| |
PARAMETER ESTIMATION IN A GEARED FLEXIBLE MECHANICAL SYSTEM USING A COMPOSITE NEURAL NETWORK-CONVENTIONAL APPROACH
| [View]
[Download]
|
| Author(s) |
J.G. O' Donovan; RC. Kavanagh; J.M.D. Murphy; P.J. Roche; M.G. Egan |
| Abstract |
This paper introduces a novel approach to parameter identification for nonlinear systems. A typical system to which the technique may apply is a drive motor coupled to a load in a nonlinear fashion. The algorithm includes iterative use of recursive least squares and neural net filtering of sensor quantisation noise. The effectiveness of the proposed
method is tested by considering a geared mechanical system with flexible transmission dynamics in which the gear coupling exhibits a bard non-linearity in the form of backlash. The governing system parameters, which include those of the gear, drive motor, shaft and load parameters are determined sequentially. |
| Download |
| Filename: | Unnamed file |
| Filesize: | 430 KB |
|
| Type |
Members Only |
| Date |
Last modified 2018-04-10 by System |
| |
 |