%0 Journal Article %A Volosov, Victor V. %D 1998 %I Begell House %N 4-5 %P 21-35 %R 10.1615/JAutomatInfScien.v30.i4-5.30 %T To the Construction of Parametric Families of Ellipsoidal Estimates and their Optimization in Problems of the Nonstochastic Identification of Parameters and State of Many-Dimensional Discrete Control Objects %U https://www.dl.begellhouse.com/journals/2b6239406278e43e,573e6fed09bf9833,5c9bcb1b4dec2e68.html %V 30 %X The methods for obtaining the parametric families of ellipsoids containing the intersection of the initial ellipsoid with a cylinder or with an arbitrary number of layers, each of which is bounded by two parallel hyperplanes are proposed. The algorithms for estimation parameters and the initial state of multivariable control objects are developed under the assumption that the vector of errors of observation satisfies the given quadratic or geometric constraints. The algorithms are not sensitive (robust) to the errors in the a priori estimates and computational errors. %8 1998-10-01