Linear MIMO System Identification and Order Estimation Using Canonical Form

Abstract

In this paper, we present identification algorithms for parametric canonical-formed state space models and also for determining the order of models, based on maximum-likelihood and Akaike's Information Criteria (AIC). For obtaining maximum-likelihood estimates of model parameters, we apply the Expectation-Maximization (EM) algorithm which is an iterative method such that choice of the initial estimate is most important. The initial estimate is obtained by N4SID or MOESP methods and the similar transforms to canonical-formed state space models.