System identification techniques

In this survey paper we present a number of results, examples and applications of parameter estimation techniques that are relevant to structural identification. The overall topic of system identification and parameter estimation is quite extensive and continually growing in its scope. Thus, we can only bring forth a portion of the fundamental ideas that have appeared in the literature. Included in this survey, are estimation procedures as well as discussions of the quality of the specific estimators. Asymptotic convergence and asymptotic statistical properties are basic to the understanding of the nature of the errors in the estimators from finite sets of data. We discuss estimators from time domain models, both in continuous and discrete time, as well as frequency domain estimators, i.e., eigenvalues and modal properties. Theorems are stated without proof, and examples of various of the techniques are included. Finally, a reasonably comprehensive bibliography is included.