A new second order sliding modes type-2 fuzzy controller for nonlinear uncertain perturbed systems is developed in this paper. To overcome the constraint on the knowledge of the system model, local models related to some operating points were used to synthesize a type-2 nominal fuzzy global model. For sliding modes, two adaptive fuzzy type-2 systems have been introduced to generate the two super twisting signals to avoid both the chattering and the constraint on the knowledge of upper bounds disturbances and uncertainties. These adaptive fuzzy type-2 systems are adjusted on-line by two adaptation laws deduced from the stability analysis in Lyapunov sense. Many simulation results for a robot arm actuated by a DC motor are given to illustrate the good tracking performances and the smooth signal control obtained by the proposed approach.