Virtual lab for system identification of an electromechanical system

TitleVirtual lab for system identification of an electromechanical system
Publication TypeConference Proceedings
Year of Conference2005
AuthorsO'Malley, MK, McStravick, DM
Conference NameAmerican Society of Mechanical Engineers, Dynamic Systems and Control Division (Publication) DSC
Volume74 DSC
Pagination705 - 712
Conference LocationOrlando, FL, United States
KeywordsComputer software; Data acquisition; Mathematical models; Mechanical engineering; Real time systems; Students
Abstract

A stand-alone virtual instrument (vi) has been developed to augment an experimental system identification laboratory exercise in a required mechanical engineering course on system dynamics. The Virtual Lab (VL) was used productively as a post-lab exercise in conjunction with an existing laboratory experiment for system identification. The VL can be formatted as a standalone file, which the students can download and access at their convenience, without the need for LabVIEW software. The virtual lab presented in this paper used the experimental identification of a transfer function for an xy recorder developed at Rose-Hulman Institute of Technology. In the original Rose-Hulman experiment, students view a video of the acquisition of frequency response data for an X-Y recorder. Then, students complete a detailed optimization procedure using Microsoft Excel in order to determine system parameters for two transfer function models. This paper describes using the Virtual Lab to extend the original lab exercise into an interactive mode. The students complete the Microsoft Excel part of the exercise, but then repeat the optimization using brute force via the LabVIEW based VL developed by the authors, rather than using the optimization toolbox in Excel. With the VL, students can see in real-time the effects of each unknown parameter on the frequency response plot, thus providing additional insight into the relationships between these parameters and the behavior of the electromechanical system. This feature is notably absent in the Microsoft Excel portion of the exercise. Although this exercise uses simple dynamic models, the combination of Excel and LabVIEW approaches provide an insightful introduction to experimental system identification. In this paper, details of the VL are presented, including the functionality of the VL and methodologies for disseminating the VL as a stand-alone piece of software. Finally some assessment results for the original (Excel version) and VL methods of presenting the laboratory exercise are discussed. Copyright © 2005 by ASME.

File attachments: