Detection of power transformer windings looseness based on vibration analysis

Jie Guo,Hai Huang
2011-01-01
Abstract:Throug the coi feature two asy feature based o is estab the glob which /bottom simplif compon vibratio to exam the exp comple gh the mode il vibration s which can ymmetry fe which rep on the classi blished first bal asymme is only cou m asymmetr fied SDOF nents and th on signals s mine the pro periments a ement to the elling, simu n signals un n help disclo eatures, the presents the ical spring-Tly, then the etry feature. unting the s ry feature a nonlinear he entropy sampled fro oposed featu and verify e fault diagn lating, expe nder differe osing the tra e axial vibr disorder o mass mode e multiple d . Combined tatic nonlin and axial v model, the of the recur om two simi ures; the res the propos nosis of the erimental re ent clampin ansformer w ration ampl of the vibra el and count dimensiona d with the w nearity of th vibration dis e nonlinear rrence histo ilar 35KV p sults show a ed features winding by esearch of th ng forces a winding loo litudes distr ation signal ts the hard n l MODF m winding inne he winding stribution f feature inc ogram are d power trans a consistenc s. The prop y vibration m he power tr are analyse oseness are p ribution fea . The MDO nonlinearitie model was in er properties is simulate features. Fu cluding the derived. At l sformers are cy between posed featur methods. ransformer w ed; some im proposed, in ature and n OF model w es of the pre ntroduced t s, the MDO ed to derive urther, by u e 200Hz fr last, the exp e analysed the simulat ures can be winding, mportant ncluding nonlinear which is essboard to derive OF model e the top using the requency periment in detail tions and a good mathods. Copyright © (2011) by the International Institute of Acoustics & Vibration.
What problem does this paper attempt to address?