A method, apparatus and storage medium for identifying weak nodes

Abstract

This invention discloses a method, device, and storage medium for identifying weak nodes, belonging to the field of power system quality identification technology. It addresses the problem that existing weak node identification methods cannot utilize limited node state information to obtain the state information of all nodes in the entire network for identifying weak nodes. The method includes the following steps: constructing a distribution system model to obtain the PMU (Power Management Unit) configuration matrix and performing network topology layering; obtaining the state information of all nodes in the network for a first static stability assessment; constructing complex voltage change curves for each node and calculating the node radius of curvature and the rate of change of the radius of curvature; and using an improved K-means clustering method to perform a second static voltage stability assessment on all nodes, identifying weak nodes in the distribution system model. This method can quickly obtain the state information of all nodes in the entire network using limited node state information to identify weak nodes, thereby determining the stability of the distribution system.

Description

Technical Field

[0001] This invention relates to the field of power system quality identification technology, and in particular to a method, device and storage medium for identifying weak nodes. Background Technology

[0002] With the development of power systems, the load changes and complexity of distribution networks are constantly increasing, making voltage stability issues increasingly prominent. Existing online monitoring systems cannot accurately assess the stability of the power grid and provide reliable early warning information because they lack multi-level comprehensive perception and multi-dimensional precise coordinated control of the real physical behavior of the power grid. This deficiency becomes particularly evident during periods of gradual power grid deterioration. Therefore, researching data-driven power system stability analysis methods is of great significance for understanding power grid behavior and taking appropriate measures.

[0003] Chinese Patent Publication No. CN112103995A, published on December 18, 2020, entitled "A Method and Device for Identifying Weak Nodes in an Active Distribution Network," discloses a method and device for identifying weak nodes in an active distribution network. The method includes the following steps: using data from a predicted date as input to a wavelet neural network model to obtain the active power and reactive power of each phase of the distributed power source at each node on the predicted date; and determining the weak nodes of each node based on the active power and reactive power of each phase of the distributed power source at each node on the predicted date and the mean prediction error of the pre-established wavelet neural network model. The invention obtains the affine values ​​of the active power and reactive power of each phase of the distributed power source on the prediction date; based on these values, the affine values ​​of the phase voltage of each node on the prediction date are obtained; and weak nodes are identified based on these affine values. This invention considers the prediction error of the distributed power source and improves the accuracy of identifying weak nodes. However, its drawback is that this weak node identification method cannot utilize limited node state information to obtain the state information of all nodes in the network to identify weak nodes. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing weak node identification methods, which cannot utilize limited node state information to obtain the state information of all nodes in the network for weak node identification. This invention provides a weak node identification method, device, and storage medium. The method obtains a PMU (Power Management Unit) configuration matrix based on a constructed power distribution system model. Using the nodes where the PMUs are located as initial nodes, the power distribution system model is layered into network topologies. Based on voltage drop, the state of each non-PMU configuration node in each layer of the network topology is estimated to obtain the voltage and phase angle information of all nodes in the network. A first static voltage stability judgment is performed based on the voltage and phase angle information of all nodes to obtain the initial stable state of all nodes. The system constructs complex voltage change curves for nodes, calculates the node curvature radius, and calculates the arithmetic mean of the node curvature radius and the rate of change of the node curvature radius over a fixed time period based on the node curvature radius. K-means clustering is then performed on the arithmetic mean of the node curvature radius, and a second static voltage stability assessment is conducted on all nodes based on the clustering results. Based on the results of the two static voltage stability assessments, the K-means clustering results, and the matched rate of change of the node curvature radius, weak nodes in the distribution system model are identified. This allows for timely monitoring of the changing trends and rates of change of all nodes in the network, and rapid identification of weak nodes, reducing computation time and providing a basis for rapid dispatching by staff. Furthermore, the stability of the distribution system can be determined by assessing the changing states and voltage stability of all nodes in the network.

[0005] The objective of this invention is achieved through the following technical solution:

[0006] A method for identifying weak nodes includes the following steps:

[0007] Based on the constructed power distribution system model, the configuration matrix of PMU measurement devices is obtained, and the network topology of the power distribution system model is layered using the node where the PMU measurement device is located as the initial node.

[0008] Based on the voltage drop, state estimation is performed on the configuration nodes of non-PMU measurement devices in each layer of the network topology layer to obtain the voltage and phase angle information of all network nodes.

[0009] The first static voltage stability judgment is made based on the total network node voltage and the total network node voltage phase angle information to obtain the initial stable state of all nodes;

[0010] Based on the initial steady state, construct the node complex voltage variation curve and calculate the node radius of curvature.

[0011] Calculate the arithmetic mean of the node curvature radius over a fixed time and the rate of change of the node curvature radius over a fixed time based on the node curvature radius.

[0012] K-means clustering was performed on the arithmetic mean of the node curvature radius, and a second static voltage stability assessment was performed on all nodes based on the clustering results.

[0013] Based on the results of two static voltage stability assessments, K-means clustering results, and the matched rate of change of node curvature radius, weak nodes in the distribution system model are identified. A PMU (Power Management Unit) configuration matrix is ​​obtained based on the constructed distribution system model. Using the nodes where the PMUs are located as initial nodes, the distribution system model is layered into network topologies. State estimation is performed on each non-PMU configuration node in each layer of the network topology based on voltage drop, obtaining the total network node voltage and phase angle information. A first static voltage stability assessment is performed based on this information to obtain the initial stable state of all nodes. Complex voltage change curves for each node are constructed based on these initial stable states, and the node curvature radius is calculated. Finally, the node curvature radius over a fixed time period is calculated based on the node curvature radius. The arithmetic mean of the radius of curvature and the rate of change of the node curvature radius over a fixed time are used to perform K-means clustering on the arithmetic mean of the node curvature radius. Based on the clustering results, a second static voltage stability assessment is performed on all nodes. Based on the results of the two static voltage stability assessments, the K-means clustering results, and the matched rate of change of the node curvature radius, weak nodes in the power distribution system model are identified. This allows for timely monitoring of the changing trends and rates of change of all nodes in the network, and rapid identification of weak nodes, reducing computation time and providing a basis for rapid dispatching by staff. At the same time, the stability of the power distribution system can be judged by the changing state of all nodes in the network and the voltage stability state.

[0014] Furthermore, the PMU (Power Management Unit) configuration matrix is ​​obtained based on the constructed power distribution system model, and the network topology of the power distribution system model is layered using the node where the PMU is located as the initial node, including:

[0015] Read the network topology of the power distribution system model to obtain the PMU measurement device configuration matrix;

[0016] The network topology layering process is determined based on the PMU (Power Measurement Unit) configuration matrix. If the layering is complete, the process ends and the result is output. If not, the process continues, constructing an adjacency matrix of the PMU configuration matrix and updating it. The new PMU configuration matrix is ​​then used to determine if the network topology layering is complete. If so, the process ends and the result is output. If not, the process is repeated. Alternatively, the PMU configuration matrix is ​​obtained by reading the network topology of the distribution system model. The process then determines if the network topology layering is complete. If so, the process ends and the result is output. If not, the process continues until the layering is finished. This approach saves time in estimating the state of nodes with unknown voltage information and accelerates the identification of weak points.

[0017] Further, the step of performing state estimation on the configuration nodes of non-PMU measurement devices in each layer of the network topology layer based on voltage drop, and obtaining the voltage and phase angle information of all nodes in the network, includes:

[0018] By layering the network topology, a topology network layer is obtained with each PMU configuration node as the initial node;

[0019] Calculate the voltage drop based on the PMU measurement information and load information of the network topology layer;

[0020] Based on the voltage drop, obtain the voltage information of non-PMU configured nodes and calculate the total network node voltage of the distribution system model;

[0021] The phase angle information of the voltage at all nodes in the network is calculated based on voltage drop and the total voltage at all nodes. Through network topology layering, a topology network layer is obtained with each PMU configuration node as the initial node. Voltage drop is calculated using PMU measurement information and load information. Voltage information of non-PMU configuration nodes is obtained, and the total voltage at all nodes in the distribution system model is calculated. Finally, the phase angle information of the voltage at all nodes in the network is calculated based on voltage drop and the total voltage at all nodes. This allows for the rapid acquisition of the total voltage phase angle information of all nodes in the entire distribution system using limited node information, thereby improving the efficiency of obtaining the total node status information.

[0022] Further, the step of performing the first static voltage stability judgment based on the total network node voltage and the total network node voltage phase angle information to obtain the initial stable state of all nodes includes:

[0023] Calculate the static voltage stability margin;

[0024] The initial node stability position is determined based on the static voltage stability margin. By calculating the static voltage stability margin, the initial node stability position can be determined. Furthermore, when the voltage of all nodes in the network changes, complex voltage values ​​can be calculated and node complex voltage curves can be plotted, providing a basis for subsequent accurate identification of weak nodes.

[0025] Further, the step of constructing the node complex voltage variation curve based on the initial steady state and calculating the node radius of curvature includes:

[0026] Using the voltage phase angle of each node as the abscissa and the voltage amplitude as the ordinate, construct the node complex voltage curve;

[0027] The node curvature radius is calculated based on the node complex voltage curve. Using PMU measurement information and pseudo-measurement information, the node status information of the entire network is obtained, including the total network node voltage and the total network node voltage phase angle. The node complex voltage curve is plotted with the voltage phase angle of each node as the abscissa and the voltage amplitude as the ordinate. Changes in distribution system load, photovoltaic output, and electric vehicle charging load fluctuations all affect each node differently, resulting in different complex voltage change curves for each node. This allows for the accurate acquisition of the node voltage change situation.

[0028] Further, the calculation of the arithmetic mean of the node curvature radius over a fixed time and the rate of change of the node curvature radius over a fixed time, based on the node curvature radius, includes:

[0029] Calculate the arithmetic mean of the node curvature radii over a fixed time period based on the node curvature radius to form a set of curvature radii;

[0030] The rate of change of node curvature radius over a fixed time period is calculated based on the node curvature radius, forming a set of curvature radius change rates. By calculating the arithmetic mean of the node curvature radii over a fixed time period, a set of curvature radii can be formed, providing input data for clustering. This allows the clustering results to accurately fit the power distribution system and characterize the magnitude of node state fluctuations. Furthermore, the set of curvature radius change rates, calculated based on the node curvature radius over a fixed time period, can characterize the speed of node state fluctuations and provide a basis for identifying vulnerable nodes.

[0031] Further, the step of performing K-means clustering on the arithmetic mean of the node curvature radii and then performing a second static voltage stability assessment on all nodes based on the clustering results includes:

[0032] K-means clustering was performed using historical power distribution network operation data as input data to obtain k initial cluster centers m. i m1, m2, m3, ..., m k ;

[0033] The set of curvature radii of the complex voltage change curvature of the distribution network nodes is used as the input data for clustering, and the number of clusters is set to k.

[0034] Calculate the radius of curvature of each node. With the initial cluster center m i distance Where h is the number of iterations, and the minimum distance is min(d) i (This is used as the clustering result;)

[0035] Calculate the new cluster centers: N i r is the sum of one class in the first clustering result. i The radius of curvature in this class;

[0036] Calculate the difference in cluster centers between two iterations. If the difference in cluster centers between two iterations satisfies |m i (h+1)-m i When (h)|<ε, the iteration ends; otherwise, the distance between the radius of curvature and the new cluster center is recalculated, and the iteration continues.

[0037] A second static voltage stability assessment is performed on all nodes based on the clustering results. The initial cluster centers of the improved k-means clustering method are not randomly selected without any basis, but are obtained from historical data, which avoids the clustering results getting trapped in local optima and improves the convergence speed and stability.

[0038] Furthermore, the step of identifying weak nodes in the power distribution system model based on the results of two static voltage stability assessments, K-means clustering results, and the matched rate of change of node curvature radius includes:

[0039] Based on the initial static voltage stability assessment, all network nodes are categorized into safe and unsafe classes.

[0040] The fluctuation state of all nodes in the network is obtained by clustering the nodes' curvature radii.

[0041] A second static voltage stability assessment is performed on all nodes in the network to identify stable nodes, relatively weak nodes, and weak nodes in the entire network.

[0042] Calculate the rate of change of the radius of curvature of the weaker nodes and the rate of change of the radius of curvature of the weaker nodes.

[0043] Weak nodes across the entire network are identified by measuring the rate of change of the radius of curvature of both weaker and weaker nodes. Based on the initial static voltage stability assessment, all network nodes are categorized into safe and unsafe classes. The fluctuation state of all network nodes is determined by the clustering results of their radius of curvature. A second static voltage stability assessment is then performed to identify stable, relatively weak, and weak nodes. The rates of change of the radius of curvature of both weaker and weaker nodes are then calculated, and these rates are used to identify the weakest nodes in the entire network. This process improves the accuracy of weak node identification and avoids requiring assessment of node static voltage stability for each load change. Combined with the topology hierarchical structure, this accelerates the identification of weak nodes, saving scheduling time for staff.

[0044] The present invention also provides a weak node identification device, which includes the following modules:

[0045] Information acquisition module: used to collect status information of all nodes in the power distribution system model;

[0046] Information processing module: Used to process the status information of all nodes in the information acquisition module and identify weak nodes in the power distribution system model;

[0047] Information Display Module: Used to display the weak nodes identified in the information processing module. The information acquisition module transmits the collected status information of all network nodes to the information processing module. The information processing module performs the first static voltage stability judgment based on the status information of all network nodes to obtain the initial stable state of all nodes. Then, based on the initial stable state of all nodes, it constructs the node complex voltage change curve, calculates the node radius of curvature, and then calculates the arithmetic mean of the node radius of curvature and the rate of change of the node radius of curvature over a fixed time. K-means clustering is performed on the arithmetic mean of the node radius of curvature, and a second static voltage stability judgment is performed on all nodes based on the K-means clustering results. Then, based on the results of the two static voltage stability judgments, the K-means clustering results, and the matched rate of change of the node radius of curvature, the module processes the data to identify weak nodes in the power distribution system model and transmits the processing results to the information display module. The information display module receives the instructions and finally displays the weak nodes in the power distribution system model, which can improve the efficiency of identifying weak nodes in the power distribution system model.

[0048] This invention also provides a storage medium, which is a computer-readable storage medium storing a computer program thereon. When the computer program is executed by a processor, it implements the aforementioned weak node identification method. By writing each step of the weak node identification method into a computer program and storing it in the storage medium, when the computer program is executed by a processor, the weak node identification method executes each step, enabling rapid and efficient identification of weak nodes in the power distribution system model.

[0049] The beneficial effects of this invention are as follows: Based on the constructed power distribution system model, a PMU (Power Management Unit) configuration matrix is ​​obtained. Using the nodes where the PMUs are located as initial nodes, the power distribution system model is layered into network topologies. Based on voltage drop, state estimation is performed on each non-PMU configuration node in the network topology layer. The voltage and phase angle information of all network nodes are obtained. Based on the voltage and phase angle information of all network nodes, a first static voltage stability judgment is performed to obtain the initial stable state of all nodes. Based on the initial stable state, a node complex voltage change curve is constructed, and the node curvature radius is calculated. Based on the node curvature radius, a fixed-time time calculation is performed. The arithmetic mean of the node curvature radius and the rate of change of the node curvature radius over a fixed time period are used to perform K-means clustering on the arithmetic mean of the node curvature radius. Based on the clustering results, a second static voltage stability assessment is performed on all nodes. By combining the results of the two static voltage stability assessments, the K-means clustering results, and the matched rate of change of the node curvature radius, weak nodes in the distribution system model are identified. This allows for timely monitoring of the changing trends and rates of change of all nodes in the network, and rapid identification of weak nodes, reducing computation time and providing a basis for rapid dispatching by staff. Furthermore, the stability of the distribution system can be judged by the changing states of all nodes and the voltage stability state. The improved K-means clustering method avoids the clustering results getting trapped in local optima, while also improving convergence speed and stability. Attached Figure Description

[0050] Figure 1 This is a flowchart of a weak node identification method according to an embodiment of the present invention;

[0051] Figure 2 This is a diagram illustrating the network topology layering steps according to an embodiment of the present invention;

[0052] Figure 3 This is a diagram illustrating the first scenario in the static voltage stability determination process according to an embodiment of the present invention.

[0053] Figure 4 This is a diagram illustrating the second scenario in the static voltage stability determination process according to an embodiment of the present invention.

[0054] Figure 5This is a diagram illustrating the third scenario in the static voltage stability determination process according to an embodiment of the present invention.

[0055] Figure 6 This is a diagram illustrating the fourth scenario in the static voltage stability determination process according to an embodiment of the present invention.

[0056] Figure 7 This is a diagram illustrating the fifth scenario in the static voltage stability determination process according to an embodiment of the present invention.

[0057] Figure 8 This is a diagram illustrating the weak node identification process according to an embodiment of the present invention;

[0058] Figure 9 This is a schematic diagram of the weak node identification device according to an embodiment of the present invention. Detailed Implementation

[0059] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0060] Example 1: A method for identifying weak nodes, such as Figure 1 As shown, it includes the following steps:

[0061] Based on the constructed power distribution system model, the configuration matrix of PMU measurement devices is obtained, and the network topology of the power distribution system model is layered using the node where the PMU measurement device is located as the initial node.

[0062] Specifically, the steps for constructing the power distribution system model are as follows: Based on the K-means clustering method, process the historical annual photovoltaic output data to obtain the overall representative index of photovoltaic power output under three weather conditions, and the average photovoltaic output. Based on the K-means clustering method, historical photovoltaic (PV) output data was processed to obtain the overall representative index of PV output for the entire day under three weather conditions, and the average PV output.

[0063]

[0064] Where T0 is the time period during which the photovoltaic power station can generate electricity; P represents the total time during which the photovoltaic power station can generate electricity; t Let t be the photovoltaic output of the photovoltaic power station at time t.

[0065] Considering the charging load of electric vehicles based on user travel characteristics, it is necessary to group the vehicle trajectory data package provided by the Gaia platform for D days into "days" and then integrate the data from orders within the same day:

[0066] Remove trajectory data that is not within the scope of the study, remove trajectories where the time interval between adjacent trajectory coordinates of the same vehicle on GPS exceeds 100 seconds, and remove average speed V. iFor trajectories with speeds greater than 120 km / h, trajectories with a starting straight-line distance of less than 500 meters are removed. The remaining qualifying trajectory data undergoes traffic origin-end data (OD) extraction. Using the Leuven map matching tool in Python, GPS data is matched with the road network of the study area. Simultaneously, based on the changes in the passenger load parameter P (0, 1), empty and loaded states are distinguished to form the trajectory network for day x.

[0067] The traffic network electric vehicle trajectory dataset contains the vehicle's ID N. i Vehicle running time T i Vehicle GPS positioning (X) i Y i Vehicle speed V i And whether it is carrying passengers P (1 for carrying passengers, 0 for not carrying passengers), and the battery status S at the time of electric vehicle travel. i,o Information. By using trajectory information and battery status at the time of electric vehicle travel, the system predicts user charging times and durations, forming an electric vehicle charging load curve that takes into account user travel characteristics.

[0068] The network topology of the power distribution system model is layered based on the constructed power distribution system model.

[0069] Specifically, it includes the following steps:

[0070] Read the network topology of the power distribution system model to obtain the PMU measurement device configuration matrix;

[0071] The network topology layering process is determined based on the PMU (Physical Measurement Unit) configuration matrix. If the network topology layering is complete, the process ends and the result is output. If the network topology layering is not complete, the process continues, constructing an adjacency matrix of the PMU configuration matrix, updating the PMU configuration matrix, and obtaining a new PMU configuration matrix. The network topology layering process is then determined based on the new PMU configuration matrix. If the network topology layering is complete, the process ends and the result is output. If the network topology layering is not complete, the process is repeated.

[0072] Specifically, such as Figure 2 The diagram shows an IEEE 7-node power distribution network, with PMU (Power Management Unit) devices configured at nodes 2 and 4. The mathematical expression for the PMU configuration matrix A1 is:

[0073]

[0074] In the formula, 0 indicates that the node is not configured with a PMU device, and 1 indicates that a PMU device is configured. The number of "1"s in A1 is equal to the total number of nodes N in the network to determine whether topology layering has been completed. In the above formula, the number of "1"s is 2, which is less than the total number of nodes N = 7, so further layering of the network is required. Specifically, the adjacency matrix X1 of A1 is constructed. Since the impedance between nodes 2 and 3 is less than the impedance between nodes 3 and 4, node 3 is assigned to the layer containing node 2.

[0075] The mathematical expression for the adjacent matrix X1 is:

[0076]

[0077] Thus, we obtain the new A2:

[0078]

[0079] At this point, the number of "1"s in A2 is 6, which is still less than the total number of network nodes N=7. Therefore, it is necessary to continue layering the network topology to obtain a new adjacency matrix X2 for A2.

[0080]

[0081] Update PMU configuration matrix A3 again:

[0082]

[0083] At this point, the number of "1"s in the PMU configuration node matrix A3 is 7, which is equal to the total number of nodes N=7. This completes all the steps of topology layering, dividing the original 7-node network structure into three layers. This saves time for subsequent state estimation of nodes with unknown voltage information and accelerates the identification of weak nodes.

[0084] Based on the voltage drop, state estimation is performed on the configuration nodes of non-PMU measurement devices in each layer of the network topology to obtain the voltage and phase angle information of all nodes in the network.

[0085] Specifically, it includes the following steps:

[0086] By layering the network topology, a topology network layer is obtained with each PMU configuration node as the initial node;

[0087] Calculate the voltage drop based on the PMU measurement information and load information of the network topology layer;

[0088] Based on the voltage drop, obtain the voltage information of non-PMU configured nodes and calculate the total network node voltage of the distribution system model;

[0089] The phase angle information of the voltage at all nodes in the network is calculated based on the voltage drop and the voltage at all nodes in the network.

[0090] The first static voltage stability assessment is performed based on the voltage and phase angle information of all nodes in the network to obtain the initial stable state of all nodes.

[0091] Specifically, it includes the following steps:

[0092] Calculate the static voltage stability margin;

[0093] The initial node stability position is determined based on the static voltage stability margin.

[0094] Specifically, the quantitative formula for static voltage stability margin is as follows:

[0095] f(θ,V)+λ·b=0;

[0096] In the formula, θ is the node voltage phase angle; v is the node voltage amplitude; λ is the load and generator growth parameter; and b is the load growth mode constant of each node in the system.

[0097] like Figures 3 to 7 As shown, the change region of the node voltage from the initial state to the end of the iteration has the following five situations. Currently, my country's static voltage stability evaluation standard is based on the regional load active power margin K. p and bus load reactive power margin K q definition.

[0098] In the formula, P max P0 is the active power value at the maximum transmission power critical point; Q is the initial transmission active power; max Q0 is the reactive power value at the critical point of maximum transmission power; Q0 is the reactive power of the initial transmission.

[0099] This application considers the impact of active power growth at load nodes on system voltage stability, therefore K is selected. p =8% as the critical state, in Figures 3 to 7 In the middle, when K p <8% is considered an unsafe state, when K p A voltage ≥8% is considered a safe state. Before calculating and plotting the complex voltage change curve as the load changes, the initial node stability position should be determined. Figures 3 to 7 The five scenarios are categorized. It is also necessary to re-categorize the load when it reaches its maximum. Figures 3 to 7 The five situations are judged and categorized, and a second judgment result is given to provide a basis for subsequent accurate judgment of weak points.

[0100] Based on the initial steady state, construct the node complex voltage variation curve and calculate the node curvature radius.

[0101] Specifically, it includes the following steps:

[0102] Using the voltage phase angle of each node as the abscissa and the voltage amplitude as the ordinate, construct the node complex voltage curve;

[0103] Calculate the node curvature radius based on the node complex voltage curve.

[0104] Specifically, the complex voltage curves of each node are plotted with the voltage phase angle as the abscissa and the voltage magnitude as the ordinate. Below is the radius of curvature r of the complex voltage. n,i The calculation formula is as follows:

[0105]

[0106] In the formula, r n,i Let v be the radius of curvature of the complex voltage curve at time i of node n. n,i Let θ be the voltage magnitude at node n at time i. n,i Let be the voltage phase angle at time i of node n.

[0107] Calculate the arithmetic mean of the node curvature radius over a fixed time period and the rate of change of the node curvature radius over a fixed time period based on the node curvature radius.

[0108] Specifically, it includes the following steps:

[0109] Calculate the arithmetic mean of the node curvature radii over a fixed time period based on the node curvature radius to form a set of curvature radii;

[0110] Calculate the rate of change of the node curvature radius over a fixed time period based on the node curvature radius, and form a set of curvature radius change rates.

[0111] K-means clustering was performed on the arithmetic mean of the node curvature radius, and a second static voltage stability assessment was performed on all nodes based on the clustering results.

[0112] Specifically, it includes the following steps:

[0113] K-means clustering was performed using historical power distribution network operation data as input data to obtain k initial cluster centers m. i m1, m2, m3, ..., m k ;

[0114] The set of curvature radii of the complex voltage change curvature of the distribution network nodes is used as the input data for clustering, and the number of clusters is set to k.

[0115] Calculate the radius of curvature of each node. With the initial cluster center m idistance Where h is the number of iterations, and the minimum distance is min(d) i (This is used as the clustering result;)

[0116] Calculate the new cluster centers: N i r is the sum of one class in the first clustering result. i The radius of curvature in this class;

[0117] Calculate the difference in cluster centers between two iterations. If the difference in cluster centers between two iterations satisfies |m i (h+1)-m i When (h)|<ε, the iteration ends; otherwise, the distance between the radius of curvature and the new cluster center is recalculated, and the iteration continues.

[0118] A second static voltage stability assessment is performed on all nodes based on the clustering results.

[0119] Based on the results of two static voltage stability assessments, K-means clustering results, and the rate of change of the node curvature radius, weak nodes in the power distribution system model are identified.

[0120] Specifically, it includes the following steps:

[0121] Based on the initial static voltage stability assessment, all network nodes are categorized into safe and unsafe classes.

[0122] The fluctuation state of all nodes in the network is obtained by clustering the nodes' curvature radii.

[0123] A second static voltage stability assessment is performed on all nodes in the network to identify stable nodes, relatively weak nodes, and weak nodes in the entire network.

[0124] Calculate the rate of change of the radius of curvature of the weaker nodes and the rate of change of the radius of curvature of the weaker nodes.

[0125] Weak nodes in the entire network are identified by the rate of change of the radius of curvature of the weaker nodes and the rate of change of the radius of curvature of the weakest nodes.

[0126] Specifically, such as Figure 8 As shown, based on the first static voltage stability assessment, the nodes are divided into two categories. The first category is the "safe" category, which is K in the static voltage stability assessment. p ≥8%, the second category is "unsafe", that is, K in the static voltage stability judgment. p <8%.

[0127] Then, based on the clustering results of the curvature radius of the node's changing state, they are classified into two categories: the first category is ①②, which means the node fluctuates less and has a larger curvature radius; the second category is ③, which means the node fluctuates more and has a smaller curvature radius.

[0128] The third step involves a second static voltage stability assessment, similar to the first, categorizing the voltage into two types. One type is the "safe" type, which corresponds to K in the static voltage stability assessment. p ≥8%, the other category is "unsafe", that is, K in the static voltage stability judgment. p <8%.

[0129] Based on the above judgment results, according to Figure 8 This allows us to obtain preliminary assessments of node vulnerabilities, which can be categorized into three types: stable nodes, relatively weak nodes, and weak nodes.

[0130] Finally, the rate of change of the radius of curvature of the weaker nodes and the weak nodes themselves is calculated. Based on the node fluctuation rate, the weaker nodes and the weak nodes are reclassified. If the rate is fast, the improved K-means clustering method from S7 is used to classify the node fluctuation rate into two categories: ① fast and ② slow. If a node was originally a weaker node, it is upgraded to a weaker node; if the rate is slow, it remains unchanged. After these steps, the final distribution of weak nodes is obtained.

[0131] This judgment process can improve the accuracy of weak node identification and assess the static voltage stability of nodes with each load change. Combined with the topology layering structure, it can improve the speed of weak node identification and save scheduling time for staff.

[0132] Example 2:

[0133] The present invention also provides a weak node identification device 10, such as Figure 9 As shown, it includes an information acquisition module 11, an information processing module 12, and an information display module 13. The information acquisition module 11 is used to acquire the status information of all nodes in the power distribution system model; it is used to process the status information of all nodes in the power distribution system model and identify weak nodes; the information display module 13 is used to display the weak nodes identified by the information processing module 12.

[0134] Specifically, the information acquisition module 11 transmits the acquired network node status information to the information processing module 12. The information processing module 12 performs the first static voltage stability judgment based on the network node status information to obtain the initial stable state of all nodes. Then, based on the initial stable state of all nodes, it constructs the node complex voltage change curve, calculates the node radius of curvature, and then calculates the arithmetic mean of the node radius of curvature and the rate of change of the node radius of curvature over a fixed time based on the radius of curvature. K-means clustering is performed on the arithmetic mean of the node radius of curvature, and a second static voltage stability judgment is performed on all nodes based on the K-means clustering results. Then, based on the results of the two static voltage stability judgments, the K-means clustering results, and the matched rate of change of the node radius of curvature, the module processes the data to identify weak nodes in the power distribution system model and transmits the processing results to the information display module 13. The information display module 13 receives the instructions and finally displays the weak nodes in the power distribution system model, which can improve the efficiency of identifying weak nodes in the power distribution system model.

[0135] This embodiment provides a storage medium, which is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements a weak node identification method. By writing each step of the weak node identification method into a computer program and storing it in the storage medium, when the computer program is executed by the processor, the weak node identification method executes each step, enabling it to quickly and efficiently identify weak nodes in the power distribution system model.

[0136] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented in various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0137] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0138] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0139] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0140] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0141] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for identifying weak nodes, characterized in that, Includes the following steps: Based on the constructed power distribution system model, the configuration matrix of PMU measurement devices is obtained, and the network topology of the power distribution system model is layered using the node where the PMU measurement device is located as the initial node. Based on the voltage drop, state estimation is performed on the configuration nodes of non-PMU measurement devices in each layer of the network topology layer to obtain the voltage and phase angle information of all network nodes. The first static voltage stability judgment is made based on the total network node voltage and the total network node voltage phase angle information to obtain the initial stable state of all nodes; Based on the initial steady state, construct the node complex voltage variation curve and calculate the node radius of curvature. Calculate the arithmetic mean of the node curvature radius over a fixed time and the rate of change of the node curvature radius over a fixed time based on the node curvature radius. K-means clustering was performed on the arithmetic mean of the node curvature radius, and a second static voltage stability assessment was performed on all nodes based on the clustering results. Based on the results of two static voltage stability assessments, K-means clustering results, and the rate of change of the radius of curvature of the matched nodes, weak nodes in the power distribution system model are identified. The step of performing state estimation on the non-PMU measurement device configuration nodes in each layer of the network topology based on voltage drop, and obtaining the total network node voltage and total network node voltage phase angle information, includes: By layering the network topology, a topology network layer is obtained with each PMU configuration node as the initial node; Calculate the voltage drop based on the PMU measurement information and load information of the network topology layer; Based on the voltage drop, obtain the voltage information of non-PMU configured nodes and calculate the total network node voltage of the distribution system model; The phase angle information of the voltage at all nodes in the network is calculated based on the voltage drop and the voltage at all nodes in the network. The process of constructing the node complex voltage variation curve based on the initial steady state and calculating the node radius of curvature includes: Using the voltage phase angle of each node as the abscissa and the voltage amplitude as the ordinate, construct the node complex voltage curve; Calculate the node radius of curvature based on the node complex voltage curve; The process of identifying weak nodes in the power distribution system model based on two static voltage stability assessments, K-means clustering results, and the matched rate of change of node curvature radius includes: Based on the initial static voltage stability assessment, all network nodes are categorized into safe and unsafe classes. The fluctuation state of all nodes in the network is obtained by clustering the nodes' curvature radii. A second static voltage stability assessment is performed on all nodes in the network to identify stable nodes, relatively weak nodes, and weak nodes in the entire network. Calculate the rate of change of the radius of curvature of the weaker nodes and the rate of change of the radius of curvature of the weaker nodes. Weak nodes in the entire network are identified by the rate of change of the radius of curvature of the weaker nodes and the rate of change of the radius of curvature of the weakest nodes.

2. The weak node identification method according to claim 1, characterized in that, The configuration matrix of PMU measurement devices is obtained based on the constructed power distribution system model, and the network topology of the power distribution system model is layered using the nodes where the PMU measurement devices are located as initial nodes, including: Read the network topology of the power distribution system model to obtain the PMU measurement device configuration matrix; The network topology layering process is determined based on the PMU (Physical Measurement Unit) configuration matrix. If the network topology layering is complete, the process ends and the result is output. If the network topology layering is not complete, the process continues, constructing an adjacency matrix of the PMU configuration matrix, updating the PMU configuration matrix, and obtaining a new PMU configuration matrix. The network topology layering process is then determined based on the new PMU configuration matrix. If the network topology layering is complete, the process ends and the result is output. If the network topology layering is not complete, the process is repeated.

3. The weak node identification method according to claim 1, characterized in that, The first static voltage stability assessment based on the total network node voltage and the total network node voltage phase angle information, to obtain the initial stable state of all nodes, includes: Calculate the static voltage stability margin; The initial node stability position is determined based on the static voltage stability margin.

4. The weak node identification method according to claim 1, characterized in that, The calculation of the arithmetic mean of the node curvature radius over a fixed time and the rate of change of the node curvature radius over a fixed time includes: Calculate the arithmetic mean of the node curvature radii over a fixed time period based on the node curvature radius to form a set of curvature radii; Calculate the rate of change of the node curvature radius over a fixed time period based on the node curvature radius, and form a set of curvature radius change rates.

5. The weak node identification method according to claim 4, characterized in that, The arithmetic mean of the node curvature radii is used for K-means clustering, and a second static voltage stability assessment is performed on all nodes based on the clustering results, including: K-means clustering was performed using historical power distribution network operation data as input to obtain k initial cluster centers. : ; The set of curvature radii of the complex voltage change curvature of the distribution network nodes is used as the input data for clustering, and the number of clusters is set to k. Calculate the radius of curvature of each node. with the initial cluster centers distance Where h is the number of iterations, and the minimum distance is taken. As a clustering result; Calculate the new cluster centers: , This is the sum of one class from the first clustering results. The radius of curvature in this class; Calculate the difference in cluster centers between two iterations. If the difference in cluster centers between two iterations satisfies... If the condition is met, the iteration ends; otherwise, the distance between the radius of curvature and the new cluster center is recalculated, and the iteration continues. A second static voltage stability assessment is performed on all nodes based on the clustering results.

6. A weak node identification device, applicable to the weak node identification method according to any one of claims 1-5, characterized in that, Includes the following modules: Information acquisition module: used to collect status information of all nodes in the power distribution system model; Information processing module: Used to process the status information of all nodes in the information acquisition module and identify weak nodes in the power distribution system model; Information display module: Used to display the weak points identified in the information processing module.

7. A storage medium, said storage medium being a computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements a weak node identification method as described in any one of claims 1-5.

Images