Merging of User's Knowledge into Identification Part of Self-Tuners

Abstract

The problem of incorporating user's knowledge - possibly uncertain and/or contradictory - is inspected. Bayesian methodology together with a technique of generating fictitious data are used for computing appropriate initial conditions of recursive least squares for estimating parameters of Gaussian ARX model. Resulting algorithms respect different uncertainty of particular pieces of available information. From engineer's view point, the paper presents algorithms which translate "technological'' knowledge of the user into probabilistic language which is usually foreign to him.