A comprehensive evaluation of three robust adaptive control methodologies

Using Monte-Carlo simulations, the performance of three recently developed multiple-controller robust adaptive control methods is compared for different time-varying uncertain parameter waveforms, using a mass-spring-dashpot benchmark example. We further examine the performance of these different adaptive methods with that of the best possible robust nonadaptive design using the same physical example. Whenever possible, we also compare the adaptive methods' performance with that of the (unrealizable) “Perfect Model Identification (PM.ID)”. In order to have a “fair” comparison of the performance obtained with the different control laws we used the same bank of local controllers and tuned each approach for its best performance. The Monte-Carlo simulations highlight the strength and aptitude of each method as well as its shortcomings and drawbacks.