شناسايي سيستم ها در نرم بي نهايت

SYSTEM IDENTIFICATION IN (H)

In this paper linear and a nonlinear algorithm for system identification in H are presented. Error bounds are derived for both algorithms and it is shown that the linear algorithm does not have robust convergence. Conversely, the nonlinear algorithm is robustly convergent and the worst bound identification error (in Hsense) is derived. Linear methods are easier than nonlinear algorithms but for robust control design this method is not suitable. However, for lower order models this method can be utilized for identification purposes. On the other hand, due to robustness of nonlinear identification algorithm it can be used for control design. This method creates less identification errors compared with the linear algorithms.