DESIGN OF ROBUST SELF-TUNING CONTROL SCHEMES FOR STOCHASTIC SYSTEMS DESCRIBED BY INPUT-OUTPUT MATHEMATICAL MODELS

摘要

This paper deals with the robust self-tuning control schemes for stochastic systems, which can be described by the input-output Auto-Regressive Auto-Regressive Moving Average with exogenous (ARARMAX) mathematical model with unknown parameters in the presence of unmodelled dynamics. This explicit self-tuning control scheme is based on the proposed modified filtering recursive least squares algorithm with dead zone (m-F-RLS) in the step of estimation. We have applied the developed generalized minimum variance self-tuning regulator to a numerical simulation of a climate control in building, in order to test its performances. The obtained numerical simulation results are satisfactory.