A New Hierarchical Fuzzy Control Method for 2-DOF Underactuated Robots

A new hierarchical fuzzy control method for endpoint position control of planar 2-DOF underactuated robots with a passive revolute joint is presented in this paper. The control objective is to move the links from rest to a target-configuration with zero-velocity. This kind of underactuated mechanical system which is subject to the nonholonomic second-order constraints is controlled to gain the desired position by three fuzzy logic controllers. The genetic algorithm is introduced to achieve the global optimization. Fuzzy rules are automatically generated by the way of genetic algorithm from the system model according to the defined fitness function. Numerical simulation results show the validity of the proposed method.