Robust optimization scheduling method and system for virtual power plant under multiple uncertainties

Abstract

This invention provides a robust optimization scheduling method and system for virtual power plants considering multiple uncertainties, relating to the field of power market dispatching technology. The method includes: establishing a virtual power plant dispatching model based on ARIMA-SVR, where uncertainties include distributed energy generation, electricity prices, and load demand; extracting features of the uncertainties, constructing an uncertainty set, and transforming the virtual power plant dispatching model into a robust optimization scheduling model; and using a particle swarm optimization algorithm to solve the robust optimization scheduling model and optimize the dispatching strategies for each energy resource within the virtual power plant. This invention, by constructing a robust optimization scheduling model considering multiple uncertainties and combining it with advanced solving algorithms, achieves optimized dispatching of virtual power plants in complex market environments, improving the dispatching efficiency, economic benefits, and ability to cope with uncertainties in complex market environments.

Description

Technical Field

[0001] This invention relates to the field of power market dispatching technology, specifically to a robust optimization dispatching method and system for virtual power plants considering multiple uncertainties. Background Technology

[0002] With the widespread integration of distributed energy resources and the advancement of electricity market reforms, virtual power plants (VPS), as a new type of entity integrating distributed power sources, energy storage systems, and controllable loads, are playing an increasingly important role in the electricity market. However, VPS face numerous uncertainties when participating in market dispatch. First, distributed energy generation exhibits significant randomness and volatility. For example, solar photovoltaic power generation is affected by sunlight intensity and weather conditions, while wind power generation is constrained by wind speed and direction changes, making it difficult to accurately predict power generation. Second, electricity market prices fluctuate frequently, influenced by various factors such as supply and demand, energy policies, and market competition. This price uncertainty increases the difficulty of predicting the revenue of VPS. Furthermore, load demand is also uncertain; changes in different seasons, time periods, and user behavior can all lead to load fluctuations.

[0003] While research on the dispatching of virtual power plants has made some progress, it still falls short in addressing multiple uncertainties. Some studies consider only a single uncertainty factor, failing to fully reflect the complexities faced by virtual power plants in actual operation. For example, some studies focus only on the uncertainty of distributed energy generation, ignoring fluctuations in electricity prices and load demand, leading to poor performance of dispatching strategies in practical applications. Studies considering multiple uncertainties mostly employ stochastic programming or probability distribution methods, but these methods suffer from high model complexity and difficulty in solving complex uncertainty scenarios, and their descriptions of uncertainty factors are not precise enough to provide effective dispatching decision support for virtual power plants.

[0004] In summary, existing technologies cannot effectively address the robust optimization scheduling problem of virtual power plants under multiple uncertainties, and are insufficient in improving the competitiveness and operational stability of virtual power plants in the market. Summary of the Invention

[0005] To address the aforementioned problems, this invention proposes a robust optimization scheduling method and system for virtual power plants that considers multiple uncertainties. By constructing a robust optimization scheduling model that considers multiple uncertainties and combining it with advanced solution algorithms, the invention enables optimized scheduling of virtual power plants in complex market environments, thereby improving the scheduling efficiency, economic benefits, and ability to cope with uncertainties of virtual power plants in complex market environments.

[0006] According to some embodiments, the present invention adopts the following technical solution:

[0007] Robust optimization scheduling methods for virtual power plants under multiple uncertainties include:

[0008] Based on ARIMA-SVR, a virtual power plant dispatch model is established that considers multiple uncertainties, including distributed energy generation, electricity prices, and load demand.

[0009] Extract the characteristics of uncertainty factors, construct an uncertainty set, and transform the virtual power plant scheduling model into a robust optimization scheduling model;

[0010] The particle swarm optimization algorithm is used to solve the robust optimization scheduling model and optimize the scheduling strategy of each energy resource in the virtual power plant.

[0011] According to some embodiments, the present invention adopts the following technical solution:

[0012] Consider a robust optimization scheduling system for virtual power plants under multiple uncertainties, including:

[0013] The model building module is configured to: build a virtual power plant dispatch model based on ARIMA-SVR, considering multiple uncertainties, including distributed energy generation, electricity price and load demand;

[0014] The model transformation module is configured to: extract the features of uncertainty factors, construct an uncertainty set, and transform the virtual power plant scheduling model into a robust optimization scheduling model.

[0015] The model solving module is configured to use the particle swarm optimization algorithm to solve the robust optimization scheduling model and optimize the scheduling strategy of each energy resource in the virtual power plant.

[0016] According to some embodiments, the present invention adopts the following technical solution:

[0017] A computer program product includes a computer program that, when executed by a processor, implements the robust optimization scheduling method for virtual power plants considering multiple uncertainties.

[0018] According to some embodiments, the present invention adopts the following technical solution:

[0019] A non-transitory computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the robust optimization scheduling method for virtual power plants considering multiple uncertainties.

[0020] According to some embodiments, the present invention adopts the following technical solution:

[0021] An electronic device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the robust optimization scheduling method for virtual power plants considering multiple uncertainties.

[0022] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0023] (1) Comprehensive consideration of uncertainty: Taking into account the multiple uncertainties of distributed energy generation, electricity price and load demand, a virtual power plant scheduling model that is closer to reality is constructed. It can more accurately reflect the operation of virtual power plants in complex market environments and provide a more reliable basis for optimizing scheduling.

[0024] (2) Enhance robustness: Through the analysis of uncertainty factors, the robust optimization method is used to transform the uncertainty problem into a deterministic problem, and a robust optimization scheduling model is obtained. This effectively enhances the scheduling strategy's ability to cope with uncertainty, improves the stability and reliability of the virtual power plant operation, and reduces the risks caused by uncertainty.

[0025] (3) Efficient solution and optimization: The robust optimization scheduling model is solved by using the particle swarm optimization algorithm, which improves the solution efficiency and can quickly obtain the optimized scheduling strategy that meets the actual needs. Attached Figure Description

[0026] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0027] Figure 1 This is a flowchart of the method in Example 1. Detailed Implementation

[0028] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0029] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0030] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0031] Example 1

[0032] One embodiment of the present invention provides a robust optimization scheduling method for virtual power plants considering multiple uncertainties, including:

[0033] Step S1: Based on ARIMA-SVR, establish a virtual power plant dispatch model that considers multiple uncertainties, including distributed energy generation, electricity price, and load demand.

[0034] Step S2: Extract the features of the uncertainty factors, construct the uncertainty set, and transform the virtual power plant scheduling model into a robust optimization scheduling model;

[0035] Step S3: Use the particle swarm optimization algorithm to solve the robust optimization scheduling model and optimize the scheduling strategy of each energy resource in the virtual power plant.

[0036] As one embodiment, the robust optimization scheduling method for virtual power plants under multiple uncertainties in this invention constructs a robust optimization scheduling model considering multiple uncertainties and combines it with advanced solution algorithms to achieve optimal scheduling of virtual power plants in complex market environments. This improves the scheduling efficiency, economic benefits, and ability to cope with uncertainties of virtual power plants in complex market environments. Figure 1 As shown, the specific implementation process is as follows:

[0037] I. Establishing a virtual power plant scheduling model based on ARIMA-SVR that considers multiple uncertainties;

[0038] To address the shortcomings of existing scheduling models that cannot fully consider various uncertainties, this embodiment constructs a virtual power plant scheduling model that comprehensively considers the uncertainties of distributed energy generation, electricity prices, and load demand. The specific steps are as follows:

[0039] Step 1: Model the internal resources of the virtual power plant.

[0040] Detailed modeling is performed on the internal resources of the virtual power plant, including distributed power sources (such as photovoltaic power plants (PV), wind farms (WT), energy storage systems (ESS), and controllable loads (CL):

[0041] (1) Distributed power sources

[0042] Distributed power sources establish power generation prediction models based on their power generation characteristics. In this embodiment, prediction models for photovoltaic power plants (PV) and wind farms (WT) are provided.

[0043] Photovoltaic power plant power generation capacity The prediction model is expressed by the formula:

[0044] (1)

[0045] in, Rated power of photovoltaic modules G This represents the actual light intensity. The light intensity under standard test conditions. For power temperature coefficient, T For the temperature of photovoltaic modules, Temperature under standard test conditions.

[0046] Wind farm power generation Based on wind speed The relationship with the power curve is predicted, and expressed by the formula:

[0047] (2)

[0048] in, , , These are the cut-in wind speed, rated wind speed, and cut-out wind speed, respectively. This indicates the rated power (maximum output power) of the wind power. This represents the actual wind speed.

[0049] (2) Energy Storage System (ESS)

[0050] Regarding energy storage systems (ESS), considering the limitations of charging and discharging power and state of charge (SOC) constraints, the formula is as follows:

[0051] (3)

[0052] (4)

[0053] (5)

[0054] (6)

[0055] in, , The states of charge of the energy storage system at time t and t-1 are respectively. , These represent charging and discharging efficiencies, respectively. , The charging and discharging power at time t. For time intervals, These are the minimum and maximum values ​​of the state of charge, respectively. These are the minimum and maximum charging power, respectively. These represent the minimum and maximum discharge power, respectively.

[0056] (3) Controllable load CL

[0057] Controllable load power Based on its adjustable range and adjustment cost The model is constructed, and expressed by the formula:

[0058] (7)

[0059] (8)

[0060] in, To adjust costs, These are the minimum and maximum values ​​of the controllable load power, respectively. , This is the cost coefficient.

[0061] Step 2: Market environment modeling.

[0062] Considering the uncertainty of electricity market prices, a price prediction model is constructed using a combination of time series analysis and machine learning. Specifically:

[0063] First, the historical time series data of electricity prices are analyzed and predicted using the autoregressive integral moving average (ARIMA) model.

[0064] The general form of an ARIMA model is ARIMA(p,d,q), where p is the number of autoregressive terms, d is the difference order, and q is the number of moving average terms. The mathematical expression for the general form of an ARIMA model is:

[0065] (9)

[0066] (10)

[0067] (11)

[0068] in, Let B be the electricity price at time t, and B be the lag operator. It is an autoregressive polynomial. It is a moving average polynomial. It is a white noise sequence. These are the autoregressive coefficients. This is the moving average coefficient.

[0069] By differentiating the historical time series data of electricity prices, a suitable value of d is determined, and then the autoregressive coefficients are estimated. and moving average coefficient .

[0070] However, the ARIMA model only considers the time-series characteristics of electricity prices and is insufficient for some complex influencing factors (such as weather and policies). Therefore, a Support Vector Regression (SVR) model is introduced to correct the prediction results of the ARIMA model. The SVR model maps the input data to a high-dimensional feature space through nonlinear mapping, and performs linear regression in this space. Its objective function is:

[0071] (12)

[0072] The constraints are:

[0073] (13)

[0074] in, ω Let b be the weight vector, and b be the bias term. and Let be the slack variable, n be the number of samples, C be the penalty factor, and ϵ be the parameters of the insensitive loss function. To input data Functions that map (such as historical electricity prices, weather, etc.) to a high-dimensional feature space.

[0075] By solving the above optimization problem, the prediction function of the SVR model is obtained:

[0076] (14)

[0077] in, For input The corresponding predicted electricity price, and For Lagrange multipliers, For the new input sample, As training samples, The kernel function is expressed by the formula:

[0078] (15)

[0079] This paper synthesizes the prediction results of the ARIMA and SVR models. The ARIMA model is used to analyze historical time-series electricity price data, capturing its linear trends and periodic patterns to obtain preliminary electricity price forecasts. However, because the ARIMA model does not adequately account for complex nonlinear factors such as weather and policies, the SVR model is introduced to learn and correct the prediction errors of the ARIMA model. The SVR model processes complex influencing factors through nonlinear mapping and outputs corrected values. The preliminary ARIMA forecast and the corrected SVR value are then combined to obtain the final electricity price forecast. and prediction interval [ This approach takes into account both the regularity of time series data and the influence of nonlinear factors.

[0080] Using a combined ARIMA-SVR method similar to electricity price forecasting, a load forecasting model is established based on historical load data and relevant influencing factors (such as temperature and date type) to obtain the predicted load demand. and load demand forecast range [ ].

[0081] The core logic of the ARIMA-SVR combined method is "linear trend capture + non-linear factor correction". For electricity prices, non-linear factors include weather, policies, etc., while for load, non-linear factors include temperature, date type, etc. The two differ only in specific input features (e.g., using policy factors for electricity prices and temperature for load), but the core framework (ARIMA handles linearity + SVR handles non-linearity) is universal.

[0082] Step 3: Construct the objective function.

[0083] The objective function is to maximize the expected return of virtual power plants participating in the market. Expressed as a formula:

[0084]

[0085] (16)

[0086] Where T is the scheduling period. Let be the electricity market price at time t. Let be the power generation capacity of the distributed power source at time t. , Let t be the discharge and charging power of the energy storage system. Let be the power of the controllable load at time t. This is the cost function for distributed generation. For a general energy storage system operating cost function, The adjustment cost for controllable load.

[0087] Step 4: Set constraints.

[0088] (1) Power balance constraint: Ensure that the power generation of the virtual power plant is in balance with the load demand and the charging and discharging power of the energy storage at any time, which can be expressed by the formula:

[0089] (17)

[0090] DG stands for Distributed Generation Group. Let be the power generation of the i-th distributed power source at time t, and ESS be the set of energy storage systems. Let be the discharge power of the j-th energy storage system at time t. The charging power of the k-th energy storage system at time t, where CL is the set of controllable loads. The power of the l-th controllable load at time t. Let t be the load demand at time t.

[0091] (2) Equipment operation constraints: including distributed power generation power limits, energy storage system charging and discharging power and state of charge constraints, and controllable load adjustment range constraints, etc.

[0092] (3) Electricity market transaction constraints: Limiting the power purchase and sale capacity of virtual power plants, expressed by the formula:

[0093] (18)

[0094] (19)

[0095] in, Let t be the power output of the virtual power plant purchased from the market (a positive number indicates power purchase). This is the minimum power purchase capacity (usually 0 or a negative number, with a negative number indicating that power sales are permitted). The maximum power purchase capacity (subject to market trading rules or line capacity restrictions). Let t be the power output of the virtual power plant sold to the market (a positive number indicates power sales). This is the minimum electricity sales capacity (usually 0 or a negative number, with a negative number indicating that electricity purchase is permitted). This is the maximum power output for sale (subject to market trading rules or line capacity limitations).

[0096] II. Establish a robust optimization scheduling model based on the virtual power plant scheduling model;

[0097] Through uncertainty analysis, a robust optimization method is used to transform the uncertain problem into a deterministic problem, resulting in a robust optimized scheduling model. The specific steps are as follows:

[0098] Step 1: Extract the characteristics of uncertainty factors through uncertainty factor analysis.

[0099] To address the uncertainties in distributed energy generation, electricity prices, and load demand, key characteristics are extracted to comprehensively describe the changing patterns of these uncertainties:

[0100] (1) For distributed energy generation, in addition to the mean and variance of power generation, its changing trend and correlation characteristics are also considered;

[0101] (2) For electricity prices, analyze their fluctuation range, seasonal variation patterns, etc.;

[0102] (3) For load demand, study its peak and valley characteristics and its correlation with meteorological factors.

[0103] Step 2: Constructing the set of uncertainties.

[0104] Based on the characteristics of the uncertainty factors, an uncertainty set is constructed, which covers all possible values ​​of the uncertainty factors within the scheduling period:

[0105] (1) For the uncertainty of distributed energy generation power, the possible range of power generation values ​​is determined by analyzing historical data and prediction errors, forming a set of power generation uncertainty. ;

[0106] (2) For electricity price uncertainty, an electricity price uncertainty set is constructed based on the electricity price prediction range and fluctuation characteristics. ;

[0107] (3) For load demand uncertainty, establish a load demand uncertainty set. .

[0108] Step 3: Robust optimization conversion.

[0109] By employing robust optimization methods, the uncertain problem is transformed into a deterministic problem, and the virtual power plant dispatch model is modified by introducing robust adjustment parameters. Γ By adjusting the objective function and constraints, a robust optimization scheduling model is obtained.

[0110] The objective function aims to maximize the expected revenue of the virtual power plant, taking into account the impact of uncertainties such as electricity prices, distributed generation capacity, and load demand on the revenue, and introducing... Γ Transforming "expected return" into "worst-case return floor" ensures that the basic return is guaranteed even within a range of uncertain fluctuations. The adjusted result is:

[0111] (20)

[0112] (twenty one)

[0113] in, It is a set of uncertainties (including fluctuations in electricity prices, power generation, load, etc.). As a robust penalty term, the larger Γ is, the more "reservations" are made for uncertainty, ensuring that there are still basic returns even in the worst-case scenario. The magnitude of return fluctuations due to uncertainty. Controlling the degree of "penalty" for fluctuations ( The larger the value, the higher the priority is given to ensuring the lower limit of returns. , , These are the uncertainties (actual values) of electricity price, distributed power generation capacity, and load demand at time t, respectively. , , These are the predicted means of the variables mentioned above;

[0114] The power balance constraint in the constraint conditions is transformed into a robust constraint as follows:

[0115] (twenty two)

[0116] in, This is the average value of the load demand forecast. Γ Used to control the degree of robustness Γ The larger the value, the more robust the scheduling strategy is to uncertainty, but this may come at the cost of some economic benefits.

[0117] By adjusting Γ The value seeks a balance between economic benefits and robustness.

[0118] III. Solving the robust optimization scheduling model;

[0119] In this embodiment, the improved particle swarm optimization algorithm (IPSO) is used to solve the robust optimization scheduling model. The specific steps are as follows:

[0120] Step 1: Particle initialization.

[0121] Within the constraints, particle positions and velocities are randomly generated. Particle positions represent the scheduling strategies for various energy resources within the virtual power plant, such as the power allocation of distributed generation sources (corresponding to variables). Energy storage system charge and discharge plan (corresponding variables) , , ), controllable load adjustment (corresponding variable) Assuming the particle dimension is n, the position vector of the kth particle... and velocity vector Represented as:

[0122] (twenty three)

[0123] (twenty four)

[0124] Step 2: Fitness value calculation.

[0125] The fitness value of each particle is calculated based on the objective function of the robust optimization scheduling model, which involves substituting the particle position vector into the objective function. In the process, the fitness value is obtained. The higher the fitness value, the higher the expected return of the virtual power plant under the scheduling strategy.

[0126] Step 3: Particle position and velocity update.

[0127] In traditional particle swarm optimization algorithms, the formulas for updating particle velocity and position are:

[0128] (25)

[0129] (26)

[0130] Where ω is the inertia weight, , As a learning factor, , A random number between [0,1] This represents the historical optimal position of particle k. This is the globally optimal position.

[0131] In the improved particle swarm optimization algorithm, dynamic inertia weights are introduced. :

[0132] (27)

[0133] in, , These are the maximum and minimum inertia weights, respectively. The maximum number of iterations is denoted by t, and the current iteration number is denoted by t, in order to balance the algorithm's global search and local search capabilities.

[0134] Meanwhile, to improve the efficiency of the algorithm in handling robust optimization problems, the particle position is corrected based on the uncertainty set to ensure that the particle always searches within the feasible region that satisfies the robust constraints. Specifically:

[0135] After the particle updates its position, through a pre-constructed set of uncertainties (such as power generation) Electricity price Load demand The possible range of values ​​is used to check the feasibility of the particle position for robust constraints (such as power balance, equipment operation limitations, etc.). If the particle position does not meet the constraints (such as power imbalance), the particle position is forced to return to the feasible region that meets the robust constraints based on the fluctuation range of the uncertainty set. This ensures that the scheduling strategy generated in each iteration can not only cope with multiple uncertainties, but also has practical operability, thereby dynamically balancing robustness and economic benefits in the algorithm search process.

[0136] Step 4: Iteration termination condition.

[0137] The algorithm terminates when the maximum number of iterations or the fitness value converges to a certain accuracy after iteratively executing steps 2-3, and outputs the global optimal solution, which is the robust optimization scheduling strategy of the virtual power plant under multiple uncertainties.

[0138] IV. Scheduling strategy evaluation and feedback mechanism.

[0139] This embodiment establishes a scheduling strategy evaluation and feedback mechanism, thereby achieving continuous optimization of the scheduling strategy and further improving the economic benefits and overall competitiveness of the virtual power plant. Specifically:

[0140] (1) Construction of scheduling strategy evaluation index system: Establish a comprehensive scheduling strategy evaluation index system, including economic benefit indicators (such as expected benefits and cost reduction rate), robustness indicators (such as the ability to cope with uncertainty and strategy stability) and environmental benefit indicators (such as carbon emission reduction). Through these indicators, the scheduling strategies obtained by the solution are comprehensively evaluated to fully measure the advantages and disadvantages of the scheduling strategies.

[0141] (2) Strategy Evaluation and Feedback: The obtained scheduling strategy is applied to the actual operation simulation of the virtual power plant, and the strategy is evaluated according to the evaluation index system. If the scheduling strategy performs poorly in actual operation, such as the expected benefits not meeting expectations or insufficient robustness, the evaluation results are fed back to the robust optimization scheduling model solution stage to adjust relevant parameters (such as robust adjustment parameter Γ, algorithm parameters, etc.). , , , , , (etc.), re-solve the model to obtain a better scheduling strategy. Through continuous evaluation and feedback, the scheduling strategy is continuously optimized.

[0142] Example 2

[0143] One embodiment of the present invention provides a robust optimization scheduling system for virtual power plants considering multiple uncertainties, comprising:

[0144] The model building module is configured to: build a virtual power plant dispatch model based on ARIMA-SVR, considering multiple uncertainties, including distributed energy generation, electricity price and load demand;

[0145] The model transformation module is configured to: extract the features of uncertainty factors, construct an uncertainty set, and transform the virtual power plant scheduling model into a robust optimization scheduling model.

[0146] The model solving module is configured to use an improved particle swarm optimization algorithm to solve the robust optimization scheduling model and optimize the scheduling strategy of each energy resource in the virtual power plant.

[0147] Example 3

[0148] One embodiment of the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the robust optimization scheduling method for virtual power plants considering multiple uncertainties.

[0149] Example 4

[0150] In one embodiment of the present invention, a non-transitory computer-readable storage medium is provided for storing computer instructions. When the computer instructions are executed by a processor, they implement the robust optimization scheduling method for virtual power plants considering multiple uncertainties.

[0151] Example 5

[0152] One embodiment of the present invention provides an electronic device, including: a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the robust optimization scheduling method for virtual power plants considering multiple uncertainties.

[0153] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0154] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0155] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A robust optimization scheduling method for virtual power plants under multiple uncertainties, characterized in that, include: Based on ARIMA-SVR, a virtual power plant dispatch model is established that considers multiple uncertainties, including distributed energy generation, electricity prices, and load demand. The characteristics of uncertain factors are extracted, an uncertainty set is constructed, and the virtual power plant scheduling model is transformed into a robust optimization scheduling model. The specific steps of the robust optimization transformation are as follows: By employing robust optimization methods, the uncertain problem is transformed into a deterministic problem, and the virtual power plant dispatch model is modified by introducing robust adjustment parameters. Γ By adjusting the objective function and constraints, a robust optimization scheduling model is obtained. The objective function aims to maximize the expected revenue of the virtual power plant, taking into account the impact of uncertainties such as electricity price, distributed power generation, and load demand on the revenue. Γ Transforming "expected return" into "worst-case return floor" ensures that the basic return is guaranteed even within a range of uncertain fluctuations. The adjusted value is: in, For an uncertain set, As a robust penalty term, the larger Γ is, the more "reservations" are made for uncertainty, ensuring that there is still a basic return even in the worst case. The magnitude of return fluctuations due to uncertainty. Control the degree of "punishment" for fluctuations; , , These are the uncertainties related to electricity price, distributed power generation, and load demand at time t, respectively. , , The predicted mean of the above variables, the power balance constraint in the constraints, and the transformed robust constraint are as follows: in, This is the average value of the load demand forecast. Γ Used to control the degree of robustness Γ The larger the value, the more robust the scheduling strategy is to uncertainty, but this comes at the cost of some economic benefits. This can be addressed by adjusting... Γ The value of [the product / service] should be balanced between economic efficiency and robustness. The particle swarm optimization algorithm is used to solve the robust optimization scheduling model and optimize the scheduling strategy of each energy resource in the virtual power plant.

2. The robust optimization scheduling method for virtual power plants considering multiple uncertainties as described in claim 1, characterized in that, The specific steps for establishing the virtual power plant scheduling model are as follows: Virtual power plant internal resource modeling includes modeling distributed power sources, energy storage systems (ESS), and controllable loads (CL) within the virtual power plant; Market environment modeling, including building an electricity price forecasting model based on ARIMA-SVR; Objective function construction: The objective function is constructed with the goal of maximizing the expected returns of virtual power plants participating in the market. Constraint settings include power balance constraints, equipment operation constraints, and electricity market transaction constraints.

3. The robust optimization scheduling method for virtual power plants considering multiple uncertainties as described in claim 2, characterized in that, The electricity price forecasting model based on ARIMA-SVR is specifically as follows: Considering the time-series characteristics of electricity prices, an ARIMA model is constructed using historical time-series data of electricity prices; The SVR model is introduced to correct the prediction results of the ARIMA model, resulting in the final electricity price prediction model.

4. The robust optimization scheduling method for virtual power plants considering multiple uncertainties as described in claim 1, characterized in that, The extraction of features of uncertainty factors and the construction of an uncertainty set are specifically as follows: Analyze uncertain factors and extract features that comprehensively describe the changing patterns of these uncertain factors; Calculate the possible value range of the characteristics of the uncertainty factors within the scheduling period, and form an uncertainty set.

5. The robust optimization scheduling method for virtual power plants considering multiple uncertainties as described in claim 1, characterized in that, The transformation of the virtual power plant scheduling model into a robust optimization scheduling model involves using robust optimization methods to convert an uncertain problem into a deterministic problem. Specifically: Robust adjustment parameters are introduced to adjust the objective function and constraints of the virtual power plant dispatch model.

6. The robust optimization scheduling method for virtual power plants considering multiple uncertainties as described in claim 1, characterized in that, The particle swarm optimization algorithm is used to solve the robust optimization scheduling model, specifically as follows: The particle position represents the scheduling strategy of each energy resource in the virtual power plant. Within the range of constraints, the particle position and velocity are randomly generated to complete the particle initialization. Substitute the particle position vector into the objective function to calculate the fitness value; Based on the optimal position, the particle position and velocity of the current iteration are updated. After the particle position is updated, the particle position is corrected using the uncertainty set to ensure that the particle always searches within the feasible region that satisfies the robust constraints. The process iteratively executes two steps: fitness value calculation and particle position and velocity update, until the iteration termination condition is met, resulting in the global optimal solution.

7. A robust optimization scheduling system for virtual power plants under multiple uncertainties, characterized in that, include: The model building module is configured to: build a virtual power plant dispatch model based on ARIMA-SVR, considering multiple uncertainties, including distributed energy generation, electricity price and load demand; The model transformation module is configured to: extract the features of uncertainty factors, construct an uncertainty set, and transform the virtual power plant scheduling model into a robust optimization scheduling model. The specific steps of the robust optimization transformation are as follows: By employing robust optimization methods, the uncertain problem is transformed into a deterministic problem, and the virtual power plant dispatch model is modified by introducing robust adjustment parameters. Γ By adjusting the objective function and constraints, a robust optimization scheduling model is obtained. The objective function aims to maximize the expected revenue of the virtual power plant, taking into account the impact of uncertainties such as electricity price, distributed power generation, and load demand on the revenue. Γ Transforming "expected return" into "worst-case return floor" ensures that the basic return is guaranteed even within a range of uncertain fluctuations. The adjusted value is: in, For an uncertain set, As a robust penalty term, the larger Γ is, the more "reservations" are made for uncertainty, ensuring that there is still a basic return even in the worst case. The magnitude of return fluctuations due to uncertainty. Control the degree of "punishment" for fluctuations; , , These are the uncertainties related to electricity price, distributed power generation, and load demand at time t, respectively. , , The predicted mean of the above variables, the power balance constraint in the constraints, and the transformed robust constraint are as follows: in, This is the average value of the load demand forecast. Γ Used to control the degree of robustness Γ The larger the value, the more robust the scheduling strategy is to uncertainty, but this comes at the cost of some economic benefits. This can be addressed by adjusting... Γ The value of [the product / service] should be balanced between economic efficiency and robustness. The model solving module is configured to use the particle swarm optimization algorithm to solve the robust optimization scheduling model and optimize the scheduling strategy of each energy resource in the virtual power plant.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the robust optimization scheduling method for virtual power plants considering multiple uncertainties as described in any one of claims 1-6.

9. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement the robust optimization scheduling method for virtual power plants considering multiple uncertainties as described in any one of claims 1-6.

10. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the robust optimization scheduling method for virtual power plants considering multiple uncertainties as described in any one of claims 1-6.

Images