Title :
A necessary condition for effective performance of the multiple model adaptive estimator
Author_Institution :
Dept. of Eng., Hofstra Univ., Hempstead, NY
fDate :
7/1/1995 12:00:00 AM
Abstract :
Effective adaptive estimation for a general linear system driven by an input modeled by a randomly switching Gaussian process is considered. The performance of the multiple model adaptive estimator (MMAE) is, in some cases, unexpectedly hampered by a necessary condition not satisfied by the linear system. This key dependency for effective MMAE performance is based on a particular property of the DC gain of the linear system
Keywords :
Gaussian processes; adaptive Kalman filters; adaptive estimation; adaptive signal processing; DC gain; effective performance; general linear system; linear system; multiple model adaptive estimator; randomly switching Gaussian process; Adaptive algorithm; Adaptive estimation; Adaptive filters; Gaussian noise; Gaussian processes; Linear systems; Parallel processing; Performance gain; Random processes; State estimation;
Journal_Title :
Aerospace and Electronic Systems, IEEE Transactions on
