Date of Award
Summer 1967
Document Type
Thesis - Restricted
Degree Name
Master of Science (MS)
Department
Mechanical Engineering
First Advisor
Blank, Gary L.
Abstract
This thesis is concerned with the identification of multiparameter systems by the use of adaptive multiparameter learning models. Multiparameter adjustment signals based on the differences between the corresponding model and system parameters are obtained by cross-correlating two signals within the adaptive system. The adjustment signals are fed back to the model to adjust its parameters such that the Mean Square Error is minimized along its path of steepest descent in parameter space. The Mean Square Error is a minimum when the corresponding system and model parameters are matched and thus indicates identification of the system by the model. The results obtained for the multiparameter case are applied to the identification of the parameters corresponding to the damping ratio and natural frequency of a second order system by adjustment of the corresponding parameters of a second order model. Analog simulation studies of the second order example are conducted in order to demonstrate and validate the theory.
Recommended Citation
Rygiewicz, Carl P., "An Adaptive Learning System Using Correlation and Steepest Descent Optimization Techniques" (1967). Master's Theses (1922-2009) Access restricted to Marquette Campus. 1681.
https://epublications.marquette.edu/theses/1681