Source-network-load multi-resource coordinated power grid toughness improving and reinforcing method

Abstract

The invention relates to a power grid toughness improving and reinforcing method based on source-network-load multi-resource cooperation. The method comprises the steps of obtaining an operation data sequence in a power grid; analyzing the amplitude dispersion in the power load data sequence, the fluctuation amplitude of the power load data sequence and the average recovery duration of recovering each frequency in the power grid frequency data sequence to a preset frequency after deviation to obtain a load frequency response coefficient; the recovery rate for recovering each power in the active power data sequence to the reference power after deviation, the symmetry degree between a rising recovery section and a falling recovery section after each power in the active power data sequence is deviated, and the fluctuation degree of the power factor data sequence are calculated; analyzing the similarity between the new energy generating capacity data sequence and the new energy generating capacity prediction sequence to obtain a fault recovery coefficient; and in response to a fault of the power grid, performing power recovery by taking the target node as an initial node. The overall recovery efficiency of the power grid after the fault can be effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power grid resilience improvement reinforcement, in particular to a power grid resilience improvement reinforcement method based on source-grid-load multi-resource cooperation. BACKGROUND

[0002] Power grid resilience refers to the power grid's prevention, resistance, response and recovery capability under extreme events such as natural disasters or network attacks. With the large-scale access of distributed power sources, large-scale power outages caused by extreme events occur frequently. Traditional power grid reinforcement methods are costly and passive in response, and are difficult to cope with renewable energy fluctuations and multiple faults. In the prior art, traditional structural indicators such as average path length, network diameter, node degree distribution, betweenness centrality and connectivity are used as important evaluation tools to measure the anti-disturbance and recovery capability of the power grid system. They mainly focus on the global characteristics or a single node of the network, and do not consider the demand response and load adjustable capacity of each node of the power grid system. Moreover, the static structural indicators of the power grid system cannot cover the dynamic distribution of the power flow of each node, resulting in inaccurate evaluation of the actual resilience value of each node of the power grid system. Especially in the actual black start scenario, the power grid system expands the recovery range step by step through the root node. The area of the farthest root node often determines the longest time required for overall recovery, which increases the risk of cascading overload caused by faults and restricts the overall recovery efficiency of the power grid system.

[0003] When evaluating the resilience value of each node in the power grid system using traditional structural indicators, the demand response and load adjustable capacity of each node of the power grid system are not considered, and the dynamic distribution of the power flow of each node cannot be accurately evaluated, resulting in inaccurate evaluation of the actual resilience value of each node of the power grid system. The power recovery time in the area of the farthest root node is prolonged, which restricts the overall recovery efficiency of the power grid system. SUMMARY

[0004] To solve the above technical problems, the purpose of the present application is to provide a power grid resilience improvement reinforcement method based on source-grid-load multi-resource cooperation. The technical solution adopted is as follows: In a first aspect, a power grid resilience improvement reinforcement method based on source-grid-load multi-resource cooperation is provided. The method comprises: Obtaining the operation data sequence of each node in the power grid; the operation data sequence includes the power load data sequence, the power grid frequency data sequence, the active power data sequence, the power factor data sequence, and the new energy power generation data sequence; Analyzing the amplitude dispersion in the power load data sequence of each node, the fluctuation amplitude of the power load data sequence, and the average recovery time of each frequency deviation in the power grid frequency data sequence to recover to the preset frequency to obtain the load frequency response coefficient of the node; the load frequency response coefficient represents the response degree of the node to recover through load adjustment under frequency anomaly; The recovery rate of each power deviation in the active power data sequence of each node, the symmetry between the rising recovery section and the falling recovery section after each power deviation in the active power data sequence, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain a fault recovery coefficient of the node; the fault recovery coefficient represents the comprehensive adjustment degree of the node after the fault; and the new energy power generation prediction sequence is obtained by predicting the new energy power generation data sequence. In response to a fault occurring in the power grid, power recovery is performed with the target node as a starting node; and the target node screens a plurality of nodes according to the load frequency response coefficient and the fault recovery coefficient.

[0005] Optionally, the operation data sequence of each node in the power grid is obtained, including: In each monitoring interval, a plurality of types of operation data of each node in the power grid are obtained, and each type of operation data is sorted in time sequence to obtain a plurality of operation data sequences of the node in the monitoring interval.

[0006] Optionally, the load frequency response coefficient of each node is obtained by analyzing the amplitude dispersion in the power load data sequence of the node, the fluctuation amplitude of the power load data sequence, and the average recovery time of each frequency deviation in the power grid frequency data sequence to a preset frequency. All power load maximum points and all power load minimum points in the power load data sequence of each node are obtained, and each power load maximum point and the closest power load minimum point are determined as a pair of associated extreme points. The absolute difference between each pair of associated extreme points in the power load data sequence of each node is calculated to obtain the absolute difference of each pair of associated extreme points. The absolute differences of the plurality of pairs of associated extreme points in the power load data sequence of each node are sorted in time sequence to obtain an absolute difference sequence. The load frequency response coefficient of each node is obtained by analyzing the coefficient of variation of the absolute difference sequence in the power load data sequence of the node, the fluctuation amplitude of the power load data sequence, and the average recovery time of each frequency deviation in the power grid frequency data sequence to a preset frequency; and the amplitude dispersion in the power load data sequence of each node includes the coefficient of variation of the absolute difference sequence in the power load data sequence of each node.

[0007] Optionally, the load frequency response coefficient of each node is obtained by analyzing the coefficient of variation of the absolute difference sequence in the power load data sequence of the node, the fluctuation amplitude of the power load data sequence, and the average recovery time of each frequency deviation in the power grid frequency data sequence to a preset frequency. a difference between the maximum value and the minimum value in the power load data sequence of each node, to obtain a fluctuation amplitude of the node; the fluctuation amplitude of the power load data sequence of the node includes the fluctuation amplitude of the node; a product of the coefficient of variation of the absolute difference sequence in the power load data sequence of each node and the fluctuation amplitude of the node, to obtain a load adjustment smoothing coefficient of the node; the load adjustment smoothing coefficient represents a smoothing degree of load adjustment of the node; a load frequency response coefficient of the node is obtained by analyzing the load adjustment smoothing coefficient of each node and an average recovery time length of each frequency deviation in the power grid frequency data sequence recovering to the preset frequency.

[0008] Optionally, the load frequency response coefficient of the node is obtained by analyzing the load adjustment smoothing coefficient of each node and an average recovery time length of each frequency deviation in the power grid frequency data sequence recovering to the preset frequency, including: Based on the BG sequence segmentation algorithm, a plurality of power grid frequency mutation points of the power grid frequency data sequence of each node and a plurality of power load mutation points of the power load data sequence are obtained, and the plurality of power grid frequency mutation points are sorted in time sequence to obtain a power grid frequency mutation point sequence, and the plurality of power load mutation points are sorted in time sequence to obtain a power load mutation point sequence; a shortest time length of each power grid frequency mutation point in the power grid frequency data sequence of each node recovering to the preset frequency is calculated, to obtain a recovery time length of the power grid frequency mutation point; an average recovery time length of the plurality of power grid frequency mutation points in the power grid frequency data sequence of each node is calculated, to obtain an average recovery time length of the node; the average recovery time length of each frequency deviation in the power grid frequency data sequence recovering to the preset frequency includes the average recovery time length of the node; sequence decomposition is performed on the power grid frequency mutation point sequence and the power load mutation point sequence, to obtain a power grid frequency mutation point trend item and a power load mutation point trend item; a cosine similarity between the power grid frequency mutation point trend item and the power load mutation point trend item of each node is calculated, to obtain a coordination coefficient of the node; the coordination coefficient represents a consistency degree of load mutation and frequency mutation of the node in variation; a ratio of the coordination coefficient of each node to the average recovery time length of the node is calculated, to obtain a frequency recovery coefficient of the node; the frequency recovery coefficient represents a contribution degree of the node to power grid frequency recovery; a load frequency response coefficient of the node is obtained according to a product of the frequency recovery coefficient of each node and the load adjustment smoothing coefficient of the node.

[0009] Optionally, the average recovery rate of each power deviation in the active power data sequence of each node, the symmetry between the rising recovery segment and the falling recovery segment after each power deviation, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node, including: The mean value of a plurality of active power data in the active power sequence of each node is calculated to obtain the normal active power; and the absolute difference between the reference power and the normal active power is less than a preset difference value. For each extreme point in the active power sequence of each node, a plurality of active power data between the extreme point and the corresponding reference power point are obtained to obtain an active recovery sub-sequence corresponding to the extreme point; the reference power point corresponding to each extreme point indicates that the first active power data point after the extreme point is the reference power; A coordinate system is established with the time corresponding to the sampling point as the horizontal axis and each active power in the active recovery sub-sequence corresponding to each extreme point as the vertical coordinate, and the absolute value of the slope of the fitted straight line is calculated according to the least square method to obtain the absolute slope of the active recovery sub-sequence corresponding to each extreme point; the recovery rate of each power deviation to the reference power in the active power data sequence of each node includes the absolute slope of the active recovery sub-sequence corresponding to each extreme point. The sum of the absolute slopes of the active recovery sub-sequences corresponding to a plurality of extreme points in the active power data sequence of each node, the symmetry between the rising recovery segment and the falling recovery segment after each power deviation, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node.

[0010] Optionally, the sum of the absolute slopes of the active recovery sub-sequences corresponding to a plurality of extreme points in the active power data sequence of each node, the symmetry between the rising recovery segment and the falling recovery segment after each power deviation, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node, including: The absolute difference between each peak and the corresponding reference power point in the active power data sequence of each node is calculated respectively, and the absolute difference between a plurality of peaks and the corresponding reference power is sorted in the time order of the peaks to obtain a difference sequence corresponding to the peaks; the extreme points in the active power data sequence include peaks and troughs. Calculate the absolute difference between each wave trough in the active power data sequence of each node and the corresponding reference power point, and sort the absolute differences between multiple wave troughs and the corresponding reference power according to the time sequence of the wave troughs to obtain a difference sequence corresponding to the wave troughs; Calculate the DTW distance between the difference sequence corresponding to the wave peak and the difference sequence corresponding to the wave trough of each node to obtain the symmetry degree of the node; the symmetry degree of the node indicates the symmetry degree between the rising recovery segment and the falling recovery segment after the power deviation in the active power data sequence of the node; Analyze the product of the sum of the absolute slopes of the active recovery subsequences corresponding to the multiple extreme points in the active power data sequence of each node and the symmetry degree of the node, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence to obtain the fault recovery coefficient of the node.

[0011] Optionally, the analysis of the product of the sum of the absolute slopes of the active recovery subsequences corresponding to the multiple extreme points in the active power data sequence of each node and the symmetry degree of the node, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence to obtain the fault recovery coefficient of the node comprises: Calculate the average and variance of the power factor data sequence of each node, and calculate the product of the average and the variance to obtain the fluctuation degree of the power factor data sequence; Extract a reference subsequence from the new energy power generation data sequence of each node; Based on the reference subsequence, generate prediction data through an exponential moving average algorithm to construct a new energy power generation prediction sequence according to the prediction data; Calculate the cosine similarity between the new energy power generation data sequence and the new energy power generation prediction sequence to obtain the power generation similarity; the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence includes the power generation similarity; According to the product of the sum of the absolute slopes of the active recovery subsequences corresponding to the multiple extreme points in the active power data sequence of each node and the symmetry degree of the node, the fluctuation degree of the power factor data sequence, and the power generation similarity, obtain the fault recovery coefficient of the node.

[0012] Optionally, before the power restoration starting from the target node in response to the failure of the power grid, the method further comprises: Based on the TOPSIS advantage and disadvantage solution distance method, calculate the resilience and reliability coefficient of each node according to the load frequency response coefficient, the fault recovery coefficient, and the power index of each node; the power index includes the average path length, the betweenness centrality, and the connectivity of the node; According to the maximum inter-class variance method, the resilience reliability coefficients of multiple nodes are segmented to obtain a segmentation threshold; Nodes with a resilience reliability coefficient greater than or equal to the segmentation threshold are determined as target nodes.

[0013] Optionally, the power recovery is performed with the target nodes as starting nodes in response to the failure of the power grid, and the method comprises the following steps of: In response to the failure of the power grid, a recovery path of each target node is calculated according to a depth-first search algorithm; A recovery priority coefficient of the recovery path of the target node is calculated according to a sum of active power of all nodes in the recovery path of the target node, a total load power before the failure, a number of nodes in the recovery path, and a number of nodes with a failure in the power grid; The recovery path with the largest recovery priority coefficient is taken as a target path, and the power recovery is performed.

[0014] On the basis of common knowledge in the art, the above-mentioned preferred conditions can be combined arbitrarily, that is, the preferred examples of the present application are obtained.

[0015] The present application has the following beneficial effects: by obtaining power load data sequences, power grid frequency data sequences, active power data sequences, power factor data sequences and new energy power generation data sequences of each node in the power grid, the dynamic characteristics of each sequence are analyzed in multiple dimensions. Based on the amplitude dispersion and fluctuation amplitude of the power load data sequence, the average recovery time after the frequency deviation in the power grid frequency data sequence, the load frequency response coefficient representing the load adjustment response degree is calculated. Based on the recovery rate and recovery segment symmetry of the active power data sequence, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence, the failure recovery coefficient representing the comprehensive adjustment degree is calculated. Then, according to the load frequency response coefficient and the failure recovery coefficient, the target nodes are selected from the multiple nodes, and the power recovery is performed with the target nodes as starting nodes when the power grid fails. Thus, the recovery strategy failure or the risk of cascading overload caused by the misjudgment of the node recovery ability is avoided, the accuracy and efficiency of the failure recovery are significantly improved, the resilience of the power grid is improved through the source-grid-load multi-resource cooperation, the accurate evaluation of the real resilience recovery ability of the node is realized, the defects of the traditional method of ignoring the dynamic adjustment and recovery potential of the node are overcome, and thus the optimal recovery path starting point can be quickly and accurately located, and the overall recovery efficiency and resilience level of the power grid after the failure are effectively improved. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the accompanying drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the accompanying drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0017] Figure 1 A flow chart of a source network load multi-resource coordination-based power grid resilience improvement and reinforcement method in an embodiment; Figure 2 A structural schematic diagram of a source network load multi-resource coordination-based power grid resilience improvement and reinforcement system in an embodiment; Figure 3 A structural schematic diagram of an electronic device in an embodiment. DETAILED DESCRIPTION

[0018] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined purposes, the specific embodiments, structures, features and effects of the source network load multi-resource coordination-based power grid resilience improvement and reinforcement method according to the present application are described in detail below in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0020] The specific scheme of the source network load multi-resource coordination-based power grid resilience improvement and reinforcement method provided by the present application is specifically described below in combination with the accompanying drawings. As shown in Figure 1 The method comprises the following steps: S11, obtaining a running data sequence of each node in a power grid.

[0021] The running data sequence comprises a power load data sequence, a power grid frequency data sequence, an active power data sequence, a power factor data sequence and a new energy power generation data sequence.

[0022] In one embodiment, obtaining the running data sequence of each node in the power grid comprises: In each monitoring interval, a plurality of types of running data of each node in the power grid are obtained, and each type of running data is sorted in time sequence to obtain a plurality of running data sequences of the node in the monitoring interval.

[0023] The application acquires the power load, power grid frequency, active power, power factor and new energy power station power generation data of each node in the power network based on source network load multi-resource cooperation by deploying power factor test sensors, smart meters and AGC automatic power generation control systems at each system node in the power grid, acquires the phase difference between voltage and current in the power transmission line at each node and the rated frequency of each node power grid at each sampling time, aligns the time stamps of the related operation data of each node in the power network to the UTC (Coordinated Universal Time) index, and normalizes the related operation data by Z-Score standardization processing.

[0024] In order to prevent data loss caused by network fluctuations or electromagnetic interference during data transmission, the missing value filling processing is performed on the related operation data of the power grid by the polynomial interpolation method. Z-Score standardization and polynomial interpolation method are both known technologies, and the specific means will not be described again.

[0025] The application divides each 1h as a monitoring interval, acquires the power load, power grid frequency, active power, power factor and new energy power station power generation data of each node in the power grid in each monitoring interval, and sorts each type of operation data in time sequence respectively to obtain an operation data sequence corresponding to each type of operation data, so as to obtain multiple operation data sequences of the node in the monitoring interval, i.e. power load data sequence, power grid frequency data sequence, active power data sequence, power factor data sequence and new energy power generation data sequence. The operation data sequence in the following steps is in a monitoring interval, i.e. the multiple nodes in a monitoring interval are analyzed.

[0026] S12, the amplitude dispersion in the power load data sequence of each node, the fluctuation amplitude of the power load data sequence, the average recovery time length of each frequency deviation in the power grid frequency data sequence to the preset frequency is analyzed, and the load frequency response coefficient of the node is obtained.

[0027] The load frequency response coefficient represents the response degree of the node to recover through load adjustment under frequency anomaly. The preset frequency is 50Hz. The frequency anomaly indicates that the absolute difference between the preset frequency is greater than or equal to 0.5Hz.

[0028] In one embodiment, the amplitude dispersion in the power load data sequence of each node, the fluctuation amplitude of the power load data sequence, the average recovery time length of each frequency deviation in the power grid frequency data sequence to the preset frequency is analyzed, and the load frequency response coefficient of the node is obtained, including: all power load maximum points and all power load minimum points in the power load data sequence of each node are obtained, and each power load maximum point and the interval nearest power load minimum point are determined as a pair of associated extreme points; The absolute difference value between each pair of associated extreme points in the power load data sequence of each node is calculated respectively to obtain the absolute difference value of each pair of associated extreme points; The absolute difference values of multiple pairs of associated extreme points in the power load data sequence of each node are sorted in time sequence to obtain an absolute difference value sequence; The variation coefficient of the absolute difference value sequence in the power load data sequence of each node, the fluctuation amplitude of the power load data sequence, and the average recovery time length of each frequency deviation to the preset frequency in the power grid frequency data sequence are analyzed to obtain the load frequency response coefficient of the node; the amplitude dispersion in the power load data sequence of each node includes the variation coefficient of the absolute difference value sequence in the power load data sequence of each node.

[0029] Specifically, the variation coefficient of the absolute difference value sequence in the power load data sequence of each node, the fluctuation amplitude of the power load data sequence, and the average recovery time length of each frequency deviation to the preset frequency in the power grid frequency data sequence are analyzed to obtain the load frequency response coefficient of the node, including: The difference between the maximum value and the minimum value in the power load data sequence of each node is calculated to obtain the fluctuation amplitude of the node; the fluctuation amplitude of the power load data sequence of the node includes the fluctuation amplitude of the node; The product of the variation coefficient of the absolute difference value sequence in the power load data sequence of each node and the fluctuation amplitude of the node is calculated to obtain the load adjustment smoothing coefficient of the node; the load adjustment smoothing coefficient represents the smoothing degree of the load adjustment of the node; The load adjustment smoothing coefficient of each node and the average recovery time length of each frequency deviation to the preset frequency in the power grid frequency data sequence are analyzed to obtain the load frequency response coefficient of the node.

[0030] Specifically, the load adjustment smoothing coefficient of each node and the average recovery time length of each frequency deviation to the preset frequency in the power grid frequency data sequence are analyzed to obtain the load frequency response coefficient of the node, including: Based on the BG sequence segmentation algorithm, multiple power grid frequency mutation points of the power grid frequency data sequence of each node and multiple power load mutation points of the power load data sequence are obtained, and the multiple power grid frequency mutation points are sorted in time sequence to obtain a power grid frequency mutation point sequence, and the multiple power load mutation points are sorted in time sequence to obtain a power load mutation point sequence; recovering time of each power grid frequency mutation point in the power grid frequency data sequence in each node is calculated to obtain the recovering time of the power grid frequency mutation point; The average recovering time of each node is calculated by averaging the recovering times of the multiple power grid frequency mutation points in the power grid frequency data sequence of the node. The power grid frequency mutation point sequence and the power load mutation point sequence are decomposed to obtain a power grid frequency mutation point trend item and a power load mutation point trend item. The cosine similarity between the power grid frequency mutation point trend item and the power load mutation point trend item of each node is calculated to obtain a coordination coefficient of the node. The ratio of the coordination coefficient of each node to the average recovering time of the node is calculated to obtain a frequency recovering coefficient of the node. The load frequency response coefficient of each node is obtained according to the product of the frequency recovering coefficient of the node and the load adjustment smoothing coefficient of the node.

[0031] In the power network under the source-grid-load multi-resource coordination, the demand response capability and the load adjustable capability of each distributed power system node can reflect the output adjustment status of the power system node to the new energy power and the recovery status of the power fluctuation, effectively represent the tolerance of the power system to the power grid regulation pressure and the fault risk cascading, and are important resilience resource evaluation indexes for evaluating the recovery capability of the global power network under the disturbance and fault state. The power system node with strong source-load regulation capability should be established as a regional resilience hub to support the rapid and efficient local self-healing and power supply recovery of the surrounding area.

[0032] Specifically, in the power network under the source-grid-load multi-resource coordination, the stronger the demand response capability and the load adjustable capability of the power system node are, the more significant the peak-valley smoothness of the power load curve is due to the superior peak clipping and valley filling capability of the power system node, and the wider the power load adjustment range is. Meanwhile, the higher the response load resource capability of the power system node is, the shorter the time for recovering to the rated frequency when the power grid frequency fluctuates, and the more obvious the power system negative feedback mechanism is, that is, the higher the correlation between the node load and the power grid frequency fluctuation trend is.

[0033] The application divides each 1h as a monitoring interval, constructs a load frequency response condition, and is used for characterizing the load adjustment smoothness and frequency recovery correlation of any power system node in a power network. The subsequent analysis is performed by taking any monitoring interval in the power network operation process as an example.

[0034] All maximum points and minimum points in the power load data sequence corresponding to each power system node are obtained, each power load maximum value in the power load data sequence and the nearest power load minimum value in the time interval are recorded as a pair of associated extreme points, the absolute difference value between each pair of associated extreme points in the power load data sequence of each power system node is calculated, and the absolute difference values of multiple pairs of associated extreme points in the power load data sequence of each node are sorted in time sequence to obtain an absolute difference value sequence.

[0035] The power load data sequence and the grid frequency data sequence of any power system node in the power network operation process are respectively taken as inputs, the BG (Bernaola Galvan) sequence segmentation algorithm is used to obtain the power load mutation points and the grid frequency mutation points of each power system node, multiple grid frequency mutation points are sorted in time sequence to obtain a grid frequency mutation point sequence, multiple power load mutation points are sorted in time sequence to obtain a power load mutation point sequence, the time interval between each grid frequency mutation point and the first preset frequency (rated grid frequency) data after the grid frequency mutation point is recorded as the shortest time interval between the recovery of each grid frequency mutation point to the rated grid frequency, and the recovery duration of the grid frequency mutation point is obtained. The average of the recovery durations of multiple grid frequency mutation points in the grid frequency data sequence of each node is calculated to obtain the average recovery duration of the node.

[0036] The grid frequency mutation point sequence and the power load mutation point sequence are subjected to sequence decomposition by the STL (Seasonal and Trend decomposition using Loess) sequence decomposition algorithm to obtain a grid frequency mutation point trend item and a power load mutation point trend item, the cosine similarity between the grid frequency mutation point trend item and the power load mutation point trend item of each node is calculated to obtain the coordination coefficient of the node.

[0037] The calculation formula of the load frequency response coefficient of the i th node is: . Wherein, is the load frequency response coefficient of the i th node, is the load adjustment smoothing coefficient of the i th node in the power network operation process, The calculation formula of the load frequency response coefficient of the i th node is: ​​is the coefficient of variation of the absolute difference sequence in the power load data sequence corresponding to the i-th node, and the coefficient of variation is calculated as the ratio of the standard deviation to the mean of the absolute difference sequence, is the fluctuation amplitude of the i-th node, i.e., the difference between the maximum and minimum values in the power load data sequence corresponding to the i-th node. is the frequency recovery coefficient of the i-th node during the operation of the power network, The calculation formula is: wherein is the average recovery time of the i-th node, i.e., the shortest time interval between the recovery of each power grid frequency mutation point to the rated power grid frequency in the power grid frequency data sequence of the i-th node, is the coordination coefficient of the i-th node, i.e., the cosine similarity between the trend item of the power load mutation point and the trend item of the power grid frequency mutation point; norm() is a normalization function, so that the value of is in the range of [0, 1].

[0038] The load frequency response coefficient reflects the load adjustment smoothness and frequency recovery correlation of any power system node in each monitoring interval during the operation of the power network. The load adjustment smoothness coefficient reflects the peak-valley smoothing feature and power load adjustment range condition of any power system node in each monitoring interval during the operation of the power network. The frequency recovery coefficient represents the time-consuming speed of the recovery of the power grid frequency and the positive correlation between the power system load and the frequency during the operation of the power network.

[0039] In the power network under the coordination of multiple resources of source, network and load, when the demand response capability and load adjustable capability of the power system node are stronger, the peak-valley data change smoothness of the power system node load data is more significant, the power load adjustment range is more extensive, i.e., the load adjustment smoothness coefficient becomes larger. At the same time, the power support of the power system node responding to the load resource capability and the rapid frequency deviation is stronger, the time-consuming of the recovery of the power grid frequency to the rated frequency is shorter, and the correlation between the power load and the power grid frequency fluctuation trend item is stronger, i.e., the frequency recovery coefficient becomes larger.

[0040] There are still certain technical drawbacks in evaluating the power recovery resilience value by only running the load frequency response status of each power system node in the power network, lacking analysis of the active power, power factor and new energy output status of the power system node, ignoring the consideration of power flow imbalance, energy efficiency loss and new energy output uncertainty of the power system node, and being difficult to evaluate the recovery potential and vulnerability of the power system node, resulting in misjudgment of the resilience value of each node in the power network under the source-grid-load multi-resource coordination, leading to failure of subsequent power network recovery strategy, and increasing the risk of power network operation recovery.

[0041] Specifically, in the process of power network operation under the source-grid-load multi-resource coordination, when the recovery potential and recovery support capacity of the power system node are stronger, the peak-valley recovery rate of the active power data of the power system node is faster, and the power shortage and surplus status after the failure of the power node can be quickly responded, and the peak-valley of the active power data presents high symmetry; at the same time, the node with strong recovery support capacity can provide reactive voltage support after failure, so the power factor of the power system node presents high level and stable condition, and the predictability of the new energy generation is higher.

[0042] S13, analyzing the recovery rate of each power deviation in the active power data sequence of each node to the reference power, the symmetry between the rising recovery segment and the falling recovery segment after each power deviation in the active power data sequence, the fluctuation degree of the power factor data sequence, and the similarity between the new energy generation data sequence and the new energy generation prediction sequence, to obtain the fault recovery coefficient of the node.

[0043] The fault recovery coefficient represents the comprehensive adjustment degree of the node after failure.

[0044] The new energy generation prediction sequence is obtained by predicting the new energy generation data sequence.

[0045] In one embodiment, the average recovery rate of each power deviation in the active power data sequence of each node to the preset power, the symmetry between the rising recovery segment and the falling recovery segment after each power deviation in the active power data sequence, the fluctuation degree of the power factor data sequence, and the similarity between the new energy generation data sequence and the new energy generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node, including: The mean value of a plurality of active power data in the active power sequence of each node is calculated to obtain the normal active power; the absolute difference between the reference power and the normal active power is less than the preset difference value; For each extreme point in the active power sequence of each node, a plurality of active power data between the extreme point and the corresponding reference power point is obtained, to obtain an active power recovery sub-sequence corresponding to the extreme point; the reference power point corresponding to each extreme point indicates that the first active power data point after the extreme point is a reference power point; A coordinate system is established with the time corresponding to the sampling point as the horizontal axis and each active power in the active power recovery sub-sequence corresponding to each extreme point as the vertical coordinate, and a straight line fitting is performed according to the least square method, to calculate the absolute value of the slope of the fitted straight line, to obtain the absolute slope of the active power recovery sub-sequence corresponding to each extreme point; the recovery rate of each power after the power deviates in the active power data sequence of each node to the reference power includes the absolute slope of the active power recovery sub-sequence corresponding to each extreme point; The sum of the absolute slopes of the active power recovery sub-sequences corresponding to the plurality of extreme points in the active power data sequence of each node, the symmetry between the rising recovery section and the falling recovery section after the power deviates in the active power data sequence, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node.

[0046] Specifically, the sum of the absolute slopes of the active power recovery sub-sequences corresponding to the plurality of extreme points in the active power data sequence of each node, the symmetry between the rising recovery section and the falling recovery section after the power deviates in the active power data sequence, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node, including: The absolute difference between each peak and the corresponding reference power point in the active power data sequence of each node is calculated respectively, and the absolute difference between the plurality of peaks and the corresponding reference power is sorted according to the time sequence of the peaks to obtain a difference sequence corresponding to the peaks; the extreme points in the active power data sequence include peaks and troughs; The absolute difference between each trough and the corresponding reference power point in the active power data sequence of each node is calculated respectively, and the absolute difference between the plurality of troughs and the corresponding reference power is sorted according to the time sequence of the troughs to obtain a difference sequence corresponding to the troughs; The DTW distance between the difference sequence corresponding to the peaks and the difference sequence corresponding to the troughs of each node is calculated to obtain the symmetry of the node; the symmetry of the node indicates the symmetry between the rising recovery section and the falling recovery section after the power deviates in the active power data sequence of the node; The product of the sum of the absolute slopes of the active power recovery subsequences corresponding to the multiple extreme points in the active power data sequence of each node and the symmetry of the node, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node.

[0047] Specifically, the product of the sum of the absolute slopes of the active power recovery subsequences corresponding to the multiple extreme points in the active power data sequence of each node and the symmetry of the node, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node, including: The average and variance of the power factor data sequence of each node are calculated, and the product of the average and the variance is calculated to obtain the fluctuation degree of the power factor data sequence. A reference subsequence is extracted from the new energy power generation data sequence of each node. Based on the reference subsequence, a prediction data is generated by an exponential moving average algorithm to construct a new energy power generation prediction sequence according to the prediction data. The cosine similarity between the new energy power generation data sequence and the new energy power generation prediction sequence is calculated to obtain the power generation similarity; the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence includes the power generation similarity. The product of the sum of the absolute slopes of the active power recovery subsequences corresponding to the multiple extreme points in the active power data sequence of each node and the symmetry of the node, the fluctuation degree of the power factor data sequence, and the power generation similarity is obtained to obtain the fault recovery coefficient of the node.

[0048] The preset difference is set by itself according to the actual situation, for example, 1% to 5% of the normal active power.

[0049] The application constructs a fault recovery coefficient for representing the active power recovery response condition of each power system node in the process of power network operation under the source-network-load multi-resource cooperation and the reactive new energy support condition. All extreme points, i.e., all peaks and troughs, in the active power data sequence of each power system node in a monitoring interval are obtained, the average of multiple active power data of each power system node in the monitoring interval is calculated to obtain the normal active power.

[0050] Each extreme point in the active power data sequence of the power system node and the first data point after the extreme point with the active power as the reference power are determined as the reference power point corresponding to the extreme point, all active power data between each extreme point and the corresponding reference power point are obtained, and an active power recovery sub-sequence corresponding to the extreme point is obtained. A coordinate system is established with the time corresponding to the sampling point as the horizontal axis and each active power in the active power recovery sub-sequence corresponding to each extreme point as the vertical coordinate, and the absolute value of the slope of the fitted straight line is calculated according to the least square method, to obtain the absolute slope of the active power recovery sub-sequence corresponding to each extreme point.

[0051] A reference sub-sequence is extracted from the new energy power generation data sequence of each node. Specifically, the first 1 / 5 to 1 / 3 of the new energy power generation data sequence of each power system node in the monitoring interval is extracted as the reference sub-sequence. The more data elements in the reference sub-sequence, the more accurate the new energy power generation data prediction result. In this application, the first 1 / 3 of the data sub-section is extracted from the new energy power generation data sequence of the power system node as the reference sub-sequence, the Exponential Moving Average (EMA) algorithm is used to model the reference sub-sequence, and the prediction data of the next time is predicted based on the model. The obtained prediction data is appended to the end of the reference sub-sequence, thereby forming an expanded sequence. The updated sequence is used as the new input, and the above EMA prediction and sequence expansion operations are repeatedly performed, that is, the power generation at the next time step is predicted by the EMA algorithm using the current reference sub-sequence (always 1 / 3 of the original sequence length or a dynamic sliding window), and the prediction result is sequentially appended to the end of the sequence, gradually generating a complete prediction sequence that is time-aligned with the original new energy power generation data sequence and has the same length. The initial measured new energy power generation data sequence (i.e., the first 1 / 3 part) of the node and all subsequent prediction data generated by EMA recursion are spliced in time sequence to form a complete new energy power generation prediction sequence, which is used for subsequent calculation of the fault recovery coefficient and resilience evaluation.

[0052] The fault recovery coefficient of the i-th node The calculation formula of the fault recovery coefficient of the i-th node is: . Wherein, is the fault recovery coefficient of the i-th node, is the active power recovery response coefficient of the i-th node, The calculation formula of the fault recovery coefficient of the i-th node is: , is the sum of the absolute slopes of the active power recovery subsequence corresponding to the multiple extreme points in the active power data sequence corresponding to the ith node. The absolute difference between each peak and the corresponding reference power point in the active power data sequence of the ith node is calculated, and the absolute differences between the multiple peaks and the corresponding reference power are sorted in the chronological order of the peaks to obtain a peak difference value sequence. The absolute difference between each valley and the corresponding reference power point in the active power data sequence of the ith node is calculated, and the absolute differences between the multiple valleys and the corresponding reference power are sorted in the chronological order of the valleys to obtain a valley difference value sequence, is the DTW distance between the peak difference value sequence and the valley difference value sequence in the active power data sequence of the ith node. is the reactive power new energy support coefficient of the ith node in the process of operation of the power network, The calculation formula of is as follows: , wherein is the fluctuation degree of the ith node, that is, the product of the average value and the variance of the multiple data in the power factor data sequence corresponding to the ith node, is the power generation similarity of the ith node, that is, the cosine similarity between the new energy power generation data sequence of the ith node and the new energy power generation prediction sequence, and norm() is a normalization function, so that the value range of is within the range of [0, 1].

[0053] The fault recovery coefficient reflects the active power recovery response condition and the reactive power new energy support condition of each power system node in the process of operation of the power network in the monitoring interval. The active power recovery response coefficient reflects the rapid recovery of the peak and valley of the active power data and the high symmetry of the peak and valley of the active power data of each power system node in the process of operation of the power network in the monitoring interval, and the reactive power new energy support coefficient reflects the high level stability of the power factor and the predictability of the new energy power generation of each power system node in the process of operation of the power network in the monitoring interval. In the process of operation of the power network under the source-grid-load multi-resource cooperation, when the recovery potential and the recovery support ability of the power system node are stronger, the peak and valley of the active power of the power system node recover faster. Due to the power shortage and excess after the failure of the power node, the peak and valley of the active power present a high symmetry, that is, the active power recovery response coefficient becomes larger, at the same time, the power factor of the power system node presents a high level and stable condition, and the predictability of the new energy power generation is stronger, that is, the reactive power new energy support coefficient becomes larger.

[0054] ​In the process of power network operation under the source-network-load multi-resource coordination, when the load frequency response and adjustable capacity of the power system node are stronger, the power recovery potential and support capacity when the fault occurs are stronger, the power system node should be regarded as a reliable node in the process of power network resilience recovery, and the reliable node is gradually radiated to the surrounding as the root node in the process of power recovery of each power system node in the power network, helping the surrounding power system nodes to recover power normal operation faster and more efficiently.

[0055] S14, in response to the power grid fault, performing power recovery with the target node as a starting node.

[0056] The target node is selected from the plurality of nodes according to the load frequency response coefficient and the fault recovery coefficient.

[0057] In one embodiment, before the power recovery with the target node as a starting node in response to the power grid fault, the method further comprises: Based on the TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) method, the resilience reliable coefficient of each node is calculated according to the load frequency response coefficient, the fault recovery coefficient and the power index of each node. The power index includes the average path length, the betweenness centrality and the connectivity of the node. According to the maximum inter-class variance method, the resilience reliable coefficients of the plurality of nodes are segmented to obtain a segmentation threshold. The node with the resilience reliable coefficient greater than or equal to the segmentation threshold is determined as the target node.

[0058] The resilience reliable coefficient is constructed to represent the resilience reliable degree of each power system node in the process of fault recovery in the process of power network operation. The resilience reliable coefficient can be obtained by the load frequency response coefficient, the fault recovery coefficient and the power index in the process of power grid structure resilience evaluation. Specifically, the power index of each power system node in the power network under the source-network-load multi-resource coordination is normalized, the load frequency response coefficient, the fault recovery coefficient and each power index of each power system node in the process of power network operation are taken as inputs, and the TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) method is used to obtain the resilience reliable coefficient of each power system node.

[0059] The power indicators can be an average path length (an average of power transmission line lengths of the rest of the power system), a betweenness centrality (a number of shortest paths passing through each power system node in the power network), and a connectivity (a sum of in-degree and out-degree of the power system node in the power network, and when there is a power transmission line between one power system node and another power system node in the power network, it is considered that there is an edge), a load frequency response coefficient, a fault recovery coefficient, and a calculation weight of each structural indicator can be obtained by an entropy weight method, and the TOPSIS optimal and poor solution distance method and the entropy weight method are both known technologies, and specific means will not be described again.

[0060] The resilience reliability coefficients of all power system nodes in the power network during operation in a single monitoring interval are taken as inputs, a maximum interclass variance method is used to obtain a segmentation threshold, and a node with a resilience reliability coefficient greater than or equal to the segmentation threshold is taken as a resilience recovery root node of the power network during failure resilience in the monitoring interval, that is, a target node.

[0061] In one embodiment, in response to a power grid failure, power recovery is performed with the target node as a starting node, including: In response to the power grid failure, a recovery path of each target node is calculated according to a depth-first search algorithm; A recovery priority coefficient of the recovery path of each target node is calculated according to a sum of active power of all nodes in the recovery path of each target node, a total load power before the failure, a number of nodes in the recovery path, and a number of nodes of the failure in the power grid; A recovery path with a maximum recovery priority coefficient is taken as a target path, and power recovery is performed.

[0062] In response to a power grid failure, the power network topology is obtained as an input, and in the order from high to low of the resilience reliability coefficient, the recovery path of each target node after one iteration of the depth-first search algorithm (DFS) is obtained in turn, the sum of the current active power of all power system nodes in the recovery path corresponding to each target node is calculated, the total load power before the failure is calculated, the ratio of the sum of the current active power of all power system nodes in the recovery path to the total load power before the failure is taken as the load recovery coefficient of each target node, the ratio between the total number of nodes in the recovery path corresponding to each target node and the total number of nodes affected by the failure in the power network is obtained, and the connected node ratio of the target node is obtained. The product of the load recovery coefficient and the connected node ratio of each target node is taken as the recovery priority coefficient of the recovery path corresponding to each target node, the recovery path with the largest recovery priority coefficient is selected as the target path, power recovery is performed, and after the power of all power system nodes on the target path in the power network is recovered, the remaining nodes on all non-target paths in the power network are sorted in descending order according to the resilience reliability coefficient, and the power of the remaining nodes in the power network is recovered in turn according to the descending order sorting result.

[0063] The application obtains the power load data sequence, the power grid frequency data sequence, the active power data sequence, the power factor data sequence and the new energy power generation data sequence of each node in the power grid, and performs multi-dimensional analysis on the dynamic characteristics of each sequence. Based on the amplitude dispersion and fluctuation amplitude of the power load data sequence, the average recovery time after the frequency deviation in the power grid frequency data sequence, the load frequency response coefficient representing the load adjustment response degree is calculated, based on the recovery rate and recovery segment symmetry of the active power data sequence, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence, the fault recovery coefficient representing the comprehensive adjustment degree is calculated, and then the target node is selected from the multiple nodes according to the load frequency response coefficient and the fault recovery coefficient, and the power recovery is performed with the target node as the starting node when the power grid fails. Therefore, the recovery strategy failure or cascading overload risk caused by the misjudgment of the node recovery ability is avoided, the accuracy and efficiency of the fault recovery are significantly improved, the source network load multi-resource collaborative power grid resilience is improved, the accurate evaluation of the real resilience recovery ability of the node is realized, the defects of the traditional method of ignoring the dynamic adjustment and recovery potential of the node are overcome, and therefore the optimal recovery path starting point can be quickly and accurately located, and the overall recovery efficiency and resilience level of the power grid after the failure are effectively improved.

[0064] It should be understood that, although Figure 1The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0065] This application also provides a grid resilience enhancement and hardening system that coordinates multiple resources (source, grid, and load), such as... Figure 2 As shown, the system includes: The acquisition module 21 is used to acquire the operating data sequence of each node in the power grid; the operating data sequence includes the power load data sequence, the power grid frequency data sequence, the active power data sequence, the power factor data sequence, and the new energy power generation data sequence. The first analysis module 22 is used to analyze the amplitude dispersion, fluctuation amplitude, and average recovery time of each frequency deviation to the preset frequency in the power load data sequence of each node, and to obtain the load frequency response coefficient of the node. The load frequency response coefficient represents the responsiveness of the node to recovery through load regulation under frequency anomalies. The second analysis module 23 is used to analyze the recovery rate of each power deviation in the active power data sequence of each node to the reference power, the symmetry between the rising and falling recovery segments after each power deviation in the active power data sequence, the volatility of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence, to obtain the fault recovery coefficient of the node; the fault recovery coefficient characterizes the overall regulation degree of the node after a fault; the new energy power generation prediction sequence is obtained by predicting based on the new energy power generation data sequence; The recovery module 24 is used to restore power in response to a grid fault, starting with a target node; the target node is selected from multiple nodes based on the load frequency response coefficient and the fault recovery coefficient.

[0066] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts are referred to the part of the method embodiments. The system embodiments described above are only illustrative, wherein the units described as separate components can or can not be physically separated, and the components of the units can or can not be physical units, i.e., can be located in one place or distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purposes of the present application.

[0067] Figure 3 A structure diagram of an electronic device is shown for an example embodiment of the present application, which includes a memory, a processor, and a computer program stored in the memory and used for running on the processor, and the processor implements the method of any of the above embodiments when executing the computer program. Figure 3 The electronic device 30 shown is only an example and should not limit the functions and use range of the embodiments of the present application.

[0068] As shown in Figure 3 The electronic device 30 can be in the form of a general computing device, for example, it can be a server device. The components of the electronic device 30 can include but are not limited to: the above-mentioned at least one processor 31, the above-mentioned at least one memory 32, the bus 33 connecting different system components (including the memory 32 and the processor 31).

[0069] The bus 33 includes a data bus, an address bus, and a control bus.

[0070] The memory 32 can include volatile memory, such as a random access memory (RAM) 321 and / or a cache memory 322, and can further include a read-only memory (ROM) 323.

[0071] The memory 32 can further include a program tool 325 (or utility tool) having a set of (at least one) program modules 324, such as an operating system, one or more application programs, other program modules, and program data, each of which or some combination of which can include the implementation of a network environment.

[0072] The processor 31 performs various function applications and data processing by running the computer program stored in the memory 32, such as the method provided by any of the above embodiments.

[0073] The electronic device 30 can also communicate with one or more external devices 34 such as a keyboard, a pointing device, etc. through an input / output (I / O) interface 35. Further, the electronic device 30 can communicate with one or more networks such as a local area network (LAN), a wide area network (WAN), and / or a public network such as the Internet, through a network adapter 36. As depicted, the network adapter 36 communicates with the other modules of the electronic device 30 through the bus 33. It should be appreciated that other hardware and / or software modules can be used in conjunction with the electronic device 30, including but not limited to, microcode, device drivers, redundant processing units, external disk drive arrays, RAID (Redundant Array of Independent Disks) systems, tape drives, and data backup storage systems, etc.

[0074] It should be noted that although several units / modules or sub-units / modules of the electronic device are mentioned in the foregoing detailed description, such a division is merely exemplary and not mandatory. Indeed, according to an embodiment of the application, the features and functions of two or more units / modules described above can be embodied in one unit / module. Conversely, the features and functions of one unit / module described above can be further divided into units / modules embodied by several units / modules.

[0075] The embodiments of the present application further provide a computer readable storage medium, having stored thereon a computer program, which when executed by a processor, implements the method provided in any of the embodiments described above.

[0076] More specifically, the computer readable storage medium can include, but is not limited to, a portable disc, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory, an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0077] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0078] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the method described in any of the above embodiments.

[0079] The program code for executing the computer program product of this application can be written in any combination of one or more programming languages. The program code can be executed entirely on the user device, partially on the user device, as a standalone software package, partially on the user device and partially on a remote device, or entirely on a remote device.

[0080] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0081] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these modifications and improvements all fall within the protection scope of this application.

[0082] The various embodiments described in this specification are presented by way of example, and each embodiment is not inherently more important than any other embodiment.

Claims

1. A method for improving the resilience of a power grid by source-network-payload multi-resource coordination, characterized in that, The method comprises: obtaining a running data sequence of each node in the power grid; the running data sequence comprises a power load data sequence, a power grid frequency data sequence, an active power data sequence, a power factor data sequence, and a new energy power generation data sequence; analyzing the amplitude dispersion in the power load data sequence of each node, the fluctuation amplitude of the power load data sequence, and the average recovery time length after each frequency deviation in the power grid frequency data sequence recovers to a preset frequency to obtain a load frequency response coefficient of the node; the load frequency response coefficient represents the response degree of the node to recovery through load adjustment under frequency abnormality; analyzing the recovery rate of each power deviation in the active power data sequence of each node to a reference power, the symmetry between the rising recovery section and the falling recovery section after each power deviation in the active power data sequence, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and a new energy power generation prediction sequence to obtain a fault recovery coefficient of the node; the fault recovery coefficient represents the comprehensive adjustment degree of the node after a fault; the new energy power generation prediction sequence is obtained by prediction based on the new energy power generation data sequence; in response to a fault occurring in the power grid, performing power recovery with a target node as a starting node; the target node is selected from multiple nodes according to the load frequency response coefficient and the fault recovery coefficient.

2. The method of claim 1, wherein the method further comprises: The method comprises: in each monitoring interval, obtaining multiple types of running data of each node in the power grid, and sorting each type of running data in time sequence to obtain multiple running data sequences of the node in the monitoring interval.

3. The method of claim 1, wherein the method further comprises: The method comprises: obtaining all power load maximum points and all power load minimum points in the power load data sequence of each node, and determining each power load maximum point and the power load minimum point closest to the power load maximum point as a pair of associated extreme points; calculating the absolute difference between each pair of associated extreme points in the power load data sequence of each node to obtain the absolute difference of each pair of associated extreme points; sorting the absolute differences of multiple pairs of associated extreme points in the power load data sequence of each node in time sequence to obtain an absolute difference sequence; analyzing the coefficient of variation of the absolute difference sequence in the power load data sequence of each node, the fluctuation amplitude of the power load data sequence, and the average recovery time length after each frequency deviation in the power grid frequency data sequence recovers to a preset frequency to obtain a load frequency response coefficient of the node; the amplitude dispersion in the power load data sequence of each node comprises the coefficient of variation of the absolute difference sequence in the power load data sequence of each node.

4. The method of claim 3, wherein the method further comprises: The variation coefficient of the absolute difference sequence in the power load data sequence of each node, the fluctuation amplitude of the power load data sequence, and the average recovery time length of the frequency deviation in the power grid frequency data sequence are analyzed to obtain the load frequency response coefficient of the node, including: The difference between the maximum value and the minimum value in the power load data sequence of each node is calculated to obtain the fluctuation amplitude of the node; the fluctuation amplitude of the power load data sequence of the node includes the fluctuation amplitude of the node; The product of the variation coefficient of the absolute difference sequence in the power load data sequence of each node and the fluctuation amplitude of the node is calculated to obtain the load adjustment smoothing coefficient of the node; the load adjustment smoothing coefficient represents the smoothing degree of the load adjustment of the node; The load adjustment smoothing coefficient of each node and the average recovery time length of the frequency deviation in the power grid frequency data sequence are analyzed to obtain the load frequency response coefficient of the node.

5. The method of claim 4, wherein the method further comprises: The load adjustment smoothing coefficient of each node and the average recovery time length of the frequency deviation in the power grid frequency data sequence are analyzed to obtain the load frequency response coefficient of the node, including: Based on the BG sequence segmentation algorithm, a plurality of power grid frequency mutation points of the power grid frequency data sequence and a plurality of power load mutation points of the power load data sequence of each node are obtained, and the plurality of power grid frequency mutation points are sorted in time sequence to obtain a power grid frequency mutation point sequence, and the plurality of power load mutation points are sorted in time sequence to obtain a power load mutation point sequence; The shortest time length for each power grid frequency mutation point in the power grid frequency data sequence of each node to recover to a preset frequency is calculated to obtain the recovery time length of the power grid frequency mutation point; The average of the recovery time lengths of the plurality of power grid frequency mutation points in the power grid frequency data sequence of each node is calculated to obtain the average recovery time length of the node; the average recovery time length of the frequency deviation in the power grid frequency data sequence includes the average recovery time length of the node; The power grid frequency mutation point sequence and the power load mutation point sequence are subjected to sequence decomposition to obtain a power grid frequency mutation point trend item and a power load mutation point trend item; The cosine similarity between the power grid frequency mutation point trend item and the power load mutation point trend item of each node is calculated to obtain the coordination coefficient of the node; the coordination coefficient represents the consistency degree of the load mutation and the frequency mutation of the node in variation; The ratio of the coordination coefficient of each node to the average recovery time length of the node is calculated to obtain the frequency recovery coefficient of the node; the frequency recovery coefficient represents the contribution degree of the node to the power grid frequency recovery; The load frequency response coefficient of each node is obtained according to the product of the frequency recovery coefficient of the node and the load adjustment smoothing coefficient of the node.

6. The method of claim 1, wherein the method further comprises: The average recovery rate of each power deviation in the active power data sequence of each node, the symmetry between the rising recovery section and the falling recovery section after each power deviation in the active power data sequence, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node, including: The mean value of a plurality of active power data in the active power sequence of each node is calculated to obtain the normal active power; and the absolute difference between the reference power and the normal active power is less than a preset difference value; For each extreme point in the active power sequence of each node, a plurality of active power data between the extreme point and the corresponding reference power point are obtained to obtain an active recovery sub-sequence corresponding to the extreme point; the reference power point corresponding to each extreme point indicates that the first active power data point after the extreme point is the reference power; A coordinate system is established with the time corresponding to the sampling point as the horizontal axis and each active power in the active recovery sub-sequence corresponding to each extreme point as the vertical coordinate, and the absolute value of the slope of the fitted straight line is calculated according to the least square method to obtain the absolute slope of the active recovery sub-sequence corresponding to each extreme point; the recovery rate of each power deviation to the reference power in the active power data sequence of each node includes the absolute slope of the active recovery sub-sequence corresponding to each extreme point; The sum of the absolute slopes of the active recovery sub-sequences corresponding to a plurality of extreme points in the active power data sequence of each node, the symmetry between the rising recovery section and the falling recovery section after each power deviation in the active power data sequence, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node.

7. The method of claim 6, wherein the method further comprises: The sum of the absolute slopes of the active recovery sub-sequences corresponding to a plurality of extreme points in the active power data sequence of each node, the symmetry between the rising recovery section and the falling recovery section after each power deviation in the active power data sequence, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node, including: The absolute difference between each wave crest and the corresponding reference power point in the active power data sequence of each node is calculated respectively, and the absolute difference between a plurality of wave crests and the corresponding reference power is sorted in the time order of the wave crests to obtain a difference sequence corresponding to the wave crests; the extreme points in the active power data sequence include the wave crests and the wave troughs; The absolute difference between each wave trough and the corresponding reference power point in the active power data sequence of each node is calculated respectively, and the absolute difference between a plurality of wave troughs and the corresponding reference power is sorted in the time order of the wave troughs to obtain a difference sequence corresponding to the wave troughs; Calculate the DTW distance between the difference value sequence corresponding to the peak and the difference value sequence corresponding to the trough of each node, to obtain the symmetry of the node; the symmetry of the node indicates the symmetry between the rising recovery segment and the falling recovery segment after each power deviation in the active power data sequence of the node; The product of the sum of the absolute slopes of the active recovery subsequences corresponding to the multiple extreme points in the active power data sequence of each node and the symmetry of the node, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node.

8. The method of claim 7, wherein the method further comprises: The product of the sum of the absolute slopes of the active recovery subsequences corresponding to the multiple extreme points in the active power data sequence of each node and the symmetry of the node, the fluctuation degree of the power factor data sequence, and the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence are analyzed to obtain the fault recovery coefficient of the node, including: Calculate the average and variance of the power factor data sequence of each node, and calculate the product of the average and the variance to obtain the fluctuation degree of the power factor data sequence; Extract the reference subsequence from the new energy power generation data sequence of each node; Based on the reference subsequence, generate prediction data through the exponential moving average algorithm to construct a new energy power generation prediction sequence according to the prediction data; Calculate the cosine similarity between the new energy power generation data sequence and the new energy power generation prediction sequence to obtain the power generation similarity; the similarity between the new energy power generation data sequence and the new energy power generation prediction sequence includes the power generation similarity; According to the product of the sum of the absolute slopes of the active recovery subsequences corresponding to the multiple extreme points in the active power data sequence of each node and the symmetry of the node, the fluctuation degree of the power factor data sequence, and the power generation similarity, the fault recovery coefficient of the node is obtained.

9. The method of claim 1, wherein the method further comprises: The method further includes, before the power recovery is performed with the target node as the starting node in response to the power grid failure: Based on the TOPSIS advantage and disadvantage solution distance method, the resilience and reliability coefficient of each node is calculated according to the load frequency response coefficient, the fault recovery coefficient and the power index of each node; the power index includes the average path length, the betweenness centrality and the connectivity of the node; According to the maximum inter-class variance method, the resilience and reliability coefficients of multiple nodes are segmented to obtain a segmentation threshold; Nodes with a resilience and reliability coefficient greater than or equal to the segmentation threshold are determined as target nodes.

10. The method of claim 1, wherein the method further comprises: The method further includes, before the power recovery is performed with the target node as the starting node in response to the power grid failure: In response to the power grid failure, the recovery path of each target node is calculated according to the depth-first search algorithm; According to the sum of the active power of all nodes in the recovery path of each target node, the total load power before the failure, the number of nodes in the recovery path, and the number of nodes with failure in the power grid, the recovery priority coefficient of the recovery path of the target node is calculated; The recovery path with the largest recovery priority coefficient is taken as the target path for power recovery.

Images