Cost-function analysis of optimizing fuzzy-energy regions in the control of underactuated manipulators

Underactuated manipulators have to be controlled by following the restriction that the number of actuators is less than the number of generalized coordinates, because they have some passive joints. In this article, we propose a logic-based switching mechanism using a fuzzy-energy region. Boundary curves to separate an error energy region into several energy subregions are constructed by fuzzy logic. Therefore, fuzzy related parameters are optimized by a genetic algorithm (GA). This method is applied to an underactuated system with a drift term, such as a 2-DOF planar manipulator that has only one active joint. The effectiveness of the method is illustrated with some simulations.