A fast approach to identification using deconvolutionIn this paper, we propose a fast approach to impulse response and noise-variance identification for a finite-order, linear, time-invariant, single-input/single-output system, whose input driving noise is white (stationary or nonstationary) and measurement noise is stationary, white and Gaussian. Our algorithm is an iterative block component method that includes two stages, deconvolution and prediction-error identification. Experiences with our method indicate that it works well and saves about an order of magnitude in computation. Analyses and examples are given in this paper to support this claim.
Document ID
19850051162
Acquisition Source
Legacy CDMS
Document Type
Conference Paper
Authors
Chi, C.-Y. (California Institute of Technology, Jet Propulsion Laboratory, Pasadena CA, United States)
Mendel, J. M. (Southern California, University Los Angeles, CA, United States)