IDENTIFICATION IN AUTOMATIC CONTROL SYSTEMS

There are many concepts of system identification, and in this paper a comparison is made between different approaches based primarily on experimental identification with respect to the criteria of optimality selected for achieving identification, the mathematical models considered, the computing techniques used, and the kind and form of the input signal selected. Linear, non-linear, dynamic, and noisy systems are considered as well as adaptive models and stochastic approximations. It is concluded that although significant progress in system identification has been made, many problems remain, and among them is the determination of modeling accuracy especially for non-linear, noisy systems. © 1969.