The identification of nonlinear biological systems: LNL cascade models (Q1087484)

The paper deals with identification of systems represented by a cascade of a dynamic linear, a static nonlinear and a dynamic linear subsystem (LNL systems). Various identification schemes that have been proposed for these systems are critically reviewed. Analysis is focused on special problems arising in identification of nonlinear biological systems. An example is presented in which identification of each of the elements of a simulated LNL system from a limited duration input-output record is performed by an iterative identification scheme. Exact description of the LNL identification algorithm is provided in the conclusion.