Identifikation und Prädiktion von Fehlern bei Käfigläufer-Niederspannungs-Normmotoren [article]

Christoph Haug, Universität Stuttgart, Universität Stuttgart
2003
In this work the development of a monitoring and diagnostic system for AC-Induction-Motors is described. To identify problems of use and faults, also to get out the requests and the investment readiness of the customers and users of those induction motors, a survey was made. Due to this survey, as well as the advice of motor manufacturers and scientific considerations, a conceptual formulation for a monitoring and diagnostic system has been developed. The strategy for identification and
more » ... ication and prediction of faults is divided into three layers: First the monitoring layer collects data for analysis. This includes sensors, analogue to digital conversion etc.. Also the pre-processing of the direct measured signals with standard algorithms is included. Another part of the monitoring level are state value observers, which calculate unknown state values, like torque, current and power factor, out of the known values by courtesy of the motors equivalent. The fault identification layer appraises the data supplied by the monitoring level. The aim is to detect, localise and weight faults. This is done by attribute extraction and analysis by means of blurred ranges. The failure prediction layer uses the fault data, supplied by the fault identification layer. This data will be used to forecast a motor failure. The values, delivered by the monitoring layer, come in a heterogeneous structure. This heterogeneous structure has to be transformed into a simple homogeneous structure to be analysed and handled with a cheap microcontroller system. Therefore ways are searched and described to convert the data into state vectors of a state space with single scalar state values for every attribute and measured value or time function. Further, possible faults and defects of Induction-Motors are shown. Especially the acquisition by acceleration sensors and the analysis of motor vibrations, which are build by air gap forces, and a temperature prediction is described.
doi:10.18419/opus-2505 fatcat:ipn7bcts45hgjlnbfkcvravy5u