| |
DETECTION OF INDUCTION MACHINE INCIPIENT FAILURES USING KNOWLEDGE BASED TECHNIQUES
| [View]
[Download]
|
| Author(s) |
M. F. Cabanas; J. G. Aleixandre; M. A. Sanz; A. Neira; A. Otero; J. S. Sariego |
| Abstract |
This paper describes an expert system for the on-line diagnosis of working cage induction motors. The expert system uses spectral analysis of stator currents and motor vibrations as well as the supply voltages and currents, as input variables. The main difference between the system proposed here and a classical diagnosis expert system is the existence of models; the behaviour of the motor and its parts are defined by a hierarchical object network which incorporates a set of models where both theoretIcal data and data acquired through experience are integrated. The properties of network objects are defined once the motor-drive system has been specified. The structure is therefore dynamically updated for each particular case. Once the system has stored both values of electrical variables and the spectral analysis of currents and vibrations, the informatIon is compared with the data generated by the models, and the inference process
subsequently begins when a considerable difference between them is found. The system output consists of the identified failures, evidence found during the diagnosis process, and a maintenance plan based on the knowledge acquired from expert technicians. |
| Download |
| Filename: | Unnamed file |
| Filesize: | 575.9 KB |
|
| Type |
Members Only |
| Date |
Last modified 2017-11-06 by System |
| |
 |