Development of Fault Diagnostic Method for Induction Motors Based on Multi-Dimensional Probability Density Functions

点击翻译按钮获取中文摘要

Probabilistic diagnosis method was developed to detect turnto-turn short-circuit fault of stator winding, which is one of the most probable faults in induction motors. It was confirmed that magnitude and phase were features characterizing current following into a winding, depending on winding conditions. Fault probability, defined as internal area of the ellipsoid, was derived theoretically based on feature distributions of healthy motors and features of a target motor to be diagnosed. Examination using motors with some kinds of turnto-turn shortcircuit faults proved that the method proposed is effective under both no-load and loaded conditions. Possible applications of the method are outgoing inspections by motor manufacturers and periodical/non-periodical inspection at site