SRSM-based multi-microgrid cluster voltage risk pre-control calculation method and related equipment

Abstract

The invention discloses an SRSM-based multi-microgrid cluster voltage risk pre-control calculation method and related equipment, and belongs to the field of multi-microgrid cluster random optimization. The method comprises the steps of building and training a neural network model, and obtaining a node voltage inversion model; based on SRSM and Nataf inverse transformation, photovoltaic and load random input quantities are obtained, a micro-grid random optimization model is constructed, and a day-ahead scheduling strategy of the micro-grid and random response of interaction power are solved and obtained; forming a time sequence by the obtained random response of the interaction power and meteorological data, and inputting the time sequence into a node voltage inversion model to obtain random response of a voltage Gaussian mean value and a standard deviation; constructing a chance constraint model; if the node voltage does not meet the confidence coefficient requirement, adjusting the upper and lower limits of the PCC port power according to the port voltage opportunity constraint, and returning to optimize the micro-grid; and if yes, outputting the obtained day-ahead scheduling strategy of the micro-grid. According to the method, the optimal multi-microgrid cooperative scheduling strategy considering the port voltage risk and the source load uncertainty is obtained.

Description

Technical Field

[0001] This invention relates to the field of stochastic optimization of multi-microgrid clusters, and in particular to a method and related equipment for voltage risk pre-control calculation of multi-microgrid clusters based on SRSM. Background Art

[0002] Currently, considering the privacy constraints between microgrids and the distribution network, when formulating dispatch strategies for each microgrid, multi-microgrid aggregators treat all distribution network data except for their own port data as private data. Existing day-ahead dispatch strategies for microgrids lack effective microgrid port voltage risk assessment. Current research on microgrid optimal dispatch is mainly based on optimal microgrid unit combination and economic optimality models, employing robust optimization, stochastic optimization, and chance-constrained models to account for the impact of microgrid source-load uncertainties. Regarding multi-microgrid collaborative optimal dispatch strategies, game theory and distributed optimization methods are mainly used, but these methods are optimal for microgrids under the control of the distribution network. Multi-microgrids are interconnected with the distribution network through points of common coupling (PCCs), and the power and voltage at each microgrid port exhibit temporal correlation. Therefore, how to utilize the observable information of the microgrid cluster itself to invert the temporal patterns of each port voltage and actively suppress the trigger voltage risk boundary is a pressing problem in current research on microgrid day-ahead optimal dispatch strategies. Random fluctuations in distributed power sources and load demand in microgrids can affect the microgrid's dispatch strategy, causing fluctuations in power at microgrid ports. The impact on port voltage also needs to be assessed. The voltage risk probability is affected by both the microgrid's own fluctuations and the uncertainty of the distribution network's timing characteristics. Constructing an accurate random response model to describe the random response output of the voltage at each microgrid port is key to effectively managing the risk. Summary of the Invention

[0003] In order to at least partially solve one of the technical problems existing in the prior art, the purpose of this invention is to provide a multi-microgrid cluster voltage risk pre-control calculation method and related equipment based on SRSM.

[0004] The first technical solution adopted in this invention is:

[0005] A method for voltage risk prediction and control calculation of multi-microgrid clusters based on SRSM includes the following steps:

[0006] Model building and training: Build a neural network model with GRU as the main network, train the model, and obtain the node voltage inversion model;

[0007] Microgrid stochastic optimization: Obtaining stochastic inputs for photovoltaics and loads based on SRSM and Nataf inverse transformation. A stochastic optimization model for a microgrid, considering source-load uncertainty, is constructed and solved to obtain the day-ahead dispatch strategy and the stochastic response of the interactive power.

[0008] Node voltage inversion: the random response of the optimized interactive power. When combined with meteorological data to form a time series, it is input into the nodal voltage inversion model to obtain the random response of the voltage with Gaussian mean and standard deviation.

[0009] Constraint Iterative Optimization: An opportunity constraint model for node voltage under dual uncertainty is constructed based on the characteristic coefficients of the Hermite chaotic polynomial. If the node voltage does not meet the confidence requirement, the upper and lower limits of the PCC port power are adjusted according to the port voltage opportunity constraint, and the microgrid stochastic optimization step is returned. If the confidence requirement is met, the loop ends and the obtained day-ahead scheduling strategy of the microgrid is output.

[0010] Furthermore, the model is built and trained, including:

[0011] A neural network model with GRU as the main network is constructed; historical interaction power and meteorological data are combined into a time series, and historical node voltage is used as the training label to train the model; the neural network model is represented as follows:

[0012] y t =f nn (x t )

[0013] In the formula: f nn () represents a neural network model with GRU as the main network, x t As the input to the neural network model, y t =[U 1,t ,…,U N,t ] T The voltage at the microgrid PCC port during time period t;

[0014] A Gaussian process regression function, y, is introduced. t The Gaussian kernel function model is represented as:

[0015] y m,t ,y s,t =f gpr (y t )

[0016] Where f gpr () is derived from historical node voltage and f nn () The historical voltage inversion was obtained by fitting.

[0017] Let y t The mean and standard deviation are ym,t =[U m,1,t ,…,U m,N,t ] T ,y s,t =[U s,1,t ,…,U s,N,t ] T A node voltage inversion model based on time-series networks and Gaussian process regression f NVM () is represented as:

[0018] y t ,y m,t ,y s,t =f NVM (x t )

[0019] Where N is the number of microgrids.

[0020] Furthermore, the input to the neural network model tw represents the length of the input time series, d is the number of feature dimensions; P PCC,t Q represents the active power of PCC during time period t. PCC,t This represents the reactive power of PCC during time period t, in W. t This represents meteorological factors during time period t;

[0021] f is obtained using the automatic differentiation calculation of the PyTorch deep learning framework. nn (x t Jacobian matrix of )

[0022]

[0023] In the formula: This represents the input / output sensitivity coefficient matrix of the NVIM model, with matrix elements... This indicates the sensitivity of the power change at port j to the voltage change at port i during time period t.

[0024] Furthermore, the microgrid stochastic optimization includes:

[0025] In each microgrid, considering the two-dimensional uncertainties of photovoltaics and load as random inputs, and based on stochastic response surface theory, N is sampled. a A random sample is used as input. The superscript (m) indicates the m-th sample variable; each sample is a 2-dimensional vector representing a 2-dimensional uncertain input.

[0026] Based on SRSM, random input The standard sample ξ can be transformed using the Nataf transform:

[0027]

[0028] In the formula: Nataf -1 This represents the inverse Nataf transform;

[0029] Selecting a second-order chaotic polynomial to map the microgrid stochastic response process:

[0030]

[0031] In the formula: The random response output of microgrid i is represented by undetermined coefficients. Let ξ be the eigencoefficients of the Hermite chaotic polynomial, and H(ξ) denote the Hermite matrix.

[0032] Based on SRSM, chaotic polynomial coefficients It can characterize the probability distribution information of the random response output, where the random response output mean and standard deviation and The relationship is:

[0033]

[0034] A stochastic optimization model for microgrids is constructed, taking into account the impact of random fluctuations in source load on the operating cost of microgrids, with the minimum of the mean and standard deviation of operating costs as the optimization objective.

[0035] The gurobi solver is used to solve the microgrid optimization model, yielding the day-ahead optimization strategy and the stochastic response of the interactive power.

[0036] Furthermore, the construction of the microgrid stochastic optimization model includes:

[0037] The mathematical model for each microgrid's photovoltaic and load components, including uncertainties, is expressed as follows:

[0038]

[0039]

[0040] In the formula: These represent the random and predicted values ​​of photovoltaic active power, respectively. Indicates the cost of PV curtailment; c PV,i,t This indicates the unit price of PV curtailment; Indicates the percentage of light wasted; These represent the random and predicted values ​​of the active power of the load, respectively; Indicates reactive power load, and Related, The power factor tangent of the load is represented; Δt represents the time interval.

[0041] Energy storage mathematical model:

[0042]

[0043] In the formula: These represent the active power and reactive power of the energy storage ESS, respectively. These represent the ESS discharge and charging power, respectively. This indicates the state of charge of the ESS for the two time periods before and after it. These represent the discharge and charging efficiencies of the ESS, respectively. C represents the charge / discharge state of the ESS, and is a Boolean variable; ESS,i,t Indicates the operating loss cost of ESS; c ESS,i,t This indicates the unit price of ESS charging and discharging losses; These represent the maximum values ​​of active and reactive power in the ESS, respectively.

[0044] Distribution-microgrid PCC port mathematical model:

[0045]

[0046] In the formula: These represent the active power and reactive power supplied by the distribution network to the microgrid, respectively. These represent the active / reactive power input from the distribution network to the microgrid and the active / reactive power output from the microgrid to the distribution network, respectively. This indicates the cost of purchasing / selling electricity at the PCC port; These represent the unit price for purchasing and selling electricity, respectively. Indicates the power tangent; The power states of the distribution network to the microgrid, respectively, are Boolean variables. This indicates the power transmitted from the distribution network to the microgrid; The maximum and minimum values ​​of PCC active power transmission are respectively represented. Representing the maximum and minimum values ​​of PCC reactive power transmission respectively; Mathematical model of fast diesel generator:

[0047]

[0048] In the formula: These represent the active and reactive power outputs of FDE, respectively. These represent the maximum values ​​of active and reactive power in FDE, respectively; Indicates the operating cost of FDE; c FDE,i This represents the unit price of fuel cost for FDE;

[0049] Microgrid power balance constraints:

[0050]

[0051] Microgrid optimization objectives:

[0052]

[0053] In the formula: The stochastic response representing the operating cost of a microgrid; express The corresponding chaotic polynomial coefficients; They represent The mean and standard deviation.

[0054] Furthermore, the node voltage inversion includes:

[0055] The port power will be optimized. A time-series input node voltage inversion model was constructed using meteorological forecast data to obtain the port voltage inversion values, Gaussian mean, and Gaussian standard deviation of the stochastic response.

[0056] The mean and standard deviation of the random response with Gaussian mean and Gaussian variance of the port voltage are obtained based on the Hermite chaotic polynomial coefficients:

[0057]

[0058] In the formula: These represent the mean and standard deviation of the port voltages, respectively. H represents the chaotic polynomial coefficients corresponding to the Gaussian mean and Gaussian standard deviation of the port voltage, respectively. -1 (ξ) denotes the inverse of the Hermite matrix.

[0059] Furthermore, the constraint iterative optimization includes:

[0060] Construct port voltage opportunity constraints based on NVIM-SRSM;

[0061]

[0062] In the formula: z α U represents the z-value corresponding to a significance level of α in the normal z-value table. 1-α / 2,i,t and U α / 2,i,t These are the upper and lower bounds of the PCC voltage confidence interval for microgrid i during time period t;

[0063] When the random response of the port voltage does not meet the confidence requirement, the microgrid side needs to adjust the port power threshold and re-optimize to obtain a new microgrid optimization strategy; the PCC power limit adjustment formula for each microgrid is as follows:

[0064]

[0065] In the formula: These represent the upper and lower limits of the PCC power of microgrid i after adjustment.

[0066] The second technical solution adopted in this invention is:

[0067] An electronic device includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set, or an instruction set, and the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by the processor to implement the voltage risk pre-control calculation method for multi-microgrid clusters based on SRSM as described above.

[0068] The third technical solution adopted in this invention is:

[0069] A computer-readable storage medium storing at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the voltage risk pre-control calculation method for a multi-microgrid cluster based on SRSM as described above.

[0070] The fourth technical solution adopted in this invention is:

[0071] A computer program product or computer program includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform the aforementioned SRSM-based multi-microgrid cluster voltage risk pre-control calculation method.

[0072] The beneficial effects of this invention are as follows: Compared with existing multi-microgrid stochastic optimization methods, the proposed multi-microgrid stochastic optimization method considers the risk constraints of port voltage interaction between microgrids and distribution networks. It proposes a node voltage inverse model (NVIM) using time-series neural networks, introduces the stochastic response surface method (SRSM) to describe the stochastic response output of the multi-microgrid, and constructs port voltage opportunity constraints for NVIM-SRSM using Hermite chaotic polynomial coefficients. This invention adheres to the information privacy constraints of distribution networks, training the NVIM by sharing historical port power and voltage data among microgrids and combining it with historical meteorological information. Considering the uncertainties of source loads in each microgrid and the uncertainty of NVIM output deviation, the probability distribution function of port voltage under the action of dual uncertainties is derived using the characteristic coefficients of Hermite chaotic polynomial. Based on this, a multi-microgrid port voltage opportunity constraint based on NVIM-SRSM is established. Its confidence interval range is used to suppress the voltage risk of the microgrid cluster day-ahead scheduling strategy and guide the stochastic optimization strategy of multi-microgrid in reverse, thus obtaining the optimal multi-microgrid cooperative scheduling strategy considering port voltage risk and source load uncertainty. Attached Figure Description

[0073] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following description is provided with accompanying drawings of the relevant technical solutions in the embodiments of the present invention or the prior art. It should be understood that the accompanying drawings described below are only for the purpose of clearly illustrating some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0074] Figure 1 This is a flowchart illustrating the voltage risk pre-control calculation method for multi-microgrid clusters based on SRSM in an embodiment of the present invention;

[0075] Figure 2 This is a schematic diagram of the power distribution network topology in an embodiment of the present invention. Detailed Implementation

[0076] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application. The step numbers in the following embodiments are set only for ease of explanation, and there is no limitation on the order between the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0077] The terminology used in the embodiments of this application is for the purpose of describing specific embodiments only and is not intended to limit the embodiments of this application. The singular forms "a," "described," and "the" used in the embodiments of this application and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. Furthermore, unless otherwise expressly limited, terms such as "set," "install," and "connect" should be interpreted broadly, and those skilled in the art can reasonably determine the specific meaning of the above terms in this invention in conjunction with the specific content of the technical solution.

[0078] In the description of this application, it should be understood that the orientation descriptions, such as up, down, front, back, left, right, etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0079] In the description of this application, "several" means one or more, "more than" means two or more, "greater than," "less than," and "exceeding" are understood to exclude the stated number, while "above," "below," and "within" are understood to include the stated number. The use of "first" and "second" in the description is merely for distinguishing technical features and should not be construed as indicating or implying relative importance, or implicitly indicating the number of indicated technical features, or implicitly indicating the order of the indicated technical features.

[0080] In the description of this application, "and / or" describes the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the related objects before and after it are in an "or" relationship.

[0081] Terminology Explanation:

[0082] GRU: an abbreviation for Gated Recurrent Unit, is an improved variant of recurrent neural network (RNN). By introducing gating mechanisms such as update gates and reset gates, it effectively solves the gradient vanishing problem of traditional RNNs and is suitable for processing sequential data (such as natural language, time series, etc.).

[0083] SRSM: an abbreviation for stochastic response surface method, is a statistical method used to deal with engineering problems with uncertainty.

[0084] NVIM: an abbreviation for node voltage inverse model, a time-series neural network interval prediction model for node voltage inversion.

[0085] PCC: an abbreviation for point of common coupling, referring to the common coupling point used in multi-micronets.

[0086] GPR is an abbreviation for Gaussian process regression.

[0087] Example 1

[0088] like Figure 1 As shown, this embodiment provides a method for voltage risk pre-control calculation of multi-microgrid clusters based on SRSM, including the following steps:

[0089] Step 1: Build a neural network model with GRU as the main network, and use historical interaction power and meteorological data to form a time series, and historical node voltage as the training target for training. The training ends when the model converges.

[0090] In some embodiments, step 1 specifically includes the following steps:

[0091] Step 1.1: Assume the number of microgrids is N. Construct a neural network model with GRUs as the main network, using historical active and reactive power data at the ports and meteorological information to build a time series, and using historical port voltage as the training label for training. The specific network model is represented as follows:

[0092] y t =f nn (x t (1)

[0093] In equation (1), f nn (·) represents a network model with GRUs as the main units; x t =[P PCC,t Q PCC,t W t ]Τ Indicates the network input. tw represents the length of the input time series, with a feature dimension of d, containing active and reactive power data of PCC and strongly correlated meteorological information (temperature, humidity, solar irradiance, etc.) for the tw time period. W t This indicates the meteorological factors affecting the power distribution network during time period t. y t =[U 1,t ,…,U N,t ] Τ The voltage at the PCC port of the microgrid during time period t.

[0094] f can be obtained using the automatic differentiation calculation of the PyTorch deep learning framework. nn (x t Jacobian matrix of )

[0095]

[0096] Equation (2) represents the input / output sensitivity coefficient matrix of the NVIM model, and the matrix elements are... This indicates the sensitivity of the power change at port j to the voltage change at port i during time period t.

[0097] Step 1.2: A Gaussian process regression (GPR) function is introduced, y t The Gaussian kernel function model can be expressed as:

[0098] y m,t ,y s,t =f gpr (y t (3)

[0099] Where f gpr (·) is derived from historical node voltage and f nn (·) The voltage was obtained by historical inversion fitting.

[0100] Step 1.3: Let y t The mean and standard deviation are y m,t =[U m,1,t ,…,U m,N,t ] Τ ,y s,t =[U s,1,t ,…,U s,N,t ] Τ A node voltage inversion model based on time-series networks and Gaussian process regression f NVIM (·) can be represented as:

[0101] y t ,ym,t ,y s,t =f NVIM (x t (4)

[0102] Step 2: Using the prediction models for photovoltaics and loads and the inverse Nataf transform, the random input quantities for photovoltaics and loads can be obtained. Based on SRSM, a day-ahead optimization model for multiple microgrids that takes into account source-load uncertainties can be constructed. Solving this model yields the day-ahead scheduling strategy and the stochastic response of the interactive power of the microgrid.

[0103] In some embodiments, step 2 specifically includes the following steps:

[0104] Step 2.1: Considering the two-dimensional uncertainties of photovoltaics and load as random inputs in each microgrid, and based on the stochastic response surface theory, draw Na random samples as inputs. The superscripts "(0)", "(m)", and "(Na-1)" represent sample variables 0, m, and Na-1, respectively. Each sample is a 2-dimensional vector. This represents a two-dimensional uncertain input. Based on the stochastic response surface methodology, random input... The standard sample ξ can be transformed using the Nataf transform:

[0105]

[0106] In the formula: Nataf -1 This represents the inverse Nataf transform.

[0107] Selecting a second-order chaotic polynomial to map the microgrid stochastic response process:

[0108]

[0109] In the formula: Represents the random response output of microgrid i; undetermined coefficients. Represented as Hermite chaotic polynomial characteristic coefficients. H(ξ) is the Hermite matrix.

[0110] Based on SRSM, chaotic polynomial coefficients It can characterize the probability distribution information of the random response output, where the random response output mean and standard deviation and The relationship is:

[0111]

[0112] Step 2.2: Construct a stochastic optimization model for the microgrid.

[0113] The mathematical model for each microgrid's photovoltaic and load components, including uncertainties, can be expressed as follows:

[0114]

[0115] In the formula: These represent the random and predicted values ​​of photovoltaic active power, respectively. Indicates the cost of PV curtailment; c PV,i,t This indicates the unit price of PV curtailment; Indicates the percentage of light wasted; These represent the random and predicted values ​​of the active power of the load, respectively; Indicates reactive power load, and Related, Δt represents the power factor tangent of the load; Δt represents the time interval.

[0116] Energy storage mathematical model:

[0117]

[0118] In the formula: These represent the active and reactive power of the energy storage system (ESS), respectively. These represent the ESS discharge and charging power, respectively. This indicates the state of charge of the ESS for the two time periods before and after it. These represent the discharge and charging efficiencies of the ESS, respectively. C represents the charge / discharge state of the ESS, and is a Boolean variable; ESS,i,t Indicates the operating loss cost of ESS; c ESS,i,t This indicates the unit price of ESS charging and discharging losses; These represent the maximum values ​​of active and reactive power in the ESS, respectively.

[0119] Distribution-microgrid PCC port mathematical model:

[0120]

[0121] In the formula: These represent the active and reactive power supplied by the distribution network to the microgrid, respectively. These represent the active / reactive power input from the distribution network to the microgrid and the active / reactive power output from the microgrid to the distribution network, respectively. This indicates the cost of purchasing / selling electricity at the PCC port; These represent the unit price for purchasing and selling electricity, respectively. Indicates the power tangent; The power states of the distribution network to the microgrid, respectively, are Boolean variables. This indicates the power transmitted from the distribution network to the microgrid; The maximum and minimum values ​​of PCC active power transmission are respectively represented. These represent the maximum and minimum values ​​of reactive power transmission in PCC, respectively.

[0122] Mathematical model of fast diesel generator:

[0123]

[0124] In the formula: These represent the active and reactive power outputs of FDE, respectively. These represent the maximum values ​​of active and reactive power in FDE, respectively; Indicates the operating cost of FDE; c FDE,i This represents the unit price of fuel cost for FDE.

[0125] Microgrid power balance constraints:

[0126]

[0127] Microgrid optimization objectives:

[0128]

[0129] In the formula: The stochastic response representing the operating cost of a microgrid; express The corresponding chaotic polynomial coefficients; They represent The mean and standard deviation.

[0130] Considering the impact of random fluctuations in source load on the operating cost of microgrids, the optimization objective is to minimize the mean and standard deviation of the operating cost.

[0131]

[0132] Finally, the gurobi solver is used to solve the microgrid optimization model, and the day-ahead optimization strategy of the microgrid is obtained.

[0133] Step 3: Calculate the random response of the optimized interaction power. By combining meteorological data with a time series input into NVIM, a random response with Gaussian mean and standard deviation of voltage can be obtained.

[0134] In some embodiments, step 3 specifically includes the following steps:

[0135] Step 3.1: Derive the probability distribution coefficients of the PCC port voltage based on NVIM and SRSM.

[0136] The port power will be optimized. By constructing a time series input with meteorological forecast data and performing NVIM inversion, the port voltage inversion value, Gaussian mean, and Gaussian standard deviation of the random response are obtained:

[0137]

[0138] Step 3.2: Derive the distribution coefficient of the port voltage.

[0139] Based on the Hermite chaotic polynomial coefficients, the mean and standard deviation of the random response of the port voltage Gaussian mean and Gaussian variance can be obtained:

[0140]

[0141] The probability distribution coefficient of the port voltage can be derived through mathematical statistics as follows:

[0142]

[0143] In the formula: These represent the mean and standard deviation of the port voltage, respectively.

[0144] Step 4: Based on the characteristic coefficients of the Hermite chaotic polynomial, a chance constraint model of the node voltage under dual uncertainty can be constructed. If the node voltage does not meet the confidence requirement, the upper and lower limits of the PCC port power are adjusted according to the port voltage chance constraint, and the process returns to Step 2 to re-optimize the microgrid; if the confidence requirement is met, the loop ends, and the multi-microgrid day-ahead scheduling strategy obtained in Step 2 is output.

[0145] In some embodiments, step 4 specifically includes the following steps:

[0146] Step 4.1: Construct port voltage opportunity constraints based on NVIM-SRSM.

[0147]

[0148] In the formula: z α U represents the z-value corresponding to a significance level of α in the normal z-value table. 1-α / 2,i,t and U α / 2,i,t These are the upper and lower bounds of the PCC voltage confidence interval for microgrid i during time period t.

[0149] Step 4.2: When the random response of the port voltage does not meet the confidence requirement, the microgrid side needs to adjust the port power threshold and re-optimize to obtain a new microgrid optimization strategy. The PCC power limit adjustment formula for each microgrid is as follows:

[0150]

[0151] In the formula: These represent the upper and lower limits of the PCC power of microgrid i after adjustment.

[0152] The above method will be further explained below with reference to the accompanying drawings and specific embodiments.

[0153] The distribution network topology adopts the IEEE 33-node network, and photovoltaic power sources are connected to nodes 14, 22, and 33 of the original IEEE 33-node model. The power generation curves are referenced from publicly available data from the DKASC, Alice Springs photovoltaic power plant (data intervals of 5 minutes, including photovoltaic power generation and relevant meteorological data). Node load power is replaced with load data from a specific region publicly available on Alibaba Cloud Tianchi. Based on this, three target microgrid systems are designed and connected to nodes 10, 17, and 30 of the IEEE 33-node model, such as... Figure 2 As shown, an IEEE 33 network power flow model was developed to simulate the operation of a distribution network with a high proportion of distributed photovoltaic and microgrid cluster access. The simulation time span was 174 days, and active, reactive, and voltage data of the microgrid ports were sampled from the simulation operation output, with a calculation interval of 5 minutes. Gaussian bias was added to the meteorological data recorded by the photovoltaic power station to simulate the deviation of meteorological forecasts, thus forming an NVIM dataset.

[0154] The confidence level of the port voltage opportunity constraint is set to 95%, and 0.95 pu is set as the voltage warning value of NVIM-SRSM. The T+1 day operation strategy is optimized, and the lower limit of the 95% confidence interval of the port voltage is statistically analyzed and listed in Table 1.

[0155] Table 1 Lower limit of the 95% confidence interval for port voltage (pu)

[0156]

[0157]

[0158] As shown in Table 1, the day-ahead scheduling strategy obtained from the NVIM-SRSM model can ensure that the lower limit of the port voltage on day T+1 at a 95% confidence level does not exceed the set value of 0.95 pu, thus meeting the required confidence interval constraint. This embodiment illustrates that the multi-microgrid stochastic optimization method proposed in this invention can achieve port voltage risk pre-control.

[0159] Example 2

[0160] This invention also provides an electronic device, which includes a processor and a memory. The memory stores at least one instruction, at least one program, a code set, or an instruction set. The at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by the processor to achieve the following: Figure 1This paper presents a calculation method for voltage risk pre-control of multi-microgrid clusters based on SRSM.

[0161] It is understood that the memory may include random access memory (RAM) or read-only memory. Optionally, the memory may include non-transitory computer-readable storage medium. The memory can be used to store instructions, programs, code, code sets, or instruction sets. The memory may include a stored program area and a stored data area, wherein the stored program area may store instructions for implementing an operating system, instructions for at least one function, instructions for implementing the various method embodiments described above, etc.; the stored data area may store data created according to the use of the server, etc.

[0162] A processor may include one or more processing cores. The processor connects to various parts of the server via various interfaces and lines, executing instructions, programs, code sets, or instruction sets stored in memory, and accessing data stored in memory to perform various server functions and process data. Optionally, the processor may be implemented using at least one of the following hardware forms: Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), and Programmable Logic Array (PLA). The processor may integrate one or more of the following: Central Processing Unit (CPU) and Modem. The CPU primarily handles the operating system and applications; the modem handles wireless communication. It is understood that the modem may also be implemented as a separate chip without being integrated into the processor.

[0163] Since this electronic device is the electronic device corresponding to the voltage risk pre-control calculation method of multi-microgrid cluster based on SRSM in the embodiments of the present invention, and the principle of solving the problem by this electronic device is similar to that of the method, the implementation of this electronic device can refer to the implementation process of the above method embodiments, and the repeated parts will not be described again.

[0164] Example 3

[0165] This invention also provides a computer-readable storage medium storing at least one instruction, at least one program, a code set, or an instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to achieve the following: Figure 1 This paper presents a calculation method for voltage risk pre-control of multi-microgrid clusters based on SRSM.

[0166] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium capable of carrying or storing data.

[0167] Since the storage medium is the storage medium corresponding to the voltage risk pre-control calculation method of multi-microgrid cluster based on SRSM in the embodiment of the present invention, and the principle of the storage medium in solving the problem is similar to that of the method, the implementation of the storage medium can refer to the implementation process of the above method embodiment, and the repeated parts will not be described again.

[0168] Example 4

[0169] In some possible implementations, various aspects of the methods of the embodiments of the present invention can also be implemented as a program product, comprising program code that, when run on a computer device, causes the computer device to perform the steps of a voltage risk pre-control calculation method for a multi-microgrid cluster based on SRSM according to various exemplary embodiments of this application as described above. The executable computer program code or "code" used to perform the various embodiments can be written in high-level programming languages ​​such as C, C++, Python, Smalltalk, Java, JavaScript, Visual Basic, Structured Query Language (e.g., Transact-SQL), Perl, or in various other programming languages.

[0170] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0171] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0172] The above embodiments are merely illustrative of the technical concept and features of the present invention, and are intended to enable those skilled in the art to understand the content of the present invention and implement it accordingly. They should not be construed as limiting the scope of protection of the present invention. All equivalent changes or modifications made based on the essence of the content of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A method for voltage risk prediction and control calculation of multi-microgrid clusters based on SRSM, characterized in that, The following steps are involved: Model building and training: Build a neural network model with GRU as the main network, train the model, and obtain the node voltage inversion model; Microgrid stochastic optimization: Obtaining stochastic inputs for photovoltaics and loads based on SRSM and Nataf inverse transformation. A stochastic optimization model for a microgrid, considering source-load uncertainty, is constructed and solved to obtain the day-ahead dispatch strategy and the stochastic response of the interactive power. Node voltage inversion: the random response of the optimized interactive power. When combined with meteorological data to form a time series, it is input into the nodal voltage inversion model to obtain the random response of the voltage with Gaussian mean and standard deviation. Constraint Iterative Optimization: Construct an opportunity constraint model for node voltage under dual uncertainties; if the node voltage does not meet the confidence requirement, adjust the upper and lower limits of PCC port power according to the port voltage opportunity constraint, and return to the step of performing microgrid stochastic optimization; if the confidence requirement is met, end the loop and output the obtained day-ahead scheduling strategy of the microgrid.

2. The method for voltage risk pre-control calculation of multi-microgrid clusters based on SRSM according to claim 1, characterized in that, The model construction and training include: A neural network model with GRU as the main network is constructed; historical interaction power and meteorological data are combined into a time series, and historical node voltage is used as the training label to train the model; the neural network model is represented as follows: y t =f nn (x t ) In the formula: f nn () represents a neural network model with GRU as the main network, x t As the input to the neural network model, y t =[U 1,t ,…,U N,t ] T The voltage at the microgrid PCC port during time period t; A Gaussian process regression function, y, is introduced. t The Gaussian kernel function model is represented as: and m,t ,and s,t =f gpr (and t ) Where f gpr () is derived from historical node voltage and f nn () The historical voltage inversion was obtained by fitting. Let y t The mean and standard deviation are y m,t =[U m,1,t ,…,u m,N,t ] T ,y s,t =[u s,1,t ,…,u s,N,t ] T A node voltage inversion model based on time-series networks and Gaussian process regression f NVM () is represented as: and t ,and m,t ,and s,t =f NVM (x t ) Where N is the number of microgrids.

3. The method for voltage risk pre-control calculation of multi-microgrid clusters based on SRSM according to claim 2, characterized in that, Input of neural network model tw represents the length of the input time series, and d is the number of feature dimensions; P PCC,t Q represents the active power of PCC during time period t. PCC,t This represents the reactive power of PCC during time period t, in W. t This represents meteorological factors during time period t; f is obtained using the automatic differentiation calculation of the PyTorch deep learning framework. nn (x t Jacobian matrix of ) In the formula: This represents the input / output sensitivity coefficient matrix of the NVIM model, with matrix elements... This indicates the sensitivity of the power change at port j to the voltage change at port i during time period t.

4. The method for voltage risk pre-control calculation of multi-microgrid clusters based on SRSM according to claim 1, characterized in that, The microgrid stochastic optimization includes: In each microgrid, considering the two-dimensional uncertainties of photovoltaics and load as random inputs, and based on stochastic response surface theory, N is sampled. a A random sample is used as input. The superscript (m) indicates the m-th sample variable; Each sample is a 2-dimensional vector, representing a 2-dimensional uncertain input. Based on SRSM, random input The standard sample ξ can be transformed using the Nataf transform: In the formula: Nataf -1 This represents the inverse Nataf transform; Selecting a second-order chaotic polynomial to map the microgrid stochastic response process: In the formula: The random response output of microgrid i is represented by undetermined coefficients. Let ξ be the eigencoefficients of the Hermite chaotic polynomial, and H(ξ) denote the Hermite matrix. Based on SRSM, chaotic polynomial coefficients It can characterize the probability distribution information of the random response output, where the random response output mean and standard deviation and The relationship is: A stochastic optimization model for microgrids is constructed, taking into account the impact of random fluctuations in source load on the operating cost of microgrids, with the minimum of the mean and standard deviation of operating costs as the optimization objective. The gurobi solver is used to solve the microgrid optimization model, yielding the day-ahead optimization strategy and the stochastic response of the interactive power.

5. The method for voltage risk pre-control calculation of multi-microgrid clusters based on SRSM according to claim 4, characterized in that, The construction of the microgrid stochastic optimization model includes: The mathematical model for each microgrid's photovoltaic and load components, including uncertainties, is expressed as follows: In the formula: These represent the random and predicted values ​​of photovoltaic active power, respectively. Indicates the cost of PV curtailment; c PV,i,t This indicates the unit price of PV curtailment; Indicates the percentage of light wasted; These represent the random and predicted values ​​of the active power of the load, respectively; Indicates reactive power load, and Related, The power factor tangent of the load is represented; Δt represents the time interval. Energy storage mathematical model: In the formula: These represent the active power and reactive power of the energy storage ESS, respectively. These represent the ESS discharge and charging power, respectively. This indicates the state of charge of the ESS for the two time periods before and after it. These represent the discharge and charging efficiencies of the ESS, respectively. C represents the charge / discharge state of the ESS, and is a Boolean variable; ESS,i,t Indicates the operating loss cost of ESS; c ESS,i,t This indicates the unit price of ESS charging and discharging losses; These represent the maximum values ​​of active and reactive power in the ESS, respectively. Distribution-microgrid PCC port mathematical model: In the formula: These represent the active power and reactive power supplied by the distribution network to the microgrid, respectively. These represent the active / reactive power input from the distribution network to the microgrid and the active / reactive power output from the microgrid to the distribution network, respectively. This indicates the cost of purchasing / selling electricity at the PCC port; These represent the unit price for purchasing and selling electricity, respectively. Indicates the power tangent; The power states of the distribution network to the microgrid, respectively, are Boolean variables. This indicates the power transmitted from the distribution network to the microgrid; The maximum and minimum values ​​of PCC active power transmission are respectively represented. These represent the maximum and minimum values ​​of PCC reactive power transmission, respectively. Mathematical model of fast diesel generator: In the formula: These represent the active and reactive power outputs of FDE, respectively. These represent the maximum values ​​of active and reactive power in FDE, respectively; Indicates the operating cost of FDE; c FDE,i This represents the unit price of fuel cost for FDE; Microgrid power balance constraints: Microgrid optimization objectives: In the formula: The stochastic response representing the operating cost of a microgrid; express The corresponding chaotic polynomial coefficients; They represent The mean and standard deviation.

6. The method for voltage risk pre-control calculation of multi-microgrid clusters based on SRSM according to claim 1, characterized in that, The node voltage inversion includes: The port power will be optimized. A time-series input node voltage inversion model was constructed using meteorological forecast data. The inversion yields the port voltage inversion values, the Gaussian mean, and the Gaussian standard deviation of the random response. The mean and standard deviation of the random response with Gaussian mean and Gaussian variance of the port voltage are obtained based on the Hermite chaotic polynomial coefficients: In the formula: These represent the mean and standard deviation of the port voltages, respectively. H represents the chaotic polynomial coefficients corresponding to the Gaussian mean and Gaussian standard deviation of the port voltage, respectively. -1 (ξ) denotes the inverse of the Hermite matrix.

7. The method for voltage risk pre-control calculation of multi-microgrid clusters based on SRSM according to claim 1, characterized in that, The constraint iterative optimization includes: Construct port voltage opportunity constraints based on NVIM-SRSM; In the formula: z α U represents the z-value corresponding to a significance level of α in the normal z-value table. 1-α / 2,i,t and U α / 2,i,t These are the upper and lower bounds of the PCC voltage confidence interval for microgrid i during time period t; When the random response of the port voltage does not meet the confidence requirement, the microgrid side needs to adjust the port power threshold and re-optimize to obtain a new microgrid optimization strategy; the PCC power limit adjustment formula for each microgrid is as follows: In the formula: These represent the upper and lower limits of the PCC power of microgrid i after adjustment.

8. An electronic device, characterized in that, The electronic device includes a processor and a memory, the memory storing at least one instruction, at least one program, a code set, or an instruction set, the at least one instruction, the at least one program, the code set, or the instruction set being loaded and executed by the processor to implement the method as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The storage medium stores at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, the at least one program, the code set, or the instruction set is loaded and executed by a processor to implement the method as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, The computer program product includes computer instructions that, when executed by a processor, are used to perform the method as described in any one of claims 1 to 7.

Images