LAUSR

Abstract A minmax problem statement of a problem of the parameter identification of the input/output mapping of systems is considered, based on the quadratic Renyi entropy (in the sense) of the identification error. As well, the quadratic Renyi entropy is used to construct a measure enabling one to select input variables (structure identification as a preliminary step) of the model subject to parametric identification in an adequate manner that is justified with the properties of the measure proposed. A case study is considered, with confirming the full practical suitability of the theoretical considerations of the paper.