Towards a bench-mark standard for stochastic adaptive control

Abstract

The paper considers the proposition of standard bench-mark plant models, performance indices and testing procedures for stochastic adaptive control. Only single input-single output linear plants are considered. A set of stochastically disturbed plant models is chosen, and the proposed measures of control algorithm performance (regulability and control cost) are evaluated and discussed for the bench-mark plant models set. A set of standard modes for evaluating the performance indices for the chosen plant models completes the proposed benchmark test