On-Line Identification Algorithm with Evaluation of Model Quality

Abstract

Abstract The idea of evaluation of the statistic model quality, based on direct least-square estimator, is discussed. There is presented a special, linear transformation of the model inputs, which provides a statistic orthogonalization of the coefficients of the eqivalent process model. The model errors, transformed to the form of the statistic independent stochastic process with known variances, are used for determination a convex set of process models for any desired probability. The diameter of this set is applied as a measure of the model quality. The corresponding relations are simple and can be preformed in the iterative way. The on-line algorithm of the model identification with the sequentienal evaluation of the model quality is presented. The results of simulation of two examples of the single-output systems are described.