| |
Polynomial and neural methods for characterization and diagnostic of a converter fed synchronous machine
| [View]
[Download]
|
| Author(s) |
O. Curea; B. Dakyo; G. Barakat |
| Abstract |
For the fault detection and the diagnosis, it is useful to have a behavior model of the actual system. A
converter fed permanent magnet synchronous machine is used for experimental purpose. The
acquisition system is based on the two wattmeter principle, so two voltages and two currents are
measured on the three phase side. The representations of the behavior will be chosen simple and
relevant for a diagnosis procedure. A polynomial and a neural network approximation for
characterization will be presented and compared. For the instant these models work on a personal
computer, but is foreseeable a real-time implementation. |
| Download |
| Filename: | EPE2003-PP0864 - Curea |
| Filesize: | 337.5 KB |
|
| Type |
Members Only |
| Date |
Last modified 2003-10-17 by Unknown |
| |
 |