Application of Multiple Stepwise Regression in Identifying Critical Inter-Area Oscillation Mode

LIN Bo,L(U) Lin,LIU Junyong,LIU Youbo,WEN Lili

DOI: https://doi.org/10.13335/j.1000-3673.pst.2015.03.029

2015-01-01

Abstract:As to inter-area oscillation which frequently appears in transmission lines, a novel online WAMS identification method for critical oscillation mode is presented, which makes use of stepwise regression method. Firstly, Prony algorithm is utilized to obtain oscillation modes of active power flow from WAMS or simulation data. Then parameters of oscillation modes are introduced to construct explanatory variable set and build multiple regression model. Secondly, stepwise regression method is raised to insert and delete explanatory variable in an iteration process and acquire new explanatory variable set that meet significance test of the model. Sig-value of the specified procedures is used to evaluate importance of the obtained oscillation mode automatically. Finally, analysis results of a test model and Sichuan power grid show that the proposed stepwise regression method is able to identify critical oscillation mode and has strong robustness in noise signal, which verifies the feasibility and effectiveness of the proposed method.