Total output of new energy sources.

4. The scheduling method for a virtual power plant according to claim 1, characterized in that, The improved formula for calculating the nodal carbon emission factor is as follows: ； in, For nodes exist Improved node carbon emission factors over time periods For access nodes A collection of thermal power units, For the unit carbon emission intensity per unit of electricity generation For the unit exist The amount of effort contributed during a given period; For access nodes A collection of new energy generating units, For the first Taiwan New Energy Equivalent carbon emission intensity over a given period This will contribute to the actual absorption of this new energy source; To the node The set of upstream nodes for output power. upstream node Improved carbon emission factor For the node To the node The active power flow.

5. The scheduling method for a virtual power plant according to claim 1, characterized in that, The scheduling scheme for generating the virtual power plant includes a day-ahead scheduling phase and an intraday scheduling phase. In the intraday scheduling phase, the improved node carbon emission factor is rolled over based on the actual intraday operating data, and the corrected carbon emission factor is used as the carbon emission constraint input for the intraday scheduling model.

6. The scheduling method for a virtual power plant according to claim 5, characterized in that, The process of rolling correction of the improved node carbon emission factor includes: Distribution network side correction: The day-ahead forecast is replaced by the actual output of new energy sources, the actual amount of abandoned electricity, and the actual output of thermal power units. The improved nodal carbon emission factor is recalculated to obtain the distribution network side correction value. Virtual power plant side correction: Calculate the change in node carbon potential caused by the difference between the actual exchange power of the virtual power plant and the day-ahead planned exchange power, and use it as the virtual power plant side correction amount; Comprehensive correction: The correction value on the distribution network side is superimposed with the correction amount on the virtual power plant side to obtain the final carbon emission factor for the day.

7. The scheduling method for a virtual power plant according to claim 5, characterized in that, The rolling correction adopts a dual-mode triggering mechanism: a regular correction mode executed at a fixed cycle, and an event correction mode triggered when the deviation between the actual output of new energy and the day-ahead forecast exceeds a first predetermined threshold, or when the deviation between the actual abandoned power and the day-ahead forecast exceeds a second predetermined threshold.

8. The scheduling method for a virtual power plant according to claim 5, characterized in that, The dual-layer collaborative scheduling model includes a day-ahead dual-layer optimization model and an intraday dual-layer rolling optimization model; The upper layer of the day-ahead two-layer optimization model aims to minimize the total operating cost of the distribution network and includes carbon emission constraints. The lower layer of the day-ahead two-layer optimization model aims to minimize the internal operating cost of the virtual power plant and optimizes the output and regulation of the resources inside the virtual power plant based on the power exchange plan determined by the upper layer. The intraday two-layer rolling optimization model is re-solved based on the rolled-corrected carbon emission factor during the intraday phase, and deviation constraints are set with respect to the day-ahead scheduling plan.

9. The scheduling method for a virtual power plant according to claim 8, characterized in that, The intraday dual-layer rolling optimization model adopts a rolling window optimization architecture and sets constraints for energy storage state of charge deviation correction and adjustable load intraday response.

10. A dispatching system for a virtual power plant, characterized in that, It includes a data acquisition module, an equivalent carbon emission intensity determination module, a node carbon emission factor calculation module, and a scheduling scheme generation module, among which: The data acquisition module is used to acquire the operating data of the power system, which includes a distribution network and virtual power plants connected to the distribution network. The equivalent carbon emission intensity determination module is used to determine the equivalent carbon emission intensity of new energy based on the curtailment status of new energy units. In the period when curtailment occurs, the equivalent carbon emission intensity of new energy is negative, and in the period when curtailment does not occur, it is zero. The node carbon emission factor calculation module is used to embed the equivalent carbon emission intensity of the new energy into the calculation of the node carbon emission flow of the distribution network to obtain an improved node carbon emission factor. The scheduling scheme generation module is used to generate a scheduling scheme for the virtual power plant based on the improved node carbon emission factor through a two-layer collaborative scheduling model of the virtual power plant and the distribution network. The two-layer collaborative scheduling model includes an upper layer and a lower layer, and the upper layer and the lower layer iteratively interact by exchanging power and the improved node carbon emission factor.

Images