Feature Extraction Method for Topological Structure of Operation Mode Applicable to Multi-source Heterogeneous Power Grid

Abstract

The invention relates to a method for extracting topological structure features of an operation mode applicable to a multi-source heterogeneous power grid, and belongs to the technical field of power system operation and control. First obtain the operating status of the multi-source heterogeneous power grid, including obtaining the topology information and operation mode information of the power grid; construct a reasonable initial feature set according to the multi-source heterogeneous network structure and operation mode information; obtain the network topology of the multi-source heterogeneous power grid Edge weight adjacency matrix W and node attribute matrix M; establish an embedding representation learning model, solve to obtain a consensus embedding vector H, and then obtain a feature expression that integrates multi-source heterogeneous power grid network topology and node information; adopts the feature selection algorithm in deep learning , determine the features with high importance, and obtain the key topology features of the operation mode of the multi-source heterogeneous power grid. The method improves the efficiency and accuracy of key feature extraction of multi-source heterogeneous power grid network topology, and provides technical basis and practical method for power grid feature extraction of multi-source heterogeneous power grid.

Description

Technical field [0001] The present invention relates to an operational mode topology feature extraction method suitable for use in multi-source heterogeneous webs, which belongs to the field of power systems operation and control technology. Background technique [0002] With the scale of my country's special high pressure, the scale is expanding, the high permeability intermittent new energy power generation and massive flexible load response exacerbates the uncertainty of the bilateral bilateral side of the power grid, and the operation mode of power system and dynamic behavior are more complicated. Nature The frequency of disasters makes it more complicated in the scenario scenarios that may be generated, and higher requirements for the level of coordinating decision-making and collaborative control of dispatching operations. Therefore, the online security risk situation of the DC mixed power system is studied, and the transient stability and efficiency of large grids are impl...

AI Technical Summary

Problems solved by technology

[0003] Since the actual large-scale AC / DC hybrid system contains many variables, its data scale is large and its dimension is high, only a few features will affect the stability of the system, and most of the features are redundant. If it is used as a transient evaluation The input of the model will inevitably affect the calculation efficiency, classification effect and online application requirements, so feature dimensionality reduction is a key problem to be solved urgently

Some large disturbances will inevitably occur in the power system, which often bring short-term and large changes to the network topology or network parameters, seriously affecting the safe and stable operation of the power system.

In addition, due to the planned maintenance of lines (or other components), the commissioning of new lines or equipment may cause changes in the topology of the power grid

Some studies have discussed the impact of topology changes on transient stability assessment models, but only involve slight changes in the grid topology of pure AC systems, and have not effectively considered the impact of renewable energy and DC access on power systems

Images