Development of Explicit Model Reference Adaptive Control Using Bicriterial Optimization

Abstract

In this paper, the dual version of the explicit model reference adaptive control (MRAC) is designed using a bicriterial optimization procedure. Two criteria, which correspond to the two goals of dual control (to control the plant and to apply an active learning input signal for speeding up the estimation process at the same time) are suggested for the formulation of the synthesis problem. A special performance index with weighted control signals is introduced. The elaborated method is applicable for systems with nonminimumphase behaviour, time delay and disturbances. A computationally simple new dual control algorithm is derived. The adaptation scheme for the new system is described. The designed new dual model reference adaptive controller (DMRAC) is compared with the non-dual versions of MRAC using computer simulations. The considered examples demonstrate the improved control performance of the DMRAC.