adaptive fuzzy estimator

Flexible-joint robots have complex dynamics due to the nonlinearity, uncertainty, and flexibility in joints. The conventional sliding mode control for flexible-joint robots is based on the torque control strategy which has a few drawbacks. Because of complexity of the robot dynamics, the control law becomes complex and determining the bounding functions is difficult. In addition, dynamics of actuators are not considered in the control law; thereby the control performance is degraded. Moreover, the chattering problem occurs which excites the unmodeled dynamics. In order to overcome these problems, a novel sliding mode fuzzy control design for robots is presented in this paper. This design employs an adaptive fuzzy estimator for function approximation and has no need to calculate the bounding functions. The chattering phenomenon is reduced, as well. The controller is based on the voltage control strategy which is simpler and more effective than the torque control strategy. Another novelty is that the controller has only one control loop whereas the conventional control approach uses two control loops. Stability of the control system is proven and efficiency of the proposed control approach over the modified fuzzy proportional-integral control is shown through simulations on the flexible-joint articulated robot manipulator.