Abstract
The problem of motion control for constraint-following for uncertain mechanical systems is considered. The uncertainty in the system is (possibly) fast time-varying but is assumed to be bounded. Based on a fuzzy description of the uncertainty bound, we propose an adaptive robust control. The control is deterministic and is not if-then rule-based. An optimal design problem associated with the control is formulated, which allows us to choose the optimal gain for the adaptive law. The performance of the resulting controlled system is both deterministically guaranteed and fuzzily-optimized.