Interconnected power grid energy storage planning model construction and application method, system, equipment and medium

Abstract

The invention provides an interconnected power grid energy storage planning model construction and application method, system, device and medium. The distribution characteristics of photovoltaic output and load demand can be accurately described by using a Gaussian mixture model algorithm; secondly, a probability density function of interconnected network voltage and branch power flow can be efficiently obtained by using a cumulant method; and finally, constructing an energy storage double-layer planning model to perform energy storage site selection and capacity determination, the energy storage planning method can effectively cope with uncertainty caused by renewable energy and load demand fluctuation, can significantly reduce the voltage out-of-limit probability of each subarea power grid, can maintain the supply and demand probability balance of the subarea power grids, and can improve the energy storage capacity of the subarea power grids. And the stability, reliability and economy of the power system can be improved.

Description

Technical Field

[0001] This invention relates to the field of power system planning technology, and in particular to a method, system, equipment and medium for constructing and applying an interconnected power grid energy storage planning model. Background Technology

[0002] As the global energy structure transitions towards cleaner and lower-carbon energy, renewable energy has gained widespread application due to its environmental friendliness, and its development and utilization have become a key focus of national energy strategies. Consequently, renewable energy sources (such as photovoltaics) are being integrated into the power system on a large scale, becoming a crucial component. However, photovoltaic power generation is difficult to predict accurately, exhibiting strong randomness and volatility, which seriously threatens the supply-demand balance of the power system. Simultaneously, with the integration of emerging loads such as electric vehicles into the demand side, the volatility of their load demand has further increased, posing even more severe challenges to the power system's supply-demand balancing methods. Especially under extreme conditions, the net load demand of the power system may surge or decrease dramatically, significantly increasing the risk of supply-demand imbalance and thus threatening the economic efficiency, reliability, and stability of the power system.

[0003] Traditional stochastic programming methods often rely on parametric distribution assumptions, such as assuming that photovoltaic (PV) power output or load demand follows a normal distribution. This makes it difficult to accurately describe the nonlinear and stochastic characteristics of PV power output or load demand itself. Consequently, traditional stochastic programming methods cannot effectively address the impact of uncertainties in PV power generation or load demand on the power system's supply-demand balance. On the other hand, energy storage systems can achieve peak shaving and valley filling through energy time-shifting characteristics, ensuring the power system's supply-demand balance. However, the widespread deployment of energy storage systems is affected by investment and construction costs.

[0004] In recent years, significant progress has been made in research on energy storage planning, with many scholars and engineers proposing various methods. However, existing technologies fail to fully consider the uncertainties of photovoltaic power generation fluctuations and load fluctuations when planning energy storage, leading to discrepancies between optimized results and actual dispatch results, thus affecting the effectiveness of supply and demand balance. Furthermore, existing technologies do not consider the impact of energy storage operation on the power system's supply and demand balance, increasing the risk of power system supply and demand imbalance. Summary of the Invention

[0005] To address the problems of existing technologies failing to adequately consider the uncertainties of photovoltaic power generation fluctuations and load fluctuations during energy storage planning, leading to discrepancies between optimized and actual dispatch results and thus affecting the effectiveness of supply and demand balance; and the increased risk of power system supply and demand imbalance due to the lack of consideration for the impact of energy storage operation on the power system's supply and demand balance, this invention proposes a method for constructing an interconnected power grid energy storage planning model, comprising: Based on historical datasets of photovoltaic power output or load demand, Gaussian mixture model algorithm is used to process the uncertainty of photovoltaic power output or load demand, and the probability density function of photovoltaic power output or load demand is obtained. Based on the probability density function of photovoltaic output and the probability density function of load demand, the cumulative method is used to perform probabilistic power flow calculation to obtain the probability density function of node voltage and branch power flow of the regional power grid. Based on the probability density functions of node voltage and branch power flow in the regional power grid, a two-layer energy storage planning model considering photovoltaic and load uncertainties is constructed as an interconnected power grid energy storage planning model. The dual-layer energy storage planning model includes an upper-layer optimization model and a lower-layer optimization model. The upper-layer optimization model receives the probability density function of the node voltage and the probability density function of the branch power flow in the regional power grid, and aims to minimize the comprehensive effectiveness coefficient of energy storage installation on grid performance to obtain the optimal installation location of energy storage. The lower-layer optimization model, based on the optimal installation location of energy storage, aims to minimize the annual planning and operating costs of the interconnected power grid, optimizes the rated capacity, rated power, and charging and discharging scheduling strategy of energy storage, and returns the corresponding energy storage charging and discharging power curves to the upper-layer optimization model.

[0006] Preferably, the step of processing the uncertainty of photovoltaic output or load demand using a Gaussian mixture model algorithm based on historical datasets of photovoltaic output or load demand to obtain the probability density function of photovoltaic output or load demand includes: For photovoltaic power output or load demand, the historical dataset is divided using the K-means clustering algorithm to obtain multiple clusters, each cluster corresponding to a Gaussian component in a Gaussian mixture model algorithm; The mean and covariance matrix of the samples corresponding to each Gaussian component are calculated respectively as the initial values ​​of the mean and covariance matrix of the Gaussian component, and the weights of each Gaussian component are evenly distributed as the initial values ​​of the weights of each Gaussian component. For each Gaussian component, the current weights, mean, and covariance matrices are used respectively. The weights, mean, and covariance matrices of the Gaussian components are iteratively updated using the expectation-maximization algorithm until the weights, mean, and covariance matrices converge. Based on the weights, mean, and covariance matrices of each Gaussian component, the probability density function of photovoltaic power output or load demand is calculated. The number of clusters is determined using the Bayesian information criterion method.

[0007] Preferably, for each Gaussian component, the current weights, mean, and covariance matrix are used to iteratively update the weights, mean, and covariance matrix of the Gaussian component using the expectation-maximization algorithm until the weights, mean, and covariance matrix converge, including: For each Gaussian component, the posterior probability is calculated based on the weight, mean, and covariance matrix of the Gaussian component. Using the posterior probability, the weights, mean, and covariance matrices of each Gaussian component are recalculated; Determine whether the maximum likelihood function corresponding to the weights, mean, and covariance matrices of each Gaussian component has converged. If yes, output the weights, mean, and covariance matrices of each Gaussian component and end the process. Otherwise, jump to calculate the posterior probability for each Gaussian component based on its weights, mean, and covariance matrices.

[0008] Preferably, the posterior probability is calculated as follows:

[0009] In the formula, Indicates the first k The Gaussian component of the first n The posterior probability of each data point. Represents the first in the historical dataset n One data point, k Indicates the first k Gaussian components Indicates the first k The weights of the Gaussian components in the previous iteration. K This represents the total number of optimal Gaussian components. Representing data Gaussian distribution, Indicates the first k The mean of the Gaussian components in the previous iteration Indicates the first k The covariance matrix of the Gaussian components in the previous iteration; The formula for recalculating the weights of the Gaussian components using posterior probabilities is as follows:

[0010] The formula for recalculating the mean of the Gaussian components using posterior probability is as follows:

[0011] The formula for recalculating the covariance matrix of the Gaussian components using posterior probabilities is as follows:

[0012] In the formula, Indicates the first k The weights of the Gaussian components, Indicates the first k The mean of the Gaussian components, Indicates the first k The covariance matrix of Gaussian components,N For the first k The number of samples in each Gaussian component.

[0013] Preferably, the probability density function of the photovoltaic power output or load demand is calculated as follows:

[0014] In the formula, for The probability density function, For photovoltaic power output or load demand, K This represents the total number of optimal Gaussian components. For the first k The weights of the Gaussian components, Representing data Gaussian distribution, For the first k The mean of the Gaussian components, For the first k The covariance matrix of Gaussian components.

[0015] Preferably, the probabilistic power flow calculation based on the probability density function of photovoltaic output and the probability density function of load demand, using the cumulant method, to obtain the probability density function of voltage and the probability density function of branch power flow in the regional power grid, includes: The supply and demand balance equations of the regional power grid are rewritten in matrix form, and the power flow equations are expanded into Taylor series at the reference operating point using the Newton-Raphson method, while ignoring higher-order terms, to obtain a power flow linearization model. The power flow linearization model is rearranged and reorganized, and the probability density functions of photovoltaic output and load demand are substituted to obtain the cumulative quantities of each order of the regional grid voltage and branch power flow. Based on the cumulative quantities of the voltage and branch power flow of the regional grid, the probability density functions of the node voltage and branch power flow are calculated using the C-type Gram-Charlier series expansion method.

[0016] Preferably, the probability density function of the node voltage or branch power flow is calculated as follows:

[0017] In the formula, Let be the probability density function of node voltage or branch power flow, where y For node voltage or branch power flow; for m The series expansion coefficients of order , , m Let be the order of the series expansion. The maximum order of the series expansion. for m Hermite polynomial of order 1 m Hermite polynomial of order The calculation formula is as follows:

[0018] The m Series expansion coefficients of order By solving a system of linear equations Received; In the formula, yes The order column vector is an intermediate variable calculated from the cumulative quantities of node voltages or branch power flows. =0, The formulas for calculating the remaining components are as follows:

[0019] In the formula: It is the standard deviation of node voltage or branch power flow. T for Column vector of order, for T The m-2 order components, The other orders The calculation formula is as follows:

[0020] In the formula, q For summation index; For node voltage or branch power flow m - q Accumulative quantity; For node voltage or branch power flow m Accumulative quantity; for T of q Order component; for 1-th order symmetric matrix Middle elements The calculation formula is as follows:

[0021] In the formula, e 1 and e 2 represents the row and column indices of the matrix. s It is a summation index. express T The s Order component; This is a three-parameter function, and its calculation formula is as follows:

[0022] In the formula: For the Gamma function, a , b and c These are the three parameters of a three-parameter function.

[0023] Preferably, the objective function of the upper-level optimization model is calculated as follows:

[0024] In the formula: For the comprehensive effectiveness coefficient, Net power fluctuation of the regional power grid The efficacy coefficient, Voltage over-limit probability of the regional power grid The efficacy coefficient, Network losses for regional power grids The efficacy coefficient, , , They are respectively , , Weighting coefficients; Net power fluctuation The calculation formula is as follows:

[0025] In the formula: The sampling time interval, This represents the maximum net power within the sampling time interval. This represents the minimum net power within the sampling time interval. Tz To optimize the total duration; Voltage over-limit probability of the partitioned power grid The calculation formula is as follows:

[0026] In the formula: For regional power grid nodes j The voltage probability density function, For the voltage of the nodes in the regional power grid, This is the upper limit of the voltage. This is the lower limit of the voltage. J The total number of nodes in the regional power grid; Network losses in the regional power grid The calculation formula is as follows:

[0027] In the formula, Network loss is calculated using the probability density function of branch power flow; Efficacy coefficient The calculation formula is as follows:

[0028] In the formula, For the first d Individual efficacy coefficients of each variable , For the first d The actual value of each evaluation indicator For the first d The satisfaction value of each evaluation indicator For the first The unacceptable value of each evaluation indicator. It is the first constant. It is the second constant. It is the third constant.

[0029] Preferably, the objective function of the lower-level optimization model is calculated as follows:

[0030] In the formula: For the annual planning and operating costs of the interconnected power grid, As a capital recovery factor, For the discount rate, For the lifespan of the energy storage system, This represents the total number of energy storage units installed. J This represents the total number of nodes in the regional power grid. Cost per unit capacity of energy storage Cost per unit power of energy storage For energy storage z Rated capacity, For energy storage z Rated power, W As voltage penalty weight, For regional power grid nodes j exist t The voltage at that moment exceeds the limit. For regional power grid nodes j exist t The voltage at any given time exceeds the negative limit. This is the operation and management cost coefficient for the energy storage system. For energy storage z At any moment t The charging power, For energy storage z At any moment t The discharge power; The constraints of the lower-level optimization model are as follows:

[0031] In the formula: This is the ratio of rated energy storage capacity to rated energy storage power. For regional power grid nodes j exist t The active power injected at all times, Regional power grid nodes j exist t Load power at any given time For regional power grid nodes j exist t Photovoltaic output power at any given time For energy storage charging efficiency, For energy storage and discharge efficiency, For energy storage z exist t SOC at any moment For energy storage z exist t SOC at time +1 The minimum SOC value for energy storage. This represents the maximum SOC value for energy storage. For energy storage z Daily starting SOC For the end-of-day tolerance of SOC, For energy storage z The daily SOC value.

[0032] This invention also provides a method for applying an interconnected power grid energy storage planning model, comprising: Based on historical datasets of photovoltaic power output or load demand, Gaussian mixture model algorithm is used to process the uncertainty of photovoltaic power output or load demand, and the probability density function of photovoltaic power output or load demand is obtained. Based on the probability density function of photovoltaic output and the probability density function of load demand, the cumulative method is used to perform probabilistic power flow calculation to obtain the probability density function of node voltage and branch power flow of the regional power grid. The probability density functions of the node voltage and the branch power flow of the partitioned power grid are input into the upper planning model of the two-layer energy storage planning model to solve for the installation location of the energy storage. The installation location of the energy storage is passed into the lower planning model of the two-layer energy storage planning model, and the rated capacity, rated power, and charging and discharging scheduling strategy of the energy storage, as well as the corresponding energy storage charging and discharging power curve, are obtained by solving the problem. The energy storage charging and discharging power curve is returned to the upper-level planning model, the comprehensive efficiency coefficient is recalculated, and the installation location, rated capacity, rated power, and charging and discharging scheduling strategy of the energy storage are iteratively updated until the comprehensive efficiency coefficient no longer decreases, thus obtaining the optimal installation location, rated capacity, rated power, and charging and discharging scheduling strategy of the energy storage.

[0033] Based on the same inventive concept, this invention also provides a system for constructing an interconnected power grid energy storage planning model, comprising: The photovoltaic load probability density module is used to process the uncertainty of photovoltaic output or load demand based on historical datasets of photovoltaic power output or load demand, respectively, and obtain the probability density function of photovoltaic power output or load demand. The voltage power flow probability density module is used to perform probabilistic power flow calculations using the cumulant method based on the probability density function of photovoltaic output and the probability density function of load demand, so as to obtain the probability density function of node voltage and branch power flow of the regional power grid. The modeling module is used to construct a two-layer energy storage planning model that considers photovoltaic and load uncertainties as an interconnected grid energy storage planning model, based on the probability density function of node voltage and branch power flow in the partitioned grid. The dual-layer energy storage planning model includes an upper-layer optimization model and a lower-layer optimization model. The upper-layer optimization model receives the probability density function of the node voltage and the probability density function of the branch power flow in the regional power grid, and aims to minimize the comprehensive effectiveness coefficient of energy storage installation on grid performance to obtain the optimal installation location of energy storage. The lower-layer optimization model, based on the optimal installation location of energy storage, aims to minimize the annual planning and operating costs of the interconnected power grid, optimizes the rated capacity, rated power, and charging and discharging scheduling strategy of energy storage, and returns the corresponding energy storage charging and discharging power curves to the upper-layer optimization model.

[0034] Preferably, the photovoltaic load probability density module is specifically used for: For photovoltaic power output or load demand, the historical dataset is divided using the K-means clustering algorithm to obtain multiple clusters, each cluster corresponding to a Gaussian component in a Gaussian mixture model algorithm; The mean and covariance matrix of the samples corresponding to each Gaussian component are calculated respectively as the initial values ​​of the mean and covariance matrix of the Gaussian component, and the weights of each Gaussian component are evenly distributed as the initial values ​​of the weights of each Gaussian component. For each Gaussian component, the current weights, mean, and covariance matrices are used respectively. The weights, mean, and covariance matrices of the Gaussian components are iteratively updated using the expectation-maximization algorithm until the weights, mean, and covariance matrices converge. Based on the weights, mean, and covariance matrices of each Gaussian component, the probability density function of photovoltaic power output or load demand is calculated. The number of clusters is determined using the Bayesian information criterion method.

[0035] Preferably, the photovoltaic load probability density module, for each Gaussian component, uses the current weights, mean, and covariance matrix, and iteratively updates the weights, mean, and covariance matrix of the Gaussian component using an expectation-maximization algorithm until the weights, mean, and covariance matrix all converge, including: For each Gaussian component, the posterior probability is calculated based on the weight, mean, and covariance matrix of the Gaussian component. Using the posterior probability, the weights, mean, and covariance matrices of each Gaussian component are recalculated; Determine whether the maximum likelihood function corresponding to the weights, mean, and covariance matrices of each Gaussian component has converged. If yes, output the weights, mean, and covariance matrices of each Gaussian component and end the process. Otherwise, jump to calculate the posterior probability for each Gaussian component based on its weights, mean, and covariance matrices.

[0036] Preferably, the formula for calculating the posterior probability in the photovoltaic load probability density module is as follows:

[0037] In the formula, Indicates the first k The Gaussian component of the first n The posterior probability of each data point. Represents the first in the historical dataset n One data point, k Indicates the first k Gaussian components Indicates the first k The weights of the Gaussian components in the previous iteration. K This represents the total number of optimal Gaussian components. Representing data Gaussian distribution, Indicates the first k The mean of the Gaussian components in the previous iteration Indicates the first k The covariance matrix of the Gaussian components in the previous iteration; The formula for recalculating the weights of the Gaussian components using posterior probabilities is as follows:

[0038] The formula for recalculating the mean of the Gaussian components using posterior probability is as follows:

[0039] The formula for recalculating the covariance matrix of the Gaussian components using posterior probabilities is as follows:

[0040] In the formula, Indicates the first k The weights of the Gaussian components, Indicates the first k The mean of the Gaussian components, Indicates the first k The covariance matrix of Gaussian components, N For the first k The number of samples in each Gaussian component.

[0041] Preferably, the probability density function of photovoltaic output or load demand in the photovoltaic load probability density module is calculated as follows:

[0042] In the formula, for The probability density function, For photovoltaic power output or load demand, K This represents the total number of optimal Gaussian components. For the first k The weights of the Gaussian components, Representing data Gaussian distribution, For the first k The mean of the Gaussian components, For the first k The covariance matrix of Gaussian components.

[0043] Preferably, the voltage power flow probability density module is specifically used for: The supply and demand balance equations of the regional power grid are rewritten in matrix form, and the power flow equations are expanded into Taylor series at the reference operating point using the Newton-Raphson method, while ignoring higher-order terms, to obtain a power flow linearization model. The power flow linearization model is rearranged and reorganized, and the probability density functions of photovoltaic output and load demand are substituted to obtain the cumulative quantities of each order of the regional grid voltage and branch power flow. Based on the cumulative quantities of the voltage and branch power flow of the regional grid, the probability density functions of the node voltage and branch power flow are calculated using the C-type Gram-Charlier series expansion method.

[0044] Preferably, the probability density function of node voltage or branch power flow in the voltage power flow probability density module is calculated as follows:

[0045] In the formula, Let be the probability density function of node voltage or branch power flow, where y For node voltage or branch power flow; for m The series expansion coefficients of order , , m Let be the order of the series expansion. The maximum order of the series expansion. for m Hermite polynomial of order 1 m Hermite polynomial of order The calculation formula is as follows:

[0046] The m Series expansion coefficients of order By solving a system of linear equations Received; In the formula, yes The order column vector is an intermediate variable calculated from the cumulative quantities of node voltages or branch power flows. =0, The formulas for calculating the remaining components are as follows:

[0047] In the formula: It is the standard deviation of node voltage or branch power flow. T for Column vector of order, for T The m-2 order components, The other orders The calculation formula is as follows:

[0048] In the formula, q For summation index; For node voltage or branch power flow m - q Accumulative quantity; For node voltage or branch power flow m Accumulative quantity; for T of q Order component; for 1-th order symmetric matrix Middle elements The calculation formula is as follows:

[0049] In the formula, e 1 and e 2 represents the row and column indices of the matrix. s It is a summation index. express T The s Order component; This is a three-parameter function, and its calculation formula is as follows:

[0050] In the formula: For the Gamma function, a , b and c These are the three parameters of a three-parameter function.

[0051] Preferably, the objective function of the upper-level optimization model in the modeling module is calculated as follows:

[0052] In the formula: For the comprehensive effectiveness coefficient, Net power fluctuation of the regional power grid The efficacy coefficient, Voltage over-limit probability of the regional power grid The efficacy coefficient, Network losses for regional power grids The efficacy coefficient, , , They are respectively , , Weighting coefficients; Net power fluctuation The calculation formula is as follows:

[0053] In the formula: The sampling time interval, This represents the maximum net power within the sampling time interval. This represents the minimum net power within the sampling time interval. Tz To optimize the total duration; Voltage over-limit probability of the partitioned power grid The calculation formula is as follows:

[0054] In the formula: For regional power grid nodes j The voltage probability density function, For the voltage of the nodes in the regional power grid, This is the upper limit of the voltage. This is the lower limit of the voltage. J The total number of nodes in the regional power grid; Network losses in the regional power grid The calculation formula is as follows:

[0055] In the formula, Network loss is calculated using the probability density function of branch power flow; Efficacy coefficient The calculation formula is as follows:

[0056] In the formula, For the first d Individual efficacy coefficients of each variable , For the first d The actual value of each evaluation indicator For the first d The satisfaction value of each evaluation indicator For the first The unacceptable value of each evaluation indicator. It is the first constant. It is the second constant. It is the third constant.

[0057] Preferably, the objective function of the lower-level optimization model in the modeling module is calculated as follows:

[0058] In the formula: For the annual planning and operating costs of the interconnected power grid, As a capital recovery factor, For the discount rate, For the lifespan of the energy storage system, This represents the total number of energy storage units installed. J This represents the total number of nodes in the regional power grid. Cost per unit capacity of energy storage Cost per unit power of energy storage For energy storage z Rated capacity, For energy storage z Rated power, W As voltage penalty weight, For regional power grid nodes j exist t The voltage at that moment exceeds the limit. For regional power grid nodes jexist t The voltage at any given time exceeds the negative limit. This is the operation and management cost coefficient for the energy storage system. For energy storage z At any moment t The charging power, For energy storage z At any moment t The discharge power; The constraints of the lower-level optimization model are as follows:

[0059] In the formula: This is the ratio of rated energy storage capacity to rated energy storage power. For regional power grid nodes j exist t The active power injected at all times, Regional power grid nodes j exist t Load power at any given time For regional power grid nodes j exist t Photovoltaic output power at any given time For energy storage charging efficiency, For energy storage and discharge efficiency, For energy storage z exist t SOC at any moment For energy storage z exist t SOC at time +1 The minimum SOC value for energy storage. This represents the maximum SOC value for energy storage. For energy storage z Daily starting SOC For the end-of-day tolerance of SOC, For energy storage z The daily SOC value.

[0060] Based on the same inventive concept, this invention also provides an application system for an interconnected power grid energy storage planning model, comprising: The data input module is used to process the uncertainty of photovoltaic output or load demand based on historical datasets of photovoltaic power output or load demand, respectively, and obtain the probability density function of photovoltaic power output or load demand. The voltage and power flow calculation module is used to perform probabilistic power flow calculations using the cumulant method based on the probability density function of the photovoltaic output and the probability density function of the load demand, so as to obtain the probability density function of the node voltage and the probability density function of the branch power flow in the regional power grid. The location calculation module is used to input the probability density function of the node voltage and the probability density function of the branch power flow of the partitioned power grid into the upper planning model of the two-layer energy storage planning model, and solve for the installation location of the energy storage. The capacity strategy module is used to input the installation location of the energy storage into the lower planning model of the two-layer energy storage planning model, and solve it to obtain the rated capacity, rated power, and charging and discharging scheduling strategy of the energy storage, as well as the corresponding energy storage charging and discharging power curve. The iterative update module is used to return the energy storage charging and discharging power curve to the upper-level planning model, recalculate the comprehensive efficiency coefficient, and iteratively update the energy storage installation location, rated capacity, rated power, and charging and discharging scheduling strategy until the comprehensive efficiency coefficient no longer decreases, thus obtaining the optimal energy storage installation location, rated capacity, rated power, and charging and discharging scheduling strategy.

[0061] In another aspect, the present invention also provides an electronic device, comprising: at least one processor and a memory; The memory is used to store one or more programs; When the one or more programs are executed by the one or more processors, they implement the interconnected power grid energy storage planning model construction method or application method as described above.

[0062] In another aspect, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed, it implements the above-described method for constructing or applying an interconnected power grid energy storage planning model.

[0063] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention provides a method, system, device, and medium for constructing and applying an interconnected power grid energy storage planning model. The method includes: processing the uncertainty of photovoltaic power output or load demand using a Gaussian mixture model algorithm based on historical datasets of photovoltaic power output or load demand, to obtain the probability density function of photovoltaic power output or load demand; performing probabilistic power flow calculation using the cumulant method based on the probability density function of photovoltaic power output and the probability density function of load demand, to obtain the probability density function of node voltage and branch power flow of the regional power grid; and constructing a two-layer energy storage planning model considering the uncertainty of photovoltaic power and load as the interconnected power grid energy storage planning model based on the probability density function of node voltage and branch power flow of the regional power grid. This invention utilizes the Gaussian mixture model algorithm to accurately describe the distribution characteristics of photovoltaic power output and load demand. Secondly, it employs the cumulant method to efficiently obtain the probability density functions of interconnected grid voltage and branch power flow. Finally, it constructs a two-layer energy storage planning model for energy storage site selection and capacity determination. This energy storage planning model construction method effectively addresses the uncertainties caused by fluctuations in renewable energy and load demand through the Gaussian mixture model algorithm, obtaining the probability density functions of photovoltaic power output or load demand. Furthermore, it calculates the probability density functions of node voltage and branch power flow in the regional grid, constructing a two-layer energy storage planning model considering photovoltaic and load uncertainties as the interconnected grid energy storage planning model. Solving the interconnected grid energy storage planning model yields energy storage installation locations, rated capacity, rated power, and charging / discharging scheduling strategies that significantly reduce the voltage exceedance probability of each regional grid, maintain the supply-demand probability balance of the regional grid, and improve the stability, reliability, and economy of the power system. Attached Figure Description

[0064] Figure 1 A flowchart of a method for constructing an interconnected power grid energy storage planning model provided by the present invention; Figure 2 A flowchart of an application method for an interconnected power grid energy storage planning model provided by the present invention; Figure 3 A schematic diagram of an interconnected power grid system provided by the present invention; Figure 4 This is a schematic diagram of the annual photovoltaic power output curves of various regional power grids provided by the present invention; Figure 5 This is a schematic diagram of the annual load demand curves for each regional power grid provided by the present invention; Figure 6 A schematic diagram of the charge-discharge curve of the energy storage unit of the partitioned power grid 1 for one year provided by the present invention; Figure 7 A schematic diagram of the SOC variation curve of the energy storage unit of the partitioned power grid 1 over one year, provided by the present invention. Figure 8A schematic diagram of the charge-discharge curve of the energy storage unit of the partitioned power grid 2 provided by the present invention over one year; Figure 9 A schematic diagram of the SOC variation curve of the energy storage unit of the partitioned power grid 2 over one year, provided by the present invention; Figure 10 A schematic diagram of the charge-discharge curve of the energy storage unit of the partitioned power grid 3 provided by the present invention over one year; Figure 11 A schematic diagram of the SOC variation curve of the energy storage unit of the partitioned power grid 3 over one year, provided by the present invention; Figure 12 A comparison chart of the probability of voltage exceeding the limit at each grid node in each zone before and after the installation of energy storage, provided for the present invention; Figure 13 A comparison chart of average voltage in each regional power grid before and after the installation of energy storage, provided for the present invention; Figure 14 A comparison chart of voltage over-limit probabilities with and without energy storage in extreme scenarios provided by the present invention; Figure 15 A comparison chart of average voltage with and without energy storage under extreme scenarios provided by this invention; Figure 16 A system structure diagram for constructing an interconnected power grid energy storage planning model is provided by this invention; Figure 17 This invention provides a structural diagram of an application system for an interconnected power grid energy storage planning model. Figure 18 This invention provides a structural diagram of an electronic device. Detailed Implementation

[0065] This invention proposes a method, system, equipment, and medium for constructing and applying an energy storage planning model for interconnected power grids. It employs a Gaussian Mixture Model (GMM) to handle the uncertainties in photovoltaic (PV) output and load demand, performs probabilistic power flow calculations based on the cumulant method, and constructs a two-layer energy storage planning model that considers the uncertainties in PV output and load demand. First, the GMM method accurately describes the distribution characteristics of PV output and load demand. Second, the cumulant method efficiently obtains the probability density functions of voltage and branch power flow in the interconnected power grid. Finally, a two-layer energy storage planning model is constructed to ensure the probabilistic balance of supply and demand in the interconnected power grid. The method proposed in this invention can effectively address the uncertainties caused by fluctuations in renewable energy and load demand, reduce the probability of voltage exceedances in different regional power grids, maintain the probabilistic balance of supply and demand in regional power grids, and improve the stability, reliability, and economy of the power system.

[0066] Example 1: A method for constructing an interconnected power grid energy storage planning model, such as Figure 1 As shown, it includes: Step A1: Based on the historical datasets of photovoltaic power output or load demand, the Gaussian mixture model algorithm is used to process the uncertainty of photovoltaic power output or load demand to obtain the probability density function of photovoltaic power output or load demand. Step A2: Based on the probability density function of photovoltaic power output and the probability density function of load demand, use the cumulant method to perform probabilistic power flow calculation to obtain the probability density function of node voltage and branch power flow of the regional power grid. Step A3: Based on the probability density function of the node voltage and the probability density function of the branch power flow in the regional power grid, construct a two-layer energy storage planning model that considers the uncertainties of photovoltaics and loads as the energy storage planning model for the interconnected power grid. The two-layer energy storage planning model includes an upper-layer optimization model and a lower-layer optimization model. The upper-layer optimization model receives the probability density function of the node voltage and the probability density function of the branch power flow in the regional power grid. With the goal of minimizing the comprehensive effectiveness coefficient of energy storage installation on grid performance improvement, it obtains the optimal installation location for energy storage. The lower-layer optimization model, based on the optimal installation location of energy storage, aims to minimize the annual planning and operating costs of the interconnected power grid. It optimizes the rated capacity, rated power, and charging and discharging scheduling strategy of energy storage and returns the corresponding energy storage charging and discharging power curves to the upper-layer optimization model.

[0067] Step A1 is based on GMM's photovoltaic and load uncertainty data modeling, specifically including: This invention uses Gaussian Mixture Model (GMM) to model the uncertainty of photovoltaic (PV) power output and load demand. PV power output and load demand are inherently uncertain, and traditional parametric models struggle to accurately describe their complex distribution characteristics. Therefore, this invention uses GMM to calculate the probability density function of PV power output and load demand. The GMM method posits that the complex data distribution we observe is actually a mixture of multiple simple Gaussian distributions, and a complete probability density function can be obtained by estimating the parameters and weights of each Gaussian component. The probability density function of PV power output and load demand consists of a weighted sum of multiple Gaussian components, mathematically expressed as: (1) In the formula: for The probability density function, For photovoltaic power output or load demand. The optimal number of Gaussian components can be determined using the Bayesian Information Criterion (BIC) method, specifically the one with the smallest corresponding BIC value. K The value is the optimal Gaussian component number. The formula for the BIC method is: ,in For historical datasets X The log-likelihood function; This represents the total number of free parameters in the GMM model. For data dimensions, N For historical datasets X The total number of samples in the sample. For the first k The weights of the Gaussian components, and satisfying the constraints . For the first k The mean of the components. For the first k The covariance matrix of the components. For the first k A Gaussian distribution, the formula is: (2) In the formula: For index, For transpose, For the first k The inverse covariance matrix of the components.

[0068] The parameters of the GMM are estimated using the Expectation-Maximization (EM) algorithm and updated iteratively through EM. , , The algorithm considers convergence to occur when the model parameters obtained in the current iteration change very little compared to the parameter values ​​in the previous iteration. The objective of the EM algorithm is to maximize the maximum likelihood function. (3) In the formula: Historical datasets for photovoltaic power output or load demand. For the first n Each photovoltaic output or load demand value. The number of samples for photovoltaic output or load demand data for each regional power grid.

[0069] The EM algorithm first uses K-means clustering on historical datasets of photovoltaic power output or load demand. X Divide into, into K There are 3 clusters, each corresponding to a Gaussian component in a Gaussian Mixture Model (GMM). The optimal number of clusters is determined, and the mean and variance of the samples corresponding to each component are calculated as initial values ​​to initialize the clusters. Based on the current situation... , Calculate the posterior probability : (4) In the formula: The weights from the previous iteration; This is the mean of the previous iteration; This represents the variance of the previous iteration.

[0070] According to formula (4) Recalculate , , .

[0071] (5) (6) (7) Recalculate the maximum likelihood function and determine if it has converged (i.e., the parameter values ​​no longer change significantly). If converged, output the current value. , , If it does not converge, return to formula (4).

[0072] The GMM method is used to model the uncertainty of photovoltaic power output and load demand. The historical datasets of photovoltaic power output and load demand are transformed into corresponding probability density functions and then passed to the subsequent probabilistic power flow calculation.

[0073] Step A2, based on the probabilistic power flow calculation using the cumulant method, specifically includes: This invention uses the cumulant method for probabilistic power flow calculation, which can efficiently obtain the probability density functions of voltage and branch power flow in a regional power grid. Probabilistic Power Flow (PPF) is a method for calculating power flow in a power grid under uncertainties. Unlike traditional deterministic power flow, probabilistic power flow considers the randomness of input variables and represents the output as a probability density function. Through probabilistic power flow modeling using the cumulant method, the impact of source load fluctuations on the power supply and demand balance can be effectively assessed.

[0074] Based on the photovoltaic output and load demand obtained from GMM, the corresponding , , The photovoltaic output is calculated using the following formulas. v Step cumulative and load demand v Step cumulative : (8) (9) (10) (11) In the formula: This is a first-order cumulant, representing the mean. This is a second-order cumulant, representing the variance. This is a third-order cumulant, representing skewness. This is a fourth-order cumulant, representing kurtosis.

[0075] When performing probabilistic power flow calculations, the supply and demand balance equations and branch transmission power equations for the regional power grid are as follows: (12) (13) In the formula: J This represents the total number of nodes in each regional power grid. and Representing the nodes of the regional power grid Total active power and total reactive power, , These are the nodes of the regional power grid. and voltage amplitude, It is a regional power grid node and Voltage phase difference between For regional power grid nodes and The real part of the inter-branch admittance, For regional power grid nodes and The imaginary part of the admittance of the intermediate branch. Representative from the regional power grid node Transmit to node j The active power on the transmission line.

[0076] By rewriting the nonlinear power flow equation in formula (12) into matrix form, and expanding the power flow equation into a Taylor series at the reference operating point using the Newton-Raphson method while ignoring higher-order terms, a power flow linearization model can be obtained: (14) In the formula: To inject total power into the voltage of each node in the regional power grid. The total active power or total reactive power transmitted by the branch. For voltage disturbances at various nodes of the regional power grid. , These represent the total power injection and branch power flow at the baseline operating point, respectively. J 0 and G 0 represents the Jacobian matrix used in the last calculation of the Newton-Raphson method.

[0077] By rearranging and simplifying formula (14), the voltage disturbances at each node of the regional power grid can be obtained. With branch tidal current disturbance The Step cumulative and : (15) In the formula: This is the cumulative power injection, calculated from the cumulative amount of photovoltaic output and load demand. ; Cumulative amount of node voltage disturbance Accumulated power injection of v The first-order sensitivity matrix is... of v Power of; Cumulative power of branch circuits Accumulated power injection of v The first-order sensitivity matrix is... of v Power of 1.

[0078] Due to the reference operating point of the node voltage in the regional power grid If it is a constant, then the node voltage state quantity With node voltage disturbance The cumulative quantities of each order have the following relationship: (16) (17) In the formula, This represents the average voltage at each node in a regional power grid, which is the first-order cumulative quantity of the node voltage. . The mean of the node voltage disturbance is represented by its first-order cumulant. Representing node voltage v Step cumulative quantity, That is, the node voltage state quantities and the disturbances have the same central moments (such as variance, kurtosis, and skewness), differing only in their means. Therefore, to solve The node voltage state can be obtained by accumulating the amount. The probability distribution characteristics. Similarly, for branch power flow, the reference operating point of the branch power flow. Also a constant, the branch power flow state quantity With branch tidal current disturbance The cumulative amount also conforms to the relationships in equations (16) and (17). Through the above transformation, the cumulative amount of the node voltage of the regional power grid can be obtained. Cumulative amount of branch tidal current .

[0079] Obtain the cumulative voltage of nodes in the regional power grid. Cumulative amount of branch tidal current Then, the probability density functions of node voltage and branch power flow are calculated using the C-type Gram-Charlier series expansion method. This probability density function is the final result output by the cumulant method, and it will be used in the bi-level programming model. (18) In the formula ： Let be the probability density function of node voltage or branch power flow, where y For node voltage or branch power flow; The coefficients are the series expansion coefficients. . m Let be the order of the series expansion. This is the maximum order of the series expansion. It is an m-th order Hermite polynomial. The calculation formula is as follows: (19) Equation (18) does not directly use the obtained cumulative quantities of each order, but rather solves the system of linear equations. Solving for the series expansion coefficients involves converting the cumulative quantities of each order into series expansion coefficients. Use, here That is, in equation (18) The vector formed by these vectors. yes m max The order column vector is an intermediate variable calculated from the cumulative quantities of node voltages or branch power flows. =0, the rest can be obtained through formula (20): (20) In the formula: This is the standard deviation of the node voltage or branch power flow. Since the probability density function constructed in this invention is a standard normal distribution, therefore... . T It is an intermediate variable, for Column vector of order, for T of m -2nd order components can be derived from m Step cumulative It is calculated. ,when Only when the time comes, the method of formula (21) is used for calculation.

[0080] (twenty one) In the formula, q For summation index; For node voltage or branch power flow m - q Accumulative quantity; For node voltage or branch power flow m Accumulative quantity; for T of q Order component; In addition, middle, for m max The symmetric matrix of order 1 is obtained by combining formula (21). It can find the elements of the matrix. for: (twenty two) In the formula: e 1、 e 2 represents the row and column indices of the matrix; s It is a summation index. express T The s Order component; For a three-parameter function defined by formula (21): (twenty three) In the formula: For the Gamma function, a , b and c These are the three parameters of a three-parameter function, for positive integers. a ,have .

[0081] Unlike traditional deterministic power flow calculations, probabilistic power flow can directly handle random variables. Both inputs and outputs are probability density functions, not deterministic values, effectively addressing uncertainties in grid operation. Through probabilistic power flow calculations based on the cumulant method, the cumulants calculated from the probability density functions of photovoltaic output and load demand can be transformed into probability density functions of node voltage and branch power flow, and then passed to the subsequent two-layer energy storage planning model.

[0082] Step A3 involves constructing a two-layer energy storage planning model that considers photovoltaic and load uncertainties, specifically including: Energy storage systems can maintain a probabilistic balance between supply and demand through flexible charging and discharging operations. However, the high investment and construction costs of energy storage make large-scale application difficult. Therefore, the installation location and rated capacity of energy storage need to comprehensively consider both technical performance and economic factors. This invention proposes a two-layer planning method for energy storage to select locations and determine capacity. This invention constructs a two-layer energy storage planning model as an interconnected power grid energy storage planning model. The two-layer energy storage planning model includes an upper-layer optimization model and a lower-layer optimization model. The upper-layer optimization model receives the probability density functions of node voltages and branch power flows in the regional power grid, aiming to minimize the comprehensive effectiveness coefficient of energy storage installation on grid performance, and obtains the optimal installation location of the energy storage. The lower-layer optimization model, based on the optimal installation location of the energy storage, aims to minimize the annual planning and operating costs of the interconnected power grid, optimizes the rated capacity, rated power, and charging and discharging scheduling strategy of the energy storage, and returns the corresponding energy storage charging and discharging power curves to the upper-layer optimization model. That is, this invention determines the optimal installation location of the energy storage through upper-layer optimization, and determines the rated power, capacity, and simulated charging and discharging scheduling strategy of the energy storage through lower-layer optimization.

[0083] The two-layer model (i.e., the two-layer energy storage planning model) accepts the probability density functions of node voltage and branch power flow obtained from probabilistic power flow calculations. And the sensitivity matrix of node voltage and branch power flow with respect to input variables. For upper-level optimization, the comprehensive efficiency coefficient is used. The objective function for measuring the improvement in grid performance caused by the installation of energy storage is as follows: (twenty four) In the formula: Net power fluctuations of the regional power grid Voltage over-limit probability of regional power grid Network losses in regional power grids The efficacy coefficient. , , They are respectively , , The weighting coefficients of the three items.

[0084] Net power fluctuation It is the average peak-to-valley difference of net load power during the production simulation cycle, serving as an indicator of the peak-shaving and valley-filling capability of the selected energy storage configuration. The calculation formula is: (25) In the formula: The sampling time interval, This represents the maximum net power within the sampling time interval. This represents the minimum net power within the sampling time interval. TzTo optimize the total duration.

[0085] Voltage over-limit probability This refers to the probability that the node voltage exceeds the preset safe operating range. It is calculated based on the probability density function of the node voltage in the regional power grid, and the calculation formula is as follows: (26) In the formula: For regional power grid nodes j The voltage probability density function, The voltage of the nodes in the regional power grid; and These are the upper and lower limits of the voltage. J This represents the total number of nodes in the regional power grid.

[0086] Network loss This refers to an indicator for reducing network losses in energy storage systems. By using the probability density function of branch power flow, the mathematical expectation of the squared power of each branch can be calculated, thereby determining the network losses. .

[0087] The efficacy coefficient method can normalize multi-objective problems, transforming them into single-objective functions, and can effectively solve multi-objective optimization problems. The calculation is as follows: (27) In the formula: For the first The individual efficacy coefficient values ​​of each variable. ; For the first The actual values ​​of each evaluation indicator; For the first The satisfaction value of each evaluation indicator; For the first The unacceptable value of each evaluation indicator. It is the first constant. It is the second constant. This is the third constant. In this embodiment, The value is 40. The value is 60. The value is 100.

[0088] After the upper-level optimization model selects the energy storage location, it passes the optimal location result to the lower-level optimization model. For the lower-level optimization, the objective is to minimize the annual planning and operating costs of the interconnected power grid. The objective function is as follows: (28) In the formula: The annual planning and operating costs of the interconnected power grid consist of the annualized investment cost of energy storage, voltage over-limit penalties, and annualized operation and maintenance costs of energy storage. As a capital recovery factor, For the discount rate, For the lifespan of the energy storage system, This represents the total number of energy storage units installed. Cost per unit capacity of energy storage Cost per unit power of energy storage For energy storage z Rated capacity, For energy storage z Rated power, As voltage penalty weight, , These are the nodes of the regional power grid. j exist t The positive and negative over-limit values ​​of the voltage at any given time. This is the operation and management cost coefficient for the energy storage system. For energy storage z At any moment t The charging power, For energy storage z At any moment t The discharge power.

[0089] The constraints of the lower-level optimization model are as follows: (29) In the formula: Formula (29a) is for energy storage z The relationship between rated capacity and rated power is constrained. This is the ratio of rated energy storage capacity to rated energy storage power. Formula (29b) is the charging power limit. Formula (29c) is the discharging power limit. Formula (29d) is the charging / discharging mutual exclusion constraint. Formula (29e) is the power balance constraint. J This represents the total number of nodes in the regional power grid. For the partitioned power grid node j in t Power injected at all times , For regional power grid nodes j exist t The load demand power and photovoltaic output power at any given time. , These represent the energy storage charge and discharge efficiencies, respectively. Equation (29f) represents the state of charge (SOC) constraint of the energy storage system. For energy storage z exist t SOC at any moment , These represent the minimum and maximum SOC values ​​for energy storage, respectively. Formula (29g) represents the end-of-day SOC constraint. For energy storage z Daily starting SOC For the end-of-day tolerance of SOC, For energy storage z The end-of-day SOC value. Formula (29h) is for energy storage. z The SOC state update formula.

[0090] Example 2: This invention also provides a method for applying an interconnected power grid energy storage planning model, such as... Figure 2 As shown, it includes: Step B1: Based on the historical datasets of photovoltaic power output or load demand, the Gaussian mixture model algorithm is used to process the uncertainty of photovoltaic power output or load demand to obtain the probability density function of photovoltaic power output or load demand. Step B2: Based on the probability density function of photovoltaic power output and the probability density function of load demand, use the cumulant method to perform probabilistic power flow calculation to obtain the probability density function of node voltage and branch power flow of the regional power grid. Step B3: Input the probability density function of the node voltage and the probability density function of the branch power flow of the regional power grid into the upper planning model of the two-layer energy storage planning model, and solve for the installation location of the energy storage. Step B4: Input the installation location of the energy storage into the lower planning model of the two-layer energy storage planning model, and solve it to obtain the rated capacity, rated power and charging and discharging scheduling strategy of the energy storage, as well as the corresponding energy storage charging and discharging power curve. Step B5: Return the energy storage charging and discharging power curve to the upper-level planning model, recalculate the comprehensive efficiency coefficient, and iteratively update the energy storage installation location, rated capacity, rated power, and charging and discharging scheduling strategy until the comprehensive efficiency coefficient no longer decreases, thus obtaining the optimal energy storage installation location, rated capacity, rated power, and charging and discharging scheduling strategy.

[0091] This embodiment solves the established interconnected power grid energy storage planning model for energy storage planning. The two-layer energy storage planning model receives the probability density functions of node voltage and branch power flow obtained from probabilistic power flow calculations, which are used to calculate voltage exceedance probabilities and network losses in the upper-layer optimization model. The upper-layer optimization model solves the optimization model by minimizing the comprehensive efficiency coefficient to obtain the installation location of the energy storage system, and transmits this to the lower-layer optimization model. The lower-layer optimization model, based on the energy storage installation location given by the upper layer, uses a solver to optimize the rated capacity, rated power, and charging / discharging scheduling strategy of the energy storage system while minimizing the total cost. The resulting energy storage charging / discharging power curve is returned to the upper-layer optimization model to recalculate its comprehensive efficiency coefficient, iterating until the comprehensive efficiency coefficient no longer decreases. Through the two-layer energy storage planning model, the installation location, capacity, and scheduling strategy of the energy storage system can be obtained. The entire process relies on modeling the uncertainties of the power grid and corresponding probabilistic calculation methods.

[0092] Example 3: This embodiment proposes an energy storage planning method for interconnected power grids. It uses Gaussian Model (GMM) to model the uncertainty of photovoltaic output and load demand, and then uses the cumulant method for probabilistic power flow calculation. This constructs a two-layer energy storage planning method that considers the uncertainty of both photovoltaic output and load demand, and analyzes voltage over-limit situations under extreme line interruption scenarios.

[0093] A schematic diagram of an interconnected power grid system is shown below. Figure 3 As shown in the figure, PV1-PV4 represent photovoltaic cells 1-4. The operation of the interconnected grid before and after energy storage configuration is analyzed, considering the uncertainties in photovoltaic output and load demand, to optimize the time frame. Tz =8760, =1. Photovoltaic power sources were connected to node 10 of zone 1, nodes 3 and 10 of zone 2, and node 8 of zone 3. Historical data on annual load demand and photovoltaic output are as follows: Figure 4 and 5 As shown.

[0094] The optimization model yielded the following final site selection results for the energy storage units: node 5 of zone 1, node 1 of zone 2, and node 3 of zone 3. The optimal energy capacity and optimal power capacity of each energy storage unit are shown in Table 1. The energy storage units maintain supply and demand balance through charging and discharging behavior. Figure 6 , Figure 7 The figures show the charge / discharge curves of the energy storage unit in Zone 1 over one year and the SOC change curve of the energy storage unit in Zone 1 over one year. Figure 8 , Figure 9 The figures show the charge / discharge curves and SOC change curves of the energy storage unit 2 in the sub-grid for one year. Figure 10 , Figure 11The figures show the charge / discharge curves and SOC (State of Charge) change curves for the three energy storage units in the regional power grid over one year. The SOC change curve can simulate the charge / discharge operation of the energy storage over the next year.

[0095] Table 1 Planning Results for Each Energy Storage Unit

[0096] Figure 12 To compare the probability of voltage exceeding limits in each regional power grid before and after the installation of energy storage. Figure 13 This figure compares the average voltage of each grid zone before and after energy storage installation. Grid zone 1 did not experience any voltage exceedances before or after energy storage installation, so it is not shown in the figure. As can be seen from the figure, without energy storage, the voltage exceedance probability of grid zone 2 is 32%, and that of grid zone 3 is 47%. These high voltage exceedance probabilities severely impact power quality and supply reliability. Furthermore, before energy storage installation, the average voltage of grid zones 2 and 3 was low, both less than 0.96 pu. After energy storage was installed, the voltage exceedance probabilities of grid zones 2 and 3 decreased to 5% and 6%, respectively. This indicates that the voltage exceedance probability was significantly reduced after the energy storage unit was installed, and voltage exceedances are virtually eliminated. The average voltage of grid zones 2 and 3 also improved to some extent, both exceeding 0.96 pu. This demonstrates that the voltage exceedance problem was effectively solved after configuring the energy storage system, and grid stability was improved.

[0097] Extreme cases are simulated by disconnecting the connection between nodes 9 and 10 in partitioned power grid 2. Figure 14 This compares the probability of voltage exceeding limits with and without energy storage in extreme scenarios. Figure 15 This chart compares the average voltage with and without energy storage under extreme scenarios. Without energy storage, the voltage exceedance probability of grid zones 2 and 3 is close to 40% in extreme scenarios. However, after configuring energy storage, the voltage exceedance probability of both grid zones 2 and 3 drops to around 5%. Without energy storage, the average voltage of grid zones 2 and 3 is less than 0.96 pu in extreme scenarios, especially grid zone 3, where the average voltage approaches the lower voltage limit. After configuring energy storage, the voltage of each grid zone increases significantly, returning to a higher level, demonstrating that the energy storage system can effectively improve the voltage exceedance situation of each grid zone even under extreme conditions.

[0098] Overall, the energy storage plan has achieved remarkable results, effectively reducing the probability of voltage exceeding limits in regional power grids and maintaining the probability balance between supply and demand in regional power grids.

[0099] Example 4: Based on the same inventive concept, this invention also provides a system for constructing an interconnected power grid energy storage planning model, such as... Figure 16 As shown, it includes: The photovoltaic load probability density module is used to process the uncertainty of photovoltaic output or load demand based on historical datasets of photovoltaic power output or load demand, respectively, and obtain the probability density function of photovoltaic power output or load demand. The voltage power flow probability density module is used to perform probabilistic power flow calculations using the cumulant method based on the probability density function of photovoltaic output and the probability density function of load demand, so as to obtain the probability density function of node voltage and branch power flow of the regional power grid. The modeling module is used to construct a two-layer energy storage planning model that considers photovoltaic and load uncertainties as an interconnected grid energy storage planning model, based on the probability density function of node voltage and branch power flow in the partitioned grid. The dual-layer energy storage planning model includes an upper-layer optimization model and a lower-layer optimization model. The upper-layer optimization model receives the probability density function of the node voltage and the probability density function of the branch power flow in the regional power grid, and aims to minimize the comprehensive effectiveness coefficient of energy storage installation on grid performance to obtain the optimal installation location of energy storage. The lower-layer optimization model, based on the optimal installation location of energy storage, aims to minimize the annual planning and operating costs of the interconnected power grid, optimizes the rated capacity, rated power, and charging and discharging scheduling strategy of energy storage, and returns the corresponding energy storage charging and discharging power curves to the upper-layer optimization model.

[0100] Preferably, the photovoltaic load probability density module is specifically used for: For photovoltaic power output or load demand, the historical dataset is divided using the K-means clustering algorithm to obtain multiple clusters, each cluster corresponding to a Gaussian component in a Gaussian mixture model algorithm; The mean and covariance matrix of the samples corresponding to each Gaussian component are calculated respectively as the initial values ​​of the mean and covariance matrix of the Gaussian component, and the weights of each Gaussian component are evenly distributed as the initial values ​​of the weights of each Gaussian component. For each Gaussian component, the current weights, mean, and covariance matrices are used respectively. The weights, mean, and covariance matrices of the Gaussian components are iteratively updated using the expectation-maximization algorithm until the weights, mean, and covariance matrices converge. Based on the weights, mean, and covariance matrices of each Gaussian component, the probability density function of photovoltaic power output or load demand is calculated. The number of clusters is determined using the Bayesian information criterion method.

[0101] Preferably, the photovoltaic load probability density module, for each Gaussian component, uses the current weights, mean, and covariance matrix, and iteratively updates the weights, mean, and covariance matrix of the Gaussian component using an expectation-maximization algorithm until the weights, mean, and covariance matrix all converge, including: For each Gaussian component, the posterior probability is calculated based on the weight, mean, and covariance matrix of the Gaussian component. Using the posterior probability, the weights, mean, and covariance matrices of each Gaussian component are recalculated; Determine whether the maximum likelihood function corresponding to the weights, mean, and covariance matrices of each Gaussian component has converged. If yes, output the weights, mean, and covariance matrices of each Gaussian component and end the process. Otherwise, jump to calculate the posterior probability for each Gaussian component based on its weights, mean, and covariance matrices.

[0102] Preferably, the formula for calculating the posterior probability in the photovoltaic load probability density module is as follows:

[0103] In the formula, Indicates the first k The Gaussian component of the first n The posterior probability of each data point. Represents the first in the historical dataset n One data point, k Indicates the first k Gaussian components Indicates the first k The weights of the Gaussian components in the previous iteration. K This represents the total number of optimal Gaussian components. Representing data Gaussian distribution, Indicates the first k The mean of the Gaussian components in the previous iteration Indicates the first k The covariance matrix of the Gaussian components in the previous iteration; The formula for recalculating the weights of the Gaussian components using posterior probabilities is as follows:

[0104] The formula for recalculating the mean of the Gaussian components using posterior probability is as follows:

[0105] The formula for recalculating the covariance matrix of the Gaussian components using posterior probabilities is as follows:

[0106] In the formula, Indicates the firstk The weights of the Gaussian components, Indicates the first k The mean of the Gaussian components, Indicates the first k The covariance matrix of Gaussian components, N For the first k The number of samples in each Gaussian component.

[0107] Preferably, the probability density function of photovoltaic output or load demand in the photovoltaic load probability density module is calculated as follows:

[0108] In the formula, for The probability density function, For photovoltaic power output or load demand, K This represents the total number of optimal Gaussian components. For the first k The weights of the Gaussian components, Representing data Gaussian distribution, For the first k The mean of the Gaussian components, For the first k The covariance matrix of Gaussian components.

[0109] Preferably, the voltage power flow probability density module is specifically used for: The supply and demand balance equations of the regional power grid are rewritten in matrix form, and the power flow equations are expanded into Taylor series at the reference operating point using the Newton-Raphson method, while ignoring higher-order terms, to obtain a power flow linearization model. The power flow linearization model is rearranged and reorganized, and the probability density functions of photovoltaic output and load demand are substituted to obtain the cumulative quantities of each order of the regional grid voltage and branch power flow. Based on the cumulative quantities of the voltage and branch power flow of the regional grid, the probability density functions of the node voltage and branch power flow are calculated using the C-type Gram-Charlier series expansion method.

[0110] Preferably, the probability density function of node voltage or branch power flow in the voltage power flow probability density module is calculated as follows:

[0111] In the formula, Let be the probability density function of node voltage or branch power flow, where y For node voltage or branch power flow; for m The series expansion coefficients of order , , m Let be the order of the series expansion. The maximum order of the series expansion. for m Hermite polynomial of order 1 m Hermite polynomial of order The calculation formula is as follows:

[0112] The m Series expansion coefficients of order By solving a system of linear equations Received; In the formula, yes The order column vector is an intermediate variable calculated from the cumulative quantities of node voltages or branch power flows. =0, The formulas for calculating the remaining components are as follows:

[0113] In the formula: It is the standard deviation of node voltage or branch power flow. T for Column vector of order, for T The m-2 order components, The other orders The calculation formula is as follows:

[0114] In the formula, q For summation index; For node voltage or branch power flow m - q Accumulative quantity; For node voltage or branch power flow m Accumulative quantity; for T of q Order component; for 1-th order symmetric matrix Middle elements The calculation formula is as follows:

[0115] In the formula, e 1 and e 2 represents the row and column indices of the matrix. s It is a summation index. express TThe s Order component; This is a three-parameter function, and its calculation formula is as follows:

[0116] In the formula: For the Gamma function, a , b and c These are the three parameters of a three-parameter function.

[0117] Preferably, the objective function of the upper-level optimization model in the modeling module is calculated as follows:

[0118] In the formula: For the comprehensive effectiveness coefficient, Net power fluctuation of the regional power grid The efficacy coefficient, Voltage over-limit probability of the regional power grid The efficacy coefficient, Network losses for regional power grids The efficacy coefficient, , , They are respectively , , Weighting coefficients; Net power fluctuation The calculation formula is as follows:

[0119] In the formula: The sampling time interval, This represents the maximum net power within the sampling time interval. This represents the minimum net power within the sampling time interval. Tz To optimize the total duration; Voltage over-limit probability of the partitioned power grid The calculation formula is as follows:

[0120] In the formula: For regional power grid nodes j The voltage probability density function, For the voltage of the nodes in the regional power grid, This is the upper limit of the voltage. This is the lower limit of the voltage. J The total number of nodes in the regional power grid; Network losses in the regional power grid The calculation formula is as follows:

[0121] In the formula, Network loss is calculated using the probability density function of branch power flow; Efficacy coefficient The calculation formula is as follows:

[0122] In the formula, For the first d Individual efficacy coefficients of each variable , For the first d The actual value of each evaluation indicator For the first d The satisfaction value of each evaluation indicator For the first The unacceptable value of each evaluation indicator. It is the first constant. It is the second constant. It is the third constant.

[0123] Preferably, the objective function of the lower-level optimization model in the modeling module is calculated as follows:

[0124] In the formula: For the annual planning and operating costs of the interconnected power grid, As a capital recovery factor, For the discount rate, For the lifespan of the energy storage system, This represents the total number of energy storage units installed. J This represents the total number of nodes in the regional power grid. Cost per unit capacity of energy storage Cost per unit power of energy storage For energy storage z Rated capacity, For energy storage z Rated power, W As voltage penalty weight, For regional power grid nodes j exist t The voltage at that moment exceeds the limit. For regional power grid nodes j exist t The voltage at any given time exceeds the negative limit. This is the operation and management cost coefficient for the energy storage system. For energy storage z At any moment t The charging power, For energy storage z At any moment t The discharge power; The constraints of the lower-level optimization model are as follows:

[0125] In the formula: This is the ratio of rated energy storage capacity to rated energy storage power. For regional power grid nodes j exist t The active power injected at all times, Regional power grid nodes j exist t Load power at any given time For regional power grid nodes j exist t Photovoltaic output power at any given time For energy storage charging efficiency, For energy storage and discharge efficiency, For energy storage z exist t SOC at any moment For energy storage z exist t SOC at time +1 The minimum SOC value for energy storage. This represents the maximum SOC value for energy storage. For energy storage z Daily starting SOC This represents the end-of-day tolerance of the SOC.

[0126] Example 5: Based on the same inventive concept, this invention also provides an application system for an interconnected power grid energy storage planning model, such as... Figure 17 As shown, it includes: The data input module is used to process the uncertainty of photovoltaic output or load demand based on historical datasets of photovoltaic power output or load demand, respectively, and obtain the probability density function of photovoltaic power output or load demand. The voltage and power flow calculation module is used to perform probabilistic power flow calculations using the cumulant method based on the probability density function of the photovoltaic output and the probability density function of the load demand, so as to obtain the probability density function of the node voltage and the probability density function of the branch power flow in the regional power grid. The location calculation module is used to input the probability density function of the node voltage and the probability density function of the branch power flow of the partitioned power grid into the upper planning model of the two-layer energy storage planning model, and solve for the installation location of the energy storage. The capacity strategy module is used to input the installation location of the energy storage into the lower planning model of the two-layer energy storage planning model, and solve it to obtain the rated capacity, rated power, and charging and discharging scheduling strategy of the energy storage, as well as the corresponding energy storage charging and discharging power curve. The iterative update module is used to return the energy storage charging and discharging power curve to the upper-level planning model, recalculate the comprehensive efficiency coefficient, and iteratively update the energy storage installation location, rated capacity, rated power, and charging and discharging scheduling strategy until the comprehensive efficiency coefficient no longer decreases, thus obtaining the optimal energy storage installation location, rated capacity, rated power, and charging and discharging scheduling strategy.

[0127] Example 6 like Figure 18 As shown, the present invention also provides an electronic device, which may be a computer device, a microcontroller device, a smart mobile device, etc. The electronic device in this embodiment may include a processor, a memory, a transceiver component, etc. The memory, processor, and transceiver component are connected via a bus; the memory can be used to store executable programs, and an exemplary executable program may include instructions; the processor is used to execute the instructions stored in the memory. The memory can also be used to store data, which can be accessed and / or modified when instructions are executed.

[0128] The processor may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing and control core of the terminal, and it is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the storage medium to realize the corresponding method flow or corresponding function, so as to realize the steps of the interconnected power grid energy storage planning model construction or application method in the above embodiments.

[0129] Example 7 Based on the same inventive concept, this invention also provides a readable storage medium, specifically an electronic device readable storage medium (Memory). This readable storage medium is a memory device within an electronic device used to store programs and data. It is understood that the storage medium here can include both built-in storage media within the electronic device and extended storage media supported by the electronic device. The storage medium provides storage space, which stores the terminal's operating system. Furthermore, this storage space also stores one or more instructions suitable for loading and execution by a processor. These instructions can be one or more executable programs (including program code). It should be noted that the storage medium here can be high-speed RAM or non-volatile memory, such as at least one disk storage device. Loading and executing one or more instructions stored in the storage medium by the processor can implement the steps of the interconnected power grid energy storage planning model construction or application method described in the above embodiments.

[0130] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0131] 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.

[0132] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0133] 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.

[0134] The above are merely embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of the claims of the present invention pending approval.

Claims

1. A method for constructing an interconnected power grid energy storage planning model, characterized in that, include: Based on historical datasets of photovoltaic power output or load demand, Gaussian mixture model algorithm is used to process the uncertainty of photovoltaic power output or load demand, and the probability density function of photovoltaic power output or load demand is obtained. Based on the probability density function of photovoltaic output and the probability density function of load demand, the cumulative method is used to perform probabilistic power flow calculation to obtain the probability density function of node voltage and branch power flow of the regional power grid. Based on the probability density functions of node voltage and branch power flow in the regional power grid, a two-layer energy storage planning model considering photovoltaic and load uncertainties is constructed as an energy storage planning model for the interconnected power grid. The dual-layer energy storage planning model includes an upper-layer optimization model and a lower-layer optimization model. The upper-layer optimization model receives the probability density function of the node voltage and the probability density function of the branch power flow in the regional power grid, and aims to minimize the comprehensive effectiveness coefficient of energy storage installation on grid performance to obtain the optimal installation location of energy storage. The lower-layer optimization model, based on the optimal installation location of energy storage, aims to minimize the annual planning and operating costs of the interconnected power grid, optimizes the rated capacity, rated power, and charging and discharging scheduling strategy of energy storage, and returns the corresponding energy storage charging and discharging power curves to the upper-layer optimization model.

2. The method for constructing an interconnected power grid energy storage planning model as described in claim 1, characterized in that, The method, based on historical datasets of photovoltaic power output or load demand respectively, uses a Gaussian mixture model algorithm to process the uncertainty of photovoltaic power output or load demand, and obtains the probability density function of photovoltaic power output or load demand, including: For photovoltaic power output or load demand, the historical dataset is divided using the K-means clustering algorithm to obtain multiple clusters, each cluster corresponding to a Gaussian component in a Gaussian mixture model algorithm; The mean and covariance matrix of the samples corresponding to each Gaussian component are calculated respectively as the initial values ​​of the mean and covariance matrix of the Gaussian component, and the weights of each Gaussian component are evenly distributed as the initial values ​​of the weights of each Gaussian component. For each Gaussian component, the current weights, mean, and covariance matrices are used respectively. The weights, mean, and covariance matrices of the Gaussian components are iteratively updated using the expectation-maximization algorithm until the weights, mean, and covariance matrices converge. Based on the weights, mean, and covariance matrices of each Gaussian component, the probability density function of photovoltaic power output or load demand is calculated. The number of clusters is determined using the Bayesian information criterion method.

3. The method for constructing an interconnected power grid energy storage planning model as described in claim 2, characterized in that, For each Gaussian component, the current weights, mean, and covariance matrix are used respectively. The weights, mean, and covariance matrix of the Gaussian component are iteratively updated using the expectation-maximization algorithm until all weights, mean, and covariance matrices converge. This includes: For each Gaussian component, the posterior probability is calculated based on the weight, mean, and covariance matrix of the Gaussian component. Using the posterior probability, the weights, mean, and covariance matrices of each Gaussian component are recalculated; Determine whether the maximum likelihood function corresponding to the weights, mean, and covariance matrices of each Gaussian component has converged. If yes, output the weights, mean, and covariance matrices of each Gaussian component and end the process. Otherwise, jump to calculate the posterior probability for each Gaussian component based on its weights, mean, and covariance matrices.

4. The method for constructing an interconnected power grid energy storage planning model as described in claim 2, characterized in that, The formula for calculating the posterior probability is as follows: In the formula, Indicates the first k The Gaussian component of the first n The posterior probability of each data point. Represents the first in the historical dataset n One data point, k Indicates the first k Gaussian components Indicates the first k The weights of the Gaussian components in the previous iteration. K This represents the total number of optimal Gaussian components. Representing data Gaussian distribution, Indicates the first k The mean of the Gaussian components in the previous iteration Indicates the first k The covariance matrix of the Gaussian components in the previous iteration; The formula for recalculating the weights of the Gaussian components using posterior probabilities is as follows: The formula for recalculating the mean of the Gaussian components using posterior probability is as follows: The formula for recalculating the covariance matrix of the Gaussian components using posterior probabilities is as follows: In the formula, Indicates the first k The weights of the Gaussian components, Indicates the first k The mean of the Gaussian components, Indicates the first k The covariance matrix of Gaussian components, N For the first k The number of samples in each Gaussian component.

5. The method for constructing an interconnected power grid energy storage planning model as described in claim 2, characterized in that, The probability density function of the photovoltaic power output or load demand is calculated as follows: In the formula, for The probability density function, For photovoltaic power output or load demand, K This represents the total number of optimal Gaussian components. For the first k The weights of the Gaussian components, Representing data Gaussian distribution, For the first k The mean of the Gaussian components, For the first k The covariance matrix of Gaussian components.

6. The method for constructing an interconnected power grid energy storage planning model as described in claim 1, characterized in that, The probability density function of photovoltaic output and the probability density function of load demand are used to perform probabilistic power flow calculations using the cumulant method to obtain the probability density functions of voltage and branch power flow in the regional power grid, including: The supply and demand balance equations of the regional power grid are rewritten in matrix form, and the power flow equations are expanded into Taylor series at the reference operating point using the Newton-Raphson method, while ignoring higher-order terms, to obtain a power flow linearization model. The power flow linearization model is rearranged and reorganized, and the probability density functions of photovoltaic output and load demand are substituted to obtain the cumulative quantities of each order of the regional grid voltage and branch power flow. Based on the cumulative quantities of the voltage and branch power flow of the regional grid, the probability density functions of the node voltage and branch power flow are calculated using the C-type Gram-Charlier series expansion method.

7. The method for constructing an interconnected power grid energy storage planning model as described in claim 6, characterized in that, The probability density function of the node voltage or branch power flow is calculated as follows: In the formula, Let be the probability density function of node voltage or branch power flow, where y For node voltage or branch power flow; for m The series expansion coefficients of order , , m Let be the order of the series expansion. The maximum order of the series expansion. for m Hermite polynomial of order 1 m Hermite polynomial of order The calculation formula is as follows: The m Series expansion coefficients of order By solving a system of linear equations Obtained; In the formula, yes The order column vector is an intermediate variable calculated from the cumulative quantities of node voltages or branch power flows of various orders. =0, The formulas for calculating the remaining components are as follows: In the formula: It is the standard deviation of node voltage or branch power flow. T for Column vector of order, for T The m-2 order components, The other orders The calculation formula is as follows: In the formula, q For summation index; For node voltage or branch power flow m - q Accumulative quantity; For node voltage or branch power flow m Accumulative quantity; for T of q Order component; for 1-th order symmetric matrix Middle elements The calculation formula is as follows: In the formula, e 1 and e 2 represents the row and column indices of the matrix. s It is a summation index. express T The s Order component; This is a three-parameter function, and its calculation formula is as follows: In the formula: For the Gamma function, a , b and c These are the three parameters of a three-parameter function.

8. The method for constructing an interconnected power grid energy storage planning model as described in claim 1, characterized in that, The objective function of the upper-level optimization model is calculated as follows: In the formula: For the comprehensive effectiveness coefficient, Net power fluctuation of the regional power grid The efficacy coefficient, Voltage over-limit probability of the regional power grid The efficacy coefficient, Network losses for regional power grids The efficacy coefficient, , , They are respectively , , Weighting coefficients; Net power fluctuation The calculation formula is as follows: In the formula: The sampling time interval, This represents the maximum net power within the sampling time interval. This represents the minimum net power within the sampling time interval. Tz To optimize the total duration; Voltage over-limit probability of the partitioned power grid The calculation formula is as follows: In the formula: For regional power grid nodes j The voltage probability density function, For the voltage of the nodes in the regional power grid, This is the upper limit of the voltage. This is the lower limit of the voltage. J This represents the total number of nodes in the regional power grid. Network losses in the regional power grid The calculation formula is as follows: In the formula, Network loss is calculated using the probability density function of branch power flow; Efficacy coefficient The calculation formula is as follows: In the formula, For the first d Individual efficacy coefficients of each variable , For the first d The actual value of each evaluation indicator For the first d The satisfaction value of each evaluation indicator For the first The unacceptable value of each evaluation indicator. It is the first constant. It is the second constant. It is the third constant.

9. The method for constructing an interconnected power grid energy storage planning model as described in claim 1, characterized in that, The objective function of the lower-level optimization model is calculated as follows: In the formula: For the annual planning and operating costs of the interconnected power grid, As a capital recovery factor, For the discount rate, For the lifespan of the energy storage system, This represents the total number of energy storage units installed. J This represents the total number of nodes in the regional power grid. Cost per unit capacity of energy storage Cost per unit power of energy storage For energy storage z Rated capacity, For energy storage z Rated power, W As voltage penalty weight, For regional power grid nodes j exist t The voltage at that moment exceeds the limit. For regional power grid nodes j exist t The voltage at any given time exceeds the negative limit. This is the operation and management cost coefficient for the energy storage system. For energy storage z At any moment t The charging power, For energy storage z At any moment t The discharge power; The constraints of the lower-level optimization model are as follows: In the formula: This is the ratio of rated energy storage capacity to rated energy storage power. For regional power grid nodes j exist t The active power injected at all times, Regional power grid nodes j exist t Load power at any given time For regional power grid nodes j exist t Photovoltaic output power at any given time For energy storage charging efficiency, For energy storage and discharge efficiency, For energy storage z exist t SOC at any moment For energy storage z exist t SOC at time +1 The minimum SOC value for energy storage. This represents the maximum SOC value for energy storage. For energy storage z Daily starting SOC For the end-of-day tolerance of SOC, For energy storage z The daily SOC value.

10. A method for applying an interconnected power grid energy storage planning model, characterized in that, include: Based on historical datasets of photovoltaic power output or load demand, Gaussian mixture model algorithm is used to process the uncertainty of photovoltaic power output or load demand, and the probability density function of photovoltaic power output or load demand is obtained. Based on the probability density function of photovoltaic output and the probability density function of load demand, the cumulative method is used to perform probabilistic power flow calculation to obtain the probability density function of node voltage and branch power flow of the regional power grid. The probability density functions of the node voltage and the branch power flow of the partitioned power grid are input into the upper planning model of the two-layer energy storage planning model to solve for the installation location of the energy storage. The installation location of the energy storage is passed into the lower planning model of the two-layer energy storage planning model, and the rated capacity, rated power, and charging and discharging scheduling strategy of the energy storage, as well as the corresponding energy storage charging and discharging power curve, are obtained by solving the problem. The energy storage charging and discharging power curve is returned to the upper-level planning model, the comprehensive efficiency coefficient is recalculated, and the installation location, rated capacity, rated power, and charging and discharging scheduling strategy of the energy storage are iteratively updated until the comprehensive efficiency coefficient no longer decreases, thus obtaining the optimal installation location, rated capacity, rated power, and charging and discharging scheduling strategy of the energy storage.

11. A system for constructing an interconnected power grid energy storage planning model, characterized in that, include: The photovoltaic load probability density module is used to process the uncertainty of photovoltaic output or load demand based on historical datasets of photovoltaic power output or load demand, respectively, and obtain the probability density function of photovoltaic power output or load demand. The voltage power flow probability density module is used to perform probabilistic power flow calculations using the cumulant method based on the probability density function of photovoltaic output and the probability density function of load demand, so as to obtain the probability density function of node voltage and branch power flow of the regional power grid. The modeling module is used to construct a two-layer energy storage planning model that considers photovoltaic and load uncertainties as an interconnected grid energy storage planning model, based on the probability density function of node voltage and branch power flow in the partitioned grid. The dual-layer energy storage planning model includes an upper-layer optimization model and a lower-layer optimization model. The upper-layer optimization model receives the probability density function of the node voltage and the probability density function of the branch power flow in the regional power grid, and aims to minimize the comprehensive effectiveness coefficient of energy storage installation on grid performance to obtain the optimal installation location of energy storage. The lower-layer optimization model, based on the optimal installation location of energy storage, aims to minimize the annual planning and operating costs of the interconnected power grid, optimizes the rated capacity, rated power, and charging and discharging scheduling strategy of energy storage, and returns the corresponding energy storage charging and discharging power curves to the upper-layer optimization model.

12. The interconnected power grid energy storage planning model construction system as described in claim 11, characterized in that, The photovoltaic load probability density module is specifically used for: For photovoltaic power output or load demand, the historical dataset is divided using the K-means clustering algorithm to obtain multiple clusters, each cluster corresponding to a Gaussian component in a Gaussian mixture model algorithm; The mean and covariance matrix of the samples corresponding to each Gaussian component are calculated respectively as the initial values ​​of the mean and covariance matrix of the Gaussian component, and the weights of each Gaussian component are evenly distributed as the initial values ​​of the weights of each Gaussian component. For each Gaussian component, the current weights, mean, and covariance matrices are used respectively. The weights, mean, and covariance matrices of the Gaussian components are iteratively updated using the expectation-maximization algorithm until the weights, mean, and covariance matrices converge. Based on the weights, mean, and covariance matrices of each Gaussian component, the probability density function of photovoltaic power output or load demand is calculated. The number of clusters is determined using the Bayesian information criterion method.

13. The interconnected power grid energy storage planning model construction system as described in claim 12, characterized in that, The photovoltaic load probability density module, for each Gaussian component, uses the current weights, mean, and covariance matrices, and iteratively updates the weights, mean, and covariance matrices of the Gaussian components using an expectation-maximization algorithm until all weights, mean, and covariance matrices converge, including: For each Gaussian component, the posterior probability is calculated based on the weight, mean, and covariance matrix of the Gaussian component. Using the posterior probability, the weights, mean, and covariance matrices of each Gaussian component are recalculated; Determine whether the maximum likelihood function corresponding to the weights, mean, and covariance matrices of each Gaussian component has converged. If yes, output the weights, mean, and covariance matrices of each Gaussian component and end the process. Otherwise, jump to calculate the posterior probability for each Gaussian component based on its weights, mean, and covariance matrices.

14. The interconnected power grid energy storage planning model construction system as described in claim 12, characterized in that, The formula for calculating the posterior probability in the photovoltaic load probability density module is as follows: In the formula, Indicates the first k The Gaussian component of the first n The posterior probability of each data point. Represents the first in the historical dataset n One data point, k Indicates the first k Gaussian components Indicates the first k The weights of the Gaussian components in the previous iteration. K This represents the total number of optimal Gaussian components. Representing data Gaussian distribution, Indicates the first k The mean of the Gaussian components in the previous iteration Indicates the first k The covariance matrix of the Gaussian components in the previous iteration; The formula for recalculating the weights of the Gaussian components using posterior probabilities is as follows: The formula for recalculating the mean of the Gaussian components using posterior probability is as follows: The formula for recalculating the covariance matrix of the Gaussian components using posterior probabilities is as follows: In the formula, Indicates the first k The weights of the Gaussian components, Indicates the first k The mean of the Gaussian components, Indicates the first k The covariance matrix of Gaussian components, N For the first k The number of samples in each Gaussian component.

15. The interconnected power grid energy storage planning model construction system as described in claim 12, characterized in that, The formula for calculating the probability density function of photovoltaic output or load demand in the photovoltaic load probability density module is as follows: In the formula, for The probability density function, For photovoltaic power output or load demand, K This represents the total number of optimal Gaussian components. For the first k The weights of the Gaussian components, Representing data Gaussian distribution, For the first k The mean of the Gaussian components, For the first k The covariance matrix of Gaussian components.

16. The interconnected power grid energy storage planning model construction system as described in claim 11, characterized in that, The voltage power flow probability density module is specifically used for: The supply and demand balance equations of the regional power grid are rewritten in matrix form, and the power flow equations are expanded into Taylor series at the reference operating point using the Newton-Raphson method, while ignoring higher-order terms, to obtain a power flow linearization model. The power flow linearization model is rearranged and reorganized, and the probability density functions of photovoltaic output and load demand are substituted to obtain the cumulative quantities of each order of the regional grid voltage and branch power flow. Based on the cumulative quantities of the voltage and branch power flow of the regional grid, the probability density functions of the node voltage and branch power flow are calculated using the C-type Gram-Charlier series expansion method.

17. The interconnected power grid energy storage planning model construction system as described in claim 16, characterized in that, The probability density function of node voltage or branch power flow in the voltage power flow probability density module is calculated as follows: In the formula, Let be the probability density function of node voltage or branch power flow, where y For node voltage or branch power flow; for m The series expansion coefficients of order , , m Let be the order of the series expansion. The maximum order of the series expansion. for m Hermite polynomial of order 1 m Hermite polynomial of order The calculation formula is as follows: The m Series expansion coefficients of order By solving a system of linear equations Obtained; In the formula, yes The order column vector is an intermediate variable calculated from the cumulative quantities of node voltages or branch power flows of various orders. =0, The formulas for calculating the remaining components are as follows: In the formula: It is the standard deviation of node voltage or branch power flow. T for Column vector of order, for T The m-2 order components, The other orders The calculation formula is as follows: In the formula, q For summation index; For node voltage or branch power flow m - q Accumulative quantity; For node voltage or branch power flow m Accumulative quantity; for T of q Order component; for 1-th order symmetric matrix Middle elements The calculation formula is as follows: In the formula, e 1 and e 2 represents the row and column indices of the matrix. s It is a summation index. express T The s Order component; This is a three-parameter function, and its calculation formula is as follows: In the formula: For the Gamma function, a , b and c These are the three parameters of a three-parameter function.

18. The interconnected power grid energy storage planning model construction system as described in claim 11, characterized in that, The objective function of the upper-level optimization model in the modeling module is calculated as follows: In the formula: For the comprehensive effectiveness coefficient, Net power fluctuation of the regional power grid The efficacy coefficient, Voltage over-limit probability of the regional power grid The efficacy coefficient, Network losses for regional power grids The efficacy coefficient, , , They are respectively , , Weighting coefficients; Net power fluctuation The calculation formula is as follows: In the formula: The sampling time interval, This represents the maximum net power within the sampling time interval. This represents the minimum net power within the sampling time interval. Tz To optimize the total duration; Voltage over-limit probability of the partitioned power grid The calculation formula is as follows: In the formula: For regional power grid nodes j The voltage probability density function, For the voltage of the nodes in the regional power grid, This is the upper limit of the voltage. This is the lower limit of the voltage. J This represents the total number of nodes in the regional power grid. Network losses in the regional power grid The calculation formula is as follows: In the formula, Network loss is calculated using the probability density function of branch power flow; Efficacy coefficient The calculation formula is as follows: In the formula, For the first d Individual efficacy coefficients of each variable , For the first d The actual value of each evaluation indicator For the first d The satisfaction value of each evaluation indicator For the first The unacceptable value of each evaluation indicator. It is the first constant. It is the second constant. It is the third constant.

19. The interconnected power grid energy storage planning model construction system as described in claim 11, characterized in that, The objective function of the lower-level optimization model in the modeling module is calculated as follows: In the formula: For the annual planning and operating costs of the interconnected power grid, As a capital recovery factor, For the discount rate, For the lifespan of the energy storage system, This represents the total number of energy storage units installed. J This represents the total number of nodes in the regional power grid. Cost per unit capacity of energy storage Cost per unit power of energy storage For energy storage z Rated capacity, For energy storage z Rated power, W As voltage penalty weight, For regional power grid nodes j exist t The voltage at that moment exceeds the limit. For regional power grid nodes j exist t The voltage at any given time exceeds the negative limit. This is the operation and management cost coefficient for the energy storage system. For energy storage z At any moment t The charging power, For energy storage z At any moment t The discharge power; The constraints of the lower-level optimization model are as follows: In the formula: This is the ratio of rated energy storage capacity to rated energy storage power. For regional power grid nodes j exist t The active power injected at all times, Regional power grid nodes j exist t Load power at any given time For regional power grid nodes j exist t Photovoltaic output power at any given time For energy storage charging efficiency, For energy storage and discharge efficiency, For energy storage z exist t SOC at any moment For energy storage z exist t SOC at time +1 The minimum SOC value for energy storage. This represents the maximum SOC value for energy storage. For energy storage z Daily starting SOC For the end-of-day tolerance of SOC, For energy storage z The daily SOC value.

20. An application system for an interconnected power grid energy storage planning model, characterized in that, include: The data input module is used to process the uncertainty of photovoltaic output or load demand based on historical datasets of photovoltaic power output or load demand, respectively, and obtain the probability density function of photovoltaic power output or load demand. The voltage and power flow calculation module is used to perform probabilistic power flow calculations using the cumulant method based on the probability density function of the photovoltaic output and the probability density function of the load demand, so as to obtain the probability density function of the node voltage and the probability density function of the branch power flow in the regional power grid. The location calculation module is used to input the probability density function of the node voltage and the probability density function of the branch power flow of the partitioned power grid into the upper planning model of the two-layer energy storage planning model, and solve for the installation location of the energy storage. The capacity strategy module is used to input the installation location of the energy storage into the lower planning model of the two-layer energy storage planning model, and solve it to obtain the rated capacity, rated power, and charging and discharging scheduling strategy of the energy storage, as well as the corresponding energy storage charging and discharging power curve. The iterative update module is used to return the energy storage charging and discharging power curve to the upper-level planning model, recalculate the comprehensive efficiency coefficient, and iteratively update the energy storage installation location, rated capacity, rated power, and charging and discharging scheduling strategy until the comprehensive efficiency coefficient no longer decreases, thus obtaining the optimal energy storage installation location, rated capacity, rated power, and charging and discharging scheduling strategy.

21. An electronic device, characterized in that, include: At least one processor and memory; The memory and processor are connected via a bus; The memory is used to store one or more programs; When the one or more programs are executed by the at least one processor, they implement the method for constructing an interconnected power grid energy storage planning model as described in any one of claims 1 to 9 or the method for applying an interconnected power grid energy storage planning model as described in claim 10.

22. A computer-readable storage medium, characterized in that, It contains an execution program, which, when executed, implements a method for constructing an interconnected power grid energy storage planning model as described in any one of claims 1 to 9, or a method for applying an interconnected power grid energy storage planning model as described in claim 10.

Images