Designing experiments to estimate process dynamics

The problem of dynamic system identification is an important one to control engineers and statisticians. Much work has been done, from different starting points and with different perspectives and philosophies. One difficulty with this is that different nomenclature evolves in each field. For example, the engineering term "system identification" implies three more refined statistical ideas, namely identification (of the order of the model), estimation (of the parameters of the identified model) and diagnostic checking (of the fitted model to the data). In this paper, we consider only the problem of designing an input to a dynamic system in order to best estimate the parameters in the model, according to some design criterion.