A reverse overload and risk assessment method for solid state transformer selection and configuration

Abstract

This invention discloses a method for reverse overload and risk assessment of solid-state transformer (SST) selection and configuration, comprising: 1) collecting historical operating data of the transformer substation; 2) constructing a time-series curve of reverse load rate and extracting intraday, seasonal, and substation type distribution characteristics of reverse overload; 3) establishing a photovoltaic output Beta distribution model and a load power normal distribution model, and using Monte Carlo simulation to generate an empirical probability distribution of reverse load power; 4) defining a comprehensive risk index R, quantifying the severity, probability of occurrence, and duration of overload, and comparing it with a threshold to determine the SST mitigation boundary; and 5) finally identifying key influencing factors through sensitivity analysis. This invention overcomes the limitations of deterministic models, provides a quantitative basis for the selection and configuration of flexible mitigation devices such as SST, and has the advantages of strong versatility and good operability, making it suitable for reverse overload risk assessment and mitigation decisions in high-proportion photovoltaic distribution substations.

Description

Technical Field

[0001] This invention relates to the field of power distribution network operation analysis and control technology, and in particular to a method for reverse heavy overload and risk assessment of solid-state transformer selection and configuration. Background Technology

[0002] In recent years, with the rapid development of distributed photovoltaic (PV) power generation technology and the in-depth promotion of the "dual carbon" goal, the installed capacity of distributed PV has experienced explosive growth. The high proportion of distributed PV grid connection has led to a serious "reverse overload" problem in many distribution transformer areas: 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 10 kV lines and transformer areas, severely impacting equipment lifespan and grid safety.

[0003] In traditional distribution network planning and operation analysis, the understanding of reverse heavy overload is mostly based on empirical thresholds or deterministic power flow calculations, lacking in-depth exploration of the spatiotemporal evolution of reverse load power and quantitative risk assessment methods. On the one hand, both photovoltaic output and load power exhibit significant randomness and volatility, making it difficult for deterministic models to accurately characterize the probability distribution of reverse load power. On the other hand, the reverse heavy overload characteristics of different distribution areas are affected by multiple factors such as photovoltaic penetration rate, load characteristics, transformer capacity, and line parameters. Existing methods have failed to establish a systematic risk assessment index system, resulting in a lack of quantitative technical boundary basis when formulating mitigation solutions (such as configuring solid-state transformers, SST). Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method for reverse heavy overload and risk assessment of solid-state transformer selection and configuration. This method can analyze the spatiotemporal evolution of reverse heavy overload in high-proportion photovoltaic power generation areas, quantify the risk of reverse overload, and provide theoretical guidance for the rational selection and parameter configuration of flexible mitigation devices such as SST.

[0005] The objective of this invention is achieved as follows: a method for reverse heavy overload and risk assessment of solid-state transformer selection and configuration, comprising the following steps:

[0006] Step S1: Collect historical operating data of high-proportion photovoltaic power distribution areas. The historical operating data includes photovoltaic output time-series data, load power time-series data, and transformer high and low voltage and current data.

[0007] Step S2: Based on the collected data, construct the reverse load rate time series curve, analyze the intraday and seasonal distribution characteristics of reverse heavy overload, and extract the spatiotemporal evolution law of reverse heavy overload.

[0008] Step S3: Establish a two-way power flow probability model that considers the randomness of photovoltaic output and load, and use Monte Carlo simulation to generate the probability distribution of reverse load power;

[0009] Step S4: Define the reverse heavy overload risk index, calculate the risk value based on the probability distribution, and determine the technical boundary of SST treatment for solid-state transformers;

[0010] Step S5: Identify key influencing factors through sensitivity analysis. Repeat steps S3 and S4 to calculate the partial derivatives of the risk index with respect to each factor, quantify the influence weight of each factor, and identify the key factors affecting reverse overload.

[0011] Furthermore, the reverse load rate mentioned in step S2) is defined as:

[0012]

[0013] in, Let t be the reverse active power of the transformer area. The rated capacity of the distribution transformer; the spatiotemporal evolution of reverse heavy overload includes: identification of the intraday peak period of reverse load rate, differences in reverse characteristics in different seasons, and comparison of reverse characteristics of different transformer substation types.

[0014] Furthermore, step S3) specifically includes establishing a two-way power flow probability model that considers the randomness of photovoltaic output and load:

[0015] Photovoltaic power output is modeled using a Beta distribution, with the following probability density function:

[0016]

[0017] in , The shape parameter is obtained from historical data through maximum likelihood estimation; P pv Contribute to photovoltaic power; To achieve the theoretical maximum output of the photovoltaic array;

[0018] The load power is modeled using a normal distribution; its probability density function is:

[0019]

[0020] In the formula, For load power, for load Standard deviation This is the average load.

[0021] The Monte Carlo method is used to jointly sample photovoltaic power output and load power, only when... At that time, calculate the reverse active power of the transformer area. Otherwise, the reverse load power is 0; repeat N times to obtain the reverse load power sample set. ; Empirical probability distribution function of reverse load power based on sample set statistics and probability density function .

[0022] Furthermore, the reverse overload risk index mentioned in step S4) is defined as:

[0023]

[0024] Among them, P limit For the reverse load power safety limit, f(P) rev T is the probability density function of the reverse load power. duration The expected duration of reverse overload is determined; based on the relationship between the risk index R and the preset threshold, the technical boundaries for configuring solid-state transformer SST mitigation are determined, including SST rated capacity, reverse load power suppression capability, and voltage regulation range.

[0025] Compared with the prior art, the beneficial effects of the present invention are: (1) It proposes for the first time a method for analyzing the spatiotemporal evolution of reverse heavy overload: by constructing a reverse load rate time series curve, the distribution characteristics of reverse heavy overload are revealed from multiple dimensions such as intraday, seasonal and substation type, providing a data basis for differentiated governance.

[0026] (2) A bidirectional power flow probability model considering the randomness of source and load was established: the photovoltaic output was described by the Beta distribution and the load power was described by the normal distribution. The probability distribution of the reverse load power was obtained by Monte Carlo simulation, which overcame the limitations of the deterministic model.

[0027] (3) A comprehensive risk index for reverse heavy overload was proposed: the severity of overload, probability of occurrence and duration are uniformly quantified to form a comparable risk value, providing a quantitative basis for governance decision-making.

[0028] (4) The technical boundaries of SST control were determined: Based on the risk assessment results, a quantitative method for determining key parameters such as SST capacity and suppression capability was given, avoiding empirical selection in engineering.

[0029] (5) The method is highly versatile: it can be applied to high-proportion photovoltaic areas in different regions and of different types, and only the model parameters need to be adjusted to adapt it. Attached Figure Description

[0030] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0031] Figure 1This is an overall flowchart of the method of the present invention.

[0032] Figure 2 This is a schematic diagram illustrating the intraday distribution characteristics of the reverse load rate in an example.

[0033] Figure 3 This is a schematic diagram of the photovoltaic output Beta distribution fitting for an example (including curves under different shape parameters).

[0034] Figure 4 The reverse load power probability density distribution diagram is shown in the example (Monte Carlo simulation results).

[0035] Figure 5 The example shows the risk index variation curves under different photovoltaic penetration rates.

[0036] Figure 1 Symbol names in:

[0037]

[0038] Figure 2 Symbol names in:

[0039]

[0040] Figure 3 Symbol names in:

[0041]

[0042] Figure 4 The symbol name is the same as Figure 1 ; Figure 5 The symbol names in Figure 1 . Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] A method for reverse heavy overload and risk assessment of solid-state transformer selection and configuration includes the following steps:

[0045] Step S1) Data Acquisition and Preprocessing:

[0046] Collect historical operating data of high-proportion photovoltaic power distribution areas. The historical operating data includes photovoltaic power output time-series data, load power time-series data, and transformer high and low voltage and current data.

[0047] High-precision power quality monitoring terminals were deployed in the target area to continuously collect operational data for at least one year, with a sampling interval of 5 minutes. The collected data included:

[0048] Photovoltaic power output time series data (active power);

[0049] Load power time-series data (active power, reactive power);

[0050] Voltage and current on the high-voltage and low-voltage sides of the transformer;

[0051] Auxiliary data such as harmonic content and power factor.

[0052] The collected data is preprocessed to remove outliers and missing values, forming a standardized operational characteristic database.

[0053] Step S2) Analysis of the spatiotemporal evolution of reverse heavy overload:

[0054] Based on the collected data, a time series curve of reverse load rate is constructed to analyze the intraday and seasonal distribution characteristics of reverse heavy overload and extract the spatiotemporal evolution law of reverse heavy overload.

[0055] The reverse load factor is defined as follows:

[0056]

[0057] in, Let t be the reverse active power of the transformer area (a positive value indicates that the power flows from the low-voltage side to the high-voltage side). This refers to the rated capacity of the distribution transformer.

[0058] Based on the inverse load factor time series curve, the following method is used to extract the spatiotemporal evolution pattern:

[0059] Intraday distribution characteristics: Statistically analyze the mean, maximum, and standard deviation of the reverse load rate at each time of day to identify the peak period of reverse overload (usually 11:00-14:00).

[0060] Seasonal distribution characteristics: The distribution characteristics of reverse load rate are statistically analyzed by season (spring, summer, autumn, winter) to analyze the differences in photovoltaic output and load matching in different seasons;

[0061] Comparison of transformer substation types: Transformer substations are divided into rural, suburban, and industrial types, and the differences in reverse characteristics of different types of transformer substations are compared and analyzed.

[0062] Step S3) Two-way power flow probability modeling considering source-load randomness: Establish a two-way power flow probability model considering photovoltaic output and load randomness, and use Monte Carlo simulation to generate the probability distribution of reverse load power;

[0063] 1) Establish a photovoltaic power output probability model:

[0064] Normalize the photovoltaic output to the interval [0,1]: ,in This represents the theoretical maximum output of the photovoltaic array (usually taken as the installed capacity). The data is normalized. It follows a Beta distribution, and its probability density function is:

[0065]

[0066] Shape parameters , The maximum likelihood estimation is obtained by fitting historical photovoltaic power output data. Maximum likelihood estimation requires solving a system of equations:

[0067]

[0068] in The denoted function is the digamma function, which is solved using numerical iteration (such as the Newton-Raphson method).

[0069] 2) Establish a load power probability model:

[0070] The active power of the load adopts a normal distribution. Description. Midday load power. It approximately follows a normal distribution, and its probability density function is:

[0071]

[0072] in, This is the average load. The standard deviation is calculated from historical midday load data using the formula for sample mean and sample standard deviation:

[0073]

[0074] 3) Monte Carlo simulation method: The Monte Carlo method is used to jointly sample the photovoltaic output and load power N times. The steps are as follows:

[0075] 1) Obtaining normalized output force by sampling from Beta distribution Calculate the actual photovoltaic output ;

[0076] 2) Obtaining load power by sampling from a normal distribution ;

[0077] 3) Calculate the reverse load power (when (Otherwise, the reverse load power is 0).

[0078] 4) Repeat N times (N≥100,000) to obtain the reverse load power sample set. ;

[0079] Constructing an empirical probability density function for reverse load power based on the sample set. (This can be estimated using histograms or kernel density).

[0080] Step S4) Reverse Heavy Overload Risk Assessment and SST Governance Boundary Determination:

[0081] Define a reverse overload risk index, calculate the risk value based on the probability distribution, and determine the technical boundary for SST treatment of solid-state transformers;

[0082] The reverse overload risk index is defined as:

[0083]

[0084] in: The reverse load power safety limit is typically taken as 80% of the transformer's rated capacity. Let be the probability density function of the reverse load power; The expected duration of reverse overload (obtained from historical data statistics).

[0085] In Monte Carlo simulations, the risk index is approximated using the sample mean:

[0086]

[0087] This risk indicator comprehensively considers the severity (power excess), probability of occurrence, and duration of reverse overload, and can fully reflect the reverse overload risk faced by the transformer area.

[0088] Based on risk indicators With preset threshold (e.g.) By comparing the two approaches, the technical boundaries of SST governance can be determined:

[0089] when At that time, if the risk is considered acceptable, there is no need to configure SST;

[0090] when When this is the case, an SST needs to be configured, and the rated capacity, reverse load power suppression capability, and voltage regulation range of the SST should be determined according to the risk value.

[0091] Step S5) Sensitivity analysis of key influencing factors: By changing parameters such as photovoltaic penetration rate, load level, and transformer capacity, repeat steps S3 and S4 to calculate the partial derivatives of risk indicators with respect to each factor, quantify the influence weight of each factor, identify the key factors affecting reverse heavy overload, and provide a basis for the classified management of transformer substations.

[0092] like Figure 1 The flowchart shows a method for reverse heavy overload and risk assessment of solid-state transformer selection and configuration. First, data acquisition and preprocessing are performed (S1). Then, the spatiotemporal evolution law of reverse heavy overload is analyzed (S2). Next, a bidirectional power flow probability model is established and a reverse load power sample set is generated through Monte Carlo simulation (S3). The risk index R is calculated and compared with the threshold (S4). If R is greater than the threshold, an SST mitigation plan is formulated. Finally, sensitivity analysis is performed (S5).

[0093] The photovoltaic output data obtained in step 1 is the continuously sampled power value. (Unit: kW), first it needs to be normalized to the interval [0,1], that is, defined ,in This represents the theoretical maximum output of the photovoltaic array in this area (usually taken as the historical maximum value or installed capacity). The data is normalized. It approximately follows a Beta distribution, and its probability density function is:

[0094]

[0095] For a set of independent and identically distributed observation samples Its likelihood function is:

[0096]

[0097] Taking the logarithm yields the log-likelihood function:

[0098]

[0099] make:

[0100] The log-likelihood function then simplifies to:

[0101]

[0102] Maximum likelihood estimation requires Satisfy the following system of equations (for each) and (Calculate the partial derivative and set it to zero):

[0103]

[0104] in Let be the digamma function (the logarithmic derivative of the gamma function). This system of equations has no analytical solution and is typically solved using numerical iterative methods, such as the Newton-Raphson method or the fixed-point iteration method. The specific steps are as follows:

[0105] (1) Initialization: Calculate using the method of moments The initial value is . The sample mean is . The sample variance is ,but:

[0106]

[0107] (2) Iterative update: Using the Newton-Raphson method, the values ​​of the digamma function and its derivative (trigamma function) are calculated in each iteration, and the parameters are updated until convergence (e.g., the parameter change between two iterations is less than 100%). ).

[0108] (3) Convergence judgment: check .

[0109] The final result and This refers to the shape parameter under maximum likelihood estimation. In practical engineering applications, to simplify calculations, the method of moments can be used directly instead of maximum likelihood estimation, because for the Beta distribution, the method of moments is simple to calculate and usually has sufficient accuracy.

[0110] Load power Follows a normal distribution Its probability density function is:

[0111]

[0112] in: This represents the average load power (unit: kW). The standard deviation of load power (unit: kW); Let Variance be the variance.

[0113] Power quality monitoring terminals are deployed in the target area to continuously collect operational data for at least one year, with a sampling interval of 5 minutes. Load active power data during the midday period (11:00-14:00) is extracted to form a sample set. Assume a total of [data missing] data are collected during the midday period. Load data points Then the sample mean and sample standard deviation The calculation formula is:

[0114] Sample mean:

[0115]

[0116] Sample standard deviation (unbiased estimate):

[0117]

[0118] In practical applications, load power Normally, the values ​​are positive, but the normal distribution allows negative values ​​(with a very small probability, negligible). Alternatively, a truncated normal distribution can be used to restrict the domain to... For engineering approximations, the above formula can be used directly for calculation.

[0119] In summary, the Beta distribution parameters were fitted using the annual photovoltaic power output data. and Using load data from midday (peak photovoltaic power generation period) throughout the year, calculate the normally distributed load. It might be worthwhile to set power safety limits for transformer areas that exceed these limits. .Target:

[0120] (1) Calculate the reverse load power (Only when) The value is positive when the reverse load power is positive, otherwise it is zero. However, we are usually concerned with the case where the reverse load power exceeds zero. Calculating the difference directly here might result in a negative value, but this is important in statistical analysis of reverse load power exceeding safety limits. (Only positive values ​​are considered)

[0121] (2) Calculate the expected value and standard deviation of the reverse load power, and The probability of.

[0122] The technical solution of the present invention will be described in detail below with reference to specific embodiments;

[0123] Example: A typical rural photovoltaic power station area in Hubei Province

[0124] Basic information about the distribution area: Rated capacity of the distribution transformer Photovoltaic installed capacity Therefore, the photovoltaic penetration rate (PV installed capacity / transformer capacity) is 87.5%. The distribution area is for rural residential loads, with low loads at midday, higher air conditioning loads in summer, and lower electric heating loads in winter.

[0125] Step S1) Data Acquisition: Deploy power quality monitoring terminals in this distribution area and continuously collect data for the entire year of 2024 at 5-minute intervals, obtaining a total of 105,120 sampling points. The data includes: real-time photovoltaic output, total load power of the distribution area, and voltage and current on the low-voltage side of the transformer.

[0126] Step S2) Spatiotemporal evolution analysis: Based on the collected data, the reverse load rate at each time of day is calculated, and the statistical results are as follows:

[0127] Intraday distribution: Figure 2This is a schematic diagram illustrating the intraday distribution characteristics of reverse load rate in a typical high-proportion photovoltaic area in 2024, as shown in the example. The horizontal axis represents time (hours), and the vertical axis represents the reverse load rate η. rev The curve shows that the reverse load rate peaks between 11:00 AM and 2:00 PM, with the highest point occurring around 12:30 PM, reaching a peak of approximately 0.85. The horizontal dashed line represents the safe limit of 0.8, and the shaded area indicates the peak period for reverse overload. Based on the 2024 full-year PV output data, the average reverse load power was 248 kW, and the maximum reverse load power was 340 kW (corresponding to a PV output of 350 kW and a load trough of 10 kW). The reverse load rate was calculated using the formula... , =400kVA, therefore the calculated average reverse load rate is 0.62 and the maximum reverse load rate is 0.85.

[0128] Seasonal differences: The reverse load rate is highest in spring (March-May) and autumn (September-November), with an average peak of 0.68; followed by summer, with an average peak of 0.58; and lowest in winter, with an average peak of 0.35. This is because sunlight conditions are good and air conditioning load is low in spring and autumn, resulting in weak load absorption capacity during peak photovoltaic power generation periods.

[0129] Comparison of transformer substation types: Compared with suburban transformer substations in the same region, rural transformer substations have higher peak reverse load rates (0.62 vs 0.45), but shorter durations (3 hours vs 5 hours), reflecting the characteristics of rural loads increasing sharply at night and decreasing during the day.

[0130] Step S3) Bidirectional power flow probability modeling:

[0131] (1) Photovoltaic power output Beta distribution fitting: Using photovoltaic power output data (normalized) from midday (11:00-14:00) throughout the year, the shape parameters of the Beta distribution were obtained through maximum likelihood estimation: α=1.55, β=1.20. Specific parameter calculation process:

[0132] Sample mean Sample variance

[0133] Initial value for moment estimation: ;

[0134] After convergence of the Newton-Raphson iteration, the MLE estimate is obtained. ,

[0135] The mean of this distribution Corresponding to average photovoltaic power output .variance Standard deviation .

[0136] Figure 3 This is a schematic diagram of the Beta distribution fitting for photovoltaic power output. The horizontal axis represents the photovoltaic power output P. pv The vertical axis represents the probability density f(P). pv The figure shows the Beta distribution curves under different shape parameters: α=0.85, β=1.20 corresponds to cloudy / rainy weather (J-shaped), where the probability of 0 PV power generation is highest; α=1.55, β=1.20 corresponds to cloudy / moderate weather in the example (bell-shaped, solid line), where the probability of PV power generation is highest is 258.67 kW; α=2.5, β=1.20 corresponds to sunny weather (L-shaped), where the probability of PV power generation is highest is close to the installed capacity of 350 kW. The bar chart is the histogram of actual PV output in the example, which fits well with Beta (1.55, 1.20).

[0137] (2) Normal distribution parameters of load power: The sample mean μ is obtained from the statistical load data during the midday period. L =85kW, Sample standard deviation σ L =25kW, load probability density function:

[0138]

[0139] In this example, n = 105120 data points, and μ is calculated. L =85kW, σ L =25kW. This indicates that the average load in this area during the midday period is 85kW. This indicates the degree of load fluctuation around the mean of 85kW. According to the 68-95-99.7 rule of the normal distribution:

[0140] Approximately 68% of the load data falls on (i.e., within the 60~110kW range);

[0141] Approximately 95% of the load data falls on (i.e., within the range of 35~135kW);

[0142] Approximately 99.7% of the load data falls within (i.e., within the range of 10~160kW).

[0143] This means that the midday load is rarely below 10kW or above 160kW, which is consistent with the physical characteristics of rural transformer substations where the midday load is "low but still has basic power consumption".

[0144] (3) Monte Carlo simulation: Set the number of simulated joint samplings N=100,000.

[0145] Each simulation:

[0146] 1) From Sampling ,calculate ;

[0147] 2) From Sampling ;

[0148] 3) Calculate the reverse load power ;

[0149] Statistical results:

[0150] Expected power of reverse load: ;

[0151] Standard deviation of reverse load power: (Theoretical value) (match)

[0152] Reverse load power exceeds safety limit The probability of:

[0153]

[0154] That is, in approximately 3,200 samplings, the reverse load power exceeded 320 kW.

[0155] For any power value Kernel density estimation is used, and the expression is:

[0156]

[0157] In the formula, a Gaussian kernel is selected: , IQR is the interquartile range; a randomly selected measured sample (Uniform probability) Then generate a normal random number. ,but .

[0158] Figure 4 The empirical probability density distribution curve of the reverse load power, plotted based on kernel density estimation, is presented. This curve is derived from Monte Carlo simulations (N=100,000 times). The horizontal axis represents the reverse load power P. rev (kW), with the vertical axis representing the probability density f(P) rev (1 / kW). The peak of the curve is located at the expected value of 112kW, and the standard deviation is 93kW. The vertical dashed line represents the safety limit of 320kW, and the shaded area on the right represents the portion of the reverse load power exceeding the safety limit, corresponding to a probability of 3.2%.

[0159] Step S4) Risk Assessment and SST Governance Boundary Determination: Take the expected duration of reverse overload. (Based on historical statistics, midday overload lasts for an average of about 1 hour). Calculate the risk indicator R:

[0160]

[0161] We can obtain:

[0162]

[0163] Calculation process: There are a total of 3200 out-of-limit samples; the sum of the excess quantities of all out-of-limit samples. Therefore Set a risk threshold. ,because It was determined that an SST configuration was necessary. Based on Monte Carlo simulation and measured data, the maximum reverse load power during the midday peak photovoltaic power generation period can reach 340kW (corresponding to 350kW photovoltaic output and 10kW during off-peak load). To ensure the SST can safely handle the reverse load power under extreme conditions, a certain margin needs to be considered. Typically, the overload capacity of power electronic devices is calculated based on… The design can last for 1 minute, but in order to meet the N-1 reliability requirement (the traditional transformer is powered independently when the SST fails), the SST capacity should not be less than the original traditional transformer capacity. Therefore, it is recommended that the rated capacity of the SST required for the reverse load power overload in this area be rounded up to 500kW.

[0164] Step S5) Sensitivity Analysis: Change the photovoltaic penetration rate (from 50% to 100%, corresponding to an installed capacity of 200~400kW) and calculate the corresponding risk value:

[0165] (1) Permeability 50% (200kWp):

[0166] (2) Permeability 70% (280kWp):

[0167] (3) Permeability 87.5% (350kWp):

[0168] (4) Permeability 100% (400kWp):

[0169] Figure 5 The graph shows the variation of the reverse overload risk index R under different photovoltaic penetration rates. The horizontal axis represents photovoltaic penetration rate (%), and the vertical axis represents the risk index R (kW·h). The curve increases monotonically with increasing penetration rate; in this example, a penetration rate of 87.5% corresponds to R = 12.7 kW·h. The horizontal dashed line represents the risk threshold R. threshold=5.0kW·h, and the intersection of the curve corresponds to a critical penetration rate of about 65%. That is, when the photovoltaic penetration rate exceeds 65%, SST needs to be configured for treatment.

[0170] The results show that photovoltaic penetration rate is the most sensitive factor affecting the risk of reverse overload, followed by the midday load trough.

[0171] Experimental Verification: Based on the above analysis results, a 500kVA solid-state transformer (SST) was installed in this distribution area. After the treatment, the peak reverse load power at midday decreased from 340kW to 180kW, the voltage qualification rate increased from 92.3% to 99.6%, and the highest transformer winding temperature decreased from 118℃ to 95℃. The risk of equipment overload was effectively controlled. In three months of actual operation, no further protection actions or equipment failures caused by reverse overload occurred.

[0172] The method of this invention provides a scientific quantitative basis for the governance plan of the area, avoiding investment waste or insufficient governance caused by blindly selecting based on experience.

[0173] The above description of the embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. It should be noted that those skilled in the art can make several improvements and modifications to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A method for reverse heavy overload and risk assessment of solid-state transformer selection and configuration, characterized in that, Includes the following steps: Step S1: Collect historical operating data of high-proportion photovoltaic power distribution areas. The historical operating data includes photovoltaic output time-series data, load power time-series data, and transformer high and low voltage and current data. Step S2: Based on the collected data, construct the reverse load rate time series curve, analyze the intraday and seasonal distribution characteristics of reverse heavy overload, and extract the spatiotemporal evolution law of reverse heavy overload. Step S3: Establish a two-way power flow probability model that considers the randomness of photovoltaic output and load, and use Monte Carlo simulation to generate the probability distribution of reverse load power; Step S4: Define the reverse heavy overload risk index, calculate the risk value based on the probability distribution, and determine the technical boundary of SST treatment for solid-state transformers; Step S5: Identify key influencing factors through sensitivity analysis. Repeat steps S3 and S4 to calculate the partial derivatives of the risk index with respect to each factor, quantify the influence weight of each factor, and identify the key factors affecting reverse overload.

2. The method for reverse heavy overload and risk assessment of solid-state transformer selection and configuration according to claim 1, characterized in that, The reverse load rate mentioned in step S2) is defined as: in, Let t be the reverse active power of the transformer area. The rated capacity of the distribution transformer; the spatiotemporal evolution of reverse heavy overload includes: identification of the intraday peak period of reverse load rate, differences in reverse characteristics in different seasons, and comparison of reverse characteristics of different transformer substation types.

3. The method for reverse heavy overload and risk assessment of solid-state transformer selection and configuration according to claim 1, characterized in that, Step S3) describes establishing a two-way power flow probability model that considers both photovoltaic output and load randomness, specifically including: Photovoltaic power output is modeled using a Beta distribution, with the following probability density function: in , The shape parameter is obtained from historical data through maximum likelihood estimation; P pv Contribute to photovoltaic power; To achieve the theoretical maximum output of the photovoltaic array; The load power is modeled using a normal distribution; its probability density function is: In the formula, For load power, for load Standard deviation This is the average load. The Monte Carlo method is used to jointly sample photovoltaic power output and load power, only when... At that time, calculate the reverse active power of the transformer area. Otherwise, the reverse load power is 0; repeat N times to obtain the reverse load power sample set. Empirical probability distribution function of reverse load power based on sample set statistics and probability density function .

4. The method for reverse heavy overload and risk assessment of solid-state transformer selection and configuration according to claim 1, characterized in that, The reverse overload risk index mentioned in step S4) is defined as follows: in, For reverse load power safety limits, Let be the probability density function of the reverse load power. The expected duration of reverse overload; based on risk indicators. The relationship with preset thresholds determines the technical boundaries for configuring solid-state transformer (SST) governance, including SST rated capacity, reverse load power suppression capability, and voltage regulation range.

Images