CN-121996730-A - Distribution network overhaul topology identification system based on multi-source data fusion

CN121996730ACN 121996730 ACN121996730 ACN 121996730ACN-121996730-A

Abstract

The invention relates to a distribution network overhaul topology identification system based on multi-source data fusion, which realizes effective integration and accurate processing of multi-source distribution network data through cooperative work of a multi-source data acquisition module, a data preprocessing module and a measurement generation module, remarkably improves space-time resolution of distribution network state quantity, can quantify uncertainty of an identification result by constructing a confidence evaluation system and grading confidence of an output identification structure by utilizing an algorithm confidence evaluation system, greatly improves availability and reliability of the system in an actual complex environment, ensures applicability and effectiveness of the system in an actual overhaul scene through seamless connection and data interaction of a database, a marketing system and a scheduling system, can provide accurate and reliable topology identification support for overhaul of a distribution network, and improves overhaul efficiency and power supply reliability, and has the advantages of self-adaption, high precision and reliability.

Inventors

ZHANG CHANGQING
FAN SHIQIN
WANG YANFENG
YAN JIANKE
YAO HONGYING
CHEN XUANWEN

Assignees

国网河南省电力公司襄城县供电公司

Dates

Publication Date: 20260508
Application Date: 20251201

Claims (5)

1. The utility model provides a join in marriage net maintenance topology identification system based on multisource data fusion, includes multisource data acquisition module, data preprocessing module, measurement generation module, topology parameter identification module, algorithm optimization module, drawing mould database construction module and verification module, its characterized in that: the multi-source data acquisition module is provided with a multi-protocol data interface and is used for synchronously acquiring multi-source heterogeneous data from the SCADA system, the GIS system, the intelligent instrument and the user power consumption data system; the data preprocessing module can preprocess the acquired multi-source heterogeneous data, so that the integrity and the reliability of the data are ensured; The measurement generation module can establish a multi-view super-resolution measurement model based on the multi-source heterogeneous data after preprocessing is completed, train the multi-view super-resolution measurement model by utilizing the space topological relation provided by GIS image data and the time sequence measurement data provided by intelligent instrument data, and finally reconstruct high-resolution measurement information from low-resolution data by utilizing the trained multi-view super-resolution measurement model by adopting a deep learning technology, so that the defect of direct measurement is overcome; The topology parameter identification module can establish a power distribution network topology physical feature library, map and match physical features such as circuit connection relations, equipment type parameters and the like with hierarchical structures and activation function characteristics in a neural network structure, so that a dual-branch collaborative identification model is formed, one of the two models realizes dynamic identification of the topology structure, the other model completes accurate estimation of electrical parameters, and then collaborative optimization identification of topology and parameters is realized through cross-branch feature interaction; The algorithm optimization module can carry out deep analysis on the mechanism of the existing power distribution network parameter identification model based on the transfer learning technology, proposes a high-availability parameter identification algorithm suitable for complex sampling conditions, constructs an algorithm confidence assessment system through verification set performance monitoring, model correction and hidden mode extraction technology in the supervision training process, and carries out confidence scoring on an output identification structure by utilizing the algorithm confidence assessment system, so that the accuracy and reliability of parameter identification are enhanced; The graph model database construction module can be used for establishing a graph model database of the power distribution network and realizing seamless connection and data interaction between the database and the marketing system and the scheduling system; The verification module can test the performance of the algorithm in the actual environment based on the data provided by the graph module database, collect the field actual measurement data in real time, and then compare and analyze the field actual measurement data with the system identification result to generate a parameter identification error report.
2. The distribution network overhaul topology identification system based on multi-source data fusion of claim 1, wherein the preprocessing operation of the multi-source heterogeneous data is specifically that data noise is removed through an outlier detection model, data quality quantification assessment is carried out based on an entropy weight method, and time scale unification of the multi-source data is achieved through a time sequence alignment technology.
3. The distribution network overhaul topology identification system based on multi-source data fusion as set forth in claim 1, wherein a storage unit is provided in the data preprocessing module, and the storage unit is utilized to store and backup the preprocessed data.
4. The distribution network overhaul topology identification system based on multi-source data fusion as set forth in claim 1, wherein the multi-source data acquisition module acquires real-time operation data from the SCADA system, acquires topological geographic data from the GIS system, acquires high-precision measurement data from the intelligent instrument, and acquires historical operation data from the user electricity consumption data system.
5. The distribution network overhaul topology identification system based on multi-source data fusion as recited in claim 1, wherein a decision unit is arranged in the verification module, and a visual overhaul decision report is made by using the decision unit based on the parameter identification error report.

Description

Distribution network overhaul topology identification system based on multi-source data fusion Technical Field The invention belongs to the technical field of distribution network overhaul, and particularly relates to a distribution network overhaul topology identification system based on multi-source data fusion. Background In the maintenance work of the distribution network, accurate topological structure analysis and parameter identification are key for guaranteeing maintenance efficiency and power supply reliability, however, in practical application, due to the fact that data sources of the distribution network are various, data accuracy is uneven, time scale is uneven among data, space-time resolution of distribution network state quantity is insufficient, meanwhile, the existing parameter identification model is poor in adaptability under the condition of complex sampling, identification accuracy and reliability are difficult to meet actual maintenance requirements, accurate topological parameter support cannot be provided for maintenance work, scientificity and high efficiency of maintenance of the distribution network are affected, and therefore, in order to solve the problems, development of a distribution network maintenance topology identification system which is strong in self-adaption and is accurate and reliable and based on multi-source data fusion is necessary. Disclosure of Invention The invention aims to overcome the defects of the prior art and provides a distribution network maintenance topology identification system which is strong in self-adaption, accurate and reliable and is based on multi-source data fusion. The invention aims to realize that a distribution network overhaul topology identification system based on multi-source data fusion comprises a multi-source data acquisition module, a data preprocessing module, a measurement generation module, a topology parameter identification module, an algorithm optimization module, a graph model database construction module and a verification module, wherein the multi-source data acquisition module is provided with a multi-protocol data interface for synchronously acquiring multi-source heterogeneous data from an SCADA system, a GIS system, an intelligent instrument and a user power consumption data system; the data preprocessing module can preprocess the acquired multi-source heterogeneous data, so that the integrity and the reliability of the data are ensured; The measurement generation module can establish a multi-view super-resolution measurement model based on the multi-source heterogeneous data after preprocessing is completed, train the multi-view super-resolution measurement model by utilizing the space topological relation provided by GIS image data and the time sequence measurement data provided by intelligent instrument data, and finally reconstruct high-resolution measurement information from low-resolution data by utilizing the trained multi-view super-resolution measurement model by adopting a deep learning technology, so that the defect of direct measurement is overcome; The topology parameter identification module can establish a power distribution network topology physical feature library, map and match physical features such as circuit connection relations, equipment type parameters and the like with hierarchical structures and activation function characteristics in a neural network structure, so that a dual-branch collaborative identification model is formed, one of the two models realizes dynamic identification of the topology structure, the other model completes accurate estimation of electrical parameters, and then collaborative optimization identification of topology and parameters is realized through cross-branch feature interaction; The algorithm optimization module can carry out deep analysis on the mechanism of the existing power distribution network parameter identification model based on the transfer learning technology, proposes a high-availability parameter identification algorithm suitable for complex sampling conditions, constructs an algorithm confidence assessment system through verification set performance monitoring, model correction and hidden mode extraction technology in the supervision training process, and carries out confidence scoring on an output identification structure by utilizing the algorithm confidence assessment system, so that the accuracy and reliability of parameter identification are enhanced; The graph model database construction module can be used for establishing a graph model database of the power distribution network and realizing seamless connection and data interaction between the database and the marketing system and the scheduling system; The verification module can test the performance of the algorithm in the actual environment based on the data provided by the graph module database, collect the field actual measurement data in real time, and then compare and analyze the fie