New energy sending end power grid reactive power optimization method and device

Abstract

The invention provides a new energy sending end power grid reactive power optimization method and device, and the method comprises the steps: clustering new energy output into a multi-typical scene through K-means, and selecting the most unstable 10 nodes as candidates through employing a multi-binary table criterion; and cross-scene screening is carried out to obtain five nodes with the most frequent occurrence frequency as key compensation objects. And for each scene, candidate node transient voltage is set as a target function, important node output characteristics are combined as constraints, iterative solution is carried out through a multi-target difference algorithm, and optimal reactive configuration of each scene is obtained. The method not only deals with the uncertainty of new energy output, but also balances the transient voltage stability and economy.

Description

Technical Field

[0001] The present application belongs to the field of reactive power optimization, and in particular, relates to a reactive power optimization method and device for a new energy sending-end power grid. Background Art

[0002] With the transformation of the global energy structure and the advancement of sustainable development strategies, the proportion of new energy (such as wind power, photovoltaics, etc.) in the power system is increasing. However, new energy power generation is intermittent and volatile, and its output is significantly affected by weather conditions (such as wind speed and light intensity), which poses a huge challenge to the stable operation of the power grid. Especially in the sending-end power grid with large-scale new energy access, the fluctuation and imbalance of reactive power are particularly prominent, leading to serious consequences such as voltage fluctuations, voltage over-limit and even system instability.

[0003] In order to deal with these problems and ensure the safe and stable operation of the power grid at the renewable energy sending end, reactive power optimization has become an important research direction. Traditional reactive power optimization methods are often based on certain load and power generation conditions, which are difficult to adapt to the uncertainty of renewable energy output. Summary of the invention

[0004] The purpose of this application is to overcome the defects in the above-mentioned prior art and provide a method and device for optimizing reactive power of a new energy sending-end power grid.

[0005] The present application provides a method for optimizing reactive power of a power grid at a new energy sending end, comprising:

[0006] Use K-means to cluster typical scenarios of renewable energy output and obtain multiple typical scenarios;

[0007] For each of the typical scenarios, the transient voltage stability is evaluated using the multi-binary table criterion method, and the 10 candidate nodes with the most unstable transient voltage are selected;

[0008] Among all the candidate nodes, five candidate nodes with the highest occurrence frequency are selected as important nodes for reactive power compensation;

[0009] For each of the typical scenarios, setting the transient voltage stability of the candidate node as an objective function;

[0010] For each of the typical scenarios, setting the output characteristics and constraints of the important nodes as constraints of the objective function;

[0011] According to the objective function and the constraints of the objective function, an iterative solution is performed through a multi-objective difference algorithm to obtain an optimal reactive power configuration solution in each typical scenario.

[0012] Optionally, the important nodes include: a static VAR generator, a static VAR compensator, a phase regulator or a capacitor.

[0013] Optionally, a multi-binary table criterion method is used to evaluate transient voltage stability, including:

[0014] Obtaining a state quantity X of the transient voltage stability;

[0015] Set the binary table criterion (X cr ,T cr ), where X cr is the set value, T cr is the maximum allowed deviation time;

[0016] The transient voltage stability is judged according to the state quantity X and the binary table criterion, including: Case 1: X max <X cr , T b =0, stable state; Case 2: X max >X cr , 0 <T b <T cr Acceptable stable state; Case 3: X max >X cr , T b ≥T cr Unacceptable stable state;

[0017] Among them, the T b The state quantity X exceeds X in the dynamic process cr Time, X max is the maximum value of the state quantity X.

[0018] Optionally, the stable state, the acceptable stable state and the unacceptable stable state are described to a certain extent, including:

[0019] The degree of stability is described by the following expression:

[0020]

[0021] Among them, X N is the reference value, t N .0 means the state quantity increases and exceeds X N The moment; t N .1 is recovery below X N The degree of the stable state reflects the state quantity relative to X in the dynamic process. N Whether the cumulative offset exceeds the allowable range;

[0022] The degree of acceptable stable state and unacceptable stable state is described by the following expression:

[0023]

[0024] Among them, t cr .0 and t cr .1 respectively means the state quantity exceeds and recovers below X cr moment.

[0025] Optionally, performing iterative solving by a multi-objective difference algorithm according to the objective function and the constraints of the objective function includes:

[0026] S201, determining the population size of the multi-objective difference algorithm, the number of decision variables and the upper and lower limits of each decision variable according to the objective function and the constraint conditions of the objective function;

[0027] S202, generating a population according to the population size, the number of decision variables, and the upper and lower limits of each decision variable;

[0028] S203, randomly selecting five different integer indexes from the population, and obtaining operator individuals through mutation calculation;

[0029] S204, performing crossover calculation on the operator individual and the parent individual to obtain a test individual;

[0030] S205, calculating the fitness of the test individual, comparing the fitness of the test individual with the fitness of the parent individual, and if the fitness of the test individual is greater than the fitness of the parent individual, determining whether the decision variable of the test individual is within the upper and lower limits of the decision variable;

[0031] S206. If the decision variable of the test individual is within the upper and lower limits, the parent individual is replaced by the test individual to form a new population;

[0032] S207. Repeat the above steps S203 to S206 until the test individual no longer replaces the parent individual.

[0033] The present application also provides a reactive power optimization device for a new energy sending-end power grid, comprising:

[0034] The clustering module uses K-means clustering to cluster typical scenarios of renewable energy output and obtain multiple typical scenarios;

[0035] A screening module, for each of the typical scenarios, uses a multi-binary table criterion method to evaluate the transient voltage stability and selects 10 candidate nodes with the most unstable transient voltage;

[0036] A node module selects five candidate nodes with the highest occurrence frequency from among all the candidate nodes in all the typical scenarios as important nodes for reactive power compensation;

[0037] A function module, for each of the typical scenarios, setting the transient voltage stability of the candidate node as an objective function;

[0038] A constraint condition module, for each of the typical scenarios, sets the output characteristics and constraint conditions of the important nodes as constraint conditions of the objective function;

[0039] The calculation module performs iterative solution through a multi-objective difference algorithm according to the objective function and the constraints of the objective function to obtain the optimal reactive power configuration solution under various typical scenarios.

[0040] Optionally, the important nodes include: a static VAR generator, a static VAR compensator, a phase regulator or a capacitor.

[0041] Optionally, the screening module uses a multi-binary table criterion method to evaluate transient voltage stability, including:

[0042] Obtaining a state quantity X of the transient voltage stability;

[0043] Set the binary table criteria ((X cr ,T cr ), where X cr is the set value, T cr is the maximum allowed deviation time;

[0044] The transient voltage stability is judged according to the state quantity X and the binary table criterion, including: Case 1: X max <X cr , T b =0, stable state; Case 2: X max >X cr , 0 <T b <T cr Acceptable stable state; Case 3: X max >X cr , T b ≥T cr Unacceptable stable state;

[0045] Among them, the T b The state quantity X exceeds X in the dynamic process cr Time, X max is the maximum value of the state quantity X.

[0046] Optionally, the screening module describes the degree of the stable state, the acceptable stable state and the unacceptable stable state, including:

[0047] The degree of stability is described by the following expression:

[0048]

[0049] Among them, X N is the reference value, t N .0 and t N .1 respectively means the state quantity increases and exceeds X N and restore below X N The degree of the stable state reflects the state quantity relative to X in the dynamic process. N Whether the cumulative offset exceeds the allowable range;

[0050] The degree of acceptable stable state and unacceptable stable state is described by the following expression:

[0051]

[0052] Among them, tcr,0 and tcr,1 are the moments when the state quantity exceeds and recovers to be lower than Xcr respectively.

[0053] Optionally, the calculation module performs iterative solution through a multi-objective difference algorithm, including:

[0054] S201 determines the population size of the multi-objective difference algorithm, the number of decision variables, and the upper and lower limits of each decision variable according to the objective function and the constraint conditions of the objective function;

[0055] S202 generates a population according to the population size, the number of decision variables and the upper and lower limits of each decision variable;

[0056] S203 randomly selects five different integer indexes from the population, and obtains operator individuals through mutation calculation;

[0057] S204 performs crossover calculation on the operator individual and the parent individual to obtain a test individual;

[0058] S205 calculates the fitness of the test individual, compares the fitness of the test individual with the fitness of the parent individual, and if the fitness of the test individual is greater than the fitness of the parent individual, determines whether the decision variable of the test individual is within the upper and lower limits;

[0059] S206 If the decision variable of the test individual is within the upper and lower limits, the parent individual is replaced by the test individual to form a new population;

[0060] S207 repeats S203 to S206 until the test individual no longer replaces the parent individual.

[0061] The beneficial effects of this application are:

[0062] The present application provides a method for optimizing reactive power of a power grid at the sending end of a new energy source, including: using K-means to cluster typical scenarios of new energy output to obtain multiple typical scenarios; for each of the typical scenarios, using a multi-binary table criterion method to evaluate transient voltage stability, and selecting 10 candidate nodes with the most unstable transient voltage; among all the candidate nodes, selecting the 5 candidate nodes with the highest frequency of occurrence as important nodes for reactive power compensation; for each of the typical scenarios, setting the transient voltage stability of the candidate nodes as an objective function; for each of the typical scenarios, setting the output characteristics and constraints of the important nodes as constraints of the objective function; according to the objective function and the constraints of the objective function, iteratively solving through a multi-objective differential algorithm to obtain the optimal reactive power configuration scheme under each typical scenario. The present application takes into account the uncertainty of new energy output, while taking into account reactive power optimization and the economy of the transient voltage stability and reactive power compensation of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 This is a schematic diagram of the reactive power optimization process of the new energy sending-end power grid in this application;

[0064] Figure 2 A schematic diagram of the different situations of acceptability in this application;

[0065] Figure 3 It is a flowchart of the multi-objective difference algorithm in this application. DETAILED DESCRIPTION

[0066] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that various forms of implementing the present disclosure should not be limited by the embodiments set forth herein. On the contrary, the embodiments are provided to more thoroughly understand the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0067] Please refer to Figure 1 As shown, a method for optimizing reactive power of a power grid at a new energy sending end includes the following steps:

[0068] S101. Use K-means to cluster typical scenarios of new energy output to obtain multiple typical scenarios.

[0069] In the reactive power optimization process of the renewable energy sending-end power grid, the output characteristics of renewable energy (especially wind power and photovoltaic power) are difficult to directly predict and control due to their high volatility and randomness. This output change not only affects the stability of the power system, but also increases the complexity of reactive power optimization. In order to effectively simulate and evaluate the impact of renewable energy access on the power grid and reduce the amount of calculation, it is particularly important to use the K-means clustering algorithm to cluster the renewable energy output data into typical scenarios.

[0070] First, the real-time output data of wind power and photovoltaic power in a certain area for one year (8760 hours) is collected from the monitoring system of the renewable energy power plant or the grid dispatching center. These data are the basis for subsequent analysis.

[0071] The collected data is cleaned to remove outliers (such as extreme values ​​caused by equipment failure or data transmission errors).

[0072] Missing values ​​are filled in appropriately to ensure data integrity.

[0073] Normalization processing is performed to convert data of different dimensions to the same scale to facilitate subsequent analysis.

[0074] Secondly, the elbow method is used to determine the optimal number of clusters k. The core of the elbow method is to observe the trend of the sum of squared errors (SSE) within the cluster with the number of clusters k. The calculation formula of SSE usually involves the sum of the squares of the distances from each point to the center of the cluster to which it belongs. The SSE expression is as follows:

[0075]

[0076] Among them, x (i) is the i-th sample data, the μ (j) It refers to the center of group j.

[0077] By gradually increasing the k value and calculating the corresponding SSE, a curve of SSE changing with k is plotted. When the rate of decrease of SSE suddenly slows down, the corresponding k value is the elbow point, which is regarded as the optimal number of clusters.

[0078] Again, based on the determined k value, the K-means clustering algorithm is used to cluster the preprocessed renewable energy output data.

[0079] The algorithm works in an iterative manner, and each iteration updates the cluster center of each cluster (that is, the mean of all data points in the cluster).

[0080] The iteration process continues until the convergence condition is met (such as the cluster center no longer changes significantly or the preset number of iterations is reached).

[0081] The silhouette coefficient method is used to evaluate the clustering results. The silhouette coefficient combines the cohesion within the cluster (a(i)) and the separation between clusters (b(i)) to measure the quality of the clustering effect. The expression is as follows:

[0082]

[0083] Among them, a (i) is the average distance between the i-th sample data and other points in the cluster, b (i) is the average distance between the sample data and all points in the nearest cluster.

[0084] The calculation formula of the silhouette coefficient involves the calculation of a(i) and b(i), where a(i) is the average distance between the sample point and other points in the same cluster, and b(i) is the minimum average distance between the sample point and all points in the nearest cluster. The expressions of a(i) and b(i) are as follows:

[0085]

[0086] Among them, d(i,j) represents the Euclidean distance between data, n c and n p Indicates the number of data.

[0087] The value range of the silhouette coefficient is [-1, important node 1]. The larger the value, the better the clustering effect. By calculating the silhouette coefficient of each sample point and averaging it, the silhouette coefficient of the entire clustering result can be obtained.

[0088] According to the results of K-means clustering, the 8760 hours of new energy scenarios in a certain area in a year are clustered into 96 typical scenario sets. These typical scenarios represent the typical patterns of new energy output in different time periods.

[0089] Each typical scenario includes the key characteristics of renewable energy output during that time period, such as output range, change trend, etc., to facilitate subsequent reactive power optimization analysis and simulation.

[0090] S102: For each of the typical scenarios, the transient voltage stability is evaluated using a multi-binary table criterion method, and 10 candidate nodes with the most unstable transient voltage are selected.

[0091] The generated 96 typical scenario sets were used for indicator calculation and analysis, such as transient voltage stability, reactive power fluctuation, etc. These indicators can reflect the impact of new energy access on the power system.

[0092] The analysis results will provide basic data support for subsequent reactive power optimization work and help formulate more scientific and reasonable reactive power optimization strategies.

[0093] For each renewable energy output scenario, select the 10 nodes with the most unstable transient voltage.

[0094] For any dynamic process described by a dynamic system, the acceptability of its characterization quantity X is defined as follows:

[0095] During the transient process, the time that X deviates from the set value cannot exceed the specified requirements, such as the acceptability of transient voltage rise and transient frequency deviation in the power system. The offset can be positive or negative, corresponding to the upper and lower limits of the state respectively.

[0096] Acceptability can be judged by a binary table (X cr ,T cr ) indicates that the state quantity X exceeds the set value X cr The maximum duration T b No more than T cr , which is considered acceptable at this time.

[0097] The analysis is based on the case where the state deviation is positive. At this time, for any binary table criterion (X cr ,T cr ), there are three cases of system dynamic behavior, namely:

[0098] Scenario 1: X max <X cr ,T b =0;

[0099] Scenario 2: X max >X cr , 0 <Tb<T cr ;

[0100] Scenario 3: X max >X cr ,T b ≥T cr ;

[0101] Among them, X max is the minimum value of state X in the system dynamics.

[0102] According to the acceptable definition, under the acceptable criteria, situations 1 and 2 are acceptable, and situation 3 is unacceptable.

[0103] like Figure 2 As shown, the system binary table criteria and two dynamic processes.

[0104] Further quantify the degree of unacceptability and acceptability of the system distance. Taking case 2 as an example, the dynamic process state exceeds the set value X cr Time T b <T cr , according to the definition of acceptable, both processes are acceptable, which is obviously incorrect.

[0105] In order to determine whether the system state is acceptable, the stability margin of the system state quantity is analyzed.

[0106] For the above situation 1, the stability margin index can be defined according to whether the cumulative amount of increase exceeds the limit value:

[0107]

[0108] Among them, (X N -X cr )T cr is the maximum permissible value of the accumulated state quantity based on the binary table criterion; X N is the reference value; t N.0 and t N.1 The state quantity is higher than X during the rising process N and the recovery process is higher than X N The difference is the recovery time of the state quantity.

[0109] Please refer to Figure 2 As mentioned above, the above formula shows that for case 1, the corresponding curve offset X of the system state in the dynamic process is mainly considered. N The maximum cumulative rise is divided into two cases:

[0110] (1) When T N When it is small (for example, t N.0 ), X can quickly return to its normal level, S1 is small, corresponding to curve 1 in the figure, and the system is stable.

[0111] (2) When T N When it is large (e.g. t N.1 ), X cannot return to normal level for a long time, corresponding to curve 2 in the figure, at this time the system is unstable. The corresponding index is less than 0.

[0112] For situations 2 and 3, the margin indicator can be defined according to whether the accumulated amount of increase exceeds the limit value:

[0113]

[0114] Among them, t cr.0 and t cr.1 They are respectively higher than X in the process of state quantity increasing cr and the recovery process is higher than X cr The moment; T b =t cr.1 -t cr.0 The state in the dynamic process is higher than X cr The maximum rise time.

[0115] It can be seen from the verification that the stable state is applicable to both Case 2 and Case 3, and can be applied to the transient stability evaluation index of the power system.

[0116] S103 . Among all the candidate nodes, select five candidate nodes with the highest appearance frequency as important nodes for reactive power compensation.

[0117] Subsequently, a statistical analysis was conducted on the 960 nodes selected from the 96 scenarios, and the five points with the highest frequency of occurrence were selected as important nodes for reactive power compensation.

[0118] S104: For each of the typical scenarios, setting the transient voltage stability of the candidate node as an objective function.

[0119] For each scene η Vi The initial value of the indicator is calculated, and each scene η is selected Vi The 10 nodes with the largest index are divided into the following groups: Vi The indicator is used as the objective function.

[0120] S105. For each of the typical scenarios, the output characteristics and constraints of the important nodes are set as constraints of the objective function.

[0121] The reactive power compensation equipment in the important nodes includes: SVG (static VAR generator), SVC (static VAR compensator), phase regulator or capacitor. The reactive power constraint conditions and related motion equations of a series of reactive power compensation equipment such as SVG, SVC, phase regulator (SC) or capacitor (C) are as follows:

[0122] Q min(SVCi) ≤Q SVCi ≤Q max(SVCi)

[0123] 0≤Q SVGi ≤Q max(SVGi)

[0124] Q SCi(SCi) ≤Q SC，i ≤Q max(SCi)

[0125] 0≤N SCi ≤N max(SCi)

[0126]

[0127] Among them, Q represents reactive power, N represents rated power, max represents maximum, min represents minimum, and B represents the number of times the capacitor is switched on and off.

[0128] According to the objective function and the constraints of the objective function, an iterative solution is performed through a multi-objective difference algorithm to obtain an optimal reactive power configuration solution in each typical scenario.

[0129] Please refer to Figure 3As shown in the figure, a multi-objective algorithm is used to solve the optimal reactive power configuration scheme in each typical scenario. The steps include:

[0130] S201 determines the population size of the multi-objective difference algorithm, the number of decision variables, and the upper and lower limits of each decision variable according to the objective function and the constraint conditions of the objective function;

[0131] S202 generates a population according to the population size, the number of decision variables and the upper and lower limits of each decision variable;

[0132] S203 randomly selects five different integer indexes from the population, and obtains operator individuals through mutation calculation;

[0133] S204 performs crossover calculation on the operator individual and the parent individual to obtain a test individual;

[0134] S205 calculates the fitness of the test individual, compares the fitness of the test individual with the fitness of the parent individual, and if the fitness of the test individual is greater than the fitness of the parent individual, determines whether the decision variable of the test individual is within the upper and lower limits;

[0135] S206 If the decision variable of the test individual is within the upper and lower limits, the parent individual is replaced by the test individual to form a new population;

[0136] S207 repeats the above S203 to S206 until the test individual no longer replaces the parent individual.

[0137] Specifically, the multi-objective multi-objective algorithm is a differential evolution algorithm, which adopts a real number coding method to generate a population according to the objective function and the constraint conditions of the objective function, where NP is the size of the population, and D is the number of decision variables. The specific expression is as follows:

[0138] x i ={x i，1 , x i，2 , …, x i，D},i=1,2,...,NP,j=1,2,...,D

[0139] Among them, x i，j It represents the value of the i-th individual in the j-th dimension, and its value range is between Lj and Uj, where Lj represents the lower limit of the decision variable, Uj represents the upper limit of the decision variable, and x i，j Perform random initialization uniformly within the upper and lower bounds. The specific initialization is as follows:

[0140] x i，j =rand j (0,1)·(Uj -L j )+L j ,i=1,2,...,NP,j=1,2,...,D

[0141] where rand j (0,1) represents a random number in the interval [0,1].

[0142] Based on the above population, the mutation operation is performed, and the expression is as follows:

[0143] v i =x best +F·(x s1 -x s2 )

[0144] Randomly select five different integers x from the range [1, NP] s1 、x s2 , where the best individual x best , the scaling factor F of the mutation operation is a random number in the interval [0,1].

[0145] In the differential evolution algorithm, binomial crossover is used to generate experimental individuals. During crossover, the parent individual xi and the operator individual vi generated by mutation are combined to perform binomial crossover to generate the experimental individual u i , which is expressed as follows:

[0146]

[0147] Among them, rand j is a random number selected from 0 to 1, j is taken from the dimension where the individual is located; CR represents the crossover probability of the algorithm, and its value is selected from [0,1]; sn is a random integer selected arbitrarily from [1,D].

[0148] The differential evolution algorithm uses an elite selection strategy in the selection operation. When performing the selection operation, the parent individual x is first mutated and crossover operated. i Transform and generate new experimental individuals u i , and then compare the new and old generations of individuals, and select them by crowding order, until the offspring population cannot complete the elite selection after entering the parent population, and obtain better offspring individuals.

[0149] For the boundary constraints of the algorithm, it is necessary to ensure that the parameter values ​​of the newly generated individuals are within the preset feasible domain. There are two common methods. One is to replace the new individuals that do not meet the boundary constraints with parameter vectors randomly generated from the feasible domain. The other way to deal with it is boundary absorption, that is, to adjust the individual values ​​that exceed the boundary range to the nearest boundary value. Here we take the approach of pulling the out-of-range individuals back to the boundary value.

[0150] After the number of iterations is completed or the accuracy meets the requirements, the compromise solution is selected based on the solution set with the highest fitness after elite selection sorting.

[0151] Through the calculation of the above differential algorithm, the optimal reactive power optimization plan for the renewable energy sending-end power grid is finally obtained.

[0152] The present application also includes: a reactive power optimization device for a new energy sending-end power grid, comprising:

[0153] The clustering module uses K-means clustering to cluster typical scenarios of renewable energy output and obtain multiple typical scenarios;

[0154] A screening module, for each of the typical scenarios, uses a multi-binary table criterion method to evaluate the transient voltage stability and selects 10 candidate nodes with the most unstable transient voltage;

[0155] A node module selects five candidate nodes with the highest occurrence frequency from among all the candidate nodes in all the typical scenarios as important nodes for reactive power compensation;

[0156] A function module, for each of the typical scenarios, setting the transient voltage stability of the candidate node as an objective function;

[0157] A constraint condition module, for each of the typical scenarios, sets the output characteristics and constraint conditions of the important nodes as constraint conditions of the objective function;

[0158] The calculation module performs iterative solution through a multi-objective difference algorithm according to the objective function and the constraints of the objective function to obtain the optimal reactive power configuration solution under various typical scenarios.

[0159] Although the present invention is disclosed as above by the embodiments, it is not intended to limit the protection scope of the present invention. Any changes and modifications made by any technician familiar with the technology without departing from the concept and scope of the present invention should fall within the protection scope of the present invention.

Claims

1. A method for optimizing reactive power of a new energy sending-end power grid, characterized in that: include: Use K-means to cluster typical scenarios of renewable energy output and obtain multiple typical scenarios; For each of the typical scenarios, the transient voltage stability is evaluated using the multi-binary table criterion method, and the 10 candidate nodes with the most unstable transient voltage are selected; Among all the candidate nodes, five candidate nodes with the highest occurrence frequency are selected as important nodes for reactive power compensation; For each of the typical scenarios, setting the transient voltage stability of the candidate node as an objective function; For each of the typical scenarios, setting the output characteristics and constraints of the important nodes as constraints of the objective function; According to the objective function and the constraints of the objective function, an iterative solution is performed through a multi-objective difference algorithm to obtain an optimal reactive power configuration solution in each typical scenario.

2. The reactive power optimization method for the new energy sending-end power grid according to claim 1 is characterized in that: The important nodes include: static VAR generator, static VAR compensator, phase regulator or capacitor.

3. The method for optimizing reactive power of a new energy sending-end power grid according to claim 1 is characterized in that: The transient voltage stability is evaluated using the multi-binary table criterion method, including: Obtaining a state quantity X of the transient voltage stability; Set the binary table criterion (X cr ,T cr ), where X cr is the set value, T cr is the maximum allowed deviation time; The transient voltage stability is judged according to the state quantity X and the binary table criterion, including: Case 1: X max <X cr , T b =0, stable state; Case 2: X max >X cr , 0 <T b <T cr Acceptable stable state; Case 3: X max >X cr , T b ≥T cr Unacceptable stable state; Among them, the T b The state quantity X exceeds X in the dynamic process cr Time, X max is the maximum value of the state quantity X.

4. The method for optimizing reactive power of a new energy sending-end power grid according to claim 3 is characterized in that: A description of the degree of the stable state, acceptable stable state, and unacceptable stable state, including: The degree of stability is described by the following expression: Among them, X N is the reference value, t N .0 means the state quantity increases and exceeds X N The moment; t N .1 is recovery below X N The degree of the stable state reflects the state quantity relative to X in the dynamic process. N Whether the cumulative offset exceeds the allowable range; The degree of acceptable stable state and unacceptable stable state is described by the following expression: Among them, t cr .0 and t cr .1 respectively means the state quantity exceeds and recovers below X cr moment.

5. The method for optimizing reactive power of a new energy sending-end power grid according to claim 1 is characterized in that: The iterative solution is performed by a multi-objective difference algorithm according to the objective function and the constraint conditions of the objective function, including: S201, determining the population size of the multi-objective difference algorithm, the number of decision variables and the upper and lower limits of each decision variable according to the objective function and the constraint conditions of the objective function; S202, generating a population according to the population size, the number of decision variables, and the upper and lower limits of each decision variable; S203, randomly selecting five different integer indexes from the population, and obtaining operator individuals through mutation calculation; S204, performing crossover calculation on the operator individual and the parent individual to obtain a test individual; S205, calculating the fitness of the test individual, comparing the fitness of the test individual with the fitness of the parent individual, and if the fitness of the test individual is greater than the fitness of the parent individual, determining whether the decision variable of the test individual is within the upper and lower limits of the decision variable; S206. If the decision variable of the test individual is within the upper and lower limits, the parent individual is replaced by the test individual to form a new population; S207. Repeat the above steps S203 to S206 until the test individual no longer replaces the parent individual.

6. A reactive power optimization device for a new energy sending-end power grid, characterized in that: include: The clustering module uses K-means clustering to cluster typical scenarios of renewable energy output and obtain multiple typical scenarios; A screening module, for each of the typical scenarios, uses a multi-binary table criterion method to evaluate the transient voltage stability and selects 10 candidate nodes with the most unstable transient voltage; A node module selects five candidate nodes with the highest occurrence frequency from among all the candidate nodes in all the typical scenarios as important nodes for reactive power compensation; A function module, for each of the typical scenarios, setting the transient voltage stability of the candidate node as an objective function; A constraint condition module, for each of the typical scenarios, sets the output characteristics and constraint conditions of the important nodes as constraint conditions of the objective function; The calculation module performs iterative solution through a multi-objective difference algorithm according to the objective function and the constraints of the objective function to obtain the optimal reactive power configuration solution under various typical scenarios.

7. The reactive power optimization device for the new energy sending-end power grid according to claim 6 is characterized in that: The important nodes include: static VAR generator, static VAR compensator, phase regulator or capacitor.

8. The reactive power optimization device for the new energy sending-end power grid according to claim 6 is characterized in that: The screening module uses a multi-binary table criterion method to evaluate transient voltage stability, including: Obtaining a state quantity X of the transient voltage stability; Set the binary table criteria ((X cr ,T cr ), where X cr is the set value, T cr is the maximum allowed deviation time; The transient voltage stability is judged according to the state quantity X and the binary table criterion, including: Case 1: X max <X cr , T b =0, stable state; Case 2: X max >X cr , 0 <T b <T cr Acceptable stable state; Case 3: X max >X cr , T b ≥T cr Unacceptable stable state; Among them, the T b The state quantity X exceeds X in the dynamic process cr Time, X max is the maximum value of the state quantity X.

9. The reactive power optimization device for the new energy sending-end power grid according to claim 8 is characterized in that: The screening module describes the degree of the stable state, the acceptable stable state and the unacceptable stable state, including: The degree of stability is described by the following expression: Among them, X N is the reference value, t N .0 and t N .1 respectively means the state quantity increases and exceeds X N and restore below X N The degree of the stable state reflects the state quantity relative to X in the dynamic process. N Whether the cumulative offset exceeds the allowable range; The degree of acceptable stable state and unacceptable stable state is described by the following expression: Among them, tcr,0 and tcr,1 are the moments when the state quantity exceeds and recovers to be lower than Xcr respectively.

10. The reactive power optimization device for the new energy sending-end power grid according to claim 6 is characterized in that: The calculation module performs iterative solution through a multi-objective difference algorithm, including: S201 determines the population size of the multi-objective difference algorithm, the number of decision variables, and the upper and lower limits of each decision variable according to the objective function and the constraint conditions of the objective function; S202 generates a population according to the population size, the number of decision variables and the upper and lower limits of each decision variable; S203 randomly selects five different integer indexes from the population, and obtains operator individuals through mutation calculation; S204 performs crossover calculation on the operator individual and the parent individual to obtain a test individual; S205 calculates the fitness of the test individual, compares the fitness of the test individual with the fitness of the parent individual, and if the fitness of the test individual is greater than the fitness of the parent individual, determines whether the decision variable of the test individual is within the upper and lower limits; S206 If the decision variable of the test individual is within the upper and lower limits, the parent individual is replaced by the test individual to form a new population; S207 repeats the above S203 to S206 until the test individual no longer replaces the parent individual.

Images