Diagnostic based on high frequency resonances and information fusion

The stator insulation breakdown is a major cause of AC machine failures. Ground insulation defaults are easily detected by classical systems based on leakage current measurements, however the turn-to-turn insulation degradations are more difficult to detect. For large machines, on-line methods, based on partial discharge detection and analysis, give good results but they cannot be used for low-voltage machines fed by adjustable speed drives (ASD). A new monitoring system able to detect slight variations of high frequency resonances in the winding of a working machine fed by an industrial inverter was presented in [1]. Several measurements can be used in order to estimate the aging of an AC machine winding (HF measurements of current or magnetic field) [2]. When a measure, also called a piece of information, is precise and certain, no other measure is necessary. However, such a measure is rarely obtained in real world application. Information fusion consists then in merging, or exploiting conjointly, several imperfect sources of information to make proper decision. Various frameworks can be used to model the fusion, e.g., probability theory, possibility theory, belief functions [3,4]. In this paper, different measurements of the aging of an AC machine winding are combined in the latter frame. This approach is tested on real measurements.