An improved voltage excursion risk assessment method, system, device and medium

Abstract

The application relates to the technical field of power system operation analysis and risk assessment, and discloses an improved voltage out-of-limit risk assessment method, system, device and medium. The method comprises the following steps: a source-load prediction error probability distribution model is established to generate a wind-solar-load output scene; an electric vehicle charging load probability model is established to generate a charging load sample; an index of the severity of node voltage out-of-limit is calculated by using an exponential amplification function and is improved; objective weights are calculated by using information entropy, and subjective weights are obtained in combination with an analytic hierarchy process; an optimization model is constructed based on the consistency of subjective and objective attribute values, a distribution coefficient is solved to obtain combined weights; and the voltage out-of-limit risk of each node at each time is evaluated through power flow calculation and combined weight. The application solves the problems that the existing technology is not sensitive to the voltage out-of-limit risk under the double uncertainty of source and load, and does not take into account the objective uncertainty of voltage state distribution and the subjective factors of node importance, and improves the sensitivity and accuracy of risk assessment.

Description

Technical Field

[0001] This invention relates to the field of power system operation analysis and risk assessment technology, specifically to an improved method, system, equipment, and medium for assessing voltage over-limit risks. Background Technology

[0002] With the high proportion of distributed power generation and the large-scale integration of electric vehicles into the distribution network, the uncertainty on both the source and load sides has significantly increased, exacerbating the risk of voltage exceedance and increasing the uncertainty of voltage state distribution, posing new challenges to power grid operation planning. Currently, domestic and international power grid operation risk assessments mainly focus on component failures or uncertainties in renewable energy output, while voltage exceedance risk assessment methods that consider the dual uncertainties of distributed generation (DG) and electric vehicles (EVs) have many shortcomings. A probabilistic model system that simultaneously incorporates multiple factors including day-ahead wind and solar load prediction errors and EV charging load has not been established. Traditional voltage exceedance severity indicators use linear functions, failing to employ exponential amplification functions with varying slopes to sensitively reflect the rapid increase in risk due to increased exceedance severity. Risk assessments do not consider both objective factors of voltage state distribution uncertainty and subjective factors of human decision-making, lacking a dynamic combination weighting mechanism that aligns subjective and objective attribute values. A complete, technically closed-loop assessment method has not been formed, encompassing source-load prediction error modeling, EV load probability simulation, improved risk indicator calculation, and case simulation, making it difficult to quantify voltage exceedance risk and providing planning guidance for risk reduction. Summary of the Invention

[0003] In view of the aforementioned existing problems, the present invention provides an improved method, system, device and medium for assessing voltage over-limit risks.

[0004] Therefore, the technical problem solved by this invention is: how to combine source-load prediction error modeling, electric vehicle charging load probability modeling, exponential amplification function to improve severity indicators and subjective and objective dynamic weighting mechanism to realize distribution network voltage over-limit risk assessment that takes into account the uncertainty of distributed power sources and electric vehicles, sensitively reflect the risk increase caused by the increase in the degree of over-limit, and quantify the over-limit risk to guide power grid planning by comprehensively considering the objective uncertainty of voltage state distribution and subjective factors such as node importance.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an improved voltage over-limit risk assessment method, comprising, Establish a first model, and generate a first scene based on the first model; Establish a second model, and generate a first sample based on the second model; The severity index of node voltage over-limit is calculated using the first function, and the severity index is then improved. In response to the improvement of the severity index, a weighted operation is used to calculate objective weights and subjective weights; By combining the objective and subjective weights, a third model is established; Based on the first scenario and the first sample, and combined with the third model, the risk assessment result is obtained.

[0006] As a preferred embodiment of the improved voltage over-limit risk assessment method described in this invention, the method includes: establishing a first model and generating a first scenario based on the first model, including: The first model is established, and the prediction error is generated and superimposed on the prediction information to generate the first scenario.

[0007] As a preferred embodiment of the improved voltage over-limit risk assessment method described in this invention, the method includes: establishing a second model and generating a first sample based on the second model, including: To address the key uncertainties of electric vehicles, a second model is established, with different model parameters set to generate the first sample.

[0008] As a preferred embodiment of the improved voltage over-limit risk assessment method described in this invention, the step of using a first function to calculate the severity index of node voltage over-limit includes, Based on the degree of deviation between the node voltage amplitude and the upper and lower voltage limits, the severity index is amplified and improved through the first function.

[0009] In a preferred embodiment of the improved voltage exceedance risk assessment method described in this invention, in response to the improvement of the severity index, a weighted operation is employed to calculate objective weights and subjective weights, including: The voltage values ​​of each node in each scenario are divided according to the voltage state distribution range, and the objective weight of each node is calculated. The relative importance of each node is measured to obtain the subjective weight of each node.

[0010] As a preferred embodiment of the improved voltage over-limit risk assessment method described in this invention, a third model is established by combining the objective weights and subjective weights, including: With the goal of minimizing the sum of the deviations of the subjective and objective weighted attribute values ​​of each node, the allocation coefficients of subjective and objective weights are solved to establish a third model.

[0011] As a preferred embodiment of the improved voltage over-limit risk assessment method described in this invention, the risk assessment result is obtained based on the first scenario and the first sample, combined with the third model, including: Using the first scenario and the first sample, the voltage state matrix is ​​calculated for each scenario. Based on the third model, the risk index of each node at each time moment is calculated to obtain the risk assessment result.

[0012] This invention generates scenarios by establishing a prediction error model, generates samples by establishing an electric vehicle charging load model, and improves the severity index by using an amplification function. It can sensitively reflect the risk of voltage exceeding the limit and realize the assessment of voltage exceeding the limit risk of distribution networks that takes into account dual uncertainties.

[0013] This invention provides an improved voltage over-limit risk assessment system, comprising: The first scene generation module generates prediction errors through the established first model, and then superimposes the prediction errors onto the prediction information to generate a first scene to represent uncertainty. The first sample generation module establishes a second model to address the key uncertainties in electric vehicle charging behavior, sets model parameters for different vehicle models, and generates the first sample to characterize electric vehicles. The severity index improvement module calculates and amplifies the severity index based on the degree of deviation between the node voltage amplitude and the upper and lower voltage limits using the first function, resulting in an improved severity index. The weight calculation module divides the voltage values ​​of each node in each scenario into voltage state distribution intervals, calculates objective weights based on information entropy, and simultaneously measures the relative importance of each node to calculate subjective weights. The third model building module aims to minimize the sum of the deviations between the subjective weighted attribute values ​​and the objective weighted attribute values ​​of each node, solves the allocation coefficients of subjective weights and objective weights, and establishes a third model for determining the combined weights. The risk assessment module uses the first scenario and the first sample to perform power flow calculations for each scenario at each time point to obtain the voltage state matrix. Based on the third model, it calculates the risk indicators of each node at each time point and traverses all time periods to obtain the risk assessment results.

[0014] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of an improved voltage over-limit risk assessment method.

[0015] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of an improved voltage over-limit risk assessment method.

[0016] Compared with existing technologies, the beneficial effects of this invention are as follows: By establishing a source-load prediction error model to generate wind-solar load scenarios, establishing an electric vehicle charging load model to generate samples, using an exponential amplification function to improve severity indicators, and combining information entropy weights and the analytic hierarchy process (AHP) for dynamic weighting of subjective and objective factors, this invention solves the problems of insensitive voltage limit exceedance risk assessment and failure to consider both objective uncertainty in voltage state distribution and subjective factors of node importance when considering the dual uncertainties of DG and EV in existing technologies. Voltage state distribution is obtained through time-series Monte Carlo simulation and probabilistic power flow; an exponential amplification function is used to sensitively reflect the increase in exceedance risk with varying slope; information entropy is used to quantify the dispersion of voltage distribution to determine objective weights; the AHP is used to measure node importance to determine subjective weights; and dynamic combination weighting is achieved based on the consistency of subjective and objective attribute values. This invention realizes a full-process technical assessment of voltage limit exceedance risk considering source-load uncertainties, with higher sensitivity and more accurate assessment, providing planning guidance for reducing voltage limit exceedance risk. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is an implementation diagram of an improved voltage over-limit risk assessment method provided in one embodiment of the present invention.

[0019] Figure 2 This is a schematic block diagram of the source-load prediction error probability distribution model of an improved voltage over-limit risk assessment method provided in one embodiment of the present invention.

[0020] Figure 3 This is a schematic block diagram of a distribution network day-ahead voltage over-limit risk assessment method provided in an embodiment of the present invention.

[0021] Figure 4 This is a schematic diagram of an IEEE-33 example system for an improved voltage over-limit risk assessment method provided in an embodiment of the present invention.

[0022] Figure 5 This is a schematic diagram of the photovoltaic day-ahead prediction curve for an improved voltage over-limit risk assessment method provided in one embodiment of the present invention.

[0023] Figure 6 This is a schematic diagram of the wind power day-ahead prediction curve for an improved voltage over-limit risk assessment method provided in one embodiment of the present invention.

[0024] Figure 7 This is a schematic diagram of the day-ahead load prediction curve for an improved voltage over-limit risk assessment method provided in one embodiment of the present invention.

[0025] Figure 8 This is a schematic block diagram illustrating the effectiveness analysis of an improved voltage over-limit risk assessment method according to an embodiment of the present invention.

[0026] Figure 9 The voltage distribution state diagram of node 3 at time 15 in various scenarios is provided as an embodiment of the present invention for an improved voltage over-limit risk assessment method.

[0027] Figure 10 The voltage distribution state diagram of node 3 at time 16 in various scenarios is provided as an embodiment of the present invention for an improved voltage over-limit risk assessment method. Detailed Implementation

[0028] To make the above-mentioned objects, features, and advantages of the present invention more readily understood, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0029] Example 1, referring to Figure 1 As a first embodiment of the present invention, an improved voltage over-limit risk assessment method is provided, comprising: S1: Establish the first model and generate the first scene based on the first model.

[0030] S2: Build a second model and generate the first sample based on the second model.

[0031] S3: Using the first function, calculate the severity index of node voltage exceeding the limit and improve the severity index.

[0032] S4: In response to the improvement of the severity metric, a weighted operation is used to calculate objective weights and subjective weights.

[0033] S5: Combine objective weights and subjective weights to establish a third model.

[0034] S6: Based on the first scenario and the first sample, combined with the third model, the risk assessment results are obtained.

[0035] It should be noted that the high proportion of distributed power sources and the large-scale integration of electric vehicles have significantly increased the risk of voltage exceedances in distribution networks and exacerbated the uncertainty of voltage state distribution. Existing risk assessment methods often focus on component failures or uncertainties in new energy sources, lacking a systematic consideration of the combined impact of uncertainties on both the source and load sides. This results in assessments that fail to accurately reflect the true risk of voltage exceedances. Therefore, proposing a method that considers source and load uncertainties and improves the characterization and weighting mechanism of voltage exceedance risks is crucial for enhancing the accuracy of power grid operation risk assessment and its planning guidance value.

[0036] This embodiment also establishes a wind-solar load prediction error and EV charging load probability model, introduces an amplification function to improve the severity index to enhance the sensitivity to the degree of over-limit, and constructs a dynamic combination weight that is consistent with both subjective and objective factors based on information entropy and the analytic hierarchy process, thereby comprehensively assessing the uncertainty and importance of node voltage state, and realizing the accurate quantification and effective identification of voltage over-limit risk in source-load uncertainty scenarios.

[0037] Example 2, refer to Figure 1 — Figure 10 As an embodiment of the present invention, an improved voltage over-limit risk assessment method is provided based on the above embodiment.

[0038] In this embodiment of the application, step S1 involves establishing a first model and generating a first scene based on the first model, specifically including the following steps A1-A3: like Figure 2 As shown, a source-load prediction error probability distribution model is established as the first model: a day-ahead "wind-solar-load" prediction error probability distribution model is established, and prediction errors are simulated through time-series Monte Carlo simulation and superimposed on day-ahead prediction information to generate a large number of wind and solar load output scenarios.

[0039] A1: Establish a probability distribution model for photovoltaic power prediction errors.

[0040] It should be noted that due to the influence of many factors such as the accuracy of the prediction method, the reliability of the data, and changes in actual meteorological conditions, there will inevitably be a deviation between the predicted and actual photovoltaic power values. Among these factors, solar irradiance is the main factor affecting photovoltaic output, and the relationship between the two is shown in the following formula: in, For photovoltaic output power, For photoelectric conversion efficiency, The total area of ​​the photovoltaic array Light intensity.

[0041] Furthermore, there is a linear relationship between photovoltaic output power and light intensity, while existing studies use a mean of 0 and a standard deviation of... If the normal distribution model represents the prediction error of light intensity, then the prediction error of photovoltaic power also follows a normal distribution, and its probability density function is: in, Let be the probability density function of the photovoltaic power prediction error. Let the value of the random variable representing the photovoltaic power prediction error be taken. The standard deviation of the photovoltaic power prediction error. The photovoltaic power prediction error coefficient; This refers to the installed capacity of photovoltaic (PV) units.

[0042] A2: Establish a probability distribution model for wind power prediction errors.

[0043] It should be noted that the probability distribution model for wind power prediction error is described using a continuous distribution model that most closely approximates the actual wind power prediction error distribution of the wind farm. Given the inherent uncertainty in actual wind speed, there will be a certain deviation between the predicted and actual wind power values. Therefore, a thorough study of the probability distribution of wind power prediction error is particularly important. In this embodiment, an unconventional normal distribution is used to simulate the wind power prediction error, as detailed below: in, Let be the probability density function of the wind power prediction error. Let be the value of the random variable representing the wind power prediction error. The standard deviation of the wind power prediction error. and The prediction parameters depend on the prediction duration and the size of the wind farm. For wind power unit predicted power, This refers to the rated installed capacity of the wind turbine.

[0044] A3: Establish a probability distribution model for load forecasting errors.

[0045] It should be noted that a probability distribution model for load forecasting errors is established. Electricity load is influenced by numerous factors such as user demand, climate conditions, and economic activities, exhibiting certain regularities over time. Furthermore, historical data can provide a reference for future load forecasting, thereby improving its accuracy. Using past data, load can be predicted for a future period, allowing for more rational and economical scheduling of generating units to ensure the stable and safe operation of the power grid. However, due to the incompleteness of load forecasting methods, the inaccuracy of historical data, and the unevenness of actual demands, load forecasting inherently involves errors. Analyzing the probability distribution of load forecasting errors serves as a basis for our rational analysis and decision-making.

[0046] Furthermore, a non-standard normal distribution with a mean of 0 is used to simulate the probability distribution of load power prediction error, as shown below: in, Let be the probability density function of the load forecasting error. Let the random variable representing load forecasting error take its values. This represents the standard deviation of the load forecasting error. The load power prediction error coefficient. This is the mathematical expectation or day-ahead forecast of the load's active power.

[0047] Based on step S1, this embodiment quantifies the prediction uncertainty of photovoltaic, wind power and load, providing a random input scenario that reflects actual fluctuations for subsequent voltage over-limit risk assessment, thereby supporting the accurate characterization of voltage distribution uncertainty.

[0048] In this embodiment of the invention, step S2 involves establishing a second model and generating a first sample based on the second model, specifically including the following steps B1-B2: B1: Establish a probabilistic model of electric vehicle charging load as the second model.

[0049] It should be noted that a probability distribution model considering multiple uncertainties is established for electric vehicle charging load. Based on this model, a method for planning and selecting electric vehicle charging stations considering uncertainties is proposed, and the effectiveness and practicality of the proposed algorithm are verified through numerical examples. Electric vehicles, as mobile charging devices, are affected by multiple uncertain variables such as battery capacity, charging equipment, daily driving distance, and grid connection and disconnection times, thus exhibiting considerable randomness in both time and space.

[0050] B2: Select the key factors affecting the charging load and establish mathematical probability and statistical models for each.

[0051] It should be noted that the mathematical probability and statistical model includes, but is not limited to, the initial charging time, daily mileage, and charging time, as modeled as follows: Starting charging time: The starting time is an important parameter in user charging behavior. Users are likely to start charging after the current trip to ensure normal power supply for the next trip. This embodiment obtained the power consumption and average mileage of family vehicles used by households in various cities at different times, thus proving that using the total daily power consumption of households as the starting time is reasonable. Users' daily travel habits are closely related to the starting time. To ensure the next trip, users generally start charging after the current trip. Through research on family vehicles and data fitting, it was found that the starting charging time of vehicles follows a normal distribution, as shown in the following formula: in, The probability density function for the initial charging moment of the EV. The initial charging time for the EV The expected value at the start of charging. The standard deviation of the initial charging time is denoted by , and the values ​​of both are related to vehicle travel patterns.

[0052] Daily mileage: The daily mileage of an electric vehicle not only determines its electricity consumption during driving but is also an important indicator of vehicle performance and affects charging time. By analyzing the motion and energy consumption characteristics of electric vehicles under different operating conditions, a driving behavior model and a method for predicting electric vehicle fuel consumption were established. Combining data from NHTS and literature, the daily mileage of electric vehicles follows the distribution as follows: in, Let be the probability density function of the daily driving mileage of the EV. Daily mileage, in km. for Expected value for Standard deviation; and The value of varies depending on the EV type.

[0053] The State of Charge (SOC) at the moment of initial grid connection can be calculated based on the daily mileage. The specific calculation formula is as follows: in, The state of charge at the initial grid connection time, The charging level during the last time the device was disconnected from the network. Daily mileage, in km. Electricity consumption per kilometer for electric vehicles This refers to the battery capacity of electric vehicles.

[0054] Charging time: Charging time depends on the vehicle's initial SOC, charging power, and the amount of electricity required to ensure the user's travel needs. The calculation formula is as follows: in, Charging time for electric vehicles, For the user's target state of charge, For electric vehicle battery capacity, Improve the charging efficiency of electric vehicles. Charging power for electric vehicles.

[0055] The off-grid time of the EV can be obtained from the start charging time and charging duration: in, Charging time for electric vehicles, For the time when the vehicle connects to the grid, This refers to the time when the vehicle leaves the network.

[0056] Based on step S2, this embodiment solves the technical problem of difficulty in quantifying distribution network voltage exceedance caused by the randomness of EV charging behavior by establishing a probability model of electric vehicle charging load and selecting the start charging time, daily driving mileage, and charging duration as influencing factors, and provides a scientific source-load uncertainty sample for voltage exceedance risk assessment.

[0057] In this embodiment of the invention, step S3 uses a first function to calculate the severity index of node voltage exceeding the limit, and improves the severity index, specifically including the following steps: like Figure 3 As shown, the severity index is amplified and improved by the first function based on the degree of deviation between the node voltage amplitude and the upper and lower voltage limits.

[0058] It should be noted that, in order to address the shortcomings of traditional voltage over-limit risk indicators, an amplification function is used as the first function to improve the severity indicator and more sensitively reflect the degree of node voltage over-limit.

[0059] For the severity index of voltage exceedance, an exponential function is used to amplify it with a varying slope as the degree of voltage exceedance increases; this is the amplification function, and its calculation formula is as follows: in, It is a natural constant. for Time Node In the scene The voltage severity index under the following conditions for Time Node In the scene The extent to which the voltage exceeds the limit; for Time Node In the scene Voltage amplitude under, This is the upper limit of the voltage amplitude. This is the lower limit of the voltage amplitude.

[0060] After considering the magnification factor, Time Node Risk indicators As shown below: in, for Time Node In the scene The voltage severity index is as follows: Number of scenes for Time Node In the scene The probability of the corresponding voltage state.

[0061] Based on step S3, this embodiment uses an exponential amplification probability function to amplify the voltage over-limit severity index with a variable slope, which solves the problem of the insensitivity of the traditional linear index to the increase of over-limit severity, making the severity index more able to reflect the sharp increase in operational risk caused by the increase in the degree of over-limit.

[0062] In this embodiment of the application, in response to the improvement of the severity index, step S4 employs a weighted operation to calculate the objective weight and the subjective weight, specifically including the following steps C1-C5: C1: Divide the voltage values ​​of each node in each scenario according to the voltage state distribution range, calculate the objective weight of each node, measure the relative importance of each node, and obtain the subjective weight of each node.

[0063] It should be noted that information entropy can reflect the dispersion of data. The greater the information entropy, the greater the dispersion of data, indicating that there is a lot of uncertainty in this set of data. When assessing voltage risk, if the distribution of voltage status is more dispersed, the corresponding risk will be higher, resulting in a larger corresponding weight value. This achieves the purpose of using information entropy to determine the actual weight of the risk indicator of voltage exceeding the limit.

[0064] C2: Calculate the number of voltage states and the probability of each state.

[0065] It should be noted that, in Time period For nodes The voltage state matrix contains Voltage values ​​under various scenarios. Divided according to voltage state distribution intervals Divide into equal parts, and let the part fall into the first part. The voltage values ​​for each voltage state range are: If there are 1, then the node Voltage state number Given the number of intervals containing voltage data, the formula for calculating the probability corresponding to the voltage state is as follows: in, for Time Node The The probability corresponding to each voltage state Number of scenes The number of nodes for The moment fell into the first The number of scenarios in each voltage state range. Divide the voltage state interval into several categories. Number the voltage state interval.

[0066] C3: Calculate the voltage distribution information entropy.

[0067] It should be noted that the formula for calculating the information entropy of voltage distribution is as follows: in, for Time Node Voltage information entropy, The number of voltage state intervals is determined. Numbering the voltage state intervals, for Time Node The The probability corresponding to each voltage state.

[0068] C4: Calculate the objective weights.

[0069] It should be noted that the entropy weight is used as an objective weight, and the calculation formula is as follows: in, for Time Node The objective weight of voltage risk indicators, for Time Node Voltage information entropy, The number of nodes.

[0070] C5: Measure the relative importance of each node to obtain the subjective weight of each node.

[0071] It should be noted that the analytic hierarchy process (AHP) is used for both qualitative and quantitative analysis, and decision-makers determine the subjective weights used to calculate the risk assessment indicators by assessing the relative importance of each node based on experience. .

[0072] Based on step S4, this embodiment obtains the objective weight by calculating the dispersion of the voltage state distribution through information entropy, and obtains the subjective weight reflecting the importance of the node by the analytic hierarchy process. This solves the problem that traditional risk indicators consider voltage uncertainty but do not consider the distribution of voltage uncertainty and the differences between nodes, making the risk assessment more reasonable.

[0073] In this embodiment of the application, step S5 combines objective weights and subjective weights to establish a third model, specifically including the following steps D1-D2: D1: With the goal of minimizing the sum of the deviations of the subjective and objective weighted attribute values ​​of each node, solve for the allocation coefficients of subjective and objective weights to establish the third model.

[0074] It should be noted that subjective weights play a crucial role in making decisions that align with expected goals, but the selection of these weights carries significant uncertainty. Determining subjective weights requires consideration of factors such as the decision-maker's level of understanding and preferences, making it highly subjective. While we fully utilize indicator data when obtaining objective weights, which helps eliminate the influence of human factors, we neglect people's decision-making intentions, potentially leading to inconsistencies between the evaluation results and our expected goals. Therefore, we propose a method that combines subjective and objective weights, using a combination of subjective and objective approaches to determine the weights of each evaluation factor. To ensure the objectivity and subjectivity of decisions during the evaluation process, we employ a method based on the consistency of subjective and objective attribute values, combining subjective and objective weights in a certain proportion to obtain the combined weights, as shown in the following formula: in, for Time Node The combined weight value, for Time Node The objective weight of voltage risk indicators, for Time Node The subjective weight value of the voltage risk index, This is the allocation coefficient between subjective and objective weights.

[0075] Furthermore, nodes Deviation between subjective and objective decision-making information for: in, For nodes The degree of deviation between subjective and objective decision-making information. for Time Node Risk indicators for Time Node The objective weight of voltage risk indicators, for Time Node The subjective weight value of the voltage risk index, This is the allocation coefficient between subjective and objective weights.

[0076] D2: Establish the third model.

[0077] It should be noted that for each node, there exists the above expression for the degree of deviation. When the sum of these deviations is small, it means that the choice of weight allocation coefficients is more conducive to balancing subjective and objective decision information, thus making this information more consistent. Based on this, the optimal weights of each objective function are determined by calculating the relative closeness between each evaluation index and the ideal solution. To ensure that both subjective and objective decision information are fully reflected, we construct an optimization model based on the consistency of subjective and objective weighted attribute values, as the third model, as shown below: in, The objective function is the sum of the deviations between the subjective and objective information of all nodes. For nodes The degree of deviation between objective and subjective decision-making information. This is the allocation coefficient between subjective and objective weights.

[0078] By solving this model, we can obtain the weighting coefficients that make the subjective and objective weighted attribute values ​​tend to be consistent. and .

[0079] Based on step S5, this embodiment constructs an optimization model constrained by minimizing the deviation of subjective and objective weighted attribute values, and obtains the optimal allocation coefficient of subjective and objective weights. This solves the problem that it is difficult to coordinate and match the subjective and objective weights by simply adding them together, and realizes the organic integration of subjective and objective information, making the voltage over-limit risk assessment more reasonable.

[0080] In this embodiment of the application, step S6, based on the first scenario and the first sample, and combined with the third model, yields the risk assessment result, specifically including the following steps: Using the first scenario and the first sample, the voltage state matrix is ​​calculated for each scenario. Based on the third model, the risk index of each node at each time moment is calculated to obtain the risk assessment result.

[0081] It should be noted that the overall evaluation process is as follows: Determine the basic parameters of the distribution network topology, DG, base load, EV, and the parameters of the prediction error probability distribution model; Obtain the current forecast output of "wind-solar-lotus"; Based on the day-ahead forecast information and the probability distribution model of the forecast error, N wind and solar power output scenarios are generated using the time-series Monte Carlo method. Based on the electric vehicle charging load probability model, the Monte Carlo method is used to simulate N EV charging load samples. exist Power flow calculations are performed on all scenarios at all times to obtain the voltage state matrix of each node at this time; calculate At any given moment Improve the severity metrics and the extent of limit violations in all scenarios, until all nodes have been traversed; calculate Time Node Risk indicators; compute nodes The voltage distribution entropy value is obtained, and after traversing all nodes, the objective weight of each node is calculated. To measure the relative importance of each node, the subjective weight of each node is obtained based on the analytic hierarchy process (AHP). Based on the consistency of subjective and objective attribute values, an optimization model is constructed to obtain the allocation coefficients of subjective and objective weights, and then the combined weights are calculated. Based on the combined weights, calculate Risk indicators at each node at any given time; The voltage over-limit risk assessment index based on the amplification function and subjective / objective weighting is as follows: in, As an indicator for assessing the risk of voltage exceeding limits, for Time Node The combined weight value, for Time Node Risk indicators.

[0082] By iterating through all time periods, the risk assessment results are obtained.

[0083] Furthermore, in comprehensive evaluation theory, sensitivity is typically used to measure the distinguishing ability of an evaluation indicator relative to the evaluation object. Higher sensitivity indicates a stronger distinguishing ability of the evaluation indicator, making it more effective in differentiating differences between evaluation objects, and also providing higher resistance to interference and accuracy. Sensitivity can be used to analyze the performance gap between improved indicators and traditional indicators; its calculation formula is as follows: in, This is the sensitivity value. To evaluate the optimal value of the result, This is the second-best value for the evaluation result.

[0084] Based on step S6, this embodiment calculates power flow by generating a large number of wind and solar power output scenarios and EV charging samples, and calculates the risk indicators of each node by combining subjective and objective weights. This solves the problem of difficulty in quantifying and assessing voltage over-limit risk under multi-source uncertainty and achieves complete prediction of day-ahead risk.

[0085] Example 3 is the third embodiment of the present invention, which differs from the previous two embodiments in that: This embodiment also provides an improved voltage over-limit risk assessment system, including: The first scene generation module generates prediction errors through the established first model, and then superimposes the prediction errors onto the prediction information to generate a first scene to represent uncertainty. The first sample generation module establishes a second model to address the key uncertainties in electric vehicle charging behavior, sets model parameters for different vehicle models, and generates the first sample to characterize electric vehicles. The severity index improvement module calculates and amplifies the severity index based on the degree of deviation between the node voltage amplitude and the upper and lower voltage limits using the first function, resulting in an improved severity index. The weight calculation module divides the voltage values ​​of each node in each scenario into voltage state distribution intervals, calculates objective weights based on information entropy, and simultaneously measures the relative importance of each node to calculate subjective weights. The third model building module aims to minimize the sum of the deviations between the subjective weighted attribute values ​​and the objective weighted attribute values ​​of each node, solves the allocation coefficients of subjective weights and objective weights, and establishes a third model for determining the combined weights. The risk assessment module uses the first scenario and the first sample to perform power flow calculations for each scenario at each time point to obtain the voltage state matrix. Based on the third model, it calculates the risk indicators of each node at each time point and traverses all time periods to obtain the risk assessment results.

[0086] Example 4, the fourth embodiment of the present invention, provides an implementation scenario for an improved voltage over-limit risk assessment method, including: The effectiveness of the proposed improved indicators and risk assessment methods was verified through numerical simulations.

[0087] This embodiment uses an IEEE-33 node distribution network as a simulation example. The system structure of the simulation example is as follows: Figure 4 As shown, node 1 is the balancing node, with a constant voltage of 1.05 (per unit). Photovoltaic cell arrays are connected to nodes 3 and 9, and wind turbines are connected to nodes 14 and 28.

[0088] The parameters for wind turbines and photovoltaic systems are shown in the table below: Table 1 Fan Parameters

[0089] Table 2 Photovoltaic parameters

[0090] The day-ahead power forecast values ​​for photovoltaic, wind power, and base load are as follows: Figure 5 — Figure 7 As shown, the day-ahead forecast error is set to 20%. Electric vehicle cluster access node 26, and the parameters obtained for each type of electric vehicle based on their travel patterns are shown in Table 3. The numbers of private cars, buses, taxis, and commercial vehicles are 148, 10, 11, and 27, respectively.

[0091] Table 3. Parameters of Travel Patterns for Different Types of EVs

[0092] For the calculation of subjective weights of risk indicators, in the judgment matrix of relative importance of nodes, nodes 3, 9, 14, 26 and 28, which are more likely to exceed the limit, are set to be slightly more important than other nodes. The subjective and objective weight allocation coefficients for each time period obtained based on the consistency of subjective and objective weighted attribute values ​​are shown in Table 4.

[0093] Table 4 Weighting Coefficients

[0094] Indicator effectiveness analysis: such as Figure 8 As shown, the analysis mainly includes sensitivity analysis, voltage state uncertainty distribution, and node importance.

[0095] Sensitivity analysis: An analysis was conducted on any time period. In this implementation, time 10 was used as an example. The evaluation results based on the improved risk indicators and the traditional risk indicators are shown in Table 5 below.

[0096] Table 5 Calculation Results of Improved Risk Indicators and Traditional Risk Indicators

[0097] Based on the sensitivity calculation method, the optimized risk assessment index achieved sensitivities of 0.334 and 0.068 compared to the traditional risk assessment index. Numerical examples verified that the model can accurately identify the three most likely severity levels and corresponding voltage levels after a power grid fault. Clearly, the optimized risk assessment index has higher sensitivity, mainly because the traditional severity index increases linearly with the degree of voltage exceedance. Furthermore, the voltage values ​​obtained using the per-unit representation method are relatively low, making the traditional index less sensitive in identifying the severity at different exceedance levels. Therefore, the improved index outperforms the traditional index in calculating the probability of system losses caused by cascading trips after a fault at a critical node in the power grid. The improved index uses an exponential amplification function and amplifies the severity index for different voltage exceedance levels with a variable slope, thus more sensitively reflecting the rapid increase in operational risk caused by increased voltage exceedance.

[0098] Voltage state uncertainty distribution: This embodiment takes the voltage distribution of node 3 at times 15 and 16 under various scenarios as an example, such as... Figure 9 — Figure 10 As shown, at times 15 and 16, the number of voltage states at node 3 is 259 and 298, respectively, and their voltage distribution standard deviations are 0.0111 and 0.0187, respectively. This indicates that the voltage state distribution at node 3 is more dispersed at time 16, suggesting a greater risk. The improved indices for these two time periods are 0.0162 and 0.0221, respectively, while the traditional indices are 0.0041 and 0.0039. This demonstrates that the improved indices, after considering the uncertainty of voltage state distribution, yield more reasonable evaluation results.

[0099] Node Importance: To more accurately describe the impact of the improved metrics on node importance, we selected two nodes, n1 and n2, with completely identical voltage distributions, and assumed that node n1 was more important than node n2. The improved and traditional metrics related to these two nodes are shown in the table below. By comparing and analyzing the node importance ranking results obtained by the two methods, we found that the weights of all nodes in the improved network are higher than those of the traditional algorithm. Clearly, since the traditional evaluation metrics do not fully consider the importance of nodes, the evaluation results for the two nodes are consistent. Therefore, ranking all nodes in the network according to their importance is unreasonable. The improved nodes fully consider the relative importance of nodes when configuring their weights. The evaluation results, shown in Table 6, indicate that the risk index of node n1 is higher than that of n2, thus verifying the rationality of the improved metrics in this study.

[0100] Table 6 Comparison of Risk Indicators for Nodes of Different Importance

[0101] This embodiment also provides an electronic device applicable to an improved voltage over-limit risk assessment scenario, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the improved voltage over-limit risk assessment method proposed in the above embodiment.

[0102] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements an improved voltage over-limit risk assessment method as proposed in the above embodiments.

[0103] The storage medium proposed in this embodiment and the method for implementing an improved voltage over-limit risk assessment proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0104] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0105] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. An improved method for assessing voltage over-limit risk, characterized in that: include, Establish a first model, and generate a first scene based on the first model; Establish a second model, and generate a first sample based on the second model; The severity index of node voltage over-limit is calculated using the first function, and the severity index is then improved. In response to the improvement of the severity index, a weighted operation is used to calculate objective weights and subjective weights; By combining the objective and subjective weights, a third model is established; Based on the first scenario and the first sample, and combined with the third model, the risk assessment result is obtained.

2. The improved voltage over-limit risk assessment method as described in claim 1, characterized in that: Establish a first model, and generate a first scene based on the first model. include, The first model is established, and the prediction error is generated and superimposed on the prediction information to generate the first scenario.

3. The improved voltage over-limit risk assessment method as described in claim 2, characterized in that: Establish a second model, and generate a first sample based on the second model, including: To address the key uncertainties of electric vehicles, a second model is established, with different model parameters set to generate the first sample.

4. The improved voltage over-limit risk assessment method as described in claim 3, characterized in that: The first function is used to calculate the severity index of node voltage exceeding the limit, including: Based on the degree of deviation between the node voltage amplitude and the upper and lower voltage limits, the severity index is amplified and improved through the first function.

5. The improved voltage over-limit risk assessment method as described in claim 4, characterized in that: In response to the improvements to the severity metric, a weighted operation is employed to calculate objective and subjective weights, including: The voltage values ​​of each node in each scenario are divided according to the voltage state distribution range, and the objective weight of each node is calculated. The relative importance of each node is measured to obtain the subjective weight of each node.

6. The improved voltage over-limit risk assessment method as described in claim 5, characterized in that: Combining the objective and subjective weights, a third model is established, including: With the goal of minimizing the sum of the deviations of the subjective and objective weighted attribute values ​​of each node, the allocation coefficients of subjective and objective weights are solved to establish a third model.

7. An improved voltage over-limit risk assessment method as described in claim 6, characterized in that: Based on the first scenario and the first sample, and combined with the third model, the risk assessment results are obtained, including: Using the first scenario and the first sample, the voltage state matrix is ​​calculated for each scenario. Based on the third model, the risk index of each node at each time moment is calculated to obtain the risk assessment result.

8. An improved voltage over-limit risk assessment system, employing the improved voltage over-limit risk assessment method as described in any one of claims 1 to 7, characterized in that, include: The first scene generation module generates prediction errors through the established first model, and then superimposes the prediction errors onto the prediction information to generate a first scene to represent uncertainty. The first sample generation module establishes a second model to address the key uncertainties in electric vehicle charging behavior, sets model parameters for different vehicle models, and generates the first sample to characterize electric vehicles. The severity index improvement module calculates and amplifies the severity index based on the degree of deviation between the node voltage amplitude and the upper and lower voltage limits using the first function, resulting in an improved severity index. The weight calculation module divides the voltage values ​​of each node in each scenario into voltage state distribution intervals, calculates objective weights based on information entropy, and simultaneously measures the relative importance of each node to calculate subjective weights. The third model building module aims to minimize the sum of the deviations between the subjective weighted attribute values ​​and the objective weighted attribute values ​​of each node, solves the allocation coefficients of subjective weights and objective weights, and establishes a third model for determining the combined weights. The risk assessment module uses the first scenario and the first sample to perform power flow calculations for each scenario at each time point to obtain the voltage state matrix. Based on the third model, it calculates the risk indicators of each node at each time point and traverses all time periods to obtain the risk assessment results.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the improved voltage over-limit risk assessment method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the improved voltage over-limit risk assessment method according to any one of claims 1 to 7.

Images