This paper designs a novel robust sliding-mode control using nonlinear perturbation observers for wind energy conversion systems (WECS), in which a doubly-fed induction generator (DFIG) is employed to achieve an optimal power extraction with an improved fault ride-through (FRT) capability. The strong nonlinearities originated from the aerodynamics of the wind turbine, together with the generator parameter uncertainties and wind speed randomness, are aggregated into a perturbation that is estimated online by a sliding-mode state and perturbation observer (SMSPO). Then, the perturbation estimate is fully compensated by a robust sliding-mode controller so as to provide a considerable robustness against various modelling uncertainties and to achieve a consistent control performance under stochastic wind speed variations. Moreover, the proposed approach, has an integrated structure thus only the measurement of rotor speed and reactive power is required, while the classical auxiliary dq-axis current regulation loops can be completely eliminated. Four case studies are carried out which verify that a more optimal wind power extraction and an enhanced FRT capability can be realized in comparison with that of conventional vector control (VC), feedback linearization control (FLC), and sliding-mode control (SMC).