ABSTRACT
A brief overview of the various techniques of identification for lumped linear time invariant continuous-time dynamical systems is given. Within the two-stage (primary and secondary) framework of identification the focus is towards parameterization in linearizing the estimation of parameters the models chosen for identification is discussed. Markov parameter and time moment expansions are first introduced and extended into general linear combinations of known basis functions such as Laguerre and Kautz functions. Some aspects of error quantification are discussed. These models are generalized so as to include prior knowledge of system dynamics, and to ensure that even low order models give an adequate representation of the system under test.
