Power distribution network weak node identification method and system, terminal and medium
By constructing a least-squares support vector machine model based on time scale factors, and combining the topology and electrical operation characteristics of the distribution network, the model achieves accurate identification of weak nodes of different time types, solves the problem of poor identification effect in existing technologies, and improves the safety and stability of the power grid.
Patent Information
- Application Number
- CN202411142399.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-20
- Publication Date
- 2024-11-12
AI Technical Summary
Existing technologies fail to adequately consider differences in time scales and the growth scenarios of new energy power generation when identifying weak nodes in the distribution network, resulting in poor identification performance.
By collecting historical topology and electrical operation data of the distribution network, a database of weak node characteristic parameters is constructed. A time scale factor is introduced, and a three-level classification model is established using least squares support vector machine (LSSVM) to perform differentiated analysis and identification of weak nodes of different time types.
It enables accurate identification of weak nodes in the distribution network, improves the practicality and accuracy of identification, helps dispatchers understand the weak links in the network structure, and ensures the safe and stable operation of the power grid.
Smart Images

Figure CN118940138A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of power grid security technology, and in particular to a method, system, terminal and medium for identifying weak nodes in a distribution network. Background Art
[0002] With the advancement of the construction of new power systems, the power generation of new energy in distribution networks continues to grow rapidly, and the operation of power systems has become more complicated. On the one hand, the randomness of new energy power generation has increased the difficulty of system control. On the other hand, the voltage problems caused by the connection of new energy to the grid have increased the operating risks of the distribution network. Looking back at power grid accidents at home and abroad, it is not difficult to find that voltage instability often starts from certain weak nodes, then spreads to other nodes, and finally leads to voltage collapse in the entire area. Therefore, it is of great significance to the voltage safety of the power system to accurately identify the weak nodes of the distribution network and conduct targeted monitoring and protection.
[0003] At present, there have been many studies on the identification methods of weak nodes in distribution networks. For example, the method based on power grid flow calculation identifies the weak nodes of the distribution network according to operating parameters such as flow distribution and voltage level; based on complex network theory, topological characteristic parameters such as node degree and node cohesion are used to identify weak nodes in the distribution network. Although the above methods take into account the electrical operation characteristics and topological structure characteristics of the network, they ignore the differences in the weak characteristics of the distribution network at different time scales, so they have great limitations in identifying weaknesses of different time types. In terms of application scenarios, current research mainly focuses on the identification of weak nodes in scenarios such as load growth, network attacks, and equipment failures, while there are fewer studies on the identification of weak nodes in the distribution network in the scenario of new energy power generation growth. Therefore, it is urgent to extract a distribution network weak node identification scheme that fully considers the time scale differences and is oriented to the application scenario of new energy power generation growth. Summary of the invention
[0004] In view of the deficiencies in the prior art, the present invention provides a method, system, terminal and medium for identifying weak nodes in a distribution network, deeply mines historical data and conducts a comprehensive analysis of the topological structure parameters and electrical operation parameters of the distribution network, introduces a time scale factor to perform differentiated analysis on the weak characteristics of distribution network nodes under different time scales, and finally uses a least squares support vector machine to achieve accurate identification of weak nodes of different time types, which has higher practicality than traditional weak node identification methods.
[0005] In a first aspect, a method for identifying weak nodes in a distribution network is provided, comprising the following steps:
[0006] S1: Collect historical topological data and electrical operation data of the distribution network;
[0007] S2: Build a database of weak characteristic parameters of distribution network nodes based on the historical topological data and electrical operation data of the distribution network;
[0008] S3: Construct a distribution network weak node identification model based on least squares support vector machine, and introduce a time scale factor to differentiate the weak characteristic parameters of the distribution network nodes input into the distribution network weak node identification model;
[0009] S4: using the data in the distribution network node weak characteristic parameter database to train the distribution network weak node identification model;
[0010] S5: Identify the weakness of the current distribution network nodes using the trained distribution network weak node identification model according to the weak characteristic parameters of the distribution network nodes during real-time operation.
[0011] Further, in step S2, the distribution network node weakness characteristic parameters include at least one of improved node degree, node power concentration factor, and node importance.
[0012] Furthermore, the improved node degree is calculated by the following formula:
[0013]
[0014] Where D (i) is the improved node degree of node i; D i is the node degree of node i; a ij is the element in the i-th row and j-th column of the distribution network topology adjacency matrix. When nodes i and j are adjacent, a ij The value of is 1, otherwise it is 0; avg is the average value of the node degree of the distribution network; X ij is the line e with nodes i and j as the two end points ij The reactance value; θ is the node set in the distribution network topology that does not include node i;
[0015] The node power concentration factor is calculated using the following formula:
[0016]
[0017] In the formula, S i is the node power concentration factor of node i; Adj i is the set of adjacent nodes of node i; γ ij is the line e with nodes i and j as the two end points ij The weight of line e ij The absolute value of the active power flowing through the line divided by the line e ij The reactance value;
[0018] The node importance is calculated by the following formula:
[0019]
[0020] In the formula, SI i is the node importance of node i; I i is the efficiency value of node i; L ij is the importance contribution of node j to node i, and its expression is
[0021] Further, in step S3, the distribution network weak node identification model includes a three-layer Lssvm classification model, the first layer of the Lssvm classification model is used to classify the nodes as weak or non-weak, the second layer of the Lssvm classification model is used to classify the nodes as short-term weak or medium- and long-term weak; the third layer of the Lssvm classification model is used to classify the nodes as seasonally weak or long-term weak;
[0022] The optimal linear decision function of the Lssvm classification model at each layer is expressed as follows:
[0023] y(x)=sgn[w k φ(x)+b k ]
[0024] Where x is the input distribution network node weak characteristic parameter sample, y(x) is the output corresponding to the input distribution network node weak characteristic parameter sample, and w k is the weight vector of the Lssvm classification model of the kth layer, b k is the deviation of the Lssvm classification model of the kth layer, φ(x) is the nonlinear mapping function that maps the weak characteristic parameters of the distribution network nodes from the original space to the high-dimensional space;
[0025] The objective function and constraints of each layer of the Lssvm classification model are:
[0026]
[0027] In the formula, minJ(w k ,ξ k ) is the objective function of the least squares support vector machine corresponding to the Lssvm classification model of the kth layer; x v is the input weak characteristic parameter of the vth distribution network node, y v is the output corresponding to the input weak characteristic parameter of the vth distribution network node, l is the number of samples of weak characteristic parameters of the distribution network node; ξ v k is the error variable of the Lssvm classification model of the kth layer, ξ v k≥0; c is the penalty coefficient, and c>0.
[0028] Furthermore, in step S3, a time scale factor is introduced to perform differentiated processing on the weak characteristic parameters of the distribution network nodes input into the distribution network weak node identification model, including:
[0029] Introduce the time scale factor σ, its expression is as follows:
[0030]
[0031] Where D sv,i is the number of days when the voltage of node i exceeds the limit in the historical data, D T is the total number of days of historical data;
[0032] The weak characteristic parameters of the distribution network nodes input into the Lssvm classification model at each layer of the distribution network weak node identification model are differentiated, as shown below:
[0033] x′=λ 1,lssvmk x+λ 2,lssvmk σ
[0034] Where x is the weak characteristic parameter of the distribution network node; 1,lssvmk and λ 2,lssvmk are the weight coefficients of x and σ corresponding to the Lssvm classification model of the kth layer respectively; x' is the weak characteristic parameter of the distribution network node after differentiation processing.
[0035] Furthermore, the time scale factor σ is updated online in real time.
[0036] Further, step S5 includes:
[0037] The weak characteristic parameters of the distribution network nodes in real-time operation are used as the input of the Lssvm classification model of the first layer after differentiation processing. If it is judged to be not weak, the output is the first string, and if it is judged to be weak, the output is the second string;
[0038] If it is judged to be weak, the weak characteristic parameters of the distribution network node in real-time operation are further classified by using the Lssvm classification model of the second layer after differentiation processing. If it is judged to be short-term weak, the third string is output, and if it is judged to be medium- and long-term weak, the fourth string is output;
[0039] If it is judged to be medium- to long-term weakness, the weak characteristic parameters of the distribution network nodes during real-time operation will be further classified after differentiation processing using the third-layer Lssvm classification model. If it is judged to be seasonal weakness, the fifth character string will be output; if it is judged to be long-term weakness, the sixth character string will be output.
[0040] In a second aspect, a distribution network weak node identification system is provided, comprising:
[0041] Data acquisition module, used to collect historical topological data and electrical operation data of the distribution network;
[0042] A database building module is used to build a distribution network node weakness characteristic parameter database based on the historical topological data and electrical operation data of the distribution network;
[0043] A model building module is used to build a distribution network weak node identification model based on the least squares support vector machine, and introduce a time scale factor to perform differentiated processing on the distribution network node weak characteristic parameters input into the distribution network weak node identification model;
[0044] A model training module, used to train a distribution network weak node identification model using data in a distribution network node weak characteristic parameter database;
[0045] The weak node identification module is used to identify the weakness of the current distribution network nodes using the trained distribution network weak node identification model according to the weak characteristic parameters of the distribution network nodes during real-time operation.
[0046] In a third aspect, an electronic device is provided, including:
[0047] a memory having a computer program stored thereon;
[0048] The processor is used to implement the above-mentioned method for identifying weak nodes in the distribution network when loading and executing the computer program.
[0049] In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the method for identifying weak nodes in a distribution network as described above is implemented.
[0050] The present invention proposes a method, system, terminal and medium for identifying weak nodes in a distribution network. Based on the power system analysis theory and complex network theory, the topological structure characteristics of the distribution network are deeply integrated with the electrical operation characteristics, and the weak characteristic parameters of the distribution network nodes are constructed. Moreover, the time scale factor is introduced to differentiate the weak characteristic parameters of the distribution network nodes, which can better reflect the weak time scale characteristics of the nodes. Furthermore, a distribution network multi-type weak node identification model is established based on the least squares support vector machine, which can realize the refined identification of weak nodes in the distribution network. It helps dispatchers to have a deep understanding of the weak links of the distribution network grid and further ensure the safe and stable operation of the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS
[0051] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0052] Figure 1 It is a flow chart of a method for identifying weak nodes in a distribution network provided by an embodiment of the present invention;
[0053] Figure 2 is a network topology diagram of an IEEE 33-node power distribution system provided by an embodiment of the present invention;
[0054] Figure 3 4 is a training result diagram of the first-layer Lssvm classification model provided by an embodiment of the present invention;
[0055] Figure 4 4 is a prediction result diagram of the first layer of the Lssvm classification model provided by an embodiment of the present invention;
[0056] Figure 5 is a training result diagram of the second-layer Lssvm classification model provided by an embodiment of the present invention;
[0057] Figure 6 is a prediction result diagram of the second layer Lssvm classification model provided by an embodiment of the present invention;
[0058] Figure 7 4 is a training result diagram of the third-layer Lssvm classification model provided in an embodiment of the present invention;
[0059] Figure 8 This is a prediction result diagram of the third-layer Lssvm classification model provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0060] To make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described in detail below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other implementation methods obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.
[0061] The present invention intends to start from the two aspects of the weak characteristic index of the distribution network node and the multiple weak type identification model, and provide a distribution network weak node identification method based on time scale factor and least squares support vector machine, which mainly solves the following technical problems: how to construct the weak characteristic parameters of the distribution network node that combine the topological structure characteristics of the distribution network with the electrical operation characteristics; how to analyze the differences in the weak characteristics of the distribution network nodes under different time scales; how to achieve rapid and accurate identification of multiple types of weak nodes in the distribution network. Based on the power system analysis theory and complex network theory, the present invention will deeply integrate the topological structure characteristics of the distribution network with the electrical operation characteristics, and construct weak characteristic parameters such as node improvement degree, node power concentration factor, and node importance. The time scale factor is introduced to perform differential analysis on the weak characteristics of the distribution network nodes under different time scales. Finally, the least squares support vector machine is used to quickly and accurately identify the types of weak nodes in the distribution network. The technical solution of the present invention is described in detail below in conjunction with specific embodiments.
[0062] like Figure 1 As shown, an embodiment of the present invention provides a method for identifying weak nodes in a distribution network, comprising the following steps:
[0063] S1: Collect historical topological data and electrical operation data of the distribution network. Topological data includes node type, line opening and closing status, network adjacency matrix, new energy access location, etc.; electrical operation data includes line flow, node voltage, node load, new energy power generation, etc.
[0064] S2: Construct a database of weak characteristic parameters of distribution network nodes based on the historical topological data and electrical operation data of the distribution network.
[0065] In specific implementation, the weak characteristic parameter of the distribution network node selects at least one of the improved node degree, the node power concentration factor, and the node importance. In this embodiment, the above three parameters are taken as the weak characteristic parameters of the distribution network node for explanation.
[0066] Specifically, based on the power system analysis theory and complex network theory, the topological structure characteristics and electrical operation characteristics of the distribution network are integrated to form the following distribution network node weakness characteristic parameters:
[0067] (1) Improve node degree
[0068] Considering the local and global characteristics of nodes in the distribution network, and integrating the reactance parameters of adjacent lines to form an improved node degree. The improved node degree is calculated by the following formula:
[0069]
[0070] Where D (i) is the improved node degree of node i; D iis the node degree of node i; a ij is the element in the i-th row and j-th column of the distribution network topology adjacency matrix. When nodes i and j are adjacent, a ij The value of is 1, otherwise it is 0; avg is the average value of the node degree of the distribution network; X ij is the line e with nodes i and j as the two end points ij The reactance value of the distribution network; θ is the node set in the distribution network topology that does not include node i.
[0071] (2) Node power concentration factor
[0072] The node power concentration factor is formed based on the power flow distribution of the network. The node power concentration factor is calculated by the following formula:
[0073]
[0074] In the formula, S i is the node power concentration factor of node i; Adj i is the set of adjacent nodes of node i; γ ij is the line e with nodes i and j as the two end points ij The weight of line e ij The absolute value of the active power flowing through the line divided by the line e ij The reactance value.
[0075] (3) Node importance
[0076] The node importance is formed based on the efficiency value function, and the node importance is calculated by the following formula:
[0077]
[0078] In the formula, SI i is the node importance of node i; I i is the efficiency value of node i; L ij is the importance contribution value of node j to node i, and node j is the adjacent node of node i.
[0079] The importance contribution value of node j to node i is L ij The expression is as follows:
[0080]
[0081] The efficiency value of node i is I i It is dimensionless and its expression is as follows:
[0082]
[0083] Where n is the total number of nodes in the distribution network; dij is the electrical distance between node i and node j; the electrical distance between node i and node j is defined as the equivalent impedance Z between the two points ij,equ , Z ij,equ It is numerically equal to the voltage U between nodes i and j when a unit current is injected at node i. ij , which is expressed as follows:
[0084]
[0085] Among them, I0 is the unit current.
[0086] S3: Construct a distribution network weak node identification model based on least squares support vector machine, and introduce a time scale factor to differentiate the weak characteristic parameters of the distribution network nodes input into the distribution network weak node identification model.
[0087] The weak node identification model of the distribution network includes a three-layer Lssvm classification model. The first-layer Lssvm classification model is used to classify whether the node is weak or not weak. The second-layer Lssvm classification model is used to classify whether the node is short-term weak or medium- and long-term weak. The third-layer Lssvm classification model is used to classify whether the node is seasonally weak or long-term weak.
[0088] The optimal linear decision function of the Lssvm classification model at each layer is expressed as follows:
[0089] y(x)=sgn[w k φ(x)+b k ]
[0090] Where x is the input distribution network node weak characteristic parameter sample, y(x) is the output corresponding to the input distribution network node weak characteristic parameter sample, and w k is the weight vector of the Lssvm classification model of the kth layer, b k is the deviation of the Lssvm classification model of the kth layer, φ(x) is the nonlinear mapping function that maps the weak characteristic parameters of the distribution network nodes from the original space to the high-dimensional space;
[0091] The objective function and constraints of each layer of the Lssvm classification model are:
[0092]
[0093] In the formula, minJ(w k ,ξ k ) is the objective function of the least squares support vector machine corresponding to the Lssvm classification model of the kth layer; x v is the input weak characteristic parameter of the vth distribution network node, y vis the output corresponding to the input weak characteristic parameter of the vth distribution network node, l is the number of samples of weak characteristic parameters of the distribution network node; ξ v k is the error variable of the Lssvm classification model of the kth layer, ξ v k ≥0; c is the penalty coefficient, and c>0.
[0094] Furthermore, in step S3, a time scale factor is introduced to perform differentiated processing on the weak characteristic parameters of the distribution network nodes input into the distribution network weak node identification model, including:
[0095] In order to realize the differentiated analysis of weak characteristics of distribution network nodes under different time scales, the time scale factor σ is introduced, and its expression is as follows:
[0096]
[0097] Where D sv,i is the number of days when the voltage of node i exceeds the limit in the historical data, D T is the total number of days of historical data. From the above formula, we can see that if the number of days that the voltage of node i exceeds the limit is longer, the σ value will be larger.
[0098] The weak characteristic parameters of the distribution network nodes input into the Lssvm classification model at each layer of the distribution network weak node identification model are differentiated, as shown below:
[0099] x′=λ 1,lssvmk x+λ 2,lssvmk σ
[0100] In the formula, x' is the weak characteristic parameter of the distribution network node after differentiation processing; x is the weak characteristic parameter of the distribution network node;
[0101] λ 1,lssvmk and λ 2,lssvmk are the weight coefficients of x and σ corresponding to the Lssvm classification model of the kth layer, λ 1,lssvmk and λ 2,lssvmk The value of in the three-layer Lssvm classification model must meet the following requirements: (1) 1,lssvmk +λ 2,lssvmk =1;
[0102] (2)λ 2,lssvmk It increases layer by layer, because as the time scale becomes more refined, the time features need to be highlighted;
[0103] (3)λ 1,lssvmk It decreases layer by layer, due to the constraints of (1) and (2).
[0104] S4: Use the data in the distribution network node weak feature parameter database to train the Lssvm classification model of each layer, solve the weight vector and deviation of the Lssvm classification model of each layer, and obtain the trained distribution network weak node identification model.
[0105] S5: Identify the weakness of the current distribution network nodes using the trained distribution network weak node identification model according to the weak characteristic parameters of the distribution network nodes during real-time operation.
[0106] Specifically, step S5 includes:
[0107] The weak characteristic parameters of the distribution network nodes in real-time operation are used as the input of the first-layer Lssvm classification model (Lssvm1) after differentiation processing. If it is judged to be non-weak, the output is the first string (such as -1); if it is judged to be weak, the output is the second string (such as 1);
[0108] If it is judged to be weak, the weak characteristic parameters of the distribution network node in real-time operation are further classified by using the second-layer Lssvm classification model (Lssvm2) after differentiation processing. If it is judged to be short-term weak, the output is a third string (such as 1-1). If it is judged to be medium- and long-term weak, the output is a fourth string (such as 1-2).
[0109] If it is judged to be medium- and long-term weakness, the weak characteristic parameters of the distribution network nodes in real-time operation are further classified using the third-layer Lssvm classification model (Lssvm3) after differentiation processing. If it is judged to be seasonal weakness, the fifth string (such as 1-2-1) is output, and if it is judged to be long-term weakness, the sixth string (such as 1-2-2) is output. It should be noted that the actual output of the Lssvm classification model of each layer is 1 and -1. In order to clearly express the classification results and facilitate understanding, the final output string is converted.
[0110] In order to adapt to the impact of changes in the state of the nodes in the adaptive distribution network, in some preferred embodiments, the time scale factor σ is updated online in real time, that is, the voltage exceeding limit of each node is continuously updated to update the time scale factor in real time, thereby improving the model recognition accuracy.
[0111] The above embodiment provides a method for identifying weak nodes in a distribution network. Based on the power system analysis theory and complex network theory, the topological structure characteristics of the distribution network are deeply integrated with the electrical operation characteristics, and the weak characteristic parameters of the distribution network nodes are constructed. Moreover, the time scale factor is introduced to differentiate the weak characteristic parameters of the distribution network nodes, which can better reflect the weak time scale characteristics of the nodes. Furthermore, a distribution network multi-type weak node identification model is established based on the least squares support vector machine, which can realize the refined identification of weak nodes in the distribution network. This helps dispatchers to have a deep understanding of the weak links of the distribution network grid and further ensure the safe and stable operation of the distribution network.
[0112] The following will combine the implementation of the method provided in the above embodiment on the IEEE 33-node power distribution system to enhance the understanding of the method for identifying weak nodes in the distribution network.
[0113] 1. The topology parameters are determined according to the network topology of the IEEE 33-node distribution system. Figure 2 Shown is the network topology diagram of the IEEE 33-node power distribution system.
[0114] 2. The electrical operation parameters are determined based on the historical data of load and distributed renewable energy after flow calculation. The calculation results of the time scale factor σ are shown in Table 1.
[0115] Table 1
[0116]
[0117]
[0118] 3. Calculate the weak characteristic parameters of the distribution network nodes, and then use the time scale factor σ to perform differential analysis on the weak characteristic parameters of the distribution network nodes. The weight coefficient values are shown in Table 2.
[0119] Table 2
[0120]
[0121] 4. The node weakness type is determined based on the number of days the node voltage exceeds the limit after the power flow calculation, and this data is used as the expected output of the training set.
[0122] 5. Use the above data to train the distribution network weak node identification model proposed in the present invention.
[0123] 6. The weak characteristic parameters of the distribution network nodes in the test set are calculated using the current operating data, and the time scale factor is used to perform differential analysis. The calculation results in the three-layer Lssvm classification model are shown in Tables 3-5.
[0124] Table 3
[0125]
[0126]
[0127] Table 4
[0128]
[0129] Table 5
[0130]
[0131]
[0132] 7. The data in the previous steps are input as samples to be classified into the trained distribution network weak node identification model to obtain the node weak type identification results calculated by the method proposed in the present invention. The comparison with the node weak type classification results obtained by power flow calculation is shown in Table 6.
[0133] Table 6
[0134]
[0135] The embodiment of the present invention also provides a distribution network weak node identification system, including:
[0136] Data acquisition module, used to collect historical topological data and electrical operation data of the distribution network;
[0137] A database building module is used to build a distribution network node weakness characteristic parameter database based on the historical topological data and electrical operation data of the distribution network;
[0138] A model building module is used to build a distribution network weak node identification model based on the least squares support vector machine, and introduce a time scale factor to perform differentiated processing on the distribution network node weak characteristic parameters input into the distribution network weak node identification model;
[0139] A model training module, used to train a distribution network weak node identification model using data in a distribution network node weak characteristic parameter database;
[0140] The weak node identification module is used to identify the weakness of the current distribution network nodes using the trained distribution network weak node identification model according to the weak characteristic parameters of the distribution network nodes during real-time operation.
[0141] It should be understood that the functional unit modules in various embodiments of the present invention may be concentrated in one processing unit, or each unit module may exist physically separately, or two or more unit modules may be integrated in one unit module, and may be implemented in the form of hardware or software. For other technical features that are fully described, please refer to the aforementioned embodiments, which will not be described in detail here.
[0142] An embodiment of the present invention further provides an electronic device, including:
[0143] a memory having a computer program stored thereon;
[0144] The processor is used to implement the above-mentioned method for identifying weak nodes in the distribution network when loading and executing the computer program.
[0145] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method for identifying weak nodes in a distribution network as described above is implemented.
[0146] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.
[0147] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
[0148] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. 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.
[0149] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate 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 A function specified in one or more boxes.
[0150] 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 instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0151] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.
Claims
1. A method for identifying weak nodes in a distribution network, characterized in that: The steps include: S1: Collect historical topological data and electrical operation data of the distribution network; S2: Build a database of weak characteristic parameters of distribution network nodes based on the historical topological data and electrical operation data of the distribution network; S3: Construct a distribution network weak node identification model based on least squares support vector machine, and introduce a time scale factor to differentiate the weak characteristic parameters of the distribution network nodes input into the distribution network weak node identification model; S4: using the data in the distribution network node weak characteristic parameter database to train the distribution network weak node identification model; S5: Identify the weakness of the current distribution network nodes using the trained distribution network weak node identification model according to the weak characteristic parameters of the distribution network nodes during real-time operation.
2. The method for identifying weak nodes in a distribution network according to claim 1, characterized in that: In step S2, the distribution network node weakness characteristic parameters include at least one of the improved node degree, the node power concentration factor, and the node importance.
3. The method for identifying weak nodes in a distribution network according to claim 2, characterized in that: The improved node degree is calculated by the following formula: Where D (i) is the improved node degree of node i; D i is the node degree of node i; a ij is the element in the i-th row and j-th column of the distribution network topology adjacency matrix. When nodes i and j are adjacent, a ij The value of is 1, otherwise it is 0; avg is the average value of the node degree of the distribution network; X ij is the line e with nodes i and j as the two end points ij The reactance value; θ is the node set in the distribution network topology that does not include node i; The node power concentration factor is calculated using the following formula: In the formula, S i is the node power concentration factor of node i; Adj i is the set of adjacent nodes of node i; γ ij is the line e with nodes i and j as the two end points ij The weight of line e ij The absolute value of the active power flowing through the line divided by the line e ij The reactance value; The node importance is calculated by the following formula: In the formula, SI i is the node importance of node i; I i is the efficiency value of node i; L ij is the importance contribution of node j to node i, and its expression is 4. The method for identifying weak nodes in a distribution network according to claim 1, characterized in that: In step S3, the distribution network weak node identification model includes a three-layer Lssvm classification model, the first layer of the Lssvm classification model is used to classify whether the node is weak or not weak, the second layer of the Lssvm classification model is used to classify whether the node is short-term weak or medium- and long-term weak; the third layer of the Lssvm classification model is used to classify whether the node is seasonally weak or long-term weak; The optimal linear decision function of the Lssvm classification model at each layer is expressed as follows: Where x is the input distribution network node weak characteristic parameter sample, y(x) is the output corresponding to the input distribution network node weak characteristic parameter sample, and w k is the weight vector of the Lssvm classification model of the kth layer, b k is the deviation of the Lssvm classification model of the kth layer, φ(x) is the nonlinear mapping function that maps the weak characteristic parameters of the distribution network nodes from the original space to the high-dimensional space; The objective function and constraints of each layer of the Lssvm classification model are: In the formula, minJ(w k ,ξ k ) is the objective function of the least squares support vector machine corresponding to the Lssvm classification model of the kth layer; x v is the input weak characteristic parameter of the vth distribution network node, y v is the output corresponding to the input weak characteristic parameter of the vth distribution network node, l is the number of samples of weak characteristic parameters of the distribution network node; ξ v k is the error variable of the Lssvm classification model of the kth layer, ξ v k ≥0; c is the penalty coefficient, and c>0.
5. The method for identifying weak nodes in a distribution network according to claim 4, characterized in that: In step S3, a time scale factor is introduced to perform differentiated processing on the weak characteristic parameters of the distribution network nodes input into the distribution network weak node identification model, including: Introduce the time scale factor σ, its expression is as follows: Where D sv,i is the number of days when the voltage of node i exceeds the limit in the historical data, D T is the total number of days of historical data; The weak characteristic parameters of the distribution network nodes input into the Lssvm classification model at each layer of the distribution network weak node identification model are differentiated, as shown below: x′=λ 1,lssvmk x+λ 2,lssvmk s Where x is the weak characteristic parameter of the distribution network node; 1,lssvmk and λ 2,lssvmk are the weight coefficients of x and σ corresponding to the Lssvm classification model of the kth layer respectively; x' is the weak characteristic parameter of the distribution network node after differentiation processing.
6. The method for identifying weak nodes in a distribution network according to claim 5, characterized in that: The time scale factor σ is updated online in real time.
7. The method for identifying weak nodes in a distribution network according to claim 1, characterized in that: Step S5 includes: The weak characteristic parameters of the distribution network nodes in real-time operation are used as the input of the Lssvm classification model of the first layer after differentiation processing. If it is judged to be not weak, the output is the first string, and if it is judged to be weak, the output is the second string; If it is judged to be weak, the weak characteristic parameters of the distribution network node in real-time operation are further classified by using the Lssvm classification model of the second layer after differentiation processing. If it is judged to be short-term weak, the third string is output, and if it is judged to be medium- and long-term weak, the fourth string is output; If it is judged to be medium- to long-term weakness, the weak characteristic parameters of the distribution network nodes during real-time operation will be further classified after differentiation processing using the third-layer Lssvm classification model. If it is judged to be seasonal weakness, the fifth character string will be output; if it is judged to be long-term weakness, the sixth character string will be output.
8. A distribution network weak node identification system, characterized in that: include: Data acquisition module, used to collect historical topological data and electrical operation data of the distribution network; A database building module is used to build a distribution network node weakness characteristic parameter database based on the historical topological data and electrical operation data of the distribution network; A model building module is used to build a distribution network weak node identification model based on the least squares support vector machine, and introduce a time scale factor to perform differentiated processing on the distribution network node weak characteristic parameters input into the distribution network weak node identification model; A model training module, used to train a distribution network weak node identification model using data in a distribution network node weak characteristic parameter database; The weak node identification module is used to identify the weakness of the current distribution network nodes using the trained distribution network weak node identification model according to the weak characteristic parameters of the distribution network nodes during real-time operation.
9. An electronic device, characterized in that: include: a memory having a computer program stored thereon; A processor, configured to implement the method for identifying weak nodes in a distribution network as claimed in any one of claims 1 to 7 when loading and executing the computer program.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for identifying weak nodes in a distribution network as claimed in any one of claims 1 to 7 is implemented.
Citation Information
Patent Citations
Distribution transformer operation dynamic evaluation method based on entropy method
CN111259317A
Power grid voltage feature extraction method and system based on data driving and storage medium
CN117493836A
Power distribution network weak line identification method, system and device and storage medium
CN118332418A
