Indentification of dynamic systems: a systematic approach

A systematic procedure of parameter estimation for dynamic systems is presented as a basic approach for system modeling and control. With this approach, parameters of linear discrete- and/or continuous-time, multivariable dynamic systems can be estimated. The proposed technique has advantages over most of the existing techniques, such as flexibility, computational simplicity, and straightforwardness. On the basis of this concept as well as the derived algorithms, some new algorithms of system identification can be derived. Identification practice has demonstrated that the proposed approach gives good results.<>