Flexible mechanical characteristics of power lines and poor conditions induce difficulties for the precise trajectory tracking control of a de-icing robot with an n-link robot manipulator. A robust fuzzy control based on adaptive sliding mode technique for precise trajectory tracking of a deicing robot is presented in this paper. The global controller is composed of three terms. The fuzzy inference system is utilized to approximate the unknown system structure and parameters, and the adaptive sliding mode control term is adopted to counteract the perturbations of the de-icing robot system. Meanwhile, a compensation term is designed to compensate the modeling errors, the approximation errors and others. The online adaptive laws of the controller parameters are derived via Lyapunov function method to guarantee the system stability. Finally, the experimental apparatus is provided to verify the correctness and effectiveness of the proposed scheme. Experimental results show that this scheme has more fast dynamic response and higher tracking precision than traditional feedback linearization control, and removes the control chattering phenomenon which is a main disadvantage of conventional sliding mode control in terms of robustness.