A high proportion photovoltaic substation solid state transformer capacity configuration method and system

Abstract

The application discloses a high-proportion photovoltaic substation solid-state transformer capacity configuration method, which comprises the following steps: based on historical photovoltaic output data and load power data of a target substation, a reverse power sample set is simulated and generated; based on the reverse power sample set, an over-limit probability of reverse power exceeding a safety limit value and a comprehensive risk index are calculated, wherein the comprehensive risk index is an index reflecting the severity of reverse power over-limit; whether the target substation needs to be configured with a solid-state transformer is judged based on the over-limit probability and the comprehensive risk index, and if yes, the minimum rated capacity of the solid-state transformer configured for the target substation is calculated based on the reverse power sample set. The application provides a quantitative basis for the configuration of an SST device.

Description

Technical Field

[0001] This invention relates to the field of power distribution network operation analysis and solid-state transformer capacity configuration technology, specifically to a method and system for configuring solid-state transformer capacity in high-proportion photovoltaic power distribution areas. Background Technology

[0002] With the deepening implementation of the "dual carbon" goals, the installed capacity of distributed photovoltaic (PV) power has experienced explosive growth. Currently, some provinces have a large number of overvoltage distribution areas, including several severely overvoltaged areas and a certain number of areas with reverse overload. These areas are mainly concentrated in rural and suburban areas with high PV development intensity. During peak PV power generation periods at midday, power is fed back to the main transformer in the reverse direction, causing voltage overload and overheating of 10kV lines and distribution areas, seriously affecting equipment lifespan and grid safety. As PV penetration further increases, the problems of reverse overload and voltage overload will shift from sporadic occurrences to localized, widespread issues. Without effective quantitative analysis and remediation measures, this could lead to large-scale power outages and significant economic losses.

[0003] The high proportion of photovoltaic (PV) grid connection leads to frequent voltage exceedances and reverse overloads in distribution areas. Traditional solutions (replacing with larger capacity transformers, installing on-load tap changers, etc.) are characterized by high investment costs and slow response times. Solid-state transformers (SSTs), with their integrated functions such as rapid voltage regulation, bidirectional power flow control, and harmonic suppression, have become an ideal solution to these problems. However, SST equipment is expensive (approximately 2-3 times that of traditional transformers), and its capacity configuration lacks quantitative basis. While existing probabilistic power flow methods can provide reverse power distribution, they do not address providing a direct decision-making scheme for SST selection based on the reverse power distribution. Summary of the Invention

[0004] The purpose of this invention is to overcome the problem that increasing the installed capacity of distributed photovoltaic power generation can easily cause reverse overload in the distribution area. It provides a method and system for configuring the capacity of solid-state transformers in photovoltaic distribution areas with a high proportion of capacity, so as to select equipment for managing voltage over-limit and reverse heavy overload in large-scale grid connection scenarios.

[0005] To achieve the above objectives, the technical solution of the present invention is:

[0006] In a first aspect, the present invention provides a method for configuring the capacity of solid-state transformers in high-proportion photovoltaic power distribution areas, comprising:

[0007] The method includes:

[0008] S1. Based on the historical photovoltaic power output data and load power data of the target area, simulate and generate a reverse power sample set;

[0009] S2. Based on the aforementioned reverse power sample set, calculate whether the reverse power exceeds the safety limit. Overlimit probability , and a comprehensive risk index R, wherein the comprehensive risk index R is an index reflecting the severity of reverse power over-limit;

[0010] S3, based on the aforementioned over-limit probability The comprehensive risk index R determines whether the target area needs to be equipped with a solid-state transformer. If so, proceed to step S4.

[0011] S4. Based on the reverse power sample set, calculate the minimum rated capacity of the solid-state transformer configured in the target area.

[0012] The reverse power sample set is generated by simulating historical photovoltaic output data and load power data of the target area, specifically including:

[0013] Obtain historical photovoltaic power output data and load power data for the target area; construct a probability distribution model of photovoltaic power output based on the photovoltaic power output data; and construct a probability distribution model of load power based on the load power data.

[0014] Monte Carlo simulation is used to generate reverse power samples based on the probability distribution model of photovoltaic power output and the probability distribution model of load power, thus obtaining a reverse power sample set.

[0015] The method employs Monte Carlo simulation, sampling is performed based on the probability distribution model of photovoltaic output and the probability distribution model of load power to generate reverse power samples, resulting in a reverse power sample set, specifically including:

[0016] Perform N Monte Carlo simulations, each simulation yielding a reverse power sample, resulting in a sample set containing N reverse power samples. Each simulation includes the following steps:

[0017] Choose the Copula function and estimate its correlation parameters based on the historical correlation between photovoltaic power output and load power. ;

[0018] Based on the Copula function and correlation parameters Generate two independent standard normal random variables , ,structure , ,make , ,in, It is the standard normal cumulative distribution function;

[0019] based on A photovoltaic power output sample value is obtained through inverse transformation. A load power sample value :

[0020] In the formula, Cumulative distribution function of photovoltaic power output inverse function, This is obtained based on the probability distribution model of photovoltaic power output;

[0021] In the formula, The cumulative distribution function of load power The inverse function, the This is obtained based on the probability distribution model of load power;

[0022] Calculate reverse power samples , .

[0023] The construction of the photovoltaic output distribution probability model based on the photovoltaic output data specifically includes:

[0024] Construct the Beta distribution probability density function of photovoltaic power output ;

[0025] ;

[0026] In the formula, x represents the normalized photovoltaic output. and These are the shape parameters of the Beta distribution probability model; For gamma function, , ;

[0027] Construct the maximum likelihood estimation equations for the Beta distribution parameters:

[0028] ;

[0029] In the formula, Let be the value of the i-th normalized photovoltaic power output sample, which is obtained based on historical photovoltaic power output data; n is the number of photovoltaic power output samples. It is the digamma function;

[0030] The maximum likelihood estimation equations for the Beta distribution parameters are solved using the Newton-Raphson method to obtain the shape parameters. and ;

[0031] The construction of the load power distribution probability model based on the load power data specifically includes:

[0032] Construct the normal distribution probability density function of load power ;

[0033] ;

[0034] ;

[0035] ;

[0036] In the formula, For load power; Let be the value of the i-th load power sample, which is obtained based on historical load power data; m is the number of load power samples. This represents the average load power. This represents the standard deviation of the load power.

[0037] The basis of the overlimit probability The comprehensive risk index R determines whether a target distribution area needs to be equipped with a solid-state transformer, specifically including:

[0038] Compare the above probability Compared with the preset first risk threshold And the comprehensive risk index R and the preset second risk threshold :

[0039] when ,or and It is determined at this time that a solid-state transformer needs to be configured in the target area;

[0040] The probability of exceeding the limit Calculate using the following formula:

[0041] ;

[0042] In the formula, This is the i-th reverse power sample; This is the indicator function, which takes the value 1 when the condition in parentheses is true, and takes the value 0 otherwise; N is the total number of reverse power samples;

[0043] The comprehensive risk index R is calculated according to the following formula:

[0044] ;

[0045] In the formula, This is the i-th reverse power sample; This is the expected duration of the reverse overload.

[0046] The calculation of the minimum rated capacity of the solid-state transformers configured in the target distribution area based on the reverse power sample set specifically includes:

[0047] Based on the reverse power quantile at a given confidence level Combined with the given overload margin factor Determine the minimum rated capacity of a solid-state transformer , ;

[0048] The inverse power quantile at a given confidence level Calculate using the following formula:

[0049] ;

[0050] In the formula, It is a reverse power variable; is a probability operator; p is a given confidence level.

[0051] The method further includes:

[0052] Based on the photovoltaic penetration rate and type of the transformer substation, the multiple transformer substations are divided into different levels. A representative transformer substation is selected in each level. According to steps S1 to S4, it is calculated whether the representative transformer substation needs to be configured with a solid-state transformer, and the minimum rated capacity of the solid-state transformer that needs to be configured.

[0053] For any non-representative transformer area, if the representative transformer area at its level requires a solid-state transformer, then it is determined that the non-representative transformer area requires a solid-state transformer, and the reverse power quantile of the corresponding representative transformer area is linearly scaled according to the ratio of the existing transformer capacity of the non-representative transformer area to that of the corresponding representative transformer area. Obtain the reverse power quantile of the non-representative station area And then according to the above Calculate the minimum rated capacity of the solid-state transformer required for this non-representative distribution area. , .

[0054] Step S4 further includes:

[0055] Based on historical reverse power data for the target distribution area, assess the thermal capacity requirements of the solid-state transformers configured for the target distribution area:

[0056] The historical reverse power is divided into at least three discrete states, including the normal state. Mild overload state and heavy overload conditions This forms the state space of the Markov chain. ;

[0057] Based on historical reverse power data, the frequency of various state transitions between adjacent time points is counted to construct the state transition probability matrix P of the Markov chain.

[0058] by state As the absorption state, based on the state transition probability matrix P, the Monte Carlo method is used to simulate the transition from... or From departure to first entry The number of transfer steps is used to obtain the duration of continuous reverse overload. The probability distribution;

[0059] According to the duration The probability distribution is used to obtain the duration at a given confidence level. quantiles and select the target area satisfy solid-state transformers This represents the thermal time constant of the solid-state transformer power module.

[0060] Historical photovoltaic power output and load data are all from the midday period.

[0061] The midday period includes 11:00-14:00 daily.

[0062] Secondly, the present invention provides a high-proportion photovoltaic area solid-state transformer capacity configuration system, comprising:

[0063] The simulation module is used to simulate and generate a reverse power sample set based on the historical photovoltaic output data and load power data of the target area.

[0064] The indicator calculation module is used to calculate whether the reverse power exceeds the safety limit based on the reverse power sample set. Overlimit probability , and a comprehensive risk index R, wherein the comprehensive risk index R is an index reflecting the severity of reverse power over-limit;

[0065] The judgment module is used to determine the probability of exceeding the limit. The comprehensive risk index R determines whether the target area needs to be equipped with a solid-state transformer. If so, the capacity calculation module is activated.

[0066] The capacity calculation module is used to calculate the minimum rated capacity of the solid-state transformers configured in the target area based on the reverse power sample set.

[0067] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0068] 1. In the high-proportion photovoltaic (PV) power distribution area solid-state transformer capacity configuration method of this invention, a reverse power sample set is generated based on the historical PV output data and load power data of the target power distribution area. Key indicators reflecting the risk and severity of reverse power overload are calculated based on the reverse power sample set. Based on these key indicators, it is determined whether the target power distribution area needs to be equipped with a solid-state transformer. If so, the minimum capacity of the solid-state transformer is calculated. The rapid voltage regulation and bidirectional power flow control functions of the solid-state transformer effectively suppress the problem of reverse heavy overload in the power distribution area. Therefore, this invention calculates key indicators based on the reverse power sample set, and then determines whether the target power distribution area needs to be equipped with a solid-state transformer based on these key indicators, effectively suppressing the problem of reverse heavy overload in the power distribution area by utilizing the characteristics of the solid-state transformer.

[0069] 2. In this invention's method for configuring the capacity of solid-state transformers (SSTs) in high-proportion photovoltaic (PV) distribution areas, a Beta distribution model of PV output and a normal distribution model of load power are established based on historical data. Combined with Monte Carlo simulation, samples of reverse power are indirectly obtained. This method, even with limited data, can flexibly extrapolate reverse power under various extreme scenarios through a parametric model, obtaining a smooth probability density function and expanding the data. Based on this, sensitivity analysis is performed, and probability indicators such as quantiles and out-of-limit probabilities are output, avoiding data distortion caused by limited sampling data that cannot cover all possible weather and load combinations. Therefore, this invention indirectly obtains samples of reverse power through simulation, fully utilizing historical data to fit parameters and flexibly extrapolating scenarios through parameter adjustments to obtain a smooth probability density function, providing a more accurate quantitative basis for SST capacity configuration.

[0070] 3. In the high-proportion photovoltaic (PV) area solid-state transformer capacity configuration method of this invention, during the process of obtaining a sample of reverse power through Monte Carlo simulation and sampling calculation, source-load correlation is considered. A Copula function is selected, and correlation parameters are set according to the type of the selected Copula function to characterize the relationship between PV output and load power, so that the sample set of reverse power obtained by sampling can contain the correlation information between source and load. Therefore, this invention adopts Copula joint sampling based on source-load correlation, embedding Copula correlation into the reverse power generation process, so that the sample set of reverse power obtained by sampling can contain the correlation information between source and load, thereby accurately characterizing the amplification effect of negative source-load correlation on extreme reverse power.

[0071] 4. In the high-proportion photovoltaic (PV) transformer capacity configuration method of this invention, the PV output is affected by the light intensity and exhibits a typical skewed distribution: output is concentrated in the high-value area (right-skewed) under sunny weather, concentrated in the low-value area (left-skewed) under cloudy weather, and bell-shaped with high output in the middle and low output at both ends under partly cloudy weather. The shape parameters α and β of the Beta distribution are flexible; therefore, a probabilistic model of PV output is constructed based on the Beta distribution. Simultaneously, a maximum likelihood estimation equation set for the Beta distribution parameters is constructed, and then the Newton-Raphson method is used to solve the maximum likelihood estimation equation set for the Beta distribution parameters to obtain the shape parameters. By using maximum likelihood estimation combined with Newton iteration to refine the parameters of the PV Beta distribution, the model accuracy is improved. Therefore, this invention uses maximum likelihood estimation combined with Newton iteration to refine the parameters of the PV Beta distribution, while the normal distribution of the load focuses on the low-valley fluctuations, thus improving the model accuracy.

[0072] 5. In the high-proportion photovoltaic (PV) transformer capacity configuration method of this invention, monitoring terminals are deployed in the target PV area to collect data samples during the midday period (11:00-14:00). The midday PV output peak and load trough are the only high-incidence periods for reverse overload and voltage exceeding limits. This eliminates invalid historical data from outside the midday period, which carries no reverse risk. While significantly reducing computational load and adapting to the needs of large-scale modeling across multiple PV areas, the source-load distribution fitting is more accurate, and the reverse risk characterization is more precise, leading to a more accurate mitigation solution design. Therefore, this invention focuses on data from high-risk periods, improving data fitting accuracy, increasing computational efficiency, and resulting in a more reasonable SST capacity configuration.

[0073] 6. In the method for configuring the capacity of solid-state transformers in a high-proportion photovoltaic power distribution area according to the present invention, after obtaining the reverse power sample set, the probability of the reverse power exceeding the safety limit and the comprehensive risk index are calculated based on the reverse power sample set. The comprehensive risk index reflects the severity of the reverse power exceeding the limit. Based on the probability of exceeding the limit and the comprehensive risk index, the risk of reverse power exceeding the limit in the distribution area can be accurately assessed from both the probability of risk and the degree of risk, thereby accurately determining whether the target distribution area needs to be configured with a solid-state transformer. Therefore, the present invention can accurately assess the risk of reverse power exceeding the limit in the distribution area from both the probability of risk and the degree of risk, thereby accurately determining whether the target distribution area needs to be configured with a solid-state transformer.

[0074] 7. In the method for configuring the capacity of solid-state transformers in a high-proportion photovoltaic area according to the present invention, the minimum rated capacity of the solid-state transformer is determined based on the reverse power quantile at a given confidence level and a given overload margin coefficient. Simultaneously, based on the maximum reverse power and a preset safety limit, the reverse power suppression capability and heat capacity requirements of the solid-state transformer are calculated, directly yielding the range of solid-state transformer configuration parameters and providing a quantitative decision-making basis for engineering management. Therefore, the present invention can directly calculate the range of solid-state transformer configuration parameters, providing a quantitative decision-making basis for engineering management and possessing engineering practicality.

[0075] 8. In the high-proportion photovoltaic power grid solid-state transformer capacity configuration method of this invention, based on the historical reverse power data of the target power grid, the historical reverse power is divided into three discrete states: normal state, slightly overloaded state, and heavily overloaded state, forming a Markov chain state space. The state transition frequency of adjacent time moments is statistically analyzed, and the transition probability matrix of the Markov chain is calculated. Based on the transition probability matrix, the step distribution from the slightly overloaded state or the heavily overloaded state to the first entry into the normal state is simulated to obtain the probability distribution of the duration of continuous reverse overload. Then, the quantile of the overload duration at a given confidence level is obtained. Based on this quantile, the thermal time constant of the solid-state transformer power module is determined, and the thermal capacity requirement of the solid-state transformer in the power grid is accurately assessed. Therefore, this invention establishes a Markov chain model based on the historical reverse power sequence to simulate the duration distribution of continuous reverse overload to assess the thermal capacity requirement of the solid-state transformer and provide a quantitative decision-making basis for the parameter configuration of the solid-state transformer. Attached Figure Description

[0076] Figure 1 This is a flowchart illustrating a method for configuring the capacity of solid-state transformers in high-proportion photovoltaic areas, as provided in an embodiment of the present invention.

[0077] Figure 2 This is a schematic diagram of the architecture of a high-proportion photovoltaic area provided in an embodiment of the present invention.

[0078] Figure 3 This is the normalized photovoltaic output histogram and Beta distribution fitting curve provided in the embodiments of the present invention.

[0079] Figure 4 This is the midday load distribution histogram and normal distribution fitting curve provided in the embodiments of the present invention.

[0080] Figure 5 This is the probability density estimation curve of the reverse power provided in the embodiments of the present invention.

[0081] Figure 6 This is a probability variation curve of reverse power exceeding the safety limit under different photovoltaic penetration rates provided in the embodiments of the present invention.

[0082] Figure 7 This is a comparison diagram of the reverse power probability density under the assumptions of source-load correlation and independence provided in the embodiments of the present invention.

[0083] Figure 8 This is a schematic diagram of a high-proportion photovoltaic area solid-state transformer capacity configuration system provided in an embodiment of the present invention. Detailed Implementation

[0084] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0085] In a power system, the conventional direction of current flow is from the grid to the local load; this is called forward power. When the locally generated electrical energy exceeds its own electricity demand, the excess electrical energy is fed back to the grid through the grid; this phenomenon is called reverse power, which is the power output from the local area to the grid.

[0086] In existing technologies, for transformer substations (transformer-powered areas) with a high proportion of photovoltaic (PV) grid connection, the configuration and selection of transformer capacity mainly follow traditional planning methods or calculation methods based on deterministic power flow. Traditional transformer capacity configuration primarily focuses on forward power flow. However, in high-proportion PV substations, due to the randomness and volatility of PV power generation, especially during midday when sunlight is abundant and user load is low, reverse power flow is prone to occur, resulting in reverse power, i.e., the power output from the substation to the grid. Solid-state transformers (SSTs) have integrated functions such as fast voltage regulation, bidirectional power flow control, and harmonic suppression. However, when applied to address voltage exceeding limits and reverse overload in substations, the selection and capacity configuration of SSTs lack quantitative basis.

[0087] To address the technical problems existing in the prior art, embodiments of the present invention provide a reverse power probability modeling method specifically for SST capacity configuration. This method determines whether a transformer area requires SST by using an over-limit probability threshold and comprehensive risk indicators, and directly provides the SST capacity limit range using reverse power quantiles and overload margins. Please refer to [link to relevant documentation]. Figure 1 , Figure 1 This is a flowchart illustrating a method for configuring the capacity of a high-proportion photovoltaic area solid-state transformer, as provided in an embodiment of the present invention. The method includes steps S1 to S4.

[0088] S1. Based on the historical photovoltaic power output data and load power data of the target area, simulate and generate a reverse power sample set.

[0089] To analyze reverse overload, data was primarily extracted during the midday period from 11:00 AM to 2:00 PM, a time when photovoltaic output is at its highest and load is relatively low, representing the peak reverse power period. After acquiring the data, preprocessing was performed, including removing outliers (such as photovoltaic active power output). Values ​​exceeding 120% of installed capacity or less than -5% of load active power (For negative values); missing values ​​are filled using cubic spline interpolation or linear interpolation.

[0090] In this embodiment, the step of simulating and generating a reverse power sample set based on the historical photovoltaic power output data and load power data of the target area includes the following steps in S101 to S102:

[0091] S101. Obtain historical photovoltaic power output data and load power data of the target area, construct a probability model of photovoltaic power output based on the photovoltaic power output data, and construct a probability model of load power based on the load power data;

[0092] In this embodiment, monitoring terminals are deployed in the target area to continuously collect historical operating data for at least one year at 5-15 minute intervals, including photovoltaic output. and load power We extracted samples from the midday period (11:00-14:00) and removed outliers.

[0093] The probabilistic model for photovoltaic output, constructed based on the photovoltaic output data, includes:

[0094] The collected photovoltaic output samples were normalized:

[0095] ;

[0096] In the formula, For the i-th normalized photovoltaic power output sample; For the i-th photovoltaic power output sample, This represents the maximum output power of the photovoltaic array. Typically, it refers to photovoltaic installed capacity, measured in kWp.

[0097] The normalized photovoltaic power output samples follow a Beta distribution, with the following probability density function:

[0098] ;

[0099] Its cumulative distribution function is:

[0100] ;

[0101] In the formula, A Beta distribution probability model for photovoltaic power output; Contribute to normalized photovoltaic power. Contribute to photovoltaic power This represents the maximum output power of the photovoltaic array. and These are the shape parameters of the Beta distribution probability model; For gamma function, , .when , The time function is U-shaped; when , When it forms a J-shape (leftward deviation); when , It forms an inverted J-shape (right-leaning); , It is bell-shaped.

[0102] Photovoltaic output is affected by light intensity. Its normalized value is located in the [0,1] interval and has a typical skewed distribution: under sunny weather, the output is concentrated in the high value area (right skewed), under cloudy and rainy weather, the output is concentrated in the low value area (left skewed), and under cloudy weather, it is bell-shaped with high output in the middle and low output at both ends.

[0103] Shape parameters of the probability distribution model and Solve using Maximum Likelihood Estimation (MLE).

[0104] For independent and identically distributed samples The log-likelihood function is:

[0105] ;

[0106] make: ;

[0107] right and Taking the partial derivatives and setting them to zero, we obtain the system of likelihood equations:

[0108] ;

[0109] In the formula, Let be the value of the i-th normalized photovoltaic power output sample, which is obtained based on historical photovoltaic power output data; n is the number of photovoltaic power output samples. Let be the digamma function. This system of equations has no analytical solution; however, the shape parameters can be obtained by iteratively solving it using the Newton-Raphson method. and .

[0110] In this embodiment, the Newton-Raphson iterative solution includes:

[0111] Estimate the initial value of the matrix:

[0112] ;

[0113] ;

[0114] In the formula, , These are the initial values ​​for the shape parameters. This represents the mean of the photovoltaic power output samples.

[0115] Calculate the digamma function and its derivative trigamma function .

[0116] Construct the Jacobian matrix J and the Hessian matrix H, and update the parameters:

[0117] ;

[0118] in: ;

[0119] Repeat the iteration until and .

[0120] In this embodiment, to improve computational efficiency, a pre-calculated parameter lookup table can be used when calculating for multiple transformer areas: the sample mean... and variance Discretize into grid points, and calculate the corresponding ( ) offline. , Then, online matching will be performed.

[0121] A probabilistic model of load power is constructed based on the load power data, specifically including:

[0122] Because the load power during the midday period (11:00-14:00) is influenced by the randomness of residential and commercial electricity consumption behavior, it exhibits symmetrical fluctuations around the mean. The normal distribution is a widely accepted approximation model in engineering. Calculate the sample mean. and sample standard deviation :

[0123] ;

[0124] ;

[0125] In the formula, For load power variables; Let be the value of the i-th load power sample, which is obtained based on historical load power data; m is the number of load power samples. This represents the average load power. This represents the standard deviation of the load power.

[0126] The normal distribution probability density function of the load power is:

[0127] ;

[0128] For cases where the load power might be negative (which rarely occurs in practice), a truncated normal distribution can be used, restricting the domain to [0, +∞). The truncated probability density function is:

[0129] ;

[0130] In the formula, The standard normal cumulative distribution function. When Larger, for example At that time, the truncation correction is ignored.

[0131] S102. Using Monte Carlo simulation, the photovoltaic power output distribution probability model and the load power distribution probability model are sampled to generate reverse power samples, thus obtaining a reverse power sample set.

[0132] For the reverse power distribution, directly using historical reverse power data for empirical distribution estimation has the following limitations:

[0133] (1) Historical data on photovoltaic output and load power usually only cover 1-2 years, which is insufficient and difficult to include extreme weather scenarios (such as continuous rain or sudden sunshine that occurs once every few years), resulting in insufficient estimation of tail risks.

[0134] (2) It is unable to adapt to future photovoltaic capacity expansion, load growth and other planning scenarios, and lacks extrapolation capabilities;

[0135] (3) It can only output discrete empirical probabilities, which is not conducive to calculating smooth quantiles and out-of-limit probabilities.

[0136] Based on the limitations of the direct sampling method described above, this embodiment of the invention first constructs a probabilistic model of photovoltaic power output and load power, and then uses Monte Carlo simulation to obtain the reverse power sample set. Monte Carlo simulation is the most direct and accurate method for obtaining the functional distribution of random variables. The number of simulations is usually no less than 100,000 to ensure that the relative error is less than 1% (according to the central limit theorem, the error is proportional to the load distribution). (Proportional)

[0137] In this embodiment, N Monte Carlo simulations are performed, and each simulation yields a reverse power sample, resulting in a sample set containing N reverse power samples. Specifically, the number of samplings N in the Monte Carlo simulation is ≥ 100,000, and GPU parallel acceleration is employed.

[0138] To achieve more accurate simulation results, Gaussian Copula is introduced to describe the source-load correlation for joint sampling.

[0139] Specifically, each Monte Carlo simulation includes the following steps:

[0140] In this embodiment, the Copula function is selected as Gaussian Copula.

[0141] Based on historical photovoltaic power output and load power, the correlation parameters of the Gaussian Copula function are estimated. Specifically, it can be obtained through Kendall's rank correlation coefficient conversion: Calculate the Kendall's rank correlation coefficient between photovoltaic power and load based on historical data. Using the formula Convert it to the correlation parameters of Gaussian Copula. This is used to characterize the relationship between photovoltaics and load.

[0142] Based on the Copula function and correlation parameters This generates a pair of correlated uniform variables. ;

[0143] Common methods for generating uniform variables include:

[0144] First, generate two independent standard normal random variables. , and construct , Their correlation coefficient is Then let , To obtain uniform variables ,in, It is the standard normal cumulative distribution function.

[0145] Based on uniform variables The cumulative distribution function of photovoltaic power output Cumulative distribution function of load power Perform inverse CDF transformation to obtain a photovoltaic power output sample value. A load power sample value :

[0146] In the formula, Cumulative distribution function of photovoltaic power output inverse function, This is obtained based on the probability distribution model of photovoltaic power output;

[0147] In the formula, The cumulative distribution function of load power inverse function, This is obtained based on the probability distribution model of load power;

[0148] Calculate reverse power , This yields a reverse power sample.

[0149] Repeat the above steps of simulating and generating reverse power samples N times to obtain a reverse power sample set containing correlation information between the source and the load. In the formula It is the Nth sample in the reverse power sample set.

[0150] S2. Based on the aforementioned reverse power sample set, calculate whether the reverse power exceeds the safety limit. Overlimit probability And a comprehensive risk index R, which is an index reflecting the severity of reverse power over-limit.

[0151] Probabilistic power flow (PPF) methods are frequently used in distribution network analysis. However, existing techniques are mostly geared towards transmission networks or distribution networks with a small number of distributed generation sources. For high-proportion photovoltaic (PV) areas (PV penetration exceeding 50%), the distribution characteristics of reverse power are lacking systematic research. Reverse power is a key variable for assessing reverse overload, significantly impacting SST selection and capacity configuration. While existing probabilistic power flow methods can provide reverse power distribution, they do not address providing direct decision-making schemes for SST selection based on this distribution, failing to offer guidance for selecting mitigation equipment to address voltage exceedances and reverse overloads in distribution areas. This invention proposes a method for probabilistic modeling using reverse power sets and directly outputs SST capacity configuration parameters.

[0152] Specifically, based on the aforementioned reverse power sample set, a key indicator is first calculated: the reverse power exceeds a preset safety limit. Overlimit probability Inverse power quantile at a given confidence level Comprehensive risk index R for the area.

[0153] The comprehensive risk index R is calculated according to the following formula:

[0154] ;

[0155] In other embodiments, the comprehensive risk index R can also be calculated according to the following equivalent formula:

[0156] ;

[0157] In the formula, N is the total number of reverse power samples; This is a reverse power sample; The preset safety limit; Let be the probability density function of the reverse power; The expected duration of the reverse overload, where, Take the historical average duration of exceeding the limit (e.g., 1 hour).

[0158] The probability of exceeding the limit is calculated according to the following formula:

[0159] ;

[0160] In the formula, This is the i-th reverse power sample; This is an indicator function; it takes the value 1 when the condition inside the parentheses is true, and 0 otherwise. , This represents the current rated capacity of the transformers in the distribution area.

[0161] The inverse power quantile at a given confidence level Calculate using the following formula:

[0162] ;

[0163] In the formula, is a probability operator; p is a given confidence level. In this embodiment, Take the 95th percentile, that is Given a confidence level of p=0.95.

[0164] Obtain the probability of exceeding the limit Reverse power quantile After considering the three key indicators—the comprehensive risk index R of the transformer area—it can be determined whether the corresponding transformer area needs to be configured with SST.

[0165] S3, based on the aforementioned over-limit probability The comprehensive risk index R determines whether the target area needs to be equipped with a solid-state transformer. If so, proceed to step S4.

[0166] Specifically, determine the probability of exceeding the limit. Is it greater than the preset first risk threshold? , =0.05, if Then the transformer area needs to be equipped with a solid-state transformer;

[0167] like Then, determine whether the comprehensive risk index R of the transformer area is greater than the preset second risk threshold R. th ,like Then the transformer area needs to be equipped with a solid-state transformer.

[0168] In this embodiment, the first risk threshold The value range is 0.03 to 0.05. In this embodiment, Set to 0.05; the second risk threshold R th The value is set at 0.05kW. This scheme adopts a dual-threshold judgment mechanism, which uses the over-limit probability to indicate the probability of over-limit occurrence and the comprehensive risk index to indicate the severity of over-limit occurrence. This dual judgment of probability and severity is used to determine whether the target area needs to be equipped with a solid-state transformer, avoiding the one-sidedness of a single indicator.

[0169] S4. Based on the reverse power sample set, calculate the minimum rated capacity and reverse power suppression capability of the solid-state transformers configured in the target area.

[0170] Specifically, based on the quantiles Determine the minimum rated capacity of a solid-state transformer by combining the overload margin factor k. , .

[0171] Due to certain limitations, such as transformer reverse load rate requirements, the reverse power suppression capability of the SST is not equal to its rated power. Therefore, based on preset safety limits... Calculate the lower limit of reverse power rejection capability of solid-state transformers , Take the difference between the maximum reverse power and the safety limit, that is:

[0172] ;

[0173] In the formula, This represents the maximum reverse power in the reverse power sample set.

[0174] At the same time, after calculation Based on this, and considering the energy storage and charging needs of the distribution area, the reverse power suppression capability of solid-state transformers can be further limited to no less than 200kW. When selecting solid-state transformers, both rated capacity and reverse power suppression capability must be met simultaneously. Rated capacity and reverse power suppression capability provide a quantitative basis for engineering procurement and technical specification formulation.

[0175] Finally, the voltage regulation range of the solid-state transformer was determined to be ±10%. Based on the national standard GB / T 12325 and the inverter's modulation capability, the voltage regulation range of the solid-state transformer was directly preset to ±10%.

[0176] After calculating the required capacity, reverse power suppression capability, and voltage regulation range of the solid-state transformer, a suitable solid-state transformer is selected based on the calculation results. This suitable solid-state transformer is then connected in parallel with the existing traditional transformer to jointly participate in the power supply operation of the distribution area.

[0177] When analyzing multiple transformer substations over a large area and calculating the configuration requirements of solid-state transformers, in order to reduce the amount of computation, the multiple transformer substations in the large area can be divided into different levels. A representative transformer substation is selected from each level, and the representative transformer substation is modeled to obtain the corresponding typical reverse power sample set model. After obtaining the typical parameters, the model is extended to other transformer substations in the same level by interpolation.

[0178] Specifically, based on the photovoltaic penetration rate and type of the distribution area, multiple distribution areas are divided into different levels. Distribution areas of the same type and with similar photovoltaic penetration rates are grouped into the same level. A representative distribution area is selected from each level, and steps S1 to S4 are used to calculate whether the representative distribution area needs to be equipped with a solid-state transformer, and the minimum rated capacity of the required solid-state transformer. For non-representative distribution areas, if the representative distribution area of ​​its level needs to be equipped with a solid-state transformer, then the non-representative distribution area is determined to need a solid-state transformer. Based on the ratio of the existing transformer capacity of the non-representative distribution area to that of the corresponding representative distribution area, the reverse power quantile of the corresponding representative distribution area is calculated. Perform linear scaling to obtain the inverse power quantile for this non-representative area. Then, based on the reverse power quantile of this non-representative area... Calculate the minimum rated capacity of the required solid-state transformer. ,

[0179] In some embodiments, to assess the demand of the SST heat capacity for the duration of continuous reverse overload, a Markov chain model can be established based on historical reverse power sequences.

[0180] Specifically, when a target distribution area requires the configuration of a solid-state transformer, a Markov chain model is established based on historical reverse power sequences to simulate the duration distribution of continuous reverse overloads in order to assess the thermal capacity requirements of the solid-state transformer. The thermal capacity of the solid-state transformer refers to the instantaneous thermal shock energy that its power modules can withstand during continuous reverse overloads without exceeding the safe junction temperature (typically 125℃~150℃).

[0181] The step of establishing a Markov chain model based on historical reverse power sequences to simulate the duration distribution of continuous reverse overloads includes the following steps S401 to S404:

[0182] S401. Divide the historical reverse power into at least three discrete states, including the normal state. Mild overload state and heavy overload conditions This forms the state space of the Markov chain. ;

[0183] S402. Based on historical reverse power data, count the frequency of various state transitions between adjacent time points to construct the state transition probability matrix P of the Markov chain.

[0184] S403, in state As the absorption state, based on the state transition probability matrix P, the Monte Carlo method is used to simulate the transition from... or From departure to first entry The number of transfer steps is used to obtain the duration of continuous reverse overload. The probability distribution;

[0185] When the system is overloaded or At that time, each discrete time step corresponds to an actual time unit, based on the period from entering the overload state to the first return to the normal state. The duration of the corresponding consecutive overload events can be calculated by counting the number of transition steps. Through multiple simulations using the Monte Carlo method, a sample set of overload durations can be obtained, leading to the probability distribution of the duration of consecutive reverse overload events.

[0186] S404, according to the duration The probability distribution is used to obtain the duration at a given confidence level. quantiles That is, at a set confidence level (0.95 in this embodiment), the critical threshold corresponding to the duration of reverse overload is used to select the target area. satisfy solid-state transformers The thermal time constant of the solid-state transformer power module. It is usually obtained from solid-state transformer datasheets or through engineering estimations.

[0187] In this embodiment, when configuring a solid-state transformer for the target distribution area, it is also necessary to consider calculating the required thermal capacity tolerance of the solid-state transformer, which serves as a constraint on the selection of power devices and the heat dissipation system for the solid-state transformer. In other embodiments, the average duration of continuous reverse overload obtained through multiple simulations using the Monte Carlo method can be used instead. .

[0188] Thermal time constant of solid-state transformer power modules for:

[0189] ;

[0190] In the formula, The heat capacity of the power module; This refers to the thermal conductivity of the power module. In practical engineering, the thermal time constant of the solid-state transformer can be estimated based on the module's package size and heat dissipation conditions, using thermal simulation tools (such as Icepak and Flotherm).

[0191] In other embodiments, the thermal time constant of the solid-state transformer can also be estimated by referring to typical industry values. For example, for single-tube discrete devices, the thermal time constant is generally 0.1-1 seconds; for medium-power modules (50kW-200kW level), the thermal time constant is generally 1-10 seconds; for high-power modules (above 200kW), the thermal time constant is generally 10-30 seconds. During the SST selection decision-making stage, a conservative estimate within this range (lower limit or median value) can be used to calculate the initial judgment of the SST thermal capacity requirement based on the target power level.

[0192] When selecting a SST for a target transformer area, the thermal time constant should be given priority. With overload duration Matching modules.

[0193] Traditional SST capacity configurations are based solely on instantaneous reverse power peak values ​​or probability quantiles, neglecting the duration of overload events. Prolonged, continuous overloads can lead to a cumulative increase in power module junction temperature, potentially exceeding the device's safe operating range. This invention uses a Markov chain to simulate the duration distribution, enabling the SST's thermal design (heatsink size, fan power, IGBT derating) to match the actual overload duration under real-world conditions. This avoids over-design due to short-term peak values ​​or thermal failure due to sustained overloads.

[0194] Taking a high-proportion photovoltaic distribution area as an example, this area is a typical rural distribution area with a rated capacity of 400kVA for the distribution transformer. The distribution area structure is as follows: Figure 2 As shown, the total installed capacity of distributed photovoltaic power in this distribution area will be 350 kWp in 2025 (photovoltaic penetration rate of 87.5%). This distribution area is a typical example of a severely overvoltage distribution area in a certain province.

[0195] Obtain historical operation data for a high proportion of photovoltaic power stations:

[0196] A monitoring terminal (model: iMeter 8850, accuracy class 0.5) was installed on the low-voltage side of the distribution transformer in this area. Data was collected at 5-minute intervals from January 1, 2025 to December 31, 2025, for a total of 105,120 samples. Data from the midday period (11:00-14:00) was extracted, totaling 65,700 samples. Obvious outliers (such as photovoltaic output greater than 350kW or less than 0, load power less than 0) were removed, retaining 64,892 valid samples. The mean of the midday load data was 75.2kW, and the standard deviation was 19.8kW. =75kW, =20kW is used for modeling.

[0197] Establish a probability model for photovoltaic power output distribution:

[0198] Normalize photovoltaic power output data: ;

[0199] Calculate the sample mean Sample variance .

[0200] Initial value for moment estimation:

[0201] ;

[0202] ;

[0203] Using the Newton-Raphson iteration, the results converged after 5 iterations. , .

[0204] like Figure 3 As shown, the histogram of the normalized measured photovoltaic output data for the distribution area is in high agreement with the Beta distribution fitting curve. The Kolmogorov-Smirnov test p-value is 0.23, which does not reject the hypothesis that the distribution follows a Beta distribution.

[0205] Establish a load power distribution probability model:

[0206] Calculate the sample statistics of load power during midday:

[0207] ;

[0208] The histogram of measured load distribution at midday and the fitting curve of the probability density normal distribution are shown below. Figure 4 As shown, the tail of the normal probability plot deviates slightly, which is acceptable in engineering.

[0209] Monte Carlo simulation:

[0210] Ignoring source-load correlation and using independent sampling, the Monte Carlo simulation number N=10000 is set. The Monte Carlo simulation is implemented in MATLAB to obtain the reverse power sample set. The code is as follows:

[0211] N=100000;

[0212] Ppv_max=350;

[0213] alpha=1.55; beta=1.20;

[0214] % =75kW; =20kW

[0215] mu_L=75; sigma_L=20;

[0216] Generate Beta-distributed photovoltaic output samples (using Gamma distribution conversion)

[0217] x = betarnd(alpha, beta, N, 1);

[0218] ;

[0219] % Generate normally distributed load samples

[0220] PL = normrnd(mu_L, sigma_L, N, 1);

[0221] % Calculate reverse power

[0222] Prev = Ppv - PL;

[0223] Simulation results: Mean reverse power (expected value) ; standard deviation of reverse power 95th percentile of reverse power Reverse power exceeds safety limit The number of times was 6500. Probability of exceeding limits .

[0224] Considering the source-load correlation, in order to more accurately simulate the reverse power sample data, Gaussian Copula modeling was adopted. After introducing the correlation, Monte Carlo simulation was performed again, and the probability of reverse power exceeding 320kW was obtained as 7.0%, which is 0.5 percentage points higher than the independent assumption of 6.5%. Figure 7 To compare the reverse power probability density under the assumptions of source-load correlation (using Gaussian Copula) and independence, from... Figure 7 and Figure 5The comparison shows that the correlation slightly increases the probability of the right tail, indicating that ignoring the correlation would slightly underestimate the risk, but the difference is within the allowable range of engineering (Δ=0.5 percentage points).

[0225] Figure 7 The correlation parameters are shown. Impact on tail probability. For this transformer area data, the Pearson correlation coefficient between photovoltaic output and load was calculated as follows: =0.23, indicating a weak negative correlation. A negative correlation will increase the output of photovoltaic power. and load power The difference between the two: when big, Hour, Larger; when Small, When it is big, Smaller (or more negative). Therefore, a negative correlation will make the right tail of the reverse power thicker (increasing the probability of extreme high values), thereby increasing the risk of reverse heavy overload (increasing the probability of exceeding limits).

[0226] Reverse power The kernel density estimation curve is as follows Figure 5 As shown, Figure 5 The right tail corresponding to the shaded area is longer, which corresponds to the probability of exceeding the limit. In extreme cases, the photovoltaic system is fully loaded ( And the load is extremely low. When ), reverse power .

[0227] Statistical output and risk assessment:

[0228] Therefore, it is determined that SST needs to be configured; Considering energy storage and N-1, the value is rounded to 500kVA. Reverse power suppression capability: 340-32-=20kW, but to absorb fluctuations, it is taken as 200kW. Voltage regulation range: ±10%.

[0229] Depend on Figure 5 It is known that the probability of reverse power exceeding 320kW is 3.2%, meaning that during the midday period, this transformer area is in a reverse heavy overload state for approximately 6.5% of the time each year (approximately 0.065 × 3 hours × 365 days ≈ 71 hours). The parameters used in this embodiment ( =75kW, Sensitivity calculations were performed for a 20kW photovoltaic (PV) installation at different penetration rates: with a fixed load distribution, the PV installed capacity was varied (i.e., the penetration rate was changed), and the probability of exceeding the limit was simulated. The results are as follows: at a penetration rate of 50% (200kWp), the probability of exceeding the limit is approximately 1.2%; at a penetration rate of 65% (260kWp), the probability of exceeding the limit is approximately 5.0% (critical penetration rate); at a penetration rate of 80% (320kWp), the probability of exceeding the limit is approximately 5.8%; at a penetration rate of 87.5%, the probability of exceeding the limit is approximately 6.5%; and at a penetration rate of 100%, the probability of exceeding the limit is 8.2%.

[0230] Figure 6 For different photovoltaic penetration rates, the reverse power exceeds the safety limit ( The probability change curve of (320kW) shows that the risk increases significantly when the penetration rate exceeds 65%.

[0231] The 95th percentile of reverse power is 272kW, meaning that in this type of distribution area, the reverse power is conservatively estimated to be no more than 272kW 95% of the time. If the SST capacity is designed according to this, considering a 1.2 times margin, the recommended capacity is 272 × 1.2 ≈ 326kVA. However, considering future PV capacity expansion and load growth, as well as N-1 redundancy requirements and energy storage charging needs (approximately 100kW), a capacity of 500kVA is actually chosen.

[0232] The reverse power suppression capability refers to the reverse power value that a solid-state transformer can actively limit or absorb when a reverse overload occurs during the midday peak photovoltaic power generation period. Its core function is to transform the "maximum reverse power" risk index obtained through Monte Carlo simulation and probabilistic modeling into a performance constraint that can be directly applied when selecting equipment.

[0233] The theoretical lower limit of reverse power rejection capability is , This formula guarantees that SST can theoretically cover reverse power exceeding the safety limit in extreme cases. (i.e., 80% of the transformer's rated capacity).

[0234] The theoretical lower limit of reverse power suppression capability Based on this, and considering the charging power requirements of the energy storage system in the distribution area (e.g., 100kW), the short-term fluctuation margin of photovoltaic output, and future load growth, a conservative engineering recommendation value (200kW in this embodiment) is selected. This value serves as a mandatory requirement that must be met in the SST procurement and technical specifications.

[0235] Reverse power rejection capability works in conjunction with rated capacity; SST minimum rated capacity This solves the problem of "how much power capacity is needed", while also addressing the lower limit of reverse power rejection capability. This addresses the question of how much power needs to be handled during reverse overload. After determining the minimum rated capacity and reverse power suppression capability, the duration of continuous overload is evaluated using a Markov chain model, and the thermal time constant of the SST power module is verified to meet requirements. If the suppression capability is large but the duration is too long, it is still necessary to increase the heat sink size or select a power module with higher heat capacity to ensure equipment safety. Therefore, suppression capability, rated capacity, and heat capacity complement each other, jointly guiding the final selection of the SST.

[0236] Based on the modeling results of this invention, a 500kVA power electronic transformer (SST) was configured for this high-proportion photovoltaic area, while the original 400kVA conventional transformer was retained as a backup. SST parameters: rated capacity 500kVA, reverse power suppression capability 200kW, voltage regulation range ±10%. After the treatment, it was put back into operation in January 2026. Monitoring data after three months of operation showed that the midday reverse power peak decreased from 340kW to 185kW (a reduction of 45.6%); the voltage compliance rate increased from 91.8% to 99.5%; the highest transformer winding temperature decreased from 118℃ to 96℃; and no further protection actions due to overvoltage or overload occurred. This verifies that the probabilistic modeling method of this invention provides an accurate and quantitative basis for the treatment plan.

[0237] like Figure 2 As shown, this high-proportion photovoltaic (PV) distribution area is equipped with a power electronic transformer (SST) on the basis of the original traditional transformer power distribution. The SST is connected in parallel with the original traditional transformer to transform the 10kV medium-voltage AC voltage into a 380V low-voltage voltage, thereby supplying power to the load of the distribution area. The distribution area includes PV power generation devices, which deliver PV energy to the 380VAcBus. Through the high-proportion PV distribution area solid-state transformer capacity configuration method described in this embodiment of the invention, reverse heavy overload and overvoltage assessment of the distribution area can be achieved.

[0238] For hundreds of thousands of transformer substations in a certain region, performing complete parameter fitting and Monte Carlo simulation for each one would be computationally intensive. This invention proposes a stratified sampling + clustering generalization strategy based on large-scale application:

[0239] Based on photovoltaic penetration rate (0-20%, 20-40%, 40-60%, 60-80%, >80%) and transformer substation type (including rural, suburban, industrial, and commercial substations), transformer substations are divided into 20 typical categories. For each typical category, 3-5 representative substations are selected for refined modeling to obtain the "typical parameters" (shape parameters of the photovoltaic treatment Beta distribution probability model) for that category. and and the mean of the load power normal distribution probability model and standard deviation For other distribution areas of the same category, typical parameters are used for direct calculation, or fine-tuned through linear interpolation (based on photovoltaic installed capacity). For distribution areas of the same category with different transformer capacities, the reverse power quantile is calculated based on the transformer capacity ratio. Perform linear scaling and then substitute the values ​​into the capacity calculation formula. For example, a reference station area capacity of 400kVA corresponds to... For a transformer substation with a capacity of 630kVA, its .

[0240] Meanwhile, for non-representative transformer substations at different levels, parameter lookup tables can be pre-calculated for representative substations. The probabilistic model parameters for photovoltaic output and load power of non-representative substations can then be obtained from these tables. During application, only the input parameters need to be looked up in the table, thus reducing substation modeling time and meeting the needs of real-time assessment. This invention is applicable to bidirectional power flow modeling of similar substations nationwide, thereby assisting in the analysis of substation over-limit management.

[0241] In summary, this invention employs photovoltaic Beta distribution and load normal distribution, combined with an indirect method of Monte Carlo simulation considering source-load correlation, to obtain a sample set of reverse power. This method, even with limited data, can flexibly extrapolate data through a parametric model, perform sensitivity analysis, and output indicators required for decision-making, such as quantiles and overload probability. Furthermore, based on the aforementioned indicators, the invention introduces the calculation of a risk indicator R. The risk indicator R quantifies the severity, probability, and duration of overload, enabling a more comprehensive assessment of the risks of voltage exceeding limits and reverse overload in the transformer area. This invention also determines whether a transformer area needs to be configured with a SST (Sustainable Storage and Reverse) device based on the risk indicator and overload probability. If so, it directly maps the indicators output from the aforementioned steps to the SST capacity configuration, directly outputting the range of SST capacity and suppression capability. This fills a gap in existing technologies, provides a quantitative decision-making basis for engineering management, and has extremely high engineering practicality.

[0242] Based on the methods described in the above embodiments, this embodiment will further describe a high-proportion photovoltaic area solid-state transformer capacity configuration system. This system can be implemented as a standalone entity or integrated into electronic devices, such as terminals. Please refer to... Figure 8 The system includes:

[0243] The simulation module is used to simulate and generate a reverse power sample set based on the historical photovoltaic power output data and load power data of the target area; the simulation module is used to execute step S1.

[0244] The indicator calculation module is used to calculate whether the reverse power exceeds the safety limit based on the reverse power sample set. Overlimit probability And a comprehensive risk index R, wherein the comprehensive risk index R is an index reflecting the severity of reverse power over-limit; the index calculation module is used to execute step S2.

[0245] The judgment module is used to determine the probability of exceeding the limit. The comprehensive risk index R determines whether the target area needs to be equipped with a solid-state transformer. If so, the capacity calculation module is activated. The determination module is used to execute step S3.

[0246] The capacity calculation module is used to calculate the minimum rated capacity of the solid-state transformers configured in the target area based on the reverse power sample set; the capacity calculation module is used to execute step S4.

[0247] Although embodiments of the present invention have been shown and described above, it should be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for configuring the capacity of solid-state transformers in high-proportion photovoltaic power distribution areas, characterized in that: The method includes: S1. Based on the historical photovoltaic power output data and load power data of the target area, simulate and generate a reverse power sample set; S2. Based on the aforementioned reverse power sample set, calculate whether the reverse power exceeds the safety limit. Overlimit probability , and a comprehensive risk index R, wherein the comprehensive risk index R is an index reflecting the severity of reverse power over-limit; The comprehensive risk index R is calculated according to the following formula: ； In the formula, This is the i-th reverse power sample; The expected duration of reverse overload; S3, based on the aforementioned over-limit probability The comprehensive risk index R determines whether the target area needs to be equipped with a solid-state transformer. If so, proceed to step S4. S4. Based on the reverse power sample set, calculate the minimum rated capacity of the solid-state transformers configured in the target area. The calculation of the minimum rated capacity of the solid-state transformers configured in the target distribution area based on the reverse power sample set specifically includes: Based on the reverse power quantile at a given confidence level Combined with the given overload margin factor Determine the minimum rated capacity of a solid-state transformer , .

2. The method for configuring the capacity of a high-proportion photovoltaic power station solid-state transformer according to claim 1, characterized in that: The reverse power sample set is generated by simulating historical photovoltaic output data and load power data of the target area, specifically including: Obtain historical photovoltaic power output data and load power data for the target area; construct a probability distribution model of photovoltaic power output based on the photovoltaic power output data; and construct a probability distribution model of load power based on the load power data. Monte Carlo simulation is used to generate reverse power samples based on the probability distribution model of photovoltaic power output and the probability distribution model of load power, thus obtaining a reverse power sample set.

3. The method for configuring the capacity of a high-proportion photovoltaic transformer in a distribution area according to claim 2, characterized in that: The method employs Monte Carlo simulation, sampling is performed based on the probability distribution model of photovoltaic output and the probability distribution model of load power to generate reverse power samples, resulting in a reverse power sample set, specifically including: Perform N Monte Carlo simulations, each simulation yielding a reverse power sample, resulting in a sample set containing N reverse power samples. Each simulation includes the following steps: Choose the Copula function and estimate its correlation parameters based on the historical correlation between photovoltaic power output and load power. ; Based on the Copula function, generate two independent standard normal random variables. , ,structure , ,make , ,in, It is the standard normal cumulative distribution function; based on A photovoltaic power output sample value is obtained through inverse transformation. A load power sample value : In the formula, Cumulative distribution function of photovoltaic power output inverse function, This is obtained based on the probability distribution model of photovoltaic power output; In the formula, The cumulative distribution function of load power inverse function, This is obtained based on the probability distribution model of load power; Calculate reverse power samples , .

4. The method for configuring the capacity of a high-proportion photovoltaic power station solid-state transformer according to claim 2, characterized in that: The construction of the photovoltaic output distribution probability model based on the photovoltaic output data specifically includes: Construct the Beta distribution probability density function of photovoltaic power output ; ； In the formula, x represents the normalized photovoltaic output. and These are the shape parameters of the Beta distribution probability model; For gamma function, , ; Construct the maximum likelihood estimation equations for the Beta distribution parameters: ； In the formula, Let be the value of the i-th normalized photovoltaic power output sample, which is obtained based on historical photovoltaic power output data; n is the number of photovoltaic power output samples. It is the digamma function; The maximum likelihood estimation equations for the Beta distribution parameters are solved using the Newton-Raphson method to obtain the shape parameters. and ; The construction of the load power distribution probability model based on the load power data specifically includes: Construct the normal distribution probability density function of load power ; ； ； ； In the formula, For load power; Let be the value of the i-th load power sample, which is obtained based on historical load power data; m is the number of load power samples. This represents the average load power. This represents the standard deviation of the load power.

5. The method for configuring the capacity of a high-proportion photovoltaic power station solid-state transformer according to claim 1, characterized in that: The basis of the overlimit probability The comprehensive risk index R determines whether a target distribution area needs to be equipped with a solid-state transformer, specifically including: Compare the above probability Compared with the preset first risk threshold And the comprehensive risk index R and the preset second risk threshold : when ,or and It is determined at this time that a solid-state transformer needs to be configured in the target area; The probability of exceeding the limit Calculate using the following formula: ； In the formula, This is the i-th reverse power sample; This is the indicator function, which takes the value 1 when the condition in parentheses is true, and takes the value 0 otherwise; N is the total number of reverse power samples.

6. The method for configuring the capacity of a high-proportion photovoltaic transformer in a distribution area according to claim 5, characterized in that: The inverse power quantile at a given confidence level Calculate using the following formula: ； In the formula, It is a reverse power variable; is a probability operator; p is the given confidence level.

7. The method for configuring the capacity of a high-proportion photovoltaic transformer in a distribution area according to claim 6, characterized in that: The method further includes: Based on the photovoltaic penetration rate and type of the transformer substation, the multiple transformer substations are divided into different levels. A representative transformer substation is selected in each level. According to steps S1 to S4, it is calculated whether the representative transformer substation needs to be configured with a solid-state transformer, and the minimum rated capacity of the solid-state transformer that needs to be configured. For any non-representative transformer area, if the representative transformer area at its level requires a solid-state transformer, then it is determined that the non-representative transformer area requires a solid-state transformer, and the reverse power quantile of the corresponding representative transformer area is linearly scaled according to the ratio of the existing transformer capacity of the non-representative transformer area to that of the corresponding representative transformer area. Obtain the reverse power quantile of the non-representative station area And then according to Calculate the minimum rated capacity of the solid-state transformer required for this non-representative distribution area. , .

8. The method for configuring the capacity of a high-proportion photovoltaic power station solid-state transformer according to claim 1, characterized in that: Step S4 further includes: Based on historical reverse power data for the target distribution area, assess the thermal capacity requirements of the solid-state transformers configured for the target distribution area: The historical reverse power is divided into at least three discrete states, including the normal state. Mild overload state and heavy overload conditions This forms the state space of the Markov chain. ; Based on historical reverse power data, the frequency of various state transitions between adjacent time points is counted to construct the state transition probability matrix P of the Markov chain. by state As the absorption state, based on the state transition probability matrix P, the Monte Carlo method is used to simulate the transition from... or From departure to first entry The number of transfer steps is used to obtain the duration of continuous reverse overload. The probability distribution; According to the duration The probability distribution is used to obtain the duration at a given confidence level. quantiles and select the target area satisfy solid-state transformers This represents the thermal time constant of the solid-state transformer power module.

9. The method for configuring the capacity of a high-proportion photovoltaic transformer in a distribution area according to claim 1, characterized in that: Historical photovoltaic power output and load data are all from midday hours. The midday period includes 11:00-14:00 daily.

10. A high-proportion photovoltaic distribution area solid-state transformer capacity configuration system, characterized in that, The system is used to execute the high-proportion photovoltaic area solid-state transformer capacity configuration method as described in any one of claims 1 to 9, specifically including: The simulation module is used to simulate and generate a reverse power sample set based on the historical photovoltaic output data and load power data of the target area. The indicator calculation module is used to calculate whether the reverse power exceeds the safety limit based on the reverse power sample set. Overlimit probability , and a comprehensive risk index R, wherein the comprehensive risk index R is an index reflecting the severity of reverse power over-limit; The judgment module is used to determine the probability of exceeding the limit. The comprehensive risk index R determines whether the target area needs to be equipped with a solid-state transformer. If so, the capacity calculation module is activated. The capacity calculation module is used to calculate the minimum rated capacity of the solid-state transformers configured in the target area based on the reverse power sample set.

Images