Closed-loop identification for the best asymptotic performance of adaptive robust control (Q1923047)

The paper deals with adaptive robust control of discrete-time linear systems subject to additive deterministic disturbances including both the influence of external noise (measurement inaccuracy with unknown upper bound) and small parameter variations. Two adaptive control problems are discussed: (i) ensuring stability of the closed-loop adaptive system, and (ii) ensuring the worst-case performance being arbitrarily close to the behaviour under perfect system parameter knowledge. It is shown that the design of parameter estimation algorithms must be properly subordinated to the desired global control performance of the closed-loop adaptive system. The results concerning the best achievable asymptotic performance of adpative robust control are given. An illustrative simulation example is included.