Penalty function definition for parametric identification on a finite sample

Аuthors

Shamrikov B. M., Maksin D. G.

Moscow Aviation Institute (National Research University), 4, Volokolamskoe shosse, Moscow, А-80, GSP-3, 125993, Russia

Abstract

A formalized procedure to define a performance index for parametric identification on a low number of observations is obtained. The procedure uses algebraic identification theory based on the generalized least-squares method. The performance index is formed using ortogonalization. Parametric identification algorithm offered solves the problem on a single finite sample at restricted a-priory information in a real time mode.

⁠