A rolling assessment method for operation risk of flexible distribution network

Abstract

The present invention discloses a rolling assessment method for the operation risk of a flexible distribution network, comprising the following steps: Step S1: According to the frequency and duration of typhoon extreme events and the degree of damage to the distribution network, a probability distribution model characterizing the intensity and duration of typhoon extreme disasters is established; Step S2: According to the failure mechanism of each component in the distribution network system, a relationship model between the intensity of typhoon disasters and the probability of failure of distribution network components is established; Step S3: According to the failure mode, independence and correlation of the distribution network components under typhoon disasters, a model of its impact on the distribution network is established, and it is applied to disaster simulation; Step S4: Generate a distribution network failure scenario through the vulnerability curve of the component and then dynamically evaluate the functional level of the distribution network. The scheme describes the vulnerability of components through multi-dimensional disaster intensity key parameters, providing a basic premise and support for the elastic evaluation of the distribution network.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution networks, and in particular to a rolling assessment method for operating risks of a flexible distribution network. Background Art

[0002] Resilient power grids are mainly aimed at extreme events with low probability and high risk. The behavior patterns of such events are often difficult to predict. In order to prevent such events, distribution networks must carry out detailed modeling of the frequency, location, and impact of disasters.

[0003] Traditional modeling methods are based on static probability distribution sampling or use several historical disasters as analysis scenarios, and establish component failure probability curves based on a single measure of disaster intensity. These methods are insufficient to describe the dynamic development of disasters, nor can they characterize the potential dangers of natural disasters, let alone dynamically characterize the vulnerability of components through multi-dimensional disaster intensity key parameters.

[0004] In the current resilience assessment and calculation, only the difference in time dimension is considered, the resilience research stages are divided, and the post-disaster recovery strategy is not considered in the simulation. Therefore, the indicator dimension is relatively single and not comprehensive enough. It cannot fully reflect the different application functions of system resilience assessment in power system planning, dispatching and monitoring, nor can it reflect the internal and external factors that affect the resilience of the power system, which has created certain obstacles for the formulation of distribution network resilience improvement measures.

[0005] As for the current dynamic risk assessment method, it only conducts online risk assessment from the perspective of the short-term reliability of the power system, without considering the impact of extreme events on the distribution network, and failing to integrate the load transfer during the disaster and the power supply restoration process after the disaster; the early warning indicator calculation method also only considers the power grid level, the indicators are relatively simple, and the accuracy is also lacking; the data fusion method also only processes static data and cannot realize real-time data analysis and processing. Summary of the invention

[0006] The purpose of this invention is to propose a rolling assessment method for the operating risks of a flexible distribution network. By fine-tuning the modeling of typhoon events, the impact of typhoon events on the distribution network is studied, and a theoretical basis is established for panoramic information visualization, accident simulation and recovery decision-making, and power grid planning and operation. At the same time, the key points of online monitoring, risk assessment and fine early warning of flexible distribution networks are studied, and the distribution network components and operating environment information are summarized and analyzed. The actual and forecast data are integrated to reflect and update the risks faced by the system in real time, providing a theoretical basis for risk prevention in the distribution network.

[0007] In order to achieve the above technical objectives, the present invention provides a technical solution, which is a rolling assessment method for the operation risk of a flexible distribution network, comprising the following steps:

[0008] Step S1: According to the frequency and duration of typhoon extreme events and the degree of damage to the distribution network, a probability distribution model characterizing the intensity and duration of typhoon extreme disasters is established;

[0009] Step S2: According to the failure mechanism of each component in the distribution network system, a relationship model between the typhoon disaster intensity and the failure probability of the distribution network components is established;

[0010] Step S3: According to the failure modes, independence and correlation of the distribution network components under the typhoon disaster, a model of its impact on the distribution network is established and applied to the disaster simulation;

[0011] Step S4: Generate distribution network fault scenarios through component vulnerability curves and dynamically evaluate the distribution network function level.

[0012] Preferably, step S1 comprises the following steps:

[0013] Step S11: extract key parameters that characterize the intensity of disaster-causing factors and establish their corresponding probability distribution models;

[0014] Step S12: converting the extracted key parameters into dynamic disaster scenarios based on meteorological or geographical models;

[0015] Step S13: construct a disaster simulation generator to generate dynamic simulated disaster scenarios based on the probability distribution of multi-dimensional key parameters.

[0016] As a preference, the key parameters include the annual typhoon occurrence rate Λ, the minimum distance from the research point D min , typhoon center pressure difference ΔP, typhoon moving speed V T And the typhoon's moving direction A T ;

[0017] in:

[0018] The probability distribution function corresponding to the annual typhoon occurrence rate Λ is:

[0019]

[0020] Among them, λ>0 is the parameter to be estimated in the Poisson distribution;

[0021] Minimum distance D from the research point min The probability distribution function is:

[0022]

[0023] Among them, θ1<θ2 is the parameter to be estimated in the uniform distribution;

[0024] The probability distribution function of the typhoon center pressure difference ΔP is:

[0025]

[0026] Among them, μ and σ 2 is the parameter to be estimated in the lognormal distribution;

[0027] Typhoon moving speed V T The probability distribution function is:

[0028]

[0029] Among them, μ and σ 2 is the parameter to be estimated in the lognormal distribution;

[0030] Typhoon moving direction A T The probability distribution function is:

[0031]

[0032] Among them, μ1 and is the parameter to be estimated for the first normal distribution, μ2 and is the parameter to be estimated for the second normal distribution, and ρ is the mixing degree parameter of the two normal distributions to be estimated.

[0033] Preferably, step S12 includes the following steps:

[0034] A typhoon wind field and filling model is established to describe the properties of typhoon disaster factors, and the Batts typhoon model is used in typhoon wind field simulation.

[0035] Preferably, constructing a typhoon simulation generator in step S13 comprises the following steps:

[0036] Step 1: Generate the coastline equation near the study point;

[0037] Step 2: Generate a simulated circle with the research point as the center and the given distance as the radius;

[0038] Step 3: Generate the minimum distance D between the typhoon path and the research point based on uniform probability distribution sampling min ;

[0039] Step 4: Generate the azimuth angle A of the typhoon movement based on the mixed normal probability distribution sampling T ;

[0040] Step 5: Determine the typhoon's moving path based on the minimum distance and the typhoon's moving direction. The typhoon's moving path is a spherical great circle passing through the center of the earth;

[0041] Step 6: Determine the typhoon's offshore generation point and departure point based on the intersection of the typhoon's moving path and the simulation circle;

[0042] Step 7: Determine the typhoon’s landing point based on the intersection of the typhoon’s moving path and the coastline equation;

[0043] Step 8: Generate the moving speed V of the typhoon center based on the log-normal probability distribution sampling T ;

[0044] Step 9: With hourly time resolution, determine the longitude and latitude position of the typhoon center hour by hour based on the moving speed and moving path of the typhoon center to form the dynamic moving path of the typhoon;

[0045] Step 10: Generate the typhoon's sea center pressure difference ΔP by sampling according to the log-normal probability distribution;

[0046] Step 11: Determine the hourly typhoon center pressure difference based on the Batts typhoon filling model, and detect whether the typhoon has landed. After landing, correct the land center pressure difference according to the attenuation formula;

[0047] Step 12: Determine the hourly maximum average wind speed of the typhoon based on the Batts typhoon wind field model, and then determine the typhoon level.

[0048] Preferably, in step S2, the relationship model is characterized by the distribution network component failure repair time r:

[0049]

[0050] Among them, M represents the catastrophic intensity of extreme events, M crit It represents the critical disaster intensity determined by the nature of the disaster; r0 represents the average repair time of the components when the system is operating normally; k represents the rate coefficient of increase in repair time, which is obtained by statistical analysis of historical operation data.

[0051] Preferably, in step S4, the method for generating a distribution network fault scenario comprises the following steps:

[0052] S41. Obtain disaster paths, intensity and time series changes;

[0053] S42. Grid processing is performed on the power grid on the disaster path, and each grid corresponds to a typhoon intensity;

[0054] S43, calculating the disaster intensity in each time period in each grid according to the distance between the grid and the disaster center;

[0055] S44, combining the fragility curves of the components in the grid, and calculating the failure probability of each component;

[0056] S45. Use a simulation method to simulate whether each component fails during the disaster period, obtain possible failure scenarios, and select and reduce the scenarios.

[0057] Preferably, dynamically evaluating the functional level of the distribution network includes the following steps:

[0058] The power supply restoration of the distribution network during a disaster is simulated, where the disaster repair time T follows the exponential distribution with the same parameters, that is,

[0059] Wherein, μ is the parameter of exponential distribution, and the expected value of repair time E(T) = μ;

[0060] When the number of damaged lines in the distribution network is greater than the number of repair teams that can be deployed, a greedy algorithm that guarantees single-step optimality is used to arrange the repair sequence;

[0061] Obtain the latest weather forecast data and real-time power grid data to determine the ultimate indicators that affect the resilience of the distribution network, including: the system maximum load loss rate indicator, the system full power restoration time indicator, and the system full power restoration time indicator;

[0062] In each round of rolling evaluation, the largest variance coefficient among the three ultimate indicators is selected as the basis for convergence judgment using the Monte Carlo simulation algorithm;

[0063] Utilize the information at the equipment level to infer the resilience level at the load level and the system level; use meteorological forecast data for forward prediction, use the actual power grid data for backward correction, and use the latest information to achieve rolling update of the risk assessment of the elastic distribution network.

[0064] Beneficial effects of the present invention:

[0065] 1. Extreme disaster modeling: A general modeling framework applicable to a variety of extreme natural disasters is proposed, and typhoon weather is modeled in a refined manner. By building a disaster simulation generator, dynamic simulated disaster scenarios are generated based on multidimensional probability distribution to characterize the potential risks and dynamic characteristics of natural disasters;

[0066] 2. Component failure mode modeling: Study the failure modes of elastic distribution network components under typhoon weather, establish probability analysis models of dependent failures and associated failures, and characterize the vulnerability of components through multi-dimensional disaster intensity key parameters based on a model-data hybrid drive approach, providing the basic premise and support for distribution network resilience assessment;

[0067] 3. Online risk assessment of elastic distribution networks: Use meteorological forecast data for forward prediction and real-time grid data for backward correction, comprehensively consider the load transfer during disasters and the power supply restoration process after disasters, and implement rolling updated risk assessment of elastic distribution networks based on the latest information. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1The present invention is a flow chart of a rolling assessment method for operation risk of a flexible distribution network.

[0069] Figure 2 This is a framework diagram for modeling typhoon disaster factors of the present invention.

[0070] Figure 3 It is a schematic diagram of the wind field of the Batts typhoon model of the present invention.

[0071] Figure 4 This is a schematic diagram of the extreme event and impact modeling of the present invention.

[0072] Figure 5 This is a flow chart of online risk assessment of the flexible distribution network of the present invention.

[0073] Figure 6 This is a grid processing diagram of the two-dimensional geographic image of the distribution network of the present invention.

[0074] Figure 7 The schematic diagram of scene generation of the two-dimensional geographic image of the distribution network of the present invention. DETAILED DESCRIPTION

[0075] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific implementation method described herein is only an optimal embodiment of the present invention, which is only used to explain the present invention and does not limit the scope of protection of the present invention. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0076] Example: Figure 1 As shown, a rolling assessment method for the operation risk of a flexible distribution network includes the following steps:

[0077] Step S1: According to the frequency and duration of typhoon extreme events and the degree of damage to the distribution network, a probability distribution model is established to characterize the intensity and duration of typhoon extreme disasters; extreme disasters usually include two categories: extreme natural disasters and extreme man-made attacks. Extreme natural disasters include severe weather or geological changes such as storms, thunderstorms, floods, wildfires, ice disasters, earthquakes, tsunamis, and geomagnetic storms; extreme man-made attacks include cyber attacks, electromagnetic pulses from high-altitude nuclear explosions, terrorist attacks, and other attack methods.

[0078] In step S1, the logical relationship between the three parts is as follows: Figure 2 As shown, the following steps are included:

[0079] Step S11: extract key parameters that characterize the intensity of disaster-causing factors and establish their corresponding probability distribution models;

[0080] Step S12: based on the meteorological or geographical model, the extracted key parameters are converted into dynamic disaster scenarios; a typhoon wind field and filling model describing the nature of typhoon disaster factors is established, such as Figure 3 As shown, the Batts typhoon model is applied to typhoon wind field simulation; step S13: construct a disaster simulation generator to generate dynamic simulated disaster scenarios based on the probability distribution of multi-dimensional key parameters.

[0081] Among them, disaster factor analysis is a comprehensive measurement of the two factors of disaster intensity and probability of occurrence. The hazard risk of disaster factors can be described as a binary combination of disaster intensity and corresponding probability of occurrence, that is, H =<M,P>

[0082] In the formula, H (hazard) is the hazard risk of the disaster-causing factor; M (magnitude) refers to the magnitude of the disaster-causing factor; and P (probability) refers to the probability of occurrence of a disaster-causing factor of corresponding intensity.

[0083] Key parameters include the annual typhoon occurrence rate Λ, the minimum distance from the research point D min , typhoon center pressure difference ΔP, typhoon moving speed V T And the typhoon's moving direction A T ;

[0084] The core content of typhoon disaster factor modeling is to establish the relationship between the intensity of disaster factors and probability; specifically, the simulation circle method is used for analysis. The selection of the simulation circle radius is related to the topography, meteorological environment, etc. of the study area. The present invention selects the simulation circle radius as 250km. Next, under the framework of the simulation circle method, the key parameters that can describe the typhoon disaster factors are established.

[0085] Typhoon annual occurrence rate Λ:

[0086] The number of typhoons in the simulation circle in a year is defined as the annual typhoon occurrence rate Λ; for typhoon disasters, the typhoon occurrence probability model is usually represented by a stationary Poisson distribution, and its distribution law is

[0087]

[0088] Among them, λ>0 is the parameter to be estimated in the Poisson distribution.

[0089] Minimum distance D from the research point min :

[0090] The shortest directed distance between the research point and the typhoon moving path is defined as the minimum distance D between the typhoon and the simulation point. min; Since the radius of the simulation circle is set to 250km, the minimum distance of the typhoon from the research point ranges from -250km to 250km. It is assumed that the minimum distance of the typhoon distance research follows a uniform distribution, and its probability density function is:

[0091]

[0092] Among them, θ1<θ2 is the parameter to be estimated in the uniform distribution.

[0093] Typhoon center pressure difference ΔP:

[0094] The difference between the air pressure of the undisturbed environment outside the typhoon and the air pressure at the center of the typhoon is defined as the pressure difference at the center of the typhoon ΔP. For tropical cyclones in the northwest Pacific, the outer air pressure is generally 1010.0hPa, and my country also uses 1010.0hPa. The pressure difference at the center of a typhoon is generally in the range of 0 to 135hPa, and is generally described by a log-normal distribution, with a probability density function of

[0095]

[0096] Among them, μ and σ 2 is the parameter to be estimated in the lognormal distribution.

[0097] Typhoon moving speed V T :

[0098] The moving speed of the typhoon eye is defined as the typhoon moving speed V T , whose value is calculated by the two frames of data points before and after the typhoon best path data set. The spherical distance between the typhoon center of the two frames of data before and after divided by the time interval of the two frames of data can be used to get the typhoon moving speed. The moving speed of the typhoon is generally in the range of 2km / h to 65km / h, and is described by lognormal distribution. Its probability density function is

[0099]

[0100] Among them, μ and σ 2 is the parameter to be estimated in the lognormal distribution.

[0101] Typhoon moving direction A T :

[0102] The spherical azimuth of the typhoon's moving path is defined as the typhoon's moving direction A T , whose value is calculated by the two frames of data points before and after the typhoon best path data set; the typhoon movement direction presents a bimodal trend. Here, a mixed normal distribution composed of two normal distributions is used to describe the probability model of the typhoon movement direction. Its probability density function is:

[0103]

[0104] Among them, μ1 and is the parameter to be estimated for the first normal distribution, μ2 and is the parameter to be estimated for the second normal distribution, and ρ is the mixing degree parameter of the two normal distributions to be estimated.

[0105] In step S13, the Batts typhoon model is widely used in typhoon wind field simulation, and the wind speed value of the research point is determined by an empirical formula. gx After neglecting friction, it can be expressed as:

[0106]

[0107] Where f is the Coriolis force coefficient of the earth's rotation, f = 2ωsinψ; K is an empirical coefficient, generally taken as 6.72; R max is the maximum wind speed radius, in km; ΔP is the pressure difference in the center of the typhoon, in hPa.

[0108] Usually, the average maximum wind speed of a typhoon generally occurs at the maximum wind speed radius R max The maximum wind speed radius R max The statistical fitting relationship between the typhoon center pressure difference ΔP is:

[0109] R max =1.119×10 3 ×ΔP -0.805

[0110] The distribution of the typhoon's horizontal wind field varies in different regions. The maximum wind speed value V in the typhoon wind field at a distance r from the typhoon center is r It can be expressed as

[0111]

[0112] The distribution of the typhoon's horizontal wind field varies in different regions. The maximum wind speed value V in the typhoon wind field at a distance r from the typhoon center is r It can be expressed as

[0113]

[0114] Among them, x is a parameter related to the radial attenuation of the typhoon, and its value ranges from 0.5 to 0.7. r It represents the 10-minute average wind speed at a height of 10m when the distance between the straight line OM and the typhoon center O is r. The straight line OM represents the straight line that forms an angle of 115° clockwise with the typhoon moving direction. The 10-minute average wind speed at the research point at a height of 10m on the straight line OP with an angle of α with the straight line OM and a distance of r from the typhoon center is

[0115] V r,a =V r -0.5V T (1-cosα)

[0116] After a typhoon lands on the coast, its central pressure difference will continue to decay. It is assumed that the typhoon's moving direction remains unchanged near the coast and after landing. After the typhoon lands, the central pressure difference decays according to the following formula

[0117]

[0118] Among them, t represents the time the typhoon travels after landing, in units of h; ΔP(t) is the central pressure difference at time t after the typhoon landed; ΔP0 is the central pressure difference at sea before the typhoon landed; It is the angle between the typhoon's moving direction and the coastline when it lands.

[0119] The construction of the typhoon simulation generator in step S13 includes the following steps:

[0120] Step 1: Generate the coastline equation near the study point;

[0121] Step 2: Generate a simulated circle with the research point as the center and the given distance as the radius;

[0122] Step 3: Generate the minimum distance D between the typhoon path and the research point based on uniform probability distribution sampling min ;

[0123] Step 4: Generate the azimuth angle A of the typhoon movement based on the mixed normal probability distribution sampling T ;

[0124] Step 5: Determine the typhoon's moving path based on the minimum distance and the typhoon's moving direction. The typhoon's moving path is a spherical great circle passing through the center of the earth;

[0125] Step 6: Determine the typhoon's offshore generation point and departure point based on the intersection of the typhoon's moving path and the simulation circle;

[0126] Step 7: Determine the typhoon’s landing point based on the intersection of the typhoon’s moving path and the coastline equation;

[0127] Step 8: Generate the moving speed V of the typhoon center based on the log-normal probability distribution sampling T ;

[0128] Step 9: With hourly time resolution, determine the longitude and latitude position of the typhoon center hour by hour based on the moving speed and moving path of the typhoon center to form the dynamic moving path of the typhoon;

[0129] Step 10: Generate the typhoon's sea center pressure difference ΔP by sampling according to the log-normal probability distribution;

[0130] Step 11: Determine the hourly typhoon center pressure difference based on the Batts typhoon filling model, and detect whether the typhoon has landed. After landing, correct the land center pressure difference according to the attenuation formula;

[0131] Step 12: Determine the hourly maximum average wind speed of the typhoon according to the Batts typhoon wind field model, and determine the maximum wind force level according to the national standard "Tropical Cyclone Level". The cyclone level classification standards are shown in Table 1, and then determine the typhoon level.

[0132] Tropical cyclone category Maximum wind force level near the center (level) Maximum average wind speed near the center (m / s) Tropical Depression (TD) 6～7 10.8～17.1 Tropical Storm (TS) 8～9 17.2～24.4 Severe Tropical Storm (STS) 10～11 24.5～32.6 Typhoon(TY) 12～13 32.7～41.4 Severe Typhoon (STY) 14～15 41.5～50.9 Super Typhoon (SuperTY) 16 and above ≥51.0

[0133] In this embodiment, a typhoon simulation generator model is established to analyze the hazard risk of disaster-causing factors and obtain a binary set of disaster intensity and corresponding occurrence probability {<M,P>} i∈N , used to describe typhoons of various catastrophic intensities and their probability of occurrence; the typhoon simulation generator can randomly sample the frequency of typhoon occurrence in a year according to the probability distribution model of the annual typhoon occurrence rate, and then randomly sample the typhoon transit scenario of the corresponding catastrophic intensity according to the probability distribution model of the remaining four key parameters.

[0134] The typhoon simulation generator model can generate typhoon scenarios of corresponding disaster intensity according to probability, which is equivalent to obtaining a binary set of disaster intensity and the probability of occurrence of a typhoon of corresponding intensity {<M,P>} i∈N , the hazard risk of typhoon disaster factors was characterized; the typhoon simulation generator first used the probability distribution of the typhoon's key parameters to randomly sample the key parameters of the typhoon, and then converted the key parameters into dynamic typhoon disaster scenarios through the typhoon wind field and filling model, so as to achieve the purpose of generating typhoon disaster scenarios by random sampling of the five-dimensional joint probability distribution.

[0135] Step S2: According to the failure mechanism of each component in the distribution network system, a relationship model between the typhoon disaster intensity and the failure probability of the distribution network components is established;

[0136] Furthermore, the failure modes of the main components in the distribution network are established and the relevant failure causes are extracted; then, based on a model-driven or data-driven approach, a component failure model that takes into account extreme event types and uncertain factors is established.

[0137] Furthermore, since the component repair time is easier to quantify than the failure rate, the component repair time r can generally be expressed as

[0138]

[0139] Where M represents the catastrophic intensity of extreme events, such as the maximum wind speed of storm disasters, the rainfall of rainstorm disasters, etc.; M crit It represents the critical disaster intensity determined by the nature of the disaster; r0 represents the average repair time of the components when the system is operating normally; k represents the rate coefficient of increase in repair time, which is obtained by statistical analysis of historical operation data.

[0140] Step S3: According to the failure modes, independence and correlation of the distribution network components under the typhoon disaster, a model of its impact on the distribution network is established and applied to the disaster simulation; Figure 4 shown.

[0141] Furthermore, overhead lines are the main carriers in the distribution network, and overhead lines are composed of overhead conductors, poles, towers, insulators and other components; specifically, poles and towers among the distribution network components are selected, their main failure modes are established, and their sensitivity to typhoon events is clarified.

[0142] Further, for tower modeling, pole load refers to external forces and other factors that cause internal forces and deformations in structures or structures. Generally, poles are subject to three main external forces:

[0143] Wind load on conductor w x for

[0144]

[0145] The wind load on the tower is w s for

[0146]

[0147] Wind load on insulator w z for

[0148]

[0149] In the formula, l H is the horizontal spacing of the wires, in m; w x The horizontal span of the line is l H Wind load, in kN; μ z is the wind pressure height variation coefficient; α is the wire wind pressure unevenness coefficient; μ sc is the wire shape coefficient; d is the outer diameter of the wire, in meters; is the angle between the wind direction and the line; v is the specified reference height of the line h s The design wind speed at the location is m / s; β is the wind vibration coefficient; μ s is the wind load shape system coefficient; A is the projected area of ​​the windward surface of the tower structure component, in m 2; n1 is the number of insulator strings used for one phase conductor; n2 is the number of insulators in each string; A p is the wind-exposed area of ​​each insulator. The above parameters can be obtained according to relevant design specifications.

[0150] The bending moment M at the xx position of any section of the pole x for

[0151]

[0152]

[0153]

[0154]

[0155] In the formula, w xz is the total wind load of the conductor and the wind load of the insulator, in kN; w sv Wind load on pole shaft; is the height from the xx section to the wind pressure resultant force point of the pole, in m; h1 is the distance from the xx section to the pole top, in m; h2 is the distance from the xx section to the cross arm, in m; F is the projection area of ​​the pole, in m 2 ; D0 is the pole tip diameter, in m; D x xx is the diameter, in m; m x is the additional bending moment coefficient caused by disturbance.

[0156] Bending strength of concrete pole M p Usually conforms to the normal distribution, and its probability density function f R It can be expressed as:

[0157]

[0158] Among them, μ p is the mean value of the bending strength of concrete poles; δ p is the standard deviation of the bending strength of the concrete pole; from the above formula, the strength variable and load effect of the pole can be obtained.

[0159] Furthermore, a complex fault model of the distribution network under typhoon disasters is established, and supervised learning is used to evaluate the resilience model. From a mathematical point of view, it can be described as

[0160] y=f(x)+ε

[0161] Among them, y represents the prediction quantity of the learner, which is the label of the sample; x represents the prediction factor as the input of the learner, which is the feature vector of the sample; the random additional noise ε represents the prediction error of the learner; for the fault trip rate model of the overhead line, the prediction quantity y of the learner is the fault trip rate λ of the overhead line, which is expressed as the number of trips of the overhead line per unit time. For typhoon disasters, since its transit time varies from several hours to several days, it is advisable to use hours as the time unit. Usually, the spatial span of the study area is very small compared to the diameter of the typhoon wind circle. It is believed that all overhead lines are in the same geographical environment area, the operating conditions are basically similar, and the design parameters of the overhead lines are the same.

[0162] The construction of the feature vector x depends on feature engineering, which refers to the process of using raw data to obtain learner training data in order to achieve better learning results. Feature construction is an important part of feature engineering, which refers to manually finding feature variables with practical or physical meaning from raw data based on certain experience.

[0163] Combining and splitting attributes are commonly used methods for feature construction. When constructing features, you can try to combine two or more different attributes to construct new features, or you can try to decompose or split an attribute.

[0164] Furthermore, based on the meteorological forecast data and combined with the previously established model of the impact of extreme natural disasters on the distribution network, the uncertainty of equipment failure during the disaster is considered to generate distribution network component failure scenarios.

[0165] Step S4: Generate distribution network fault scenarios through component vulnerability curves and dynamically evaluate the distribution network function level, such as Figure 5 shown.

[0166] In step S4, the method for generating a distribution network fault scenario includes the following steps:

[0167] S41. Obtain disaster paths, intensity and time series changes;

[0168] S42. Grid processing is performed on the power grid along the disaster path, such as Figure 6 As shown, each grid corresponds to a typhoon intensity;

[0169] S43, calculating the disaster intensity in each time period in each grid according to the distance between the grid and the disaster center;

[0170] S44, combining the fragility curves of the components in the grid, and calculating the failure probability of each component;

[0171] S45. Use simulation method to simulate whether each component fails during the disaster period, such as Figure 7 As shown, possible failure scenarios are obtained, and the scenarios are selected and reduced.

[0172] The optimization decision of the post-disaster repair strategy selects the breadth-first traversal method. Below, the method of simulating the load transfer during a disaster is given in the form of pseudo code. In it, G = (V, E) represents the graph composed of distribution lines including interconnecting feeders, V represents the set composed of distribution network nodes, and E represents the set composed of all distribution lines. represents the set of faulty return lines at time t. t represents the current time, t end It represents the time when the typhoon leaves the country, and Δt is the simulation time step.

[0173] Pseudo code of the algorithm for simulating load transfer during disasters.

[0174]

[0175]

[0176] Pseudocode of post-disaster power supply restoration simulation algorithm.

[0177]

[0178] When simulating load transfer during disasters, it is assumed that the lines in the distribution network have sufficient transmission capacity margin, and the upper and lower limit constraints of the voltage of each node are not considered. As long as the graph composed of all the lines in the distribution network is connected, it is assumed that the distribution network can be reconstructed and operated; when there is an isolated load node in the distribution network that is not connected to the substation, the load of the node is considered to be lost.

[0179] In the simulation of power supply restoration in disasters, it is assumed that the repair time T of each distribution line obeys the exponential distribution with the same parameters, that is,

[0180]

[0181] Wherein, μ is the parameter of the exponential distribution, and the expected value of the repair time E(T)=μ.

[0182] When the number of damaged lines in the distribution network is greater than the number of repair teams that can be deployed, the repair order needs to be arranged. In the post-disaster optimization dispatch, a "greedy" algorithm that guarantees single-step optimality is adopted. When a repair team completes the repair of a distribution line and becomes idle, the dispatcher will sort the importance of the damaged lines that have not yet been arranged for repair, and assign the most important repair line to the repair team.

[0183] Furthermore, the online risk assessment of the distribution network is realized, and the latest weather forecast data and real-time data of the power grid are obtained in each round of rolling assessment. Based on the latest data, the Monte Carlo simulation method is used to calculate the resilience index of the elastic distribution network and evaluate the risk level of the elastic distribution network.

[0184] Robustness and rapidity are the two essential characteristics of resilience. The ultimate indicator of distribution network resilience is broken down into three items, reflecting robustness, rapidity and a comprehensive consideration of the two.

[0185] The maximum load loss rate index of the system is defined as the ratio of the maximum load loss of the distribution network caused by the typhoon during the passage of the typhoon to the load of the distribution network before the typhoon entered the country, and is expressed as F r The maximum load loss rate index of the system can be calculated by the following formula:

[0186]

[0187] System maximum load loss rate index F r It represents the distribution network's ability to resist and absorb a typhoon disaster.

[0188] The system full power restoration time indicator is defined as the time required for the distribution network to restore power to all loads from the beginning of load loss during the typhoon period. In this section, it is denoted by F d The system full power restoration time index can be calculated by the following formula:

[0189] F d =t r -t0

[0190] System full power recovery time index F d Characterizes the distribution network's ability to respond to and quickly recover from a typhoon disaster.

[0191] The system power loss index is defined as the total power loss of the distribution network in a typhoon, which is denoted by F in this section. e The system power loss index can be calculated by the following formula:

[0192] F e =E r

[0193] System power loss index F e The robustness and rapidity dimensions of resilience are considered comprehensively.

[0194] In each round of rolling evaluation, the Monte Carlo simulation convergence criterion is that the variance coefficient β of the expected value of the indicator is less than the preset value. The largest variance coefficient among the three indicators is selected as the basis for judging the convergence of the algorithm, that is,

[0195]

[0196] Where N represents the number of simulation tests. When the variance coefficient β is less than the preset value, the statistical calculation and That is the risk level of the distribution network.

[0197] Therefore, using the information at the equipment level, the resilience level at the load level and the system level can be inferred; using meteorological forecast data for forward prediction and using the actual power grid data for backward correction, the latest available information can be used to achieve a rolling update of the risk assessment of the elastic distribution network.

[0198] The specific implementation described above is a preferred implementation of a rolling assessment method for operating risks of a flexible distribution network of the present invention, and is not intended to limit the specific implementation scope of the present invention. The scope of the present invention includes but is not limited to this specific implementation. All equivalent changes made in accordance with the shape and structure of the present invention are within the protection scope of the present invention.

Claims

1. A rolling assessment method for the operation risk of a flexible distribution network, characterized in that: The following steps are involved: Step S1: According to the frequency and duration of typhoon extreme disasters and the degree of damage to the distribution network, a probability distribution model characterizing the intensity and duration of typhoon extreme disasters is established; Step S2: According to the failure mechanism of each component in the distribution network system, a relationship model between the intensity of typhoon extreme disasters and the failure probability of distribution network components is established; Step S3: According to the failure modes, independence and correlation of the distribution network components under the typhoon extreme disaster, a model of its impact on the distribution network is established and applied to the disaster simulation; Step S4: Generate distribution network fault scenarios through component vulnerability curves to dynamically evaluate the distribution network function level; In step S4, the method for generating a distribution network fault scenario includes the following steps: S41. Obtain disaster paths, intensity and time series changes; S42. Grid processing is performed on the power grid on the disaster path, and each grid corresponds to a typhoon extreme disaster intensity; S43, calculating the intensity of typhoon extreme disasters in each time period in each grid according to the distance between the grid and the disaster center; S44, combining the fragility curves of the components in the grid, and calculating the failure probability of each component; S45, using a simulation method to simulate whether each component fails during the disaster period, obtain possible failure scenarios, and select and reduce the scenarios; Dynamic evaluation of the distribution network function level includes the following steps: The power supply restoration of the distribution network during a disaster is simulated, where the disaster repair time T follows the exponential distribution with the same parameters, that is, Wherein, μ is the parameter of exponential distribution, and the expected value of repair time E(T) = μ; When the number of damaged lines in the distribution network is greater than the number of repair teams that can be deployed, a greedy algorithm that guarantees single-step optimality is used to arrange the repair sequence; Obtain the latest weather forecast data and real-time power grid data to determine the ultimate indicators that affect the resilience of the distribution network, including: the system maximum load loss rate indicator, the system full power restoration time indicator, and the system full power restoration time indicator; In each round of rolling evaluation, the largest variance coefficient among the three ultimate indicators is selected as the basis for convergence judgment using the Monte Carlo simulation algorithm; Utilize the information at the equipment level to infer the resilience level at the load level and the system level; use meteorological forecast data for forward prediction, use the actual power grid data for backward correction, and use the latest information to achieve rolling update of the risk assessment of the elastic distribution network.

2. A rolling assessment method for operation risk of a flexible distribution network according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: extract key parameters that characterize the intensity of disaster-causing factors and establish their corresponding probability distribution models; Step S12: converting the extracted key parameters into dynamic disaster scenarios based on meteorological or geographical models; Step S13: construct a disaster simulation generator to generate dynamic simulated disaster scenarios based on the probability distribution of multi-dimensional key parameters.

3. A rolling assessment method for operation risk of a flexible distribution network according to claim 2, characterized in that: The key parameters include the annual typhoon occurrence rate Λ, the minimum distance from the research point D min , typhoon center pressure difference ΔP, typhoon moving speed V T And the typhoon's moving direction A T ; in: The probability distribution function corresponding to the annual typhoon occurrence rate Λ is: Among them, λ>0 is the parameter to be estimated in the Poisson distribution; Minimum distance D from the research point min The probability distribution function is: Among them, θ1<θ2 is the parameter to be estimated in the uniform distribution; The probability distribution function of the typhoon center pressure difference ΔP is: Among them, μ and σ 2 is the parameter to be estimated in the lognormal distribution; Typhoon moving speed V T The probability distribution function is: Among them, μ and σ 2 is the parameter to be estimated in the lognormal distribution; Typhoon moving direction A T The probability distribution function is: Among them, μ1 and is the parameter to be estimated for the first normal distribution, μ2 and is the parameter to be estimated for the second normal distribution, and ρ is the mixing degree parameter of the two normal distributions to be estimated.

4. A rolling assessment method for operation risk of a flexible distribution network according to claim 2, characterized in that: Step S12 includes the following steps: A typhoon wind field and filling model is established to describe the properties of typhoon disaster factors, and the Batts typhoon model is used in typhoon wind field simulation.

5. A rolling assessment method for operation risk of a flexible distribution network according to claim 2, characterized in that: The construction of the typhoon simulation generator in step S13 includes the following steps: Step 1: Generate the coastline equation near the study point; Step 2: Generate a simulated circle with the research point as the center and the given distance as the radius; Step 3: Generate the minimum distance D between the typhoon path and the research point based on uniform probability distribution sampling min ; Step 4: Generate the azimuth angle A of the typhoon movement based on the mixed normal probability distribution sampling T ; Step 5: Determine the typhoon's moving path based on the minimum distance and the typhoon's moving direction. The typhoon's moving path is a spherical great circle passing through the center of the earth; Step 6: Determine the typhoon's offshore generation point and departure point based on the intersection of the typhoon's moving path and the simulation circle; Step 7: Determine the typhoon’s landing point based on the intersection of the typhoon’s moving path and the coastline equation; Step 8: Generate the moving speed V of the typhoon center based on the log-normal probability distribution sampling T ; Step 9: With hourly time resolution, determine the longitude and latitude position of the typhoon center hour by hour based on the moving speed and moving path of the typhoon center to form the dynamic moving path of the typhoon; Step 10: Generate the typhoon's sea center pressure difference ΔP by sampling according to the log-normal probability distribution; Step 11: Determine the hourly typhoon center pressure difference based on the Batts typhoon filling model, and detect whether the typhoon has landed. After landing, correct the land center pressure difference according to the attenuation formula; Step 12: Determine the hourly maximum average wind speed of the typhoon based on the Batts typhoon wind field model, and then determine the typhoon level.

6. A rolling assessment method for operation risk of a flexible distribution network according to claim 1, characterized in that: In step S2, the relationship model is characterized by the distribution network component failure repair time r: Among them, M represents the catastrophic intensity of extreme events, M crit It represents the critical disaster intensity determined by the nature of the disaster; r0 represents the average repair time of the components when the system is operating normally; k represents the rate coefficient of increase in repair time, which is obtained by statistical analysis of historical operation data.

Images