Microgrid group cascading failure control method, electronic equipment and medium

Abstract

The invention discloses a microgrid group cascading failure control method, electronic equipment and a medium. The method comprises the following steps: establishing a longitudinal output model of the microgrid group; analyzing the distributed resource data according to a clustering algorithm to obtain microgrid group output distribution; and cascading failure control is carried out through microgrid group output distribution. According to the method, the safety and the reliability of a power system are improved through cascading failure analysis and emergency control research of the power grid containing the micro-grid group.

Description

Technical Field

[0001] The present invention relates to the field of grid auxiliary services, and more specifically, to a method for controlling cascading failures in a microgrid group, an electronic device, and a medium. Background Art

[0002] Power auxiliary services refer to services provided by grid-connected entities on the power generation side, such as thermal power, hydropower, nuclear power, wind power, photovoltaic power generation, solar thermal power generation, pumped storage, and self-owned power plants, new energy storage such as electrochemistry, compressed air, and flywheels, and adjustable loads (including those aggregated in the form of aggregators, virtual power plants, etc.) such as traditional high-energy-consuming industrial loads, industrial and commercial interruptible loads, and electric vehicle charging networks that can respond to power dispatching instructions, in order to maintain the safe and stable operation of the power system, ensure power quality, and promote the consumption of clean energy, in addition to normal power production, transmission, and use. Among power auxiliary services, the control of cascading failures is relatively important.

[0003] Cascading failures generally refer to the process in which one or more components in a system fail and are removed, triggering a series of successive component failures. Although the probability and frequency of cascading failures are very low, once they occur, they will have unpredictable impacts. With the continuous expansion of the installed capacity of renewable energy and the continuous development of renewable energy technologies, the characteristics of high complexity, strong coupling, and randomness of the power grid have become more prominent, and the consequences caused by cascading failures have become more serious. In the event of a sudden failure during the operation of the power grid, an effective emergency control strategy can be formulated based on the current power grid operation state to block the propagation of cascading failures.

[0004] The output of microgrid groups such as wind and light has characteristics such as randomness and intermittency, which increase the risk of cascading failures in the power system. Therefore, studying the output law of such microgrid groups is of great significance for analyzing cascading failures in the power grid after the extensive access of distributed energy. Regarding the uncertainty of photovoltaic and wind power output, current research mainly focuses on the output law considering the horizontal time series. The probability distribution of output is fitted based on the statistical data of photovoltaic and wind power plant output within a continuous time period, which macroscopically reflects the output fluctuations during this time period. The continuous time period is generally one quarter or one year. However, the propagation time of cascading failures is much shorter than this time period, usually several hours or dozens of minutes. When cascading failures occur at different times, the output law of the microgrid group is different during the fault propagation time, so the power flow distribution of the power grid will change, and the cascading failure propagation path may also change accordingly. Therefore, there will be certain errors in analyzing cascading failures in the power grid containing microgrid groups by considering the output law of the horizontal time series, especially when the microgrid group penetration rate is relatively high, the error will be more obvious.

[0005] Therefore, it is necessary to develop a method for controlling cascading failures in a microgrid group, an electronic device, and a medium.

[0006] The information disclosed in the background section of the present invention is only intended to enhance the understanding of the general background of the present invention and should not be regarded as an admission or any form of suggestion that such information constitutes the prior art known to those skilled in the art. Summary of the Invention

[0007] The present invention provides a method for controlling cascading failures in a microgrid group, an electronic device, and a medium, which can improve the safety and reliability of the power system through research on cascading failure analysis and emergency control of a power grid containing a microgrid group.

[0008] In a first aspect, an embodiment of the present disclosure provides a method for controlling cascading failures in a microgrid group, including:

[0009] Establish a longitudinal output model of the microgrid group;

[0010] Analyze the distributed resource data according to a clustering algorithm to obtain the output distribution of the microgrid group;

[0011] Perform cascading failure control through the output distribution of the microgrid group.

[0012] Preferably, establishing the longitudinal output model of the microgrid group includes:

[0013] Determine the longitudinal time for each time period, and use the average value of the output of the microgrid group within each time period as the output value at this longitudinal time;

[0014] Integrate the output values at the same longitudinal time within different dates of the microgrid group to obtain the output sequence samples at each longitudinal time;

[0015] Fit the probability distribution of the output at the longitudinal time of a single microgrid group, and establish the longitudinal output model of the microgrid group.

[0016] Preferably, the non-parametric kernel density estimation method is used to fit the probability distribution of the output at the longitudinal time of a single microgrid group.

[0017] Preferably, using the non-parametric kernel density estimation method to fit the probability distribution of the output at the longitudinal time of a single microgrid group includes:

[0018] According to the output sequence samples, determine the optimal window width through the overall empirical probability density and the minimum value of the integrated mean square error of the estimated probability density;

[0019] Estimate the probability density function of the output at each longitudinal time based on the Gaussian kernel function

[0020] Preferably, the optimal window width is:

[0021]

[0022] where hMISE is the optimal window width.

[0023] Preferably, the probability density function is:

[0024]

[0025] where x1, x2, …, x i , …, x n are n output samples of a microgrid cluster at a certain longitudinal moment, and K(·) is a kernel function.

[0026] Preferably, the distributed resource data is analyzed according to the agglomerative hierarchical clustering algorithm.

[0027] Preferably, analyzing the distributed resource data according to the agglomerative hierarchical clustering algorithm includes:

[0028] Taking the longitudinal power generation data at each moment as a separate class, and calculating the distance between the curves;

[0029] Merging the two classes with the smallest distance to construct a new class, and calculating the distance between the new class and other classes;

[0030] Judging whether the number of classes is less than the preset clustering quantity. If so, stop the calculation; if not, repeat the above steps;

[0031] Determining that the typical output curve formed by the average value of the power generation at the same time node in the cluster is the clustering center, and analyzing the data of the microgrid cluster by representing other curves in the cluster through the typical output curve.

[0032] In a second aspect, an embodiment of the present disclosure further provides an electronic device, which includes:

[0033] A memory storing executable instructions;

[0034] A processor that runs the executable instructions in the memory to implement the microgrid cluster cascading fault control method.

[0035] In a third aspect, an embodiment of the present disclosure further provides a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, the microgrid cluster cascading fault control method is implemented.

[0036] The method and device of the present invention have other characteristics and advantages, which will be obvious from the accompanying drawings incorporated herein and the subsequent specific embodiments, or will be described in detail in the accompanying drawings incorporated herein and the subsequent specific embodiments. These accompanying drawings and specific embodiments are used together to explain the specific principles of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The above and other objects, features, and advantages of the present invention will become more apparent by describing the exemplary embodiments of the present invention in more detail with reference to the accompanying drawings, in which, in the exemplary embodiments of the present invention, the same reference numerals generally represent the same components.

[0038] Figure 1 The flowchart showing the steps of the microgrid group cascading fault control method according to an embodiment of the present invention is shown. Detailed implementation manners

[0039] The preferred embodiments of the present invention will be described in more detail below. Although the preferred embodiments of the present invention are described below, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein.

[0040] To facilitate understanding of the solutions and effects of the embodiments of the present invention, three specific application examples are given below. Those skilled in the art should understand that this example is only for facilitating the understanding of the present invention, and any specific details are not intended to limit the present invention in any way.

[0041] Embodiment 1

[0042] Figure 1 The flowchart showing the steps of the microgrid group cascading fault control method according to an embodiment of the present invention is shown.

[0043] As Figure 1 shown, the microgrid group cascading fault control method includes:

[0044] Step 101, establishing a longitudinal output model of the microgrid group;

[0045] Step 102, analyzing the distributed resource data according to the clustering algorithm to obtain the output distribution of the microgrid group;

[0046] Step 103, performing cascading fault control through the output distribution of the microgrid group.

[0047] In one example, establishing a longitudinal output model of the microgrid group includes:

[0048] Determining the longitudinal moment for each time period, and taking the average value of the output of the microgrid group within each time period as the output value at this longitudinal moment;

[0049] Integrating the output values at the same longitudinal moment within different dates of the microgrid group to obtain the output sequence samples at each longitudinal moment;

[0050] Fitting the probability distribution of the output at the longitudinal moment of a single microgrid group to establish a longitudinal output model of the microgrid group.

[0051] In one example, the non-parametric kernel density estimation method is used to fit the probability distribution of the output power of a single microgrid group at longitudinal moments.

[0052] In one example, using the non-parametric kernel density estimation method to fit the probability distribution of the output power of a single microgrid group at longitudinal moments includes:

[0053] Based on the output power sequence samples, determine the optimal window width through the minimum of the overall empirical probability density and the integral mean square error of the estimated probability density;

[0054] Estimate the probability density function of the output power at each longitudinal moment based on the Gaussian kernel function

[0055] In one example, the optimal window width is:

[0056]

[0057] where h MISE is the optimal window width.

[0058] In one example, the probability density function is:

[0059]

[0060] where x1, x2, …, x i , …, x n are n output power samples of a microgrid group at a certain longitudinal moment, and K(·) is the kernel function.

[0061] In one example, analyze the distributed resource data according to the agglomerative hierarchical clustering algorithm.

[0062] In one example, analyzing the distributed resource data according to the agglomerative hierarchical clustering algorithm includes:

[0063] Take the longitudinal power generation data at each moment as a separate class, and calculate the distance between the curves;

[0064] Merge the two classes with the smallest distance to construct a new class, and calculate the distance between the new class and other classes;

[0065] Judge whether the number of classes is less than the preset clustering quantity. If so, stop the calculation. If not, repeat the above steps;

[0066] Determine that the typical output power curve formed by the average value of the power generation at the same time node in the class cluster is the clustering center, and represent other curves in the class cluster through the typical output power curve for data analysis of the microgrid group.

[0067] Specifically, there is a certain correlation between the output values of photovoltaic and wind power farms and the intraday time. Generally, the output of a photovoltaic power generation farm during the day is greater than that at night, and the maximum output occurs at noon. While the output of a wind power farm at night is greater than that during the day. Therefore, considering the output law of the microgrid group at longitudinal moments to analyze the cascading faults of the power grid containing the microgrid group, that is, the output law of the microgrid group at the same time every day within one year or several years, is more in line with the actual situation and can improve the reliability of cascading fault analysis. The outputs of photovoltaic and wind power farms at different geographical locations at the same time also have a certain correlation, that is, the characteristic that the output values increase or decrease simultaneously. The closer the distance between microgrid groups, the more obvious the correlation of the output.

[0068] Therefore, the present invention first establishes a longitudinal output model of the microgrid group, and then analyzes the distributed resource data according to the agglomerative hierarchical clustering algorithm to obtain the output distribution of the microgrid group considering the output law and correlation at longitudinal moments, providing an effective scenario for the generation of cascading faults and emergency control strategies.

[0069] First, divide a day evenly into 24 time periods, use the whole-hour moments within each time period as longitudinal moments, and then use the average value of the output of the microgrid group within each time period as the output value at this longitudinal moment. Sort out the statistical data of the output of the microgrid group, and integrate the output values of the microgrid group at the same longitudinal moment on different dates to form an output sequence sample for each longitudinal moment.

[0070] The probability distribution that the overall obeys can be estimated through a large number of samples. Currently, there are mainly parametric methods and non-parametric methods. The parametric method requires knowing the type of the probability distribution. Usually, the probability distribution of photovoltaic output is estimated by the Beta distribution, and the probability distribution of wind power output is estimated by the Weibull distribution. Since the output of photovoltaic and wind power is affected by many factors such as solar radiation, weather conditions, temperature, and equipment conversion efficiency, using a specific distribution to describe the output law of photovoltaic and wind power may sometimes have a large deviation. The non-parametric method does not need to know the type of the distribution that the overall obeys, but fits the probability distribution of the overall through the distribution characteristics of a large number of samples, and has strong adaptability. Therefore, the present invention uses the non-parametric kernel density estimation method to fit the probability distribution of the output of a single microgrid group at longitudinal moments, establishes a longitudinal output model of the microgrid group, and compares it with the probability distribution fitted by the conventional parametric method, and evaluates the adaptability of the model through error indicators, and evaluates the accuracy of the model through the χ 2 and K-S goodness-of-fit test method.

[0071] Let x1, x2, …, x n be n output samples of a certain longitudinal moment of the microgrid group, then the kernel density estimate of the output probability density f h (x) is defined as:

[0072]

[0073] Among them, K(·) is the kernel function; h is the window width.

[0074] There are many types of kernel functions. Common kernel functions K(·) include Gaussian, Triangle, Epanechnikov, etc. The estimated output probability density is less affected by the type of K(·). Therefore, in this invention, the Gaussian kernel function is used for kernel density estimation. The kernel function K(·) is as follows:

[0075]

[0076] The size of the window width h will affect the accuracy and smoothness of the probability density. An overly small h reflects more information, resulting in large fluctuations in the density curve. An overly large h makes the curve too smooth, leading to the loss of some characteristic information. The optimal window width can be selected using the overall empirical probability density and the minimum value of the integrated mean square error of the estimated probability density, that is:

[0077]

[0078] By solving the above equation, the optimal window width can be obtained as:

[0079]

[0080] Solve the optimal window width h of the probability density according to the samples of the longitudinal moment output sequence of the microgrid group MISE , and estimate the probability density function of the output at each longitudinal moment based on the Gaussian kernel function Establish the longitudinal moment output probability model of the microgrid group.

[0081] The goodness-of-fit test uses statistical data to test whether the estimated probability distribution reflects the characteristics of the random variable. The main goodness-of-fit tests applicable to unknown distributions are χ 2 and the K-S test method. According to Pearson's theorem, the relationship between the theoretical frequency and the sample frequency follows a specific distribution property, and its statistic is:

[0082]

[0083] Among them, k is the number of partitions of the sample; n i is the sample frequency in the i-th area; n is the sample size; p i0 is the theoretical probability value in the i-th area.

[0084] The asymptotic distribution of the above statistic is a χ 2 distribution with k - 1 degrees of freedom. Set χ 2The rejection region R of (k - 1), and then, based on the test value P on the left side of the above formula and the range of the rejection region, determine whether to accept the hypothesis that the sample data comes from the estimated probability distribution function. If Accept this hypothesis.

[0085] The K - S test method verifies the estimated probability distribution function through the continuous distribution function, sorts the sample data, and constructs an approximately continuous empirical distribution F n (x), and the constructed statistic is:

[0086]

[0087] where n is the sample size; F n (x i ) is the empirical distribution value of the i - th data xi after sorting; F0(x i ) is the estimated probability distribution value of x i . The above statistic has specific distribution properties, that is:

[0088]

[0089] where is called the Kolmogorov distribution.

[0090] Determine the rejection region through the confidence level α, R = {D n ≥D n,1-α}, obtain the statistic D n using the sample data and the estimated probability distribution. If D n ∈R, then do not accept the hypothesis that the sample data comes from the estimated probability distribution function.

[0091] There are generally multiple photovoltaic and wind farms in the power grid. The power outputs of photovoltaic and wind farms with similar geographical locations have strong correlations. It is necessary to construct corresponding models to describe this correlation and ensure the rationality of the micro - grid group power output model. The clustering algorithm is an unsupervised learning that classifies data according to similarity and can mine the massive information hidden in the data. It can divide objects with the same characteristics into the same class and divide the entire set into multiple clusters for further research. Common clustering algorithms include partitioning clustering, hierarchical clustering, density - based clustering, network clustering, etc.

[0092] Hierarchical clustering is a method of forming clusters in the form of a tree or hierarchical structure. It has high interpretability and can cluster non-spherical clusters, so it is widely used. It can be divided into divisive hierarchical clustering and agglomerative hierarchical clustering. The former uses a top-down approach, starting from a cluster that contains all data points, and then splitting the root node into some sub-clusters. Each sub-cluster continues to split recursively until a single-node cluster that contains only one data point appears, that is, each cluster contains only one data point. The latter is a bottom-up hierarchical clustering algorithm. Starting from the bottom layer, each time the most similar clusters are merged to form the clusters in the upper layer. The whole process stops when all data points are merged into one cluster or reaches a certain termination condition. The present invention uses this method for clustering.

[0093] The present invention uses the Euclidean distance as the distance metric between two curves, and performs clustering division on the standardized matrix X through the following steps:

[0094] 1) First, regard the longitudinal power generation data at each moment obtained by preprocessing as a separate class, and calculate the distance between the curves;

[0095] 2) Merge the two classes with the smallest distance to construct a new class;

[0096] 3) Calculate the distance between the new class and other classes;

[0097] 4) Determine whether the number of classes is the preset number of clusters. If so, stop the calculation. If not, return to step 2).

[0098] The desired number of clusters is determined in advance, and multiple clusters are obtained through hierarchical agglomerative clustering. The cluster center is a typical output curve formed by the average value of the power generation power at the same time node in the cluster, which can reflect the comprehensive characteristics of the curves in the cluster. These typical output curves can be used to represent other curves in the cluster for data analysis of the microgrid group, and then the cascading fault control can be carried out through the output distribution of the microgrid group.

[0099] Example 2

[0100] The present disclosure provides an electronic device, which includes: a memory storing executable instructions; a processor, and the processor runs the executable instructions in the memory to implement the above-mentioned microgrid group cascading fault control method.

[0101] The electronic device according to the embodiment of the present disclosure includes a memory and a processor.

[0102] The memory is used to store non-transitory computer-readable instructions. Specifically, the memory may include one or more computer program products, and the computer program products may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory, etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0103] The processor may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions. In one embodiment of the present disclosure, the processor is used to run the computer-readable instructions stored in the memory.

[0104] Those skilled in the art should understand that, in order to solve the technical problem of how to obtain good user experience effects, this embodiment may also include well-known structures such as communication buses, interfaces, etc., and these well-known structures should also be included in the protection scope of the present disclosure.

[0105] For the detailed description of this embodiment, reference may be made to the corresponding descriptions in the foregoing embodiments, and details will not be repeated here.

[0106] Example 3

[0107] An embodiment of the present disclosure provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the microgrid group cascading fault control method described above is implemented.

[0108] According to the computer-readable storage medium of the embodiment of the present disclosure, non-transitory computer-readable instructions are stored thereon. When the non-transitory computer-readable instructions are run by a processor, all or part of the steps of the methods of the foregoing embodiments of the present disclosure are executed.

[0109] The above computer-readable storage media include but are not limited to: optical storage media (such as: CD-ROM and DVD), magneto-optical storage media (such as: MO), magnetic storage media (such as: magnetic tape or removable hard disk), media with built-in rewritable non-volatile memory (such as: memory card) and media with built-in ROM (such as: ROM cartridge).

[0110] Those skilled in the art should understand that the purpose of the description of the embodiments of the present invention above is only to exemplarily illustrate the beneficial effects of the embodiments of the present invention, and is not intended to limit the embodiments of the present invention to any example given.

[0111] The embodiments of the present invention have been described above. The above description is exemplary and not exhaustive, and is also not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments.

Claims

1. A control method for cascading faults in a microgrid group, characterized in that, Including: Establish a longitudinal output model for the microgrid cluster; Analyze the distributed resource data according to the clustering algorithm to obtain the output distribution of the microgrid cluster; Conduct cascading fault control through the output distribution of the microgrid cluster.

2. The microgrid group cascading fault control method according to claim 1, wherein, Establishing a longitudinal output model for the microgrid cluster includes: Determine the longitudinal moment for each time period, and use the average value of the output of the microgrid cluster within each time period as the output value at this longitudinal moment; Integrate the output values at the same longitudinal moment on different dates of the microgrid cluster to obtain the output sequence samples at each longitudinal moment; Fit the probability distribution of the output at the longitudinal moment of a single microgrid cluster, and establish a longitudinal output model for the microgrid cluster.

3. The microgrid group cascading fault control method according to claim 2, wherein, Use the non-parametric kernel density estimation method to fit the probability distribution of the output at the longitudinal moment of a single microgrid cluster.

4. The microgrid group cascading fault control method according to claim 1, wherein, Using the non-parametric kernel density estimation method to fit the probability distribution of the output at the longitudinal moment of a single microgrid cluster includes: According to the output sequence samples, determine the optimal window width through the overall empirical probability density and the minimum value of the integrated mean square error of the estimated probability density; Estimate the probability density function of the output at each longitudinal moment based on the Gaussian kernel function 5. The microgrid group cascading fault control method according to claim 4, wherein, The optimal window width is: Among them, h MISE is the optimal window width.

6. The microgrid group cascading fault control method according to claim 4, wherein, Probability density function is as follows: Among them, x1, x2, …, x i , …, x n are n output samples at a certain longitudinal moment of the microgrid cluster, and K(·) is the kernel function.

7. The microgrid cluster cascading fault control method according to claim 1, wherein, Analyze the distributed resource data according to the agglomerative hierarchical clustering algorithm.

8. The microgrid group cascading fault control method according to claim 7, wherein, Analyzing the distributed resource data according to the agglomerative hierarchical clustering algorithm includes: Take the longitudinal power generation data at each moment as a separate class, and calculate the distance between the curves; Merge the two classes with the smallest distance to construct a new class, and calculate the distance between the new class and other classes; Judge whether the number of classes is less than the preset clustering quantity. If so, stop the calculation. If not, repeat the above steps; Determine that the typical output curve formed by the average value of the power generation at the same time node in the class cluster is the clustering center, and represent other curves within the class cluster through the typical output curve for data analysis of the microgrid cluster.

9. An electronic device, characterized in that, The electronic device includes: A memory storing executable instructions; A processor that runs the executable instructions in the memory to implement the cascading fault control method for the microgrid cluster according to any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the cascading fault control method for the microgrid cluster according to any one of claims 1-8.

Images