CN-122020926-A - Multi-granularity panoramic dynamic topology layering compression method and system for power distribution network

CN122020926ACN 122020926 ACN122020926 ACN 122020926ACN-122020926-A

Abstract

The invention discloses a multi-granularity power distribution network panoramic dynamic topology layering compression method and system, which belong to the technical field of power distribution network modeling and analysis, and comprise the steps of obtaining power distribution network topology and running state data, carrying out multi-scale expansion on a power distribution network topology structure, calculating fractal dimension and load distribution fractal index, combining voltage association degree index and power flow coupling coefficient among nodes to construct a topology-electric joint feature vector, adopting a global layer-partition layer-node layer three-level topology compression strategy to form a topology compression model, abstracting mapping relations among different levels into electric synaptic connection, dynamically adjusting synaptic weights based on the running state data to update the topology compression model, identifying key nodes from the topology compression model, introducing quantum entanglement state to carry out modeling, taking entanglement degree conservation as constraint conditions, carrying out calibration correction on the topology compression model by combining a digital twin model, and outputting a final characteristic-preserving multi-granularity topology model.

Inventors

ZHOU SUYANG
ZHOU AIHUA
PENG LIN
FAN JILI
GAO MINGYANG
LIU MEIZHAO

Assignees

东南大学
中国电力科学研究院有限公司
国网江苏省电力有限公司信息通信分公司

Dates

Publication Date: 20260512
Application Date: 20260128

Claims (10)

1. The panoramic dynamic topology layering compression method for the power distribution network with multiple granularities is characterized by comprising the following steps of: Acquiring topology and running state data of a power distribution network, and constructing a panoramic data set of the power distribution network; based on a panoramic dataset of the power distribution network, carrying out multi-scale expansion on a topological structure of the power distribution network, calculating fractal dimension and load distribution fractal index, and constructing a topological-electrical joint feature vector by combining voltage association index and power flow coupling coefficient among nodes; Clustering nodes with similar fractal characteristics based on topology-electric joint characteristic vectors in a global layer by adopting a global layer-partition layer-node layer three-level topology compression strategy to generate global layer supernodes; Abstracting mapping relations among different levels in the multi-granularity hierarchical topology compression model into electrical synaptic connection, and dynamically adjusting synaptic weights based on the running state data to update the multi-granularity hierarchical topology compression model; And identifying key nodes from the updated multi-granularity hierarchical topology compression model, modeling the electrical association relation among the key nodes by introducing a quantum entanglement state, taking entanglement conservation as a constraint condition in the topology compression and dynamic adjustment process, and carrying out calibration and correction on the topology compression model by combining a digital twin model to output the final characteristic-fidelity multi-granularity topology model.
2. The multi-granularity power distribution network panoramic dynamic topology layering compression method of claim 1, wherein the power distribution network panoramic dataset The expression of (2) is: Wherein, the To represent an undirected graph of the topology of the distribution network, Representing a collection of nodes in a distribution network, Representing the total number of nodes in the distribution network, Representing a collection of leg connections between nodes, Representing the total number of branches in the power distribution network; as a set of parameters of the branch, Represent the first Equivalent resistance parameters of the branches; Represent the first Equivalent reactance parameters of the branches; represent the first The switching state variable of the branch circuit is 1 when the branch circuit is closed, and 0 when the branch circuit is opened; for a set of power states of a node, Representing nodes Active load power of (a); Representing nodes Is set in the power supply system; Representing nodes The accessed distributed energy source outputs power; As a set of node load change rates, And (3) with Representing nodes respectively The active load values at two adjacent moments, The severity of the node load over time is characterized.
3. The multi-granularity power distribution network panoramic dynamic topology layering compression method according to claim 1, wherein the topology-electrical joint feature vector The expression of (2) is: Wherein, the Representing the fractal dimension of the topology, Representing scale parameters The minimum number of boxes required to down cover the topology substructure; is a fractal index of load distribution, Expressed in the load threshold scale The minimum number of coverage areas required; representing the voltage magnitudes of the adjacent nodes, Representing the equivalent impedance parameter between the corresponding nodes, Is a voltage association index; Representing the power flow coupling coefficient, Representing the active power flow of a certain branch, Representing the power flow normalization reference value at the current scale.
4. A multi-granularity power distribution network panoramic dynamic topology layering compression method as recited in claim 3, wherein topology-electrical joint feature vectors are obtained The method comprises the following steps: S21, carrying out multi-scale expansion on the topology of the power distribution network, and extracting topological substructures under different levels; S22, for the topological substructure, fractal depiction is carried out on the topological structure by adopting a box coverage idea, and the fractal dimension of the topological structure is calculated; S23, introducing fractal characteristics of node load distribution, defining a node load amplitude set, and carrying out fractal analysis on the load amplitude distribution to obtain a load distribution fractal index; S24, on the basis of the topological fractal characteristics, introducing electrical coupling information among nodes, defining a node voltage amplitude set, and constructing a voltage association index of any adjacent node pair; s25, defining a branch power flow set, and calculating a power flow coupling coefficient between node pairs; and S26, uniformly describing the topological fractal characteristics and the electric coupling characteristics, and constructing a topological-electric combined characteristic vector.
5. The multi-granularity power distribution network panoramic dynamic topology layering compression method according to claim 1, wherein the step of constructing the multi-granularity layering topology compression model comprises: S31, carrying out fractal similarity clustering on nodes of the whole network based on topology-electrical joint feature vectors, defining a node feature sample set M, carrying out similarity clustering based on the feature sample set to form a global fractal cluster set H, and enabling each fractal cluster to be equivalent to a global layer supernode; s32, constructing an equivalent electrical parameter model of each global layer supernode, and defining a supernode equivalent parameter set; S33, carrying out partition layer modeling on the inside of each super node on the basis of global layer compression to define a partition layer fractal unit set; S34, identifying non-key branches at the node layer, and defining branch importance indexes The method comprises the following steps: Wherein, the Representing the magnitude of the power flow carried by the branch, When the importance of the branch is lower than a set threshold value, judging the branch as a non-critical branch; S35, aiming at the branch which is judged to be non-critical, adopting a fractal iteration mode to compress the node layer, and defining a node compression mapping relation ; S36, after the step-by-step compression of the whole domain layer, the partition layer and the node layer is completed, a multi-granularity hierarchical topology compression model is formed : 。
6. The multi-granularity power distribution network panoramic dynamic topology layering compression method of claim 5, wherein the step of updating the multi-granularity layering topology compression model comprises: S41, carrying out hierarchical mapping modeling on the multi-granularity hierarchical topology compression model, wherein the hierarchical mapping modeling comprises a global layer, a partition layer and a node layer, and defining an electric synaptic connection set according to the mapping relation between any adjacent layers : Wherein, the Representing information transfer channels between different granularity levels; s42, constructing corresponding synaptic weights for each electrical synaptic connection Defining a set of synaptic weights The method comprises the following steps: Wherein, the An electrical coupling strength indicator representing a corresponding region of the synaptic connection; representing the topological compression ratio of the corresponding region of the synaptic connection; S43, quantifying the running state change of the power distribution network, and defining a running state change index set The method comprises the following steps: Wherein, the Representing the integrated power state value of the same region at adjacent times; Representing the dynamic fluctuation intensity of the running state of the region; s44, triggering a synaptic strengthening mechanism when the running state change strength exceeds a set threshold value, wherein the synaptic strengthening criterion function is as follows: Wherein, the Representing a synaptic strengthening trigger threshold; Indicating that synaptic boosting is effective, updating synaptic weights under synaptic boosting conditions: Wherein, the In order to synapse the enhancement gain factor, Is the highlighting weight after strengthening; S45, triggering a synaptic weakening mechanism when the running state is in a stable state for a long time, wherein the stability evaluation index is as follows: Wherein, the Representing a stability assessment time window length; representing the average fluctuation level of the operating state of the area when And when the synaptic weight is lower than a preset threshold value, performing weakening adjustment on the synaptic weight: Wherein, the For the synaptic weakening of the attenuation coefficient, Representing the synaptic weight after weakening; S46, updating the multi-granularity hierarchical topology compression model according to the real-time change result of the synaptic weight: Wherein, the Representing a dynamically adjusted set of synaptic weights; And representing the updated multi-granularity hierarchical topology compression model.
7. The multi-granularity power distribution network panoramic dynamic topology layering compression method of claim 6, wherein the process package for calibrating and correcting the topology compression model: s51, identifying key nodes with obvious influence on operation of the power distribution network from the updated multi-granularity hierarchical topology compression model to form a key node set Constructing a key association pair set according to the electrical association relation between the key nodes ; S52, aiming at the electrical association relation of each pair of key nodes, quantum entanglement is introduced to carry out modeling, and entanglement state parameter sets of the key node pairs are defined The method comprises the following steps: Wherein, the Representing the power flow association strength mapping quantity between key node pairs; Representing a voltage stability association map quantity between key node pairs; S53, introducing entanglement degree conservation constraint, and defining entanglement degree functions of key node pairs as follows: Wherein, the Representing the entanglement metric value before compression; representing the compressed entanglement metric value; a threshold value for allowable entanglement deviation; s54, constructing a digital twin mapping relation between a topological compression model and an actual power distribution network, and defining a digital twin state vector The method comprises the following steps: Wherein, the Representing voltage measurement values of the key nodes; A current measurement representing a critical leg; representing a switch state measurement value; s55, based on the output result of the digital twin model, identifying the electrical characteristic deviation of the topology compression model, and defining a characteristic deviation evaluation index The method comprises the following steps: Wherein, the Representing the electrical characteristic quantity output by the digital twin model; Representing the corresponding electrical characteristic quantity calculated by the topology compression model; s56, outputting the final characteristic fidelity multi-granularity topology model : Wherein, the Representing the dynamically adjusted topology compression model; representing a set of key nodes; representing the entangled state parameter set of key node pairs.
8. A multi-granularity panoramic dynamic topology layering compression system for a power distribution network, which performs the compression method of any one of claims 1-7, comprising: The data acquisition module is used for acquiring topology and running state data of the power distribution network and constructing a panoramic data set of the power distribution network; the multi-scale structure analysis module is used for multi-scale expansion of a topological structure of the power distribution network based on a panoramic data set of the power distribution network, calculating fractal dimension and load distribution fractal index, and constructing a topological-electrical joint feature vector by combining voltage association index and power flow coupling coefficient among nodes; The topology compression module adopts a global layer-partition layer-node layer three-level topology compression strategy, clusters nodes with similar fractal characteristics based on topology-electrical joint feature vectors in the global layer to generate global layer supernodes, reserves core electrical connection relations in fractal units in the partition layer, carries out fractal iterative compression on non-key branches in the node layer, and forms a multi-granularity hierarchical topology compression model; the dynamic adjustment module abstracts the mapping relation between different levels in the multi-granularity hierarchical topology compression model into electrical synaptic connection, and dynamically adjusts synaptic weights based on the running state data to update the multi-granularity hierarchical topology compression model; And the calibration correction module is used for identifying key nodes from the updated multi-granularity hierarchical topology compression model, modeling the electrical association relation among the key nodes by introducing a quantum entanglement state, taking entanglement conservation as a constraint condition in the topology compression and dynamic adjustment process, combining a digital twin model to calibrate and correct the topology compression model, and outputting the final characteristic fidelity multi-granularity topology model.
9. A computer storage medium storing a readable program, wherein the program, when executed, is capable of instructing a computing device to perform the multi-granularity power distribution grid panoramic dynamic topology layering compression method of any one of claims 1-7.
10. An electronic device is characterized by comprising a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete communication with each other through the communication bus; the memory is configured to store at least one executable instruction, where the executable instruction causes the processor to perform operations corresponding to the multi-granularity power distribution network panoramic dynamic topology layering compression method according to any one of claims 1-7.

Description

Multi-granularity panoramic dynamic topology layering compression method and system for power distribution network Technical Field The invention belongs to the technical field of modeling and analysis of power distribution networks, and particularly relates to a multi-granularity power distribution network panoramic dynamic topology layering compression method and system. Background In recent years, with the rapid development of new energy and the high-proportion access of distributed energy in a power distribution network, the structure of the power distribution network is gradually changed from a traditional unidirectional radiation type into a complex network form of multiple sources, multiple nodes and high coupling. Under the background, the running state of the power distribution network presents obvious dynamic and multi-scale characteristics, the topological structure is frequently changed, the number of nodes and the connection relation are continuously increased, and higher requirements are put forward on the real-time performance and the accuracy of modeling, analysis and scheduling decision of the power distribution network. Particularly, in the applications of fault analysis, power flow calculation, operation optimization and the like, how to reduce the complexity of a topology model on the premise of ensuring accurate expression of electrical characteristics becomes a key problem in intelligent operation and management of a power distribution network. In the prior art, the topology modeling and simplifying method of the power distribution network mainly comprises an equivalent modeling method based on physical partition, a hierarchical modeling method based on feeder line or station partition and a network compression method based on graph theory. The method based on physical partitioning generally divides the power distribution network according to administrative regions or equipment boundaries, has the advantage of simple implementation, and the hierarchical modeling method based on feeder lines or plant stations can reflect the network structure hierarchy to a certain extent. However, most of the methods rely on preset static division rules, and fail to fully consider self-similar characteristics of a power distribution network topological structure and complex electric coupling relations among nodes, so that a compression result is not matched with actual operation characteristics easily. In addition, the existing topology compression technology generally adopts a compression strategy with fixed granularity or fixed stride, and is difficult to adapt to dynamic changes of the running state of the power distribution network. Under the scene of fluctuation of new energy output, rapid change of load or frequent action of a switch, the compressed topology model often needs frequent reconstruction, has low updating efficiency and is not enough in real-time. Some researches attempt to introduce a data driving or self-adaptive modeling method to improve the flexibility of a model, but the method is concentrated in a single-level or unidirectional adjustment process, lacks a multi-granularity cooperative adjustment mechanism, and is difficult to consider the calculation efficiency and the model precision in different operation scenes. Meanwhile, in the topology compression process, the retention of electrical characteristics among key nodes depends on experience rules, a unified and constraint characteristic fidelity mechanism is lacked, and errors are easy to accumulate after multiple times of compression and adjustment. Therefore, in order to address the above-mentioned shortcomings, it is needed to propose a topology modeling method of a power distribution network, which can achieve the effects of simple structure, dynamic adaptation and electrical characteristics fidelity. Disclosure of Invention Aiming at the defects of the prior art, the invention aims to provide a multi-granularity panoramic dynamic topology layering compression method and system for a power distribution network, which solve the problems in the prior art. The aim of the invention can be achieved by the following technical scheme: A multi-granularity panoramic dynamic topology layering compression method for a power distribution network comprises the following steps: Acquiring topology and running state data of a power distribution network, and constructing a panoramic data set of the power distribution network; based on a panoramic dataset of the power distribution network, carrying out multi-scale expansion on a topological structure of the power distribution network, calculating fractal dimension and load distribution fractal index, and constructing a topological-electrical joint feature vector by combining voltage association index and power flow coupling coefficient among nodes; Clustering nodes with similar fractal characteristics based on topology-electric joint characteristic vectors in a global layer by adopting a gl