Electric power system transient stability assessment method and device

Abstract

Embodiments of the invention provide an electric power system transient stability assessment method and device. The method comprises the following steps of: obtaining a feature vector of a to-be-assessed point of an electric power system according to a pre-screened key state variable; respectively calculating a mahalanobis distance between the feature vector of the to-be-assessed point and a feature vector of each sample point in a sample set, and searching sample points adjacent to the to-be-assessed point according to the mahalanobis distances; and obtaining an assessment result of the to-be-assessed point according to a classification decision rule and the adjacent sample points. According to the method and device, the mahalanobis distances between the feature vector of the to-be-assessed point and the feature vectors of the sample points are calculated, and the mahalanobis distances sufficiently consider the relevancy between key state variables in the feature vectors, so that thedistance calculation is more correct, the adjacent sample points of the to-be-assessed points can be correctly searched, and the correctness of the assessment result is further improved.

Description

technical field [0001] Embodiments of the present invention relate to the field of power systems, and more specifically, to a method and device for evaluating transient stability of a power system. Background technique [0002] With the access of various new energy sources and the adoption of UHV transmission lines, the power grid has become more complex. Through the detailed modeling of the system, the method of analyzing the power system by using the traditional numerical calculation simulation method has the characteristics of complex model and slow speed. The accumulation of power grid operation data and the adoption of big data methods have brought new ideas to the method of predicting the transient stability of power systems based on data and machine learning, which can speed up the speed of power system simulation. However, if the prediction method based on machine learning directly analyzes the dynamic data of all variables in the power grid, it is not only difficul...

AI Technical Summary

Problems solved by technology

However, since the dynamic data has been processed and screened before classification or clustering, there may be a certain correlation between the attributes of the samples

If the Euclidean distance is used directly, these correlations may not be taken into account, so that the nature of the dynamic process of the power system cannot be accurately described

Images