Hierarchical least-squares based iterative identification for multivariable systems with moving average noises

A hierarchical least-squares based iterative identification algorithm is derived for multivariable systems with moving average noises (i.e., multivariable CARMA-like models). The basic idea is to combine the hierarchical identification principle and iterative identification principle and to decompose a multivariable system into two subsystems, one containing a parameter vector and the other containing a parameter matrix. To solve the difficulty of the information matrix including unmeasurable noise terms, the unknown noise terms are replaced with their iterative residuals, which are computed through the preceding parameter estimates. The algorithm performs a hierarchical computational process at each iteration. The least-squares based iterative algorithm makes full use of all data at each iteration and thus can generate highly accurate parameter estimates. The simulation results indicate that the proposed algorithm works quite well.