The paper studies the problem of robust adaptive control for a class of nonlinear systems subject to unmodeled dynamics、nonlinear uncertainties and parameter uncertainties. The more extensive unmodeled dynamics introduced is input-to-state practical stable (ISpS), and unknown nonlinearities satisfy triangular bound conditions. In the framework of Lyapunov second method and applying tuning function adaptive backstepping theory, a smooth partial state feedback controller is explicitly constructed, which guarantees global uniform ultimate boundness for all states of closed-loop systems, and renders the output converge to a small neighborhood of the origin for all admissible uncertainties. Eventually a numerical example is given to illustrate the viability of the conclusion.