A virtual power plant auxiliary service optimization scheduling method under a multi-pole extreme scenario

Abstract

This invention discloses an optimized scheduling method for ancillary services of virtual power plants under multiple extreme scenarios, belonging to the field of power system technology. It aims to solve the problems of poor adaptability to extreme scenarios, low scheduling accuracy, and failure to consider both revenue and equipment losses in existing technologies. The method includes: establishing a market model for distributed energy, virtual power plants, and ancillary services; constructing single and composite extreme scenario systems, completing scenario generation and dimensionality reduction; constructing a coupled performance degradation model for composite extreme scenarios and an adaptive correction mechanism for weather impact factors; and constructing an optimized scheduling model with the goal of synergistic optimization of revenue and equipment losses, and combining constraints to complete dynamic scheduling. Through verification in multiple sets of composite extreme scenarios, this invention can accurately adapt to extreme operating conditions, improve scheduling accuracy and robustness, balance ancillary service revenue and equipment lifespan, enhance the grid's resistance to disturbances, and demonstrates effective model verification, possessing good engineering application value.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method for optimizing the scheduling of virtual power plant ancillary services under multiple extreme scenarios. Background Technology

[0002] Virtual Power Plants (VPPs) are a core management model in new power systems that integrates distributed energy resources such as distributed photovoltaic, wind power, energy storage systems, and controllable loads. Through information and communication technologies and intelligent control algorithms, they achieve unified coordination and equivalent control of multiple types of resources. They can participate in ancillary service market transactions such as grid peak shaving, frequency regulation, spinning reserve, and non-spinning reserve. They have significant technical advantages in improving the level of renewable energy consumption, smoothing system power fluctuations, ensuring grid power balance, and optimizing power resource allocation, and have become an important technical means to support the safe and efficient operation of the power grid.

[0003] In recent years, global climate change has led to more frequent, stronger, and longer-lasting extreme weather events such as heat waves, cold waves, storms, dust storms, and extreme winds, causing severe disturbances to the safe and stable operation of power distribution networks and the output characteristics of distributed energy equipment. Under extreme weather conditions, photovoltaic modules experience a sharp drop in output due to temperature, irradiance, and dust cover; wind turbines experience output fluctuations or protective shutdowns due to excessive wind speeds and blade icing / dust accumulation; and energy storage batteries suffer from capacity decay, reduced charging and discharging efficiency, and irreversible performance degradation due to high and low temperatures and harsh environments. At the same time, rigid loads such as cooling and heating surge dramatically. The source, storage, and load exhibit strongly coupled, nonlinear, and highly random fluctuation characteristics, which can easily lead to problems such as insufficient active power, insufficient reactive power, voltage exceeding limits at end nodes, and grid frequency fluctuations. This directly weakens the ancillary service supply capacity of virtual power plants and threatens the safe operation of the power grid.

[0004] Existing technologies and research related to virtual power plant optimization scheduling have many inherent shortcomings, making it difficult to adapt to the operational needs of extreme weather scenarios: First, existing scheduling models and methods are mostly designed based on conventional operating conditions, failing to fully consider the differentiated coupling impacts of extreme weather on distributed power sources, energy storage systems, and controllable loads. They cannot accurately quantify the mechanisms of action of five typical extreme weather events: heat waves, cold waves, storms, sandstorms, and extreme winds, and have not established a physical model for the coupling performance degradation of composite extreme weather events. Second, the methods for generating extreme weather scenarios are simplistic, merely simulating weather impacts through simple parameter assignment, lacking time-series modeling, extreme event implantation, and multi-field analysis. The complete process of scene clustering and reduction suffers from poor scene representativeness and an inability to reproduce the continuous evolution of meteorological patterns, making it difficult to support long-term, high-precision scheduling simulations. Third, the scheduling model does not incorporate the irreversible degradation mechanism of equipment performance and the adaptive learning capability of weather influencing factors, relying on fixed parameter settings, resulting in large deviations between output prediction and actual operation, and insufficient robustness of the scheduling strategy. Fourth, the existing scheduling strategy takes maximizing ancillary service revenue as the sole optimization objective, without taking into account the balance between equipment wear and lifespan. In extreme scenarios, it is prone to problems such as excessive resource adjustment and accelerated equipment aging, failing to achieve coordinated optimization of power grid reliability, scheduling economy, and equipment durability.

[0005] In summary, existing virtual power plant ancillary service optimization scheduling technologies have significant shortcomings in terms of extreme weather adaptability, scenario simulation comprehensiveness, model accuracy, and optimization objective rationality. They cannot meet the needs of safe and stable operation and ancillary service guarantee of new power systems under various extreme scenarios. There is an urgent need to develop a virtual power plant ancillary service optimization scheduling method that can adapt to extreme weather and balance benefits and equipment lifespan. Summary of the Invention

[0006] The purpose of this invention is to provide a method for optimizing the scheduling of ancillary services of virtual power plants under multiple extreme scenarios, so as to achieve accurate quantification and dynamic correction of the output of virtual power plants under multiple extreme scenarios, take into account both the benefits of ancillary services and the service life of equipment, and effectively improve the stability and reliability of scheduling operation under extreme weather conditions.

[0007] To achieve the above objectives, this invention provides a method for optimizing and scheduling virtual power plant ancillary services under multiple extreme scenarios, comprising the following steps: S1: Establish a distributed energy resource model that includes photovoltaic power plants, wind farms, energy storage systems, and controllable loads, and construct a virtual power plant aggregation model and an ancillary service market model; S2: Construct a system of single extreme scenarios and composite extreme scenarios covering heat waves, cold waves, rainstorms, sandstorms, and extreme winds. Complete scenario generation and dimensionality reduction through temporal modeling, extreme event implantation, and multi-scenario clustering reduction. S3: Construct a coupled performance degradation model for composite extreme scenarios, introduce an adaptive learning mechanism for weather influencing factors, and correct the weather influencing factors in real time based on the deviation between actual output and predicted output. S4: With the goal of achieving optimal synergy between ancillary service revenue and equipment loss, a linear programming optimization scheduling model is constructed, which combines power balance, energy storage constraints, load regulation constraints, and ancillary service demand constraints to complete dynamic scheduling.

[0008] Preferably, the distributed energy model in S1 includes: (1) Photovoltaic power station model:

[0009] Where: G is solar irradiance, W / m²; A is photovoltaic panel area, m²; η is power generation efficiency; α is annual degradation rate; t is the operating life of the power station; β is temperature coefficient; T is ambient temperature, °C; Factors influencing extreme weather; (2) Wind farm model: The output power of a wind farm depends on wind speed and extreme weather conditions, as shown in the following model:

[0010] in: Actual wind speed (m / s); To cut in wind speed; Rated wind speed; To cut off the wind speed; Rated power; (3) The energy storage system model includes the available charging power model, the available discharging power model, and the state of charge (SOC) update model: Available charging power:

[0011] Available discharge power:

[0012] SOC Update:

[0013] in: This is the maximum charging power; This represents the maximum discharge power. and These are the upper and lower limits of SOC; Battery capacity; and For charge and discharge efficiency; For time step; (4) Controllable load model Reduced load:

[0014] Increased load:

[0015] in, Current load; Basic load; The load amplitude is adjustable.

[0016] Preferably, the ancillary services market model in S1 includes a demand model and a price model, both of which are affected by time and weather conditions: Demand Model:

[0017] Price model:

[0018] in, Basic requirements; Base price; The time-related factor; Weather influencing factors; This is the price elasticity coefficient.

[0019] Preferably, the influencing factor functions for the five types of single extreme weather in S2 are as follows: heat wave scene It is a piecewise linear function based on ambient temperature; Cold wave scene It is a three-segment linear function based on ambient temperature; Storm scene It is a piecewise exponential function based on rainfall intensity; Sandstorm scene It is an exponential decay function based on dust concentration; Extreme wind scenarios It is a piecewise linear function based on wind speed.

[0020] Preferably, the scene generation method in S2 takes historical meteorological data as input and executes it sequentially: Based on historical meteorological data, an autoregressive moving average model is used to generate time-series curves of basic meteorological parameters; Using the time-series curves of the basic meteorological parameters as input, extreme weather events are incorporated to construct a complete meteorological event curve that includes event intensity, peak time, and duration. The complete meteorological event curves are used to form an original scene set. The K-means clustering algorithm is used to reduce the scene set, extract typical meteorological scenes, and calculate the corresponding probability weights. The typical meteorological scenario is taken as input and combined with the distributed energy model of S1 to be coupled and mapped to the available power scenario of VPP through VPP response characteristics; Based on the time series of the typical meteorological scenarios and the available power scenarios of VPP, a Markov state transition model is established to simulate the continuous evolution process of normal weather, single extreme scenarios, and compound extreme scenarios.

[0021] Preferably, the coupled performance degradation model for combined extreme weather in S3 is as follows:

[0022] in, , These are the influencing factors of different single extreme weather events. This represents the irreversible degradation coefficient of equipment performance.

[0023] Preferably, the adaptive correction model for weather influencing factors in S3 is as follows:

[0024] Add a limiting constraint to the adaptive factor: ) Preferably, in S4, the objective function of the optimized scheduling model is:

[0025] In the formula: This represents the total number of scheduling periods; Contribute to frequency modulation service at time t; Price of frequency modulation service at time t; The rotational reserve output at time t; The price for rotating spare parts at time t; The non-rotating standby output at time t; Let t be the non-rotating spare price.

[0026] The constraints described in S4 include power balance constraints, energy storage SOC constraints, charge / discharge power constraints, controllable load regulation constraints, and ancillary service demand constraints, specifically: (1) Power balance constraint

[0027] (2) Energy storage SOC constraint

[0028] (3) Energy storage charging power constraints

[0029] (4) Energy storage discharge power constraint

[0030] (5) Controllable load reduction constraints

[0031] (6) Controllable load increases constraints

[0032] (7) Constraints on demand for ancillary services

[0033]

[0034]

[0035] in, , These are the adjustable upper and lower limits for the load, respectively. , , These are frequency modulation, spinning reserve, and non-spinning reserve requirements, respectively.

[0036] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects: (1) This invention can comprehensively cover five types of single extreme scenarios, such as heat waves, cold waves, storms, sandstorms, and extreme winds, as well as multiple types of composite extreme scenarios. By using a coupled performance degradation model, it can accurately quantify the synergistic effects of extreme weather on photovoltaics, wind power, energy storage, and controllable loads, as well as the irreversible degradation of equipment. This effectively solves the technical shortcomings of existing technologies, such as one-sided scenario simulation and the inability to accurately quantify the mechanism of extreme weather.

[0037] (2) The present invention adopts an adaptive learning mechanism for weather influence factors, and corrects parameters in real time based on the deviation between actual output and predicted output, which greatly reduces the error of resource output prediction in extreme scenarios and significantly improves the accuracy and robustness of virtual power plant scheduling strategy.

[0038] (3) The present invention achieves synergistic optimization of ancillary service revenue and equipment wear and tear. While ensuring the stable supply of ancillary services such as frequency regulation and spinning reserve in extreme scenarios, it effectively reduces excessive equipment wear and tear and extends service life, taking into account both scheduling economy and equipment durability.

[0039] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments 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.

[0041] Figure 1 This is a diagram of the virtual power plant operation structure according to Embodiment 1 of the present invention; Figure 2 The following are simulation results of extreme weather scenarios in Embodiment 1 of the present invention: (a) is the power comparison curve of the heat wave scenario, (b) is the benefit comparison curve of the heat wave scenario, (c) is the power comparison curve of the cold wave scenario, (d) is the benefit comparison curve of the cold wave scenario, (e) is the power comparison curve of the storm scenario, and (f) is the benefit comparison curve of the storm scenario. Figure 3 The diagram shows the combined extreme weather coupling performance attenuation model of Embodiment 1 of the present invention. (a) is the power characteristic curve of the sandstorm scenario, (b) is the benefit characteristic curve of the sandstorm scenario, (c) is the power characteristic curve of the extreme wind scenario, and (d) is the benefit characteristic curve of the extreme wind scenario. Figure 4 This is a schematic diagram of the adaptive weather influencing factors in Embodiment 1 of the present invention; Figure 5 This is a flowchart of a virtual power plant auxiliary service optimization scheduling method under multiple extreme scenarios according to Embodiment 1 of the present invention. Detailed Implementation

[0042] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0043] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0044] Example 1 This embodiment provides a method for optimizing the scheduling of virtual power plant ancillary services under multiple extreme scenarios, such as... Figure 5 As shown, the steps are as follows: S1: Establish a distributed energy resource model, construct a virtual power plant aggregation model and an ancillary service market model, and define the virtual power plant operation structure as follows: Figure 1 As shown.

[0045] 1. Distributed Energy Resource Model (1) Photovoltaic power station model The output power of a photovoltaic power station is affected by solar irradiance, ambient temperature, and extreme weather, as shown in the following model: (1) Where: G is solar irradiance (W / m²); A is photovoltaic panel area (m²); η is power generation efficiency; α is annual degradation rate; t is the operating life of the power station; β is temperature coefficient (approximately -0.004 / ℃); T is ambient temperature (℃). These are factors influencing extreme weather events.

[0046] (2) Wind farm model The output power of a wind farm depends on wind speed and extreme weather conditions, as shown in the following model: (2) in: Actual wind speed (m / s); To cut in wind speed; Rated wind speed; To cut off the wind speed; This is the rated power.

[0047] Under extreme weather conditions, the output of wind farms can be further affected, such as storms that may cause premature shutdowns.

[0048] (3) Energy storage system model The available power and state of charge (SOC) model of the energy storage system are as follows: Available charging power: (3) Available discharge power: (4) SOC Update: (5) in, This is the maximum charging power; This represents the maximum discharge power. and These are the upper and lower limits of SOC; Battery capacity; and For charge and discharge efficiency; For time step.

[0049] (4) Controllable load model The available flexibility model for controllable loads is as follows: Reduced load: (6) Increased load: (7) in Current load; Basic load; The load amplitude is adjustable.

[0050] 2. Virtual power plant aggregation model The virtual power plant integrates the aforementioned distributed energy resources to form a unified dispatch entity. Its available power is: (8) 3. Ancillary Services Market Model The demand and prices in the ancillary services market are affected by time and weather conditions: Demand Model: (9) Price model: (10) in: Basic requirements; Base price; The time-related factor; Weather influencing factors; This is the price elasticity coefficient.

[0051] S2. Construct an extreme scenario system to complete scenario generation and dimensionality reduction. 1. Extreme Scenario System It covers five single extreme scenarios: heat waves, cold waves, rainstorms, sandstorms, and extreme winds, as well as composite extreme scenarios combining these scenarios, clarifying the meteorological characteristics and impact mechanisms of each scenario.

[0052] (1) In the heat wave scenario, the power and benefit curves are as follows: Figure 2 As shown in (a) and (b) in the figure.

[0053] Meteorological characteristics: Ambient temperature remains above 35℃ for 6-72 hours, with a significant decrease in humidity. Impact Mechanism: Increased photovoltaic module temperature leads to decreased efficiency (temperature coefficient β = -0.004 / ℃), increased internal resistance of lithium batteries, and a 10%-25% reduction in usable capacity. Air conditioning cooling load surges, with peak load increasing by 30%-50% compared to the baseline date. Transformer and line current-carrying capacity degrades. The impact on parameters is shown in Table 1.

[0054] Table 1. Parameter baseline values ​​and heat wave impact data

[0055] Impact factor function: (11) (2) In the cold wave scenario, the power and benefit curves are as follows: Figure 2 As shown in (c) and (d) in the figure.

[0056] Meteorological characteristics: A sudden drop in ambient temperature (≥8℃ in 24 hours), sustained low temperatures (-10℃), accompanied by freezing rain and snow. Mechanism of influence: Increased viscosity of lithium battery electrolyte reduces discharge efficiency by 20%-40%, significantly decreases charging acceptance, limits start-up charging at low temperatures, causes a sharp increase in heating load, and increases peak electricity demand by 40%-60% compared to normal winter.

[0057] Ice accumulation on wind turbine blades can lead to reduced output or shutdown protection, while ice accumulation on power transmission lines can cause tower collapse and line breakage risks. The impact on parameters is shown in Table 2.

[0058] Table 2. Data on the impact of cold waves on parameters

[0059] Impact factor function: (12) (3) In the storm scenario, the power and benefit curves are as follows: Figure 2 As shown in (e) and (f) in the figure.

[0060] Meteorological characteristics: short-duration heavy rainfall (≥50mm / h), thunderstorms with strong winds (wind speed ≥17m / s), and severe lightning activity. Mechanism of influence: When solar irradiance drops sharply to 10%-30% of the normal value, photovoltaic output drops precipitously, wind speed fluctuates drastically, wind turbines frequently pass through vibration zones, mechanical fatigue intensifies, and ultra-high wind speeds trigger the wind turbines to shut down for protection (the shut-off wind speed is usually 25 m / s). Lightning strikes increase the failure rate of electrical equipment such as inverters and combiner boxes, while floods threaten the safety of power plant infrastructure. Their impact on parameters is shown in Table 3.

[0061] Table 3. Data on the impact of storms on various parameters.

[0062] Impact factor function: (13) in Rainfall intensity (mm / h).

[0063] (4) In the case of a sandstorm, the power and benefit curves are as follows: Figure 3 As shown in (a) and (b) in the figure.

[0064] Meteorological characteristics: Strong winds (≥10m / s) carrying large amounts of sand and dust, visibility <1km, and air quality index off the charts. Mechanism of Influence: Dust blocks solar radiation, reducing direct irradiance by 60%-90%. Dust deposits on the surface of photovoltaic panels cause an additional 15%-25% loss in module efficiency due to scattering effects. Dust accumulation on the surface of wind turbine blades alters aerodynamic characteristics, reducing the output coefficient by 5%-10%. Dust wears down mechanical components such as wind turbine bearings and gearboxes, and electrostatic dust accumulation affects the insulation performance of electrical equipment. The impact on parameters is shown in Table 4.

[0065] Table 4. Data on the impact of sandstorms on parameters

[0066] Impact factor function: (14) in The dust accumulation attenuation coefficient is 0.7-0.85. The concentration of dust (μg / m³) is the dust concentration. The reference concentration is 1000 μg / m³.

[0067] (5) In extreme wind scenarios, the power and benefit curves are as follows: Figure 3 As shown in (c) and (d) in the figure.

[0068] Meteorological characteristics: persistent strong winds (average wind speed ≥ 20 m / s), gusts up to 35 m / s or more, stable wind direction. Impact Mechanism: When wind speed exceeds the cut-out wind speed (25 m / s), wind turbines disconnect from the grid on a large scale. Strong winds cause grid frequency fluctuations, leading to a surge in demand for frequency regulation services. This increases the risk of wind-induced discharge and galloping on transmission lines, threatening the structural safety of distributed photovoltaic supports. Energy storage systems become the only reliable regulation resource. The impact on parameters is shown in Table 5.

[0069] Table 5. Parameters affecting extreme strong winds

[0070] Impact factor function (15) in The rated wind speed (usually 12-15 m / s). To cut off the wind speed (usually 25m / s).

[0071] 2. Scene generation and dimensionality reduction (1) Time series modeling of meteorological parameters The basic meteorological curves are generated using an autoregressive moving average (ARMA) model: (16) in Let t represent the meteorological parameters (temperature / wind speed / irradiance). and These are the autoregressive and moving average coefficients, respectively. It is white noise.

[0072] (2) Extreme weather events Superimpose extreme weather events onto the base curve: (17) in The magnitude of the event intensity. At peak time, For the rate of increase / decrease, The starting point of the event. For duration.

[0073] (3) Multi-scenario clustering and reduction The generated scene set was reduced using the K-means clustering algorithm, retaining only the most representative scenes: 1. Generate N original scenes (N=1000) 2. Extract key feature vectors: peak intensity, duration, energy deficit, ramp rate. 3. K-means clustering, setting K=5-10 typical scenarios. 4. Calculate the probability weight for each cluster. (4) VPP response characteristics coupling Map the weather scenario to the available power scenario of the VPP: (18) in The random failure factor for the equipment follows a Bernoulli distribution. Meteorological data for each scenario type are shown in Table 6.

[0074] Table 6. Statistical Table of Meteorological Data for Different Scene Types

[0075] 3. Considering the continuous evolution characteristics of extreme weather, establish a Markov state transition model: State space:

[0076] in Normal weather. to Corresponding to five single extreme weather conditions, It is a complex weather pattern.

[0077] State transition matrix: (19) Transition probability Determined based on historical meteorological data statistics. For example: Normal → Heatwave: Summer ,winter

[0078] Heatwave → Normal: (Gradually subsiding), (Transformed into a composite) Cold wave → freezing rain and snow: (When water vapor conditions are sufficient) The model supports generating a continuous weather state sequence of 8760 hours throughout the year for annual VPP operation simulation.

[0079] S3: Constructing a coupling attenuation model and adaptive correction mechanism Composite extreme weather coupled performance degradation model (20) Where k is the adaptive correction coefficient.

[0080] Adaptive Weather Influence Factors: (twenty one) Considering both degradation and adaptive VPP available power: (twenty two) To intuitively demonstrate the practical application logic of the coupled attenuation model and adaptive correction mechanism of this invention, example calculations are performed using typical combined extreme scenarios to verify the rationality of the model's calculations and its technical effectiveness. For example, in a combined scenario of heat waves and sandstorms: (twenty three) Verification results: Ancillary service revenue increased by 8.3%, equipment wear and tear decreased by 15.7%, the model is effective, and the verification is valid.

[0081] S4: Construct an optimized scheduling model to complete dynamic scheduling. 1. Optimize the scheduling objective function. The linear programming method is used to optimally schedule the resources of the virtual power plant, with the objective function being to maximize revenue. (twenty four) In the formula: —Total number of scheduling periods, annual simulation takes T=8760; —frequency modulation service output at time t; —Price of frequency modulation service at time t; —T time, rotating for standby output; —Price of spare parts at time t; —Non-rotating standby output at time t; —Price of non-rotating spare parts at time t.

[0082] 2. Constraints These include: power balance constraints, energy storage system constraints (SOC upper and lower limits, charging and discharging power limits), load regulation constraints (regulation range limits), and ancillary service demand constraints.

[0083] (1) Power balance constraint (25) (2) Energy storage SOC constraint (26) (3) Energy storage charging power constraints (27) (4) Energy storage discharge power constraint (28) (5) Controllable load reduction constraints (29) (6) Controllable load increases constraints (30) (7) Constraints on demand for ancillary services (31) (32) (33) This paper simulates five extreme scenarios to evaluate the economic viability of peak-valley arbitrage for energy storage systems. The simulated users have a voltage level of 220kV and above, an energy storage system capacity of 1000kWh, an energy storage system power of 500kW, a system lifespan of ten years, a discount rate of 6%, and a basic electricity price of 20 yuan / kW·month.

[0084] Example 2 This embodiment provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements all the steps of the virtual power plant auxiliary service optimization scheduling method under multiple extreme scenarios described in Embodiment 1.

[0085] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for optimizing and scheduling virtual power plant ancillary services under multiple extreme scenarios, characterized in that, Includes the following steps: S1: Establish a distributed energy resource model that includes photovoltaic power plants, wind farms, energy storage systems, and controllable loads, and construct a virtual power plant aggregation model and an ancillary service market model; S2: Construct a system of single extreme scenarios and composite extreme scenarios covering heat waves, cold waves, rainstorms, sandstorms, and extreme winds. Complete scenario generation and dimensionality reduction through temporal modeling, extreme event implantation, and multi-scenario clustering reduction. S3: Construct a coupled performance degradation model for composite extreme scenarios, introduce an adaptive learning mechanism for weather influencing factors, and correct the weather influencing factors in real time based on the deviation between actual output and predicted output. S4: With the goal of achieving optimal synergy between ancillary service revenue and equipment loss, a linear programming optimization scheduling model is constructed, which combines power balance, energy storage constraints, load regulation constraints, and ancillary service demand constraints to complete dynamic scheduling.

2. The virtual power plant ancillary service optimization scheduling method under multiple extreme scenarios according to claim 1, characterized in that: The distributed energy model in S1 includes: (1) Photovoltaic power station model: Where: G is solar irradiance, W / m²; A is photovoltaic panel area, m²; η is power generation efficiency; α is annual degradation rate; t is the operating life of the power station; β is temperature coefficient; T is ambient temperature, °C; Factors influencing extreme weather; (2) Wind farm model: The output power of a wind farm depends on wind speed and extreme weather conditions, as shown in the following model: in: The actual wind speed (m / s); To cut in wind speed; Rated wind speed; To cut off the wind speed; Rated power; (3) The energy storage system model includes the available charging power model, the available discharging power model, and the state of charge (SOC) update model: Available charging power: Available discharge power: SOC Update: in: This is the maximum charging power; This represents the maximum discharge power. and These are the upper and lower limits of SOC; Battery capacity; and For charge and discharge efficiency; For time step; (4) Controllable load model Reduced load: Increased load: in: Current load; Basic load; The load amplitude is adjustable.

3. The virtual power plant ancillary service optimization scheduling method under multiple extreme scenarios according to claim 1, characterized in that: The virtual power plant model in S1 is the available power model of the virtual power plant after integrating various distributed energy sources, expressed as:

4. The virtual power plant ancillary service optimization scheduling method under multiple extreme scenarios according to claim 1, characterized in that: The ancillary services market model in S1 includes a demand model and a price model, both of which are affected by time and weather conditions: Demand Model: Price model: in, Basic requirements; Base price; The time-related factor; Weather influencing factors; This is the price elasticity coefficient.

5. The virtual power plant ancillary service optimization scheduling method under multiple extreme scenarios according to claim 1, characterized in that: The influencing factor functions for the five types of single extreme weather in S2 are as follows: heat wave scene It is a piecewise linear function based on ambient temperature; Cold wave scene It is a three-segment linear function based on ambient temperature; Storm scene It is a piecewise exponential function based on rainfall intensity; Sandstorm scene It is an exponential decay function based on dust concentration; Extreme wind scenarios It is a piecewise linear function based on wind speed.

6. The virtual power plant ancillary service optimization scheduling method under multiple extreme scenarios according to claim 1, characterized in that: The scene generation method in S2 takes historical meteorological data as input and executes the following steps sequentially: Based on historical meteorological data, an autoregressive moving average model is used to generate time-series curves of basic meteorological parameters; Using the time-series curves of the basic meteorological parameters as input, extreme weather events are incorporated to construct a complete meteorological event curve that includes event intensity, peak time, and duration. The complete meteorological event curves are used to form an original scene set. The K-means clustering algorithm is used to reduce the scene set, extract typical meteorological scenes, and calculate the corresponding probability weights. The typical meteorological scenario is taken as input and combined with the distributed energy model of S1 to be coupled and mapped to the available power scenario of VPP through VPP response characteristics; Based on the time series of the typical meteorological scenarios and the available power scenarios of VPP, a Markov state transition model is established to simulate the continuous evolution process of normal weather, single extreme scenarios, and compound extreme scenarios.

7. The virtual power plant ancillary service optimization scheduling method under multiple extreme scenarios according to claim 1, characterized in that: The combined extreme weather coupling performance degradation model in S3 is as follows: in, , These are the influencing factors of different single extreme weather events. This is the irreversible degradation coefficient of equipment performance.

8. The virtual power plant ancillary service optimization scheduling method under multiple extreme scenarios according to claim 1, characterized in that: The adaptive correction model for weather influencing factors in S3 is as follows: 。 9. The virtual power plant ancillary service optimization scheduling method under multiple extreme scenarios according to claim 1, characterized in that: In S4, the objective function of the optimized scheduling model is: In the formula: This represents the total number of scheduling periods; Contribute to frequency modulation service at time t; Price of frequency modulation service at time t; The standby output is the rotational power at time t; The price for rotating spare parts at time t; The non-rotating standby output at time t; Let t be the non-rotating spare price.

10. The method for optimizing and scheduling virtual power plant ancillary services under multiple extreme scenarios according to claim 1, characterized in that: The constraints described in S4 include power balance constraints, energy storage SOC constraints, charge / discharge power constraints, controllable load regulation constraints, and ancillary service demand constraints, specifically: (1) Power balance constraint (2) Energy storage SOC constraint (3) Energy storage charging power constraints (4) Energy storage discharge power constraint (5) Controllable load reduction constraints (6) Controllable load increases constraints (7) Constraints on demand for ancillary services in, , These are the adjustable upper and lower limits for the load, respectively. , , These are frequency modulation, spinning reserve, and non-spinning reserve requirements, respectively.

Images