Affiliation of Author(s):自动化学院
Journal:Chinese Control Conf., CCC
Abstract:This study proposes a diagnosis method for IGBT open circuit fault and sensor drift fault in CRH2 through ensemble empirical mode decomposition (EEMD) and an improved kernel principal component analysis (KPCA) way. EEMD is modified to reduce noise in the current and voltage signals. Subsequently, an improved KPCA method using weight matrix to highlighting fault information is developed to detect and diagnose the composite fault. Then the three-phase current averaging method is applied to perform fault location. Numerical simulation results verify the validity and availability of the composite fault detection method. © 2019 Technical Committee on Control Theory, Chinese Association of Automation.
ISSN No.:1934-1768
Translation or Not:no
Date of Publication:2019-07-01
Co-author:Li, Ziheng,Wang, Li,Wang Liying,Jiang Bin
Correspondence Author:Fuyang Chen
Date of Publication:2019-07-01