A personal view on the development of system identification

Abstract

This paper presents the author's personal view on the development of identification theory in the control community, starting from the year 1965. We show how two landmark papers, (Ho and Kalman, 1965) and (Åström and Bohlin, 1965), gave birth to two main streams of research that have dominated the development of system identitication over the last fourty years. The Ho-Kalman paper, which gave a first solution to state-space realization theory, led to stochastic realization, and much later to subspace identification. The Åström-Bohlin paper laid the foundations for Maximum Likelihood methods based on parametric input-output models, which later became known as the highly successful Prediction Error Identification framework. The present paper also shows how the thinking in the identification community moved from a search for the "true system" to the formulation of identification as an approximation problem. This led to the view of identification as a design problem, in which the ultimate use of the model plays a paramount role in the formulation of the experiment design and in the choice of the identification criterion.