A more efficient way to estimate step response coefficients using the FSF model

Abstract

A b s t r a c t This paper describes a more efficient approach to estimating an impulse or step response model of a linear system which involves first obtaining its frequency response using the frequency sampling filter (FSF) model. We show that the FSF model has an effective or true model order which is in general lower than that of its time domain counterparts, particularly in a fast sampling environment.

Control Oriented System Identification for Performance Management in Virtualized Software System

“…where N is the number of samples corresponding to the process settling time N = T s /δt and δt is a sampling interval. The novel findings by Wang [1994] and concluded that the number of parameters associated with FSF model is much smaller than an FIR model. Referring to the second assumption that the inverse of static nonlinear element is a single-valued smooth function, polynomial where B-Spline model can be used in this context.…”