Aiming to solve the problem of nonlinear uncertainties LuGre dynamic friction and non-symmetric dead zone in the robot,this paper presents a sliding mode robust robot control algorithm,compensates for the uncertainties of robot manipulator respectively LuGre dynamic friction with fuzzy radial base function (RBF)neural network and asymmetrical dead zone with fuzzy logic,trains parameters of nonlinear dynamic friction and asymmetrical dead zone real-timely and adaptively,to achieve an accurate reproduction of the actual robot system,and proves the Lyapunov stability of the algorithm. The simulation of a two-degree-of-freedom robot manipulator proves that the algorithm improves the trajectory tracking accuracy,the control torque stability and friction torque stability of the robot. Meanwhile,it is found that control moment pulse compensation error,rhombus attractor in the kinetic properties of the friction model of robot manipulator,and lacking of dead zone compensation cause nonlinear kinetic phenomena,such as limit cycle oscillation of control system. And the estimation of ε in the dead zone inverse model plays a decisive role in the accuracy of the system.