The aim of this contribution is to describe main features in the development of system identification, in the sense of modelling from time series data. Given the restrictions in space, such an effort is necessarily fragmentary. Clearly, subjective judgements cannot be avoided. System identification has been developed in a number of different scientific communities, the most important of which are econometrics, statistics and system- and control theory. The development of the field due to the requirements of applications and due to the intrinsic dynamics of its theories, and the interactions of the different communities in contributing to this development will be briefly described as well as the basic formal features of the problem. In addition some future perspectives are given. System identification has attracted almost no interest from the part of the general public interested in the history or perspectives of other parts of science. This is explained not only be the relative importance of the subject, compared to subjects attracting a lot of attention, but also by its - in a certain sense - abstract scope and the fact that it provides an enabling technology, often hidden in wider problem solutions. What is more surprising to the author is how little interest the history of the subject has attracted for researchers in this area; a clear indication for this is the frequent lack of proper referencing to original results.
展开▼
