Resource collaborative elastic planning method considering uncertainty of new energy under typhoon disasters

Abstract

The invention discloses a resource cooperation elastic planning method considering new energy uncertainty under typhoon disasters, and the method comprises the steps: building a time-varying fault rate of a line through typhoon disaster disturbance and a line fault rate model, recognizing a fragile line, and building an evolution model considering cascading faults under the typhoon disasters; the method comprises the following steps: generating a wind and light output scene by utilizing a mixed Copula function and combining multivariate normal distribution, and then screening out a wind and light scene set by adopting a synchronous return band reduction method; the flexibility of the system is improved through flexible resource collaborative planning such as wind power photovoltaic energy storage and line reinforcement; and solving the constructed model through column and constraint generation and a Big-M method to obtain a collaborative planning scheme of various flexible resources. By utilizing the method, the problem of safe and reliable operation of a power system caused by typhoon disasters and uncertainty of new energy output can be solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system planning, and specifically relates to a resource collaborative elastic planning method taking into account the uncertainty of new energy sources under typhoon disasters. Background Art

[0002] During the critical period of energy transformation, the dual carbon goals are achieved by building a new power system with new energy as the main body. However, the power system under energy transformation shows characteristics such as strong uncertainty and complexity, making the power system more vulnerable when hit by extreme natural disasters, posing great risks to the safe and reliable operation of the power system.

[0003] Among them, typhoon disasters have spatiotemporal characteristics, which makes the time and location of failures highly uncertain. To deal with uncertainty, current research mainly uses random regularization and robust regularization. Random regularization assumes that random variables follow a certain probability distribution, but due to the low probability of extreme disasters, there is a deviation between the probability distribution fitted by historical data and the true distribution, so a robust planning method is used. Existing technologies usually use Monte Carlo simulation to generate system failure scenarios under disasters, but the Monte Carlo simulation method requires a lot of computing resources and time to ensure the accuracy of the results, and the calculation speed is slow when the number of components is large.

[0004] In addition, due to the natural changes in wind power and light, the output of wind and solar power generation has significant uncertainty, which may lead to frequency fluctuations, voltage instability and reduced power supply reliability of the power system. Therefore, how to accurately simulate the changes in wind and solar output in planning and scheduling has become a key issue in the study of modern power systems. Traditional wind and solar output modeling methods mostly rely on simple statistical models based on assumptions or independent and identically distributed random sampling methods. These methods usually ignore the interdependence and nonlinear correlation between different wind and solar power sources, resulting in limited accurate prediction capabilities of the model for wind and solar power generation. In addition, with the increase in the proportion of wind and solar power generation, the traditional scenario generation method has a large number of possible scenarios to process, and its computational cost and time complexity have also increased significantly, making it difficult to meet the real-time scheduling requirements of large-scale systems. Against this background, the present invention adopts a method combining a hybrid Copula function with a multivariate normal distribution, which can more effectively capture the complex correlations between different wind and solar power sources, thereby improving the accuracy and diversity of scenario generation. By introducing the synchronous loop reduction method for scenario screening, the dimension of the uncertainty scenario set can be effectively reduced, the amount of calculation can be reduced, and the representativeness and diversity of the scenario set can be ensured. This enables the method to provide more accurate and efficient wind and solar power output simulation when dealing with large-scale wind and solar power generation systems.

[0005] In addition, in the face of typhoon disasters and the uncertainty of renewable energy output, traditional power system planning methods usually focus on the scheduling and optimization of a single resource, such as relying solely on a single means such as energy storage systems or line reinforcement to deal with disasters. However, this method is often unable to fully cope with the complexity of disasters and the volatility of renewable energy output, especially when large-scale disasters occur, the system's resilience and recovery capabilities are insufficient. Therefore, the scheduling method of a single resource has obvious limitations in improving the overall risk resistance of the system, and cannot achieve rapid and efficient response before and after disasters.

[0006] In summary, the uncertainty of fault states and the uncertainty of renewable energy output caused by typhoon disasters pose great challenges to the safe operation of power systems. After obtaining the failure rate of each component under typhoon disasters, how to quickly and efficiently characterize the fault state of the component at each moment is an urgent problem to be solved; in addition, it is urgent to effectively capture the complex correlations between different wind and solar power sources to improve the accuracy and diversity of scenario generation; and by introducing appropriate scenario reduction methods to screen the scenarios, the dimension of the uncertainty scenario set is reduced, the amount of calculation is reduced, and a more accurate and efficient wind and solar output simulation is provided. In order to reduce the impact of extreme disasters, the uncertainty of fault states and renewable energy output caused by typhoon disasters needs to be introduced into the power system resilience planning model, and the flexible resources in the source, grid and load need to be used to plan the power system for resilience to improve the resilience level of the system, so that the power system has the ability to defend before disasters and recover after disasters, and reduce load losses under extreme disasters. Summary of the invention

[0007] In view of the above-mentioned deficiencies in the prior art, the present invention provides a resource collaborative elastic planning method taking into account the uncertainty of new energy under typhoon disasters to solve the problems caused by the uncertainty of fault states and the uncertainty of new energy output caused by typhoon disasters to the safe operation of the power system.

[0008] In order to achieve the above-mentioned invention object, the technical solution adopted by the present invention is: a resource collaborative elastic planning method taking into account the uncertainty of new energy under typhoon disasters, comprising the following steps:

[0009] S1. Based on the typhoon disaster disturbance model and the line failure rate model, a fault state uncertainty model is constructed to obtain the time-varying fault state of the line, and the final state of the line is determined by simulating the system cascading failure process;

[0010] S2, by using the mixed Copula model combined with multivariate normal distribution to generate wind and solar output scenarios, and by using the synchronous back-strip reduction method to reduce the sampled scenarios, a set of wind and solar scenarios is screened out;

[0011] S3. Based on the time-varying fault status of the line and the wind and solar scenario set, a two-stage robust elastic planning model based on a defense-attack-defense structure is constructed to consider typhoon disasters and new energy uncertainties, and its objective function and constraints are determined based on the collaborative planning of flexibility resources;

[0012] S4. Based on the determined objective function and constraints, the column and constraint generation method and the Big-M method are used to solve the two-stage robust elastic planning model, determine the fault recovery strategy, and then obtain a collaborative planning scheme for multiple flexibility resources.

[0013] Furthermore, in step S1, the typhoon disaster disturbance model is expressed as:

[0014] R max =exp(-0.1239Δp 0.6003 +5.1034)

[0015] In the formula, R max represents the maximum wind speed radius of the typhoon, Δp represents the pressure difference between the typhoon center and the area not affected by the typhoon, among which the wind speed V in the typhoon affected area r It is expressed as:

[0016]

[0017] In the formula, represents the maximum wind speed of the typhoon, r represents the wind speed in the affected area and the distance from the area to the typhoon center, and x represents the intensity attenuation coefficient;

[0018] The line fault probability model is expressed as:

[0019]

[0020] In the formula, p ij (s) represents the failure rate of the ijth branch when the typhoon intensity is s, p l (s) represents the failure rate of the conductor, p p (s) represents the failure rate of the pole, m ij Indicates the number of wires corresponding to the ijth branch, y ij Indicates the number of poles corresponding to the ijth branch;

[0021] The fault state uncertainty model is expressed as:

[0022]

[0023] Where U / u represents the fault state set / fault state of the line under typhoon disaster, p ij,t represents the failure rate of line ij at time t, z ij,t and u ij,tBoth represent 0-1 decision variables, z ij,t Indicates the fault status of line ij under the typhoon attack at time t. If the value is 0, it means that ij fails after being attacked by the typhoon. Otherwise, the value is 1. ij,t represents the actual fault state of line ij at time t after the typhoon attack. Its value is 0, indicating that the line is in a fault state, and vice versa. Γ represents the budget value of information entropy, γ represents the time of typhoon attack, T s It indicates the time when the faulty line is repaired and restored to normal state, ij indicates the line, L indicates the line set, t indicates the time, and T indicates the entire period.

[0024] Furthermore, in step S1, the method for simulating the system cascading failure process is specifically as follows:

[0025] S11. Obtain t based on line fault probability and fault state uncertainty model k Faulty lines caused by typhoon disasters at all times;

[0026] S12, determine the system status, obtain the subsystem status by identifying the system connectivity, and make power balance adjustments to ensure the power balance of each subsystem. When power imbalance occurs, adjust the output of the generator and various flexible resources to ensure the power balance of the system;

[0027] S13, calculating the power flow of each subsystem, comparing the power flow of each line with the upper limit of the normal value of the power flow, and determining whether there is an overloaded line;

[0028] If yes, the line is overloaded and the process goes to step S14;

[0029] The overload control is to control the line overload through the dispatching of generators and various flexible resources and load reduction;

[0030] If not, the simulation ends after recording the system status;

[0031] S14, calculating the overload line fault probability, using a sampling method to obtain a fault set, and determining whether it is an empty set;

[0032] If yes, the simulation ends after recording the system status;

[0033] If not, the fault set is treated as a new fault;

[0034] Among them, the overload line failure probability p lf for:

[0035]

[0036] In the formula, F l represents the active power flow of the lth line, represents the upper limit of the normal value of the power flow on the lth line, F max,l represents the power flow limit value of the lth line,

[0037] Furthermore, in step S2, the method for generating a wind-solar output scenario using a mixed Copula function combined with a multivariate normal distribution is specifically as follows:

[0038] S21. Estimate the marginal distribution function F of wind power and photovoltaic power output respectively w (x w ) and F p (x p ), use the kernel density estimation method to smooth the historical data and obtain the marginal distribution function of wind and solar power output;

[0039] S22, estimate the parameters of the mixed Copula model by EM method;

[0040] S23. Use the fitted marginal distribution function F w (x w ) and F p (x p ) and the estimated mixed Copula model to construct the joint distribution of wind and solar power output;

[0041] S24. Based on the constructed joint distribution of wind and solar power output, the time series characteristics of wind power and photovoltaic power output are further considered, and a joint distribution model with time series dynamics is established based on the hybrid Copula model;

[0042] S25, based on the Markov chain Monte Carlo sampling method, samples are extracted from the constructed joint distribution model with time series dynamics;

[0043] S26. Based on the obtained MCMC samples, an inverse cumulative distribution function is used to perform an inverse transformation, thereby converting uniformly distributed random numbers into actual output scenarios of wind power and photovoltaic power.

[0044] Furthermore, in step S2, the method of reducing the sampled scenes by the synchronous tape-back reduction method is specifically as follows:

[0045] S27, setting the initial scene set to D, the reduced scene set to J, the initialization set J to an empty set, and the initial value of the iteration number u to 1;

[0046] S28, determine the scene S that is cut at the u-th iteration u , so that the following formula is minimized;

[0047]

[0048] In the formula, p i Indicates scene Si The probability of J u is the set J at the uth iteration, c(S i ,S j ) represents scene S i and scene S j The probability distance between i ,S j )=PS i -S j P 2 ;

[0049] S29. Remove scene S from set D u , and add it to set J, increasing the number of iterations by 1;

[0050] S210, scene S to be removed u The probability is assigned to the scene with the closest probability distance, and the probability of the nearest scene is updated;

[0051] Among them, the probability of updating is:

[0052]

[0053] In the formula, J(j) represents all the scenes S in the discarded set j Alternative scenario, p' j For the final scene that remains j The probability of

[0054] S211, repeating steps S28 to S210 until the scenes in the set D are scene data that need to be retained, thereby obtaining a screened scenery scene set.

[0055] Further, in step S3, the two-stage robust elastic programming model includes a first stage and a second stage;

[0056] The first stage is the first-layer defense stage, and its planning goal is to minimize the on-duty installation cost of flexibility resources and line reinforcement cost. The first stage determines the planning decision variables and passes them to the second stage;

[0057] The second stage includes the attack stage of the second layer and the defense stage of the third layer, and its planning goal is to minimize load shedding and power system operating costs; among them, the attack stage of the second layer is the disaster occurrence stage, which determines the worst fault scenario based on the time-varying failure rate and the identified vulnerable lines under the known planning decision, and transmits it to the third layer; the third layer minimizes load shedding and power operating costs under the determined worst fault scenario, and then uses the planning decision variables of the first stage to determine the fault recovery strategy of the power system, and feeds back the worst fault scenario and fault recovery strategy of the second layer to the first stage to adjust the determined planning decision variables.

[0058] Furthermore, the objective function of the two-stage robust elastic programming model is expressed as:

[0059]

[0060] In the formula, xl Represents a 0 / 1 decision variable, indicating the state of line reinforcement. If the value is 1, it is reinforced, otherwise it is not reinforced. Both represent 0 / 1 variables, indicating whether energy storage, wind power and photovoltaic are connected to the node. If the value is 1, it is connected to node i, and if it is 0, it is not connected to node i. E ,K w ,K p Represent the installed capacity of energy storage, wind power and photovoltaic configuration respectively, C L,l represents the reinforcement cost of the lth transmission line, x l represents a 0 / 1 decision variable, indicating the state of line reinforcement. If the value is 1, it is reinforced, otherwise it is not reinforced. C p VP represents the daily installation cost of energy storage, wind power and photovoltaic power, etc. i D represents the load shedding amount of node i, P ij Represents the active power of the branch, P i G represents the active output of generator set i, P i pw , P i pv , P i ESS Respectively represent the active output of installed wind power, photovoltaic power and energy storage, V i represents the voltage at node i, C D,i represents the load shedding penalty cost of node i, C o represents the system operating cost, C ESS , C pw , C pv They represent the installation costs of energy storage, wind power and photovoltaic power, r represents the discount rate, s represents the number of years of discount, C G,i represents the operating cost of generator set i, C e,op , C w,op , C p,op are the operating cost coefficients of energy storage, wind power and photovoltaic power, Ω L ,Ω G ,Ω N They represent the collection of transmission lines, generator sets and load nodes respectively, Y is the total number of days in a year, and u / U represents a certain fault state / fault state set of the power system under the intermediate layer typhoon disaster.

[0061] Furthermore, in the two-stage robust elastic programming model, the constraints of the first stage are:

[0062]

[0063] Where N L ,N ESS ,N pw ,N pv Respectively represent the maximum number of line reinforcements, the maximum number of energy storage, wind power and photovoltaic installations, Indicates the maximum discharge power of energy storage, K E represents the installed capacity of energy storage, K E Respectively represent the upper and lower limits of energy storage installed capacity, K w represents the installed capacity of wind power, K w are the upper and lower limits of wind power installed capacity, K p represents the photovoltaic installed capacity, K p are the upper and lower limits of photovoltaic installed capacity respectively; δ pw , δ pv They represent the normalized wind power and photovoltaic power output sequences respectively. The actual output of wind and photovoltaic power is obtained by their normalized output time series and planned installed capacity.

[0064] Represent the reactive power output of wind power and photovoltaic power, Represent the maximum values ​​of wind power and photovoltaic power factor tangent angle respectively.

[0065] Furthermore, in the two-stage robust elastic planning model, the constraints of the second stage include line state constraints, power balance constraints, line flow constraints, load reduction constraints, generator output constraints, energy storage operation constraints and node voltage constraints;

[0066] The line state constraint is expressed as:

[0067]

[0068] In the formula, α ij,t represents a 0-1 decision variable, indicating the operating status of the line under typhoon attack after reinforcement. If the line is in normal operation, α ij,t =1, otherwise α ij,t =0, indicating that the lines reinforced in the first stage will not fail due to typhoon attacks, u ij,t represents the actual fault status of line ij at time t after the typhoon attack;

[0069] The power balance constraint is expressed as:

[0070]

[0071] In the formula, represents the set of lines that flow out and into power from node n, P ij,t , Q ij,t represents the active power and reactive power of line ij at time t, P i D is the load of each node;

[0072] The line power flow constraint is expressed as:

[0073]

[0074] Where P ij,max , Q ij,max They represent the upper limits of active and reactive power flows of line ij respectively;

[0075] The load reduction constraint is expressed as:

[0076]

[0077] In the formula, Respectively represent the maximum active and reactive power of each node load;

[0078] The output constraint of the generator set is expressed as:

[0079]

[0080] In the formula, They represent the upper and lower limits of the active output of the i-th generator set, They represent the upper and lower limits of reactive output of the i-th generator set respectively;

[0081] The energy storage operation constraints are expressed as:

[0082]

[0083] In the formula, represents the energy storage charge at time t, η E represents the discharge efficiency of energy storage, S oc They are the upper and lower limits of the state of charge of the energy storage respectively;

[0084] The node voltage constraint is expressed as:

[0085]

[0086] Where V i,t is the voltage at node i, V NIndicates the rated voltage, M is a parameter with a larger value, V i min and V i max Respectively represent the upper and lower limits of the node voltage.

[0087] Furthermore, the step S4 includes the following sub-steps:

[0088] S41. Convert the objective function and its constraints of the two-stage robust elastic programming model into a compact form and decompose it into a main problem and sub-problems;

[0089] The expression of the main problem is generated based on the column and constraint generation method, and the defense strategy is obtained by solving it; the sub-problem is a defense strategy determined based on the main solution problem, which determines the worst fault scenario and determines the mathematical model of the fault recovery strategy through the corresponding scheduling scheme;

[0090] S42. Solve the main problem, obtain the configuration strategy of line reinforcement and flexibility resources as the defense strategy, and update the lower bound value L B ;

[0091] S43, solving the sub-problem according to the defense strategy, determining the most serious line fault state and its fault recovery strategy under the worst fault scenario, and then updating the upper bound value;

[0092] Among them, the sub-problem is a two-level optimization problem, and its solution is:

[0093] The dual problem of inner layer minimization is solved by using strong duality theory, and then it is transformed into a single-layer problem, which is solved by linear processing of the single-layer problem using the Big-M method.

[0094] S43. Judgment Is it established? represents the convergence threshold;

[0095] If so, the fault recovery strategy is output, and then a collaborative planning scheme for multiple flexible resources is obtained;

[0096] If not, the worst failure scenario obtained by solving the sub-problem is passed to the main problem, and the process returns to step S42.

[0097] The beneficial effects of the present invention are:

[0098] 1) In view of the spatiotemporal uncertainty of typhoon disasters, the method of the present invention constructs a time-varying failure rate model of the line based on the typhoon disaster disturbance and line failure rate model, combines the information entropy method to effectively identify vulnerable lines, and considers the system cascading failure process simulation to determine the final state of the line, thereby accurately characterizing the failure state of system components in the disaster, effectively overcoming the high requirements of the traditional Monte Carlo simulation method in terms of computing resources and time, and making the prediction of the fault state more efficient and accurate.

[0099] 2) The method of the present invention generates wind and solar power output scenarios by adopting a hybrid Copula function combined with a multivariate normal distribution, and screens out the final uncertainty scenario set through the synchronous loop reduction method, thereby avoiding the over-simplification of wind and solar power output scenarios in traditional methods, and can more accurately describe the nonlinear correlation of wind and solar power output and its complex interaction with other factors (such as load demand, meteorological conditions, etc.). Compared with a single multivariate normal distribution or a simple Copula model, the hybrid Copula function can better capture the complex dependencies between different wind and solar power sources, thereby improving the accuracy and diversity of scenario generation. In addition, the synchronous loop reduction method is used for scenario screening to avoid the problems of scenario redundancy and low computational efficiency that may occur in traditional Monte Carlo simulations. The synchronous loop reduction method reduces the number of scenarios through effective scenario compression and optimization, while retaining the representativeness and uncertainty characteristics of the scenario set, thereby greatly reducing the computational complexity and improving the solution speed. This method shows better computational efficiency and more accurate risk assessment capabilities than traditional methods when dealing with large-scale wind and solar power generation uncertainties, especially in systems with a high proportion of wind and solar power generation.

[0100] 3) The method of the present invention improves the resilience of the power system in disasters through the coordinated planning of multiple flexible resources such as wind power, photovoltaics, energy storage and line reinforcement, and constructs a two-stage three-layer defense-attack-defense model for pre-disaster prevention and post-disaster recovery. This model significantly improves the system's risk resistance and recovery speed in the face of complex disasters through the efficient scheduling of flexible resources, making up for the limitations of single resource scheduling in traditional methods. It can not only effectively improve the resilience of the power system, but also reduce the economic losses caused by typhoon disasters.

[0101] 4) The method of the present invention is not only more accurate and efficient in dealing with disaster uncertainty and volatility of renewable energy output, but also significantly improves the resilience and economy of the power system through the integration of multiple flexibility resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0102] Figure 1 A flow chart of the resource collaborative elastic planning method taking into account the uncertainty of new energy under typhoon disasters provided by the present invention.

[0103] Figure 2 The present invention provides a resource coordination elastic planning framework diagram that takes into account the uncertainty of new energy under typhoon disasters.

[0104] Figure 3 This is a flow chart for simulating the system cascading failure process under typhoon disasters provided by the present invention.

[0105] Figure 4 A schematic diagram of the scenery scene set generation process provided by the present invention.

[0106] Figure 5 This is a flow chart of the C&CG algorithm provided by the present invention. DETAILED DESCRIPTION

[0107] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0108] The embodiment of the present invention provides a resource coordination elastic planning method taking into account the uncertainty of new energy sources under typhoon disasters, such as Figure 1-2 As shown, the following steps are included:

[0109] S1. Based on the typhoon disaster disturbance model and the line failure rate model, a fault state uncertainty model is constructed to obtain the time-varying fault state of the line, and the final state of the line is determined by simulating the system cascading failure process;

[0110] S2, by using the mixed Copula model combined with multivariate normal distribution to generate wind and solar output scenarios, and by using the synchronous back-strip reduction method to reduce the sampled scenarios, a set of wind and solar scenarios is screened out;

[0111] S3. Based on the time-varying fault status of the line and the wind and solar scenario set, a two-stage robust elastic planning model based on a defense-attack-defense structure is constructed to consider typhoon disasters and new energy uncertainties, and its objective function and constraints are determined based on the collaborative planning of flexibility resources;

[0112] S4. Based on the determined objective function and constraints, the column and constraint generation method and the Big-M method are used to solve the two-stage robust elastic planning model, determine the fault recovery strategy, and then obtain a collaborative planning scheme for multiple flexibility resources.

[0113] In step S1 of the embodiment of the present invention, the typhoon intensity parameters in the typhoon disaster disturbance model include maximum wind speed, wind direction, moving speed and maximum wind speed radius; in this embodiment, the Batts model is used to simulate the typhoon wind speed and wind direction changes, and the typhoon disaster disturbance model is expressed as:

[0114] R max =exp(-0.1239Δp 0.6003 +5.1034)(1)

[0115] In the formula, R max represents the maximum wind speed radius of the typhoon, Δp represents the pressure difference between the typhoon center and the area not affected by the typhoon, among which the wind speed V in the typhoon affected area r It is expressed as:

[0116]

[0117] In the formula, represents the maximum wind speed of the typhoon, r represents the wind speed in the typhoon-affected area and the distance from that area to the center of the typhoon, and x represents the intensity attenuation coefficient.

[0118] In this implementation, according to the reliability theory, when the conductor tensile strength / pole bending strength S l / p Less than conductor cross-sectional stress / pole cross-sectional bending moment L g / T Therefore, the failure rate of conductors and poles is:

[0119]

[0120] When all the wires and poles between two nodes are in normal operation, the corresponding branch can operate normally. Therefore, the branch failure probability model needs to consider connecting the wires and poles in series. The line failure probability model is expressed as:

[0121]

[0122] In the formula, p ij (s) represents the failure rate of the ijth branch when the typhoon intensity is s, p l (s) represents the failure rate of the conductor, p p (s) represents the failure rate of the pole, m ij Indicates the number of wires corresponding to the ijth branch, y ij Indicates the number of poles corresponding to the ijth branch.

[0123] Due to the spatiotemporal characteristics of typhoon disasters, the fault state of the power system is uncertain. The information entropy method can be used to characterize the uncertainty of the fault state. The fault state of the line is mainly determined based on the time-varying failure rate of each line under the typhoon disaster and the fault repair time of the system. The constructed fault state uncertainty model is:

[0124]

[0125] Where U / u represents the fault state set / fault state of the line under typhoon disaster, p ij,t represents the failure rate of line ij at time t, z ij,t and u ij,t Both represent 0-1 decision variables, z ij,t Indicates the fault status of line ij under the typhoon attack at time t. If the value is 0, it means that ij fails after being attacked by the typhoon. Otherwise, the value is 1. ij,t represents the actual fault state of line ij at time t after the typhoon attack. Its value is 0, indicating that the line is in a fault state, and vice versa. Γ represents the budget value of information entropy, γ represents the time of typhoon attack, T s It indicates the time when the faulty line is repaired and restored to normal state, ij indicates the line, L indicates the line set, t indicates the time, and T indicates the entire period.

[0126] In the above formula, formula (7) indicates that each line has only one fault in the entire period T; formula (8) indicates that the fault state of the line is determined by γ and T s It was jointly decided that normal conditions would be restored only after the line was repaired.

[0127] In step S1 of this embodiment, under the influence of a typhoon, the failure of some lines may cause a power flow shift, which in turn causes other lines to overload. If it is not adjusted in time, a chain failure may occur. Figure 3 The method for simulating the system cascading failure process is shown as follows:

[0128] S11. Obtain t based on line fault probability and fault state uncertainty model k Faulty lines caused by typhoon disasters at all times;

[0129] S12, determine the system status, obtain the subsystem status by identifying the system connectivity, and make power balance adjustments to ensure the power balance of each subsystem. When power imbalance occurs, adjust the output of the generator and various flexible resources to ensure the power balance of the system;

[0130] S13, calculating the power flow of each subsystem, comparing the power flow of each line with the upper limit of the normal value of the power flow, and determining whether there is an overloaded line;

[0131] If yes, the line is overloaded and the process goes to step S14;

[0132] The overload control is to control the line overload through the dispatching of generators and various flexible resources and load reduction;

[0133] If not, the simulation ends after recording the system status;

[0134] S14, calculating the overload line fault probability, using a sampling method to obtain a fault set, and determining whether it is an empty set;

[0135] If yes, the simulation ends after recording the system status;

[0136] If not, the fault set is treated as a new fault;

[0137] Among them, the overload line failure probability p lf for:

[0138]

[0139] In the formula, F l represents the active power flow of the lth line, represents the upper limit of the normal value of the power flow on the lth line, F max,l represents the power flow limit value of the lth line, Here ε is taken as 1.4.

[0140] In step S2 of the embodiment of the present invention, a joint distribution model of wind and solar power is established based on historical data of wind power and photovoltaic power, and a wind and solar power joint distribution sequence with dynamic correlation is further generated through multivariate normal distribution and covariance matrix; the MCMC sampling method is used to generate random distribution numbers of wind and solar power output scenarios, and the corresponding wind and solar power output scenarios are obtained through the inverse transformation method; Figure 4 The process of generating wind and solar power output scenarios using a hybrid Copula function combined with multivariate normal distribution and synchronous backlash reduction method is shown.

[0141] Specifically, in step S2, the method for generating a wind-solar output scenario using a mixed Copula function combined with a multivariate normal distribution is as follows:

[0142] S21. Estimate the marginal distribution function F of wind power and photovoltaic power output respectively w (x w ) and F p (x p ), use the kernel density estimation method (KDE) to smooth the historical data and obtain the marginal distribution function of wind and solar power output;

[0143] Among them, the kernel density estimation function is defined as:

[0144]

[0145] In the formula, Represents sample X 1 ,X 2 ,…,X n The density function at any point x, h represents the bandwidth, and K(·) represents the kernel function;

[0146] S22, estimate the parameters of the mixed Copula model by EM (Expectation Maximization) method;

[0147] Among them, the mixed Copula model is a weighted combination of different Copula functions, and the parameter estimation steps are:

[0148] Initialize the mixture weight ω k and Copula parameter θ k ;

[0149] Step E: Calculate the posterior probability of each sample belonging to each Copula component;

[0150] Step M: Update the mixing weights and Copula parameters to maximize the log-likelihood function;

[0151] Check parameter changes or changes in log-likelihood until convergence conditions are met;

[0152] Returns the estimated mixed Copula parameters;

[0153] The mixed Copula model based on this is expressed as:

[0154]

[0155] In the formula, F w (x w ) is wind power output x w The marginal cumulative distribution function value of p (x p ) is the photovoltaic output x p The marginal cumulative distribution function value of k is the kth Copula function; ω k is the mixing weight of the kth Copula function; θ k is the parameter of the kth Copula function;

[0156] S23. Use the fitted marginal distribution F w (x w ) and F p (x p) and the estimated mixed Copula model to construct the joint distribution of wind and solar power output;

[0157] Among them, the joint distribution of wind and solar power output describes the dependency structure and joint probability between wind power and photovoltaic output, which is expressed as:

[0158] F w,p (x w ,x p )=C mix (F w (x w ),F p (x p ))(12)

[0159] S24. Based on the constructed joint distribution of wind and solar power output, the time series characteristics of wind power and photovoltaic power output are further considered, and a joint distribution model with time series dynamics is established based on the hybrid Copula model;

[0160] Specifically, the wind-solar coupling output also has a certain correlation in adjacent time sections. A joint distribution model with time series dynamics is established based on the mixed Copula. Multivariate normal distribution is used to describe the dynamic correlation between wind power output and photovoltaic output at different time steps. Assume that there are T time steps, and the wind power and photovoltaic output at each time step are random variables x w (t) and x p (t), then we can construct a two-dimensional multivariate normal distribution sequence:

[0161]

[0162] In the formula, μ represents the mean vector of each time step, ∑ represents the covariance matrix, which is used to represent the correlation between wind power and photovoltaic power. The estimation of the covariance matrix is ​​usually based on historical data or inferred by the correlation model;

[0163] S25, based on the Markov Chain Monte Carlo (MCMC) sampling method, samples are extracted from the constructed joint distribution model with time series dynamics;

[0164] Specifically, the MCMC method can effectively generate random samples from complex joint distributions, ensuring that the generated wind and solar power output scenarios meet the requirements of time series correlation and joint distribution. The MCMC sampling method is used to generate random distribution data of wind power and photovoltaic output. First, the sample points are initialized, and then new sample points are generated through the transition probability. Whether to accept the new sample is determined based on the acceptance probability. The above steps are repeated until the sample converges.

[0165] S26. Based on the obtained MCMC samples, the inverse cumulative distribution function (ICDF) is used for inverse transformation to convert the uniformly distributed random numbers into actual output scenarios of wind power and photovoltaic power.

[0166] In step S2 of the embodiment of the present invention, the basic idea of ​​the synchronous rewind reduction method is to minimize the probability distance between the scene set before reduction and the scene set finally retained. i and scene S j The probability distance between them is:

[0167] c(S i ,S j )=PS i -S j P 2 (14)

[0168] Specifically, the method of reducing the sampled scenes by the synchronous tape replay reduction method in step S2 is as follows:

[0169] S27, setting the initial scene set to D, the reduced scene set to J, the initialization set J to an empty set, and the initial value of the iteration number u to 1;

[0170] S28, determine the scene S that is cut at the u-th iteration u , so that the following formula is minimized;

[0171]

[0172] In the formula, p i Indicates scene S i The probability of J u is the set J at the uth iteration, c(S i ,S j ) represents scene S i and scene S j The probability distance between i ,S j )=PS i -S j P 2 ;

[0173] S29. Remove scene S from set D u , and add it to set J, increasing the number of iterations by 1;

[0174] S210, scene S to be removed u The probability is assigned to the scene with the closest probability distance, and the probability of the nearest scene is updated;

[0175] Among them, the probability of updating is:

[0176]

[0177] In the formula, J(j) represents all the scenes S in the discarded set j Alternative scenario, p' j For the final scene that remains j The probability of

[0178] S211, repeating steps S28 to S210 until the scenes in the set D are scene data that need to be retained, thereby obtaining a screened scenery scene set.

[0179] In step S3 of the embodiment of the present invention, when constructing Figure 2 In the model framework shown, typhoon disaster uncertainty refers to the uncertainty of the time and location of typhoon-induced faults; new energy uncertainty refers to the uncertainty of wind power and photovoltaic output; pre-disaster prevention is a series of measures taken before the typhoon disaster occurs, through the planning of various flexibility resources, to enhance the ability to resist disasters and reduce the losses caused by typhoon disasters; post-disaster recovery is a series of actions and measures taken after the typhoon disaster to restore normal life and production in the affected areas as soon as possible, including the participation of maintenance teams in repairing faulty lines and dynamic scheduling of various flexibility resources to achieve rapid recovery of loads.

[0180] Specifically, the two-stage robust elastic planning model includes the first stage and the second stage;

[0181] The first stage is the first-level defense stage, and its planning goal is to minimize the daily installation costs of flexible resources (energy storage, wind power, photovoltaics, etc.) and line reinforcement costs; the first stage determines the planning decision variables (line reinforcement decisions and energy storage, wind power, photovoltaic installation locations and capacities) and passes them to the second stage;

[0182] The second stage includes the attack stage of the second layer and the defense stage of the third layer, and its planning goal is to minimize load shedding and power system operation costs; the attack stage of the second layer is the disaster occurrence stage, which determines the worst fault scenario based on the time-varying fault rate and the identified vulnerable lines under the known planning decision, and transmits it to the third layer; the third layer minimizes load shedding and power operation costs under the determined worst fault scenario, and then uses the planning decision variables of the first stage to determine the fault recovery strategy of the power system, and feeds back the worst fault scenario and fault recovery strategy of the second layer to the first stage to adjust the determined planning decision variables. Finally, the most reasonable planning decision scheme is obtained through iteration.

[0183] In step S3 of the embodiment of the present invention, in the two-stage robust elastic planning model constructed, the system elasticity is improved through the coordinated planning of flexible resources such as wind power, photovoltaic energy storage and line reinforcement. The objective function of the constructed elastic planning model is the daily installation cost and line reinforcement cost of flexible resources such as energy storage, wind power, photovoltaics, etc. in the first stage, and the load shedding and system operation cost in the second stage; the objective function of the model is expressed as:

[0184]

[0185] In the formula, xl Represents a 0 / 1 decision variable, indicating the state of line reinforcement. If the value is 1, it is reinforced, otherwise it is not reinforced. Both represent 0 / 1 variables, indicating whether energy storage, wind power and photovoltaic are connected to the node. If the value is 1, it is connected to node i, and if it is 0, it is not connected to node i. E ,K w ,K p Represent the installed capacity of energy storage, wind power and photovoltaic configuration respectively, C L,l represents the reinforcement cost of the lth transmission line, x l represents a 0 / 1 decision variable, indicating the state of line reinforcement. If the value is 1, it is reinforced, otherwise it is not reinforced. C p VP represents the daily installation cost of energy storage, wind power and photovoltaic power, etc. i D represents the load shedding amount of node i, P ij Represents the active power of the branch, P i G represents the active output of generator set i, P i pw , P i pv , P i ESS Respectively represent the active output of installed wind power, photovoltaic power and energy storage, V i represents the voltage at node i, C D,i represents the load shedding penalty cost of node i, C o represents the system operating cost, C ESS , C pw , C pv They represent the installation costs of energy storage, wind power and photovoltaic power, r represents the discount rate, s represents the number of years of discount, C G,i represents the operating cost of generator set i, C e,op , C w,op , C p,op are the operating cost coefficients of energy storage, wind power and photovoltaic power, Ω L ,Ω G ,Ω NThey represent the collection of transmission lines, generator sets and load nodes respectively, Y is the total number of days in a year, and u / U represents a certain fault state / fault state set of the power system under the intermediate layer typhoon disaster.

[0186] In the two-stage robust resilience planning model of this embodiment, before a typhoon disaster occurs, the line is reinforced, and the ESS, wind power, and photovoltaic power are optimally configured to reduce the scale of system failures and load losses, thereby enhancing the resilience of the power system. The constraints of the first stage are:

[0187]

[0188]

[0189] Where N L ,N ESS ,N pw ,N pv Respectively represent the maximum number of line reinforcements, the maximum number of energy storage, wind power and photovoltaic installations, Indicates the maximum discharge power of energy storage, K E represents the installed capacity of energy storage, K E Respectively represent the upper and lower limits of energy storage installed capacity, K w represents the installed capacity of wind power, K w are the upper and lower limits of wind power installed capacity, K p represents the photovoltaic installed capacity, K p are the upper and lower limits of photovoltaic installed capacity respectively; δ pw , δ pv They represent the normalized wind power and photovoltaic power output sequences respectively. The actual output of wind and photovoltaic power is obtained by their normalized output time series and planned installed capacity.

[0190] Represent the reactive power output of wind power and photovoltaic power, Represent the maximum values ​​of wind power and photovoltaic power factor tangent angle respectively.

[0191] In the two-stage robust elastic planning model of the embodiment of the present invention, the constraints of the second stage include line state constraints, power balance constraints, line flow constraints, load reduction constraints, generator output constraints, energy storage operation constraints and node voltage constraints;

[0192] The line state constraint is expressed as:

[0193]

[0194] In the formula, α ij,t represents a 0-1 decision variable, indicating the operating status of the line under typhoon attack after reinforcement. If the line is in normal operation, α ij,t =1, otherwise α ij,t =0, indicating that the lines reinforced in the first stage will not fail due to typhoon attacks, u ij,t represents the actual fault state of line ij at time t after the typhoon attack; Formula (28) indicates that the line reinforced in the first stage will not fail due to the typhoon attack.

[0195] The power balance constraint is expressed as:

[0196]

[0197] In the formula, represents the set of lines that flow out and into power from node n, P ij,t , Q ij,t represents the active power and reactive power of line ij at time t, P i D is the load of each node;

[0198] The line power flow constraint is expressed as:

[0199]

[0200] Where P ij,max , Q ij,max They represent the upper limits of active and reactive power flows of line ij respectively;

[0201] The load reduction constraint is expressed as:

[0202]

[0203] In the formula, Respectively represent the maximum active and reactive power of each node load;

[0204] The output constraint of the generator set is expressed as:

[0205]

[0206] In the formula, They represent the upper and lower limits of the active output of the i-th generator set, They represent the upper and lower limits of reactive output of the i-th generator set respectively;

[0207] The energy storage operation constraints are expressed as:

[0208]

[0209] In the formula, represents the energy storage charge at time t, η E represents the discharge efficiency of energy storage, S oc They are the upper and lower limits of the state of charge of the energy storage respectively;

[0210] The node voltage constraint is expressed as:

[0211]

[0212] Where V i,t is the voltage at node i, V N Indicates the rated voltage, M is a parameter with a larger value, V i min and V i max Respectively represent the upper and lower limits of the node voltage.

[0213] In step S4 of the embodiment of the present invention, the constructed two-stage robust elasticity improvement model is solved by using C&CG and Big-M methods. Figure 5 A flow chart of the C&CG algorithm is shown, specifically:

[0214] S41. Convert the objective function and its constraints of the two-stage robust elastic programming model into a compact form and decompose it into a main problem and sub-problems;

[0215] The expression of the main problem is generated based on the column and constraint generation method, and the defense strategy is obtained by solving it; the sub-problem is a defense strategy determined based on the main solution problem, which determines the worst fault scenario and determines the mathematical model of the fault recovery strategy through the corresponding scheduling scheme;

[0216] Specifically, the objective function and constraints of the two-stage robust elastic programming model are transformed into a compact form and expressed as:

[0217]

[0218] st Ax≤b (42)

[0219]

[0220] Where: x is the set of decision variables in the first stage, y is the set of decision variables in the second stage; A, D, E, F, G, H, I are coefficient matrices; c, d, b, f, g, h are constant vectors; Ω(x, u) is the feasible domain of the second stage decision variable y; Formula (42) contains the constraints (20) to (27), and Formula (43) contains the second stage constraints (28) to (40), which decomposes the original problem into the main problem and sub-problems to achieve alternating iterative solutions;

[0221] Based on the C&CG algorithm, the expression of the main problem is:

[0222]

[0223] In the formula, y l is the solution of the subproblem obtained in the lth iteration; u l is the value obtained in the worst scenario in the lth iteration, that is, the most serious line fault state; η is the cutting plane value;

[0224] Based on the defense strategy determined by the main problem, the sub-problem will determine the "worst" scenario and update the recovery strategy by taking the corresponding scheduling scheme. The mathematical model of the sub-problem can be expressed as:

[0225]

[0226] S42. Solve the main problem, obtain the configuration strategy of line reinforcement and flexibility resources as the defense strategy, and update the lower bound value L B ;

[0227] S43, solving the sub-problem according to the defense strategy, determining the most serious line fault state and its fault recovery strategy under the worst fault scenario, and then updating the upper bound value;

[0228] S43. Judgment Is it established? represents the convergence threshold;

[0229] If so, the fault recovery strategy is output, and then a collaborative planning scheme for multiple flexible resources is obtained;

[0230] If not, the worst failure scenario obtained by solving the sub-problem is passed to the main problem, and the process returns to step S42.

[0231] In step S43 of this embodiment, the sub-problem is a two-level optimization problem, and its solution is:

[0232] The dual problem of inner layer minimization is solved by adopting strong duality theory, which is then transformed into a single-layer problem and solved by linear processing of the single-layer problem using the Big-M method.

[0233] Specifically, the constraints of the subproblem include all the constraints of the second stage. Since the subproblem is a two-level optimization problem, the strong duality theory is required to find the dual problem of the inner-level minimization problem and transform the two-level problem into a single-level problem for solution. In addition, the dual variables corresponding to each constraint in formula (46) are α, β, γ, The specific form of the transformed single-layer max problem after dual transformation is:

[0234]

[0235] For the bilinear term γ contained in equation (47) T u、 The Big-M method can be used for linearization. T u, the linearization process using the Big-M method is shown as follows:

[0236]

[0237] In the formula, M is a sufficiently large positive number, and s is an introduced auxiliary variable.

[0238] From the specific implementation methods provided by the present invention, it can be seen that the two-stage robust elastic planning model under typhoon disasters provided by the present invention, through the typhoon disaster disturbance model and the line failure rate model, describes the internal mechanism of the influence of the typhoon's spatiotemporal characteristics on the line failure rate, generates the time-varying failure rate of each line, and uses the information entropy method to realize the identification of vulnerable lines, and on this basis, considers the system cascading failures to obtain the time-varying fault state of each line. The adopted C&CG algorithm realizes the iterative solution of the model and can quickly converge to the optimal solution; the coordinated regulation of multiple flexibility resources can not only effectively improve the elasticity of the power system, but also reduce the economic losses caused by typhoon disasters.

[0239] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

[0240] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A resource collaborative elastic planning method considering the uncertainty of new energy under typhoon disasters, characterized by: The following steps are involved: S1. Based on the typhoon disaster disturbance model and the line failure rate model, a fault state uncertainty model is constructed to obtain the time-varying fault state of the line, and the final state of the line is determined by simulating the system cascading failure process; S2, by using the mixed Copula model combined with multivariate normal distribution to generate wind and solar output scenarios, and by using the synchronous back-strip reduction method to reduce the sampled scenarios, a set of wind and solar scenarios is screened out; S3. Based on the time-varying fault status of the line and the wind and solar scenario set, a two-stage robust elastic planning model based on a defense-attack-defense structure is constructed to consider typhoon disasters and new energy uncertainties, and its objective function and constraints are determined based on the collaborative planning of flexibility resources; S4. Based on the determined objective function and constraints, the column and constraint generation method and the Big-M method are used to solve the two-stage robust elastic planning model, determine the fault recovery strategy, and then obtain a collaborative planning scheme for multiple flexibility resources.

2. The resource collaborative elastic planning method considering the uncertainty of new energy under typhoon disasters according to claim 1 is characterized in that: In step S1, the typhoon disaster disturbance model is expressed as: R max =exp(-0.1239Δp 0.6003 +5.1034) In the formula, R max represents the maximum wind speed radius of the typhoon, Δp represents the pressure difference between the typhoon center and the area not affected by the typhoon, among which the wind speed V in the typhoon affected area r It is expressed as: In the formula, represents the maximum wind speed of the typhoon, r represents the wind speed in the affected area and the distance from the area to the typhoon center, and x represents the intensity attenuation coefficient; The line fault probability model is expressed as: In the formula, p ij (s) represents the failure rate of the ijth branch when the typhoon intensity is s, p l (s) represents the failure rate of the conductor, p p (s) represents the failure rate of the pole, m ij Indicates the number of wires corresponding to the ijth branch, y ij Indicates the number of poles corresponding to the ijth branch; The fault state uncertainty model is expressed as: Where U / u represents the fault state set / fault state of the line under typhoon disaster, p ij,t represents the failure rate of line ij at time t, z ij,t and u ij,t Both represent 0-1 decision variables, z ij,t Indicates the fault status of line ij under the typhoon attack at time t. If the value is 0, it means that ij fails after being attacked by the typhoon. Otherwise, the value is 1. ij,t represents the actual fault state of line ij at time t after the typhoon attack. Its value is 0, indicating that the line is in a fault state, and vice versa. Γ represents the budget value of information entropy, γ represents the time of typhoon attack, T s It indicates the time when the faulty line is repaired and restored to normal state, ij indicates the line, L indicates the line set, t indicates the time, and T indicates the entire period.

3. The resource collaborative elastic planning method considering the uncertainty of new energy under typhoon disasters according to claim 1 is characterized in that: In step S1, the method for simulating the system cascading failure process is specifically as follows: S11. Obtain t based on line fault probability and fault state uncertainty model k Faulty lines caused by typhoon disasters at all times; S12, determine the system status, obtain the subsystem status by identifying the system connectivity, and make power balance adjustments to ensure the power balance of each subsystem. When power imbalance occurs, adjust the output of the generator and various flexible resources to ensure the power balance of the system; S13, calculating the power flow of each subsystem, comparing the power flow of each line with the upper limit of the normal value of the power flow, and determining whether there is an overloaded line; If yes, the line is overloaded and the process goes to step S14; The overload control is to control the line overload through the dispatching of generators and various flexible resources and load reduction; If not, the simulation ends after recording the system status; S14, calculating the overload line fault probability, using a sampling method to obtain a fault set, and determining whether it is an empty set; If yes, the simulation ends after recording the system status; If not, the fault set is treated as a new fault; Among them, the overload line failure probability p lf for: In the formula, F l represents the active power flow of the lth line, represents the upper limit of the normal value of the power flow on the lth line, F max,l represents the power flow limit value of the lth line, 4. The resource collaborative elastic planning method considering the uncertainty of new energy under typhoon disasters according to claim 1 is characterized in that: In step S2, the method for generating a wind and solar power output scenario by using a mixed Copula function combined with a multivariate normal distribution is specifically as follows: S21. Estimate the marginal distribution function F of wind power and photovoltaic power output respectively w (x w ) and F p (x p ), use the kernel density estimation method to smooth the historical data and obtain the marginal distribution function of wind and solar power output; S22, estimate the parameters of the mixed Copula model by EM method; S23. Use the fitted marginal distribution function F w (x w ) and F p (x p ) and the estimated mixed Copula model to construct the joint distribution of wind and solar power output; S24. Based on the constructed joint distribution of wind and solar power output, the time series characteristics of wind power and photovoltaic power output are further considered, and a joint distribution model with time series dynamics is established based on the hybrid Copula model; S25, based on the Markov chain Monte Carlo sampling method, samples are extracted from the constructed joint distribution model with time series dynamics; S26. Based on the obtained MCMC samples, an inverse cumulative distribution function is used to perform an inverse transformation, thereby converting uniformly distributed random numbers into actual output scenarios of wind power and photovoltaic power.

5. The resource collaborative elastic planning method considering the uncertainty of new energy under typhoon disasters according to claim 4 is characterized in that: In step S2, the method of reducing the sampled scenes by the synchronous tape replay reduction method is specifically as follows: S27, setting the initial scene set to D, the reduced scene set to J, the initialization set J to an empty set, and the initial value of the iteration number u to 1; S28, determine the scene S that is cut at the u-th iteration u , so that the following formula is minimized; In the formula, p i Indicates scene S i The probability of J u is the set J at the uth iteration, c(S i ,S j ) represents scene S i and scene S j The probability distance between i ,S j )=PS i -S j P2; S29. Remove scene S from set D u , and add it to set J, increasing the number of iterations by 1; S210, scene S to be removed u The probability is assigned to the scene with the closest probability distance, and the probability of the nearest scene is updated; Among them, the probability of update is: In the formula, J(j) represents all the scenes S in the discarded set j Alternative scenario, p' j For the final scene that remains j probability; S211, repeating steps S28 to S210 until the scenes in the set D are scene data that need to be retained, thereby obtaining a screened scenery scene set.

6. The resource collaborative elastic planning method considering the uncertainty of new energy under typhoon disasters according to claim 1 is characterized in that: In step S3, the two-stage robust elastic programming model includes a first stage and a second stage; The first stage is the first-layer defense stage, and its planning goal is to minimize the on-duty installation cost of flexibility resources and line reinforcement cost. The first stage determines the planning decision variables and passes them to the second stage; The second stage includes the attack stage of the second layer and the defense stage of the third layer, and its planning goal is to minimize load shedding and power system operating costs; among them, the attack stage of the second layer is the disaster occurrence stage, which determines the worst fault scenario based on the time-varying failure rate and the identified vulnerable lines under the known planning decision, and transmits it to the third layer; the third layer minimizes load shedding and power operating costs under the determined worst fault scenario, and then uses the planning decision variables of the first stage to determine the fault recovery strategy of the power system, and feeds back the worst fault scenario and fault recovery strategy of the second layer to the first stage to adjust the determined planning decision variables.

7. The resource collaborative elastic planning method considering the uncertainty of new energy under typhoon disasters according to claim 6 is characterized in that: The objective function of the two-stage robust elastic programming model is expressed as: In the formula, xl Represents a 0 / 1 decision variable, indicating the state of line reinforcement. If the value is 1, it is reinforced, otherwise it is not reinforced. Both represent 0 / 1 variables, indicating whether energy storage, wind power and photovoltaic are connected to the node. If the value is 1, it is connected to node i, and if it is 0, it is not connected to node i. E ,K w ,K p Represent the installed capacity of energy storage, wind power and photovoltaic configuration respectively, C L,l represents the reinforcement cost of the lth transmission line, x l represents a 0 / 1 decision variable, indicating the state of line reinforcement. If the value is 1, it is reinforced, otherwise it is not reinforced. C p VP represents the daily installation cost of energy storage, wind power and photovoltaic power, etc. i D represents the load shedding amount of node i, P ij Represents the active power of the branch, P i G represents the active output of generator set i, P i pw , P i pv , P i ESS Respectively represent the active output of installed wind power, photovoltaic power and energy storage, V i represents the voltage at node i, C D,i represents the load shedding penalty cost of node i, C o represents the system operating cost, C ESS , C pw , C pv They represent the installation costs of energy storage, wind power and photovoltaic power, r represents the discount rate, s represents the number of years of discount, C G,i represents the operating cost of generator set i, C e,op , C w,op , C p,op are the operating cost coefficients of energy storage, wind power and photovoltaic power, Ω L ,Ω G ,Ω N They represent the collection of transmission lines, generator sets and load nodes respectively, Y is the total number of days in a year, and u / U represents a certain fault state / fault state set of the power system under the intermediate layer typhoon disaster.

8. The resource collaborative elastic planning method considering the uncertainty of new energy under typhoon disasters according to claim 6 is characterized in that: In the two-stage robust elastic programming model, the constraints of the first stage are: Where N L ,N ESS ,N pw ,N pv Respectively represent the maximum number of line reinforcements, the maximum number of energy storage, wind power and photovoltaic installations, Indicates the maximum discharge power of energy storage, K E represents the installed capacity of energy storage, K E Respectively represent the upper and lower limits of energy storage installed capacity, K w represents the installed capacity of wind power, K w are the upper and lower limits of wind power installed capacity, K p represents the photovoltaic installed capacity, K p They are the upper and lower limits of photovoltaic installed capacity respectively; δ pw , δ pv They represent the normalized wind power and photovoltaic power output sequences respectively. The actual output of wind and photovoltaic power is obtained by their normalized output time series and planned installed capacity. Respectively represent the reactive power output of wind power and photovoltaic power, Represent the maximum values ​​of wind power and photovoltaic power factor tangent angle respectively.

9. The resource collaborative elastic planning method considering the uncertainty of new energy under typhoon disasters according to claim 7 is characterized in that: In the two-stage robust elastic planning model, the constraints of the second stage include line state constraints, power balance constraints, line flow constraints, load reduction constraints, generator output constraints, energy storage operation constraints and node voltage constraints; Among them, the line state constraint is expressed as: In the formula, α ij,t represents a 0-1 decision variable, indicating the operating status of the line under typhoon attack after reinforcement. If the line is in normal operation, α ij,t =1, otherwise α ij,t =0, indicating that the lines reinforced in the first stage will not fail due to typhoon attacks, u ij,t represents the actual fault status of line ij at time t after the typhoon attack; The power balance constraint is expressed as: In the formula, represents the set of lines that flow out and into power from node n, P ij,t , Q ij,t represents the active power and reactive power of line ij at time t, P i D is the load of each node; The line power flow constraint is expressed as: Where P ij,max , Q ij,max They represent the upper limits of active and reactive power flows of line ij respectively; The load reduction constraint is expressed as: In the formula, Respectively represent the maximum active and reactive power of each node load; The output constraint of the generator set is expressed as: In the formula, They represent the upper and lower limits of the active output of the i-th generator set, They represent the upper and lower limits of reactive power output of the i-th generator set respectively; The energy storage operation constraints are expressed as: In the formula, represents the energy storage charge at time t, η E represents the discharge efficiency of energy storage, S oc They are the upper and lower limits of the state of charge of the energy storage respectively; The node voltage constraint is expressed as: Where V i,t is the voltage at node i, V N Indicates the rated voltage, M is a parameter with a larger value, V i min and V i max Respectively represent the upper and lower limits of the node voltage.

10. The resource collaborative elastic planning method considering the uncertainty of new energy under typhoon disasters according to claim 6 is characterized in that: The step S4 comprises the following sub-steps: S41. Convert the objective function and its constraints of the two-stage robust elastic programming model into a compact form and decompose it into a main problem and sub-problems; The expression of the main problem is generated based on the column and constraint generation method, and the defense strategy is obtained by solving it; the sub-problem is a defense strategy determined based on the main solution problem, which determines the worst fault scenario and determines the mathematical model of the fault recovery strategy through the corresponding scheduling scheme; S42. Solve the main problem, obtain the configuration strategy of line reinforcement and flexibility resources as the defense strategy, and update the lower bound value L B ; S43, solving the sub-problem according to the defense strategy, determining the most serious line fault state and its fault recovery strategy under the worst fault scenario, and then updating the upper bound value; Among them, the sub-problem is a two-level optimization problem, and its solution is: The dual problem of inner layer minimization is solved by using strong duality theory, and then it is transformed into a single-layer problem, which is solved by linear processing of the single-layer problem using the Big-M method. S43. Judgment Is it established? represents the convergence threshold; If so, the fault recovery strategy is output, and then a collaborative planning scheme for multiple flexible resources is obtained; If not, the worst failure scenario obtained by solving the sub-problem is passed to the main problem, and the process returns to step S42.

Images