A power system transient stability assessment method and system based on depth learning

Abstract

The invention discloses a transient stability assessment method and system based on depth learning, which comprises the following steps: constructing an original feature set of transient stability assessment; establishing an off-line expected accident set; based on mutual information theory, a feature extraction model of a stack denoising automatic encoder being established adaptively, and the multi-layer feature extraction of original input being realized; a support vector machine (SVM) integrated regression model being established to judge the transient stability margin of various operationmodes under fixed expected accidents; generating a plurality of future operation modes and an on-line predicted accident set; the support vector machine integrated regression model corresponding to each expected accident being used to predict the index of transient stability margin, and the severity of the transient stability of the system under different operating modes being classified. Combinedwith the prior distributed parallel computing technology, the invention finally realizes the optimal network structure of the depth learning model with a large number of network nodes and a pluralityof hidden layers adaptively, and can quickly and accurately evaluate the transient stability degree of the power system, so as to meet the online application requirements.

Description

technical field [0001] The invention belongs to the field of dynamic security assessment of power systems, and in particular relates to a method and system for transient stability assessment of power systems based on deep learning. Background technique [0002] In recent years, the scale of my country's UHV AC / DC hybrid power grid has continued to expand. High-permeability intermittent new energy power generation and massive flexible load response have aggravated the uncertainty on both sides of the power grid. The operation mode and dynamic behavior of large power grids have become increasingly complex. The frequent occurrence of natural disasters makes the expected accident scenarios more complicated, and puts forward higher requirements for the overall decision-making level and collaborative control ability of dispatching and operation. At the same time, with the continuous deepening of smart grid construction, control centers at all levels have accumulated a large amount ...

AI Technical Summary

Problems solved by technology

However, the existing transient stability evaluation models based on deep learning have the following deficiencies: first, how to adaptively determine the optimal network structure of a deep learning model with a large number of network nodes and multiple hidden layers; assess the stability of the

Images