Digital Commons Network^™

Simultaneous System Identification And Decision-Directed Detection And Estimation Of Jump Inputs To Linear Systems, Wynn C. Stirling

Faculty Publications

A decision-directed approach is presented for analyzing linear systems with unknown jump inputs. The system model parameters are estimated using a Kalman filter, and an empirical Bayes detection procedure is introduced to set the detector parameters, resulting in a decision-directed generalized likelihood ratio test coupled with recursive system parameter estimation. Monte Carlo results are presented to validate the performance of the algorithm.