A comparison of some subspace identification methods

Picci and Katayama (1996) and Katayama and Picci (1997) derived stochastic realization methods for a system with exogenous inputs and the relevance of stochastic realization to subspace identification of state-space systems has been shown in Picci and Katayama (1996). In this paper, we briefly review the basis of stochastic subspace identification algorithms and present some simulations results to compare the performance and computational loads of the realization based algorithms, the CLS algorithm, and the basic 4SID.