Electric power system transient stability evaluation method

Abstract

The invention discloses an electric power system transient stability evaluation method, which comprises the steps: building a to-be-tested electric power system by using electric power system time domain simulation software PSD-BPA; extracting a characteristic quantity capable of representing the state of an electric power system based on large amount of sample data, and performing normalization processing on characteristic data; completing the construction of a multi-radial basis function extreme learning machine integration model, wherein the output result of the integration model is the output after the prediction results of all sub-models are integrated; and after a fault occurs, inputting the acquired characteristic quantity of the electric power system at the steady-state moment andthe fault occurrence moment of the electric power system and after the fault removing into the trained multi-radial basis function extreme learning machine integration model, and predicting the transient stability of the electric power system. According to the method, the stability rules learned by different sub-models are utilized, so that important engineering application value is achieved for online transient stability evaluation of the electric power system on the premise that high evaluation accuracy is kept.

Description

technical field [0001] The invention relates to the technical field of electric power industry, in particular to a transient stability evaluation method of a power system. Background technique [0002] At present, the evaluation methods of power system transient stability can be divided into three types: time-domain simulation method, direct method and artificial intelligence method. The time-domain simulation method is to perform numerical integration operations on the differential equations and algebraic equations reflecting the transient process of the power system, and judge the transient stability of the power system according to the change process and trend of the relative rotor angle of each synchronous generator , this method has good stability and high calculation accuracy, and can be adapted to various nonlinear models. However, due to its high calculation cost and long time, it can only be applied to offline calculation and cannot meet the calculation speed requir...

AI Technical Summary

Problems solved by technology

In addition, when using a single machine learning model for transient stability evaluation, usually only the set of attributes with the highest accuracy in the test sample is selected for further discussion, but in actual selection of different attributes for the training and testing of the machine learning model In the process, there are often many machine learning models with different attributes showing similar accuracy, but the fitted function forms are quite different.

Images