摘要

We consider the control design for under-actuated manipulator systems. The task is to drive the system to be close to a prescribed constraint. The system contains uncertainty. It is bounded where the bounding information is prescribed by a fuzzy set (e.g., the bound is close to 1). The initial condition is also prescribed by a fuzzy set. A class of robust control is proposed, which guarantees a deterministic performance. On top of that, the choice of a control design parameter is cast into a fuzzy-theoretic setting. A performance index, consisting of accumulated fuzzy-based system performance and control cost, is proposed. The optimal control design parameters, which minimize the performance index, can be obtained by solving two algebraic quartic (fourth-order) equations. As a result, the control design problem, which addresses both fuzzy and optimal characteristics, is completely solved.