Title :
Maximum likelihood estimation with reduced sensitivity functions
Author_Institution :
Systems Control, Inc., Palo Alto, CA
Abstract :
A new approach is proposed for maximum likelihood estimation in linear constant coefficient dynamic systems with general structure. The formulation is developed in terms of the minimum number of variables required to specify the parameter sensitivity of state variables, called the reduced sensitivity functions. An efficient computation procedure is obtained for parameter estimation from time domain or frequency domain data. The importance of end conditions in the frequency domain approach is analyzed.
Keywords :
Approximation algorithms; Autoregressive processes; Control system synthesis; Frequency domain analysis; Frequency estimation; Maximum likelihood estimation; Milling machines; Parameter estimation; State estimation; Stochastic processes;
Conference_Titel :
Decision and Control including the 14th Symposium on Adaptive Processes, 1975 IEEE Conference on
Conference_Location :
Houston, TX, USA
DOI :
10.1109/CDC.1975.270649
