A carrier-based unmanned aerial vehicle (UAV) landing control system based on nonlinear expanded state observer (NESO) and model predictive controller (MPC) is designed. To compensate for the impact of uncertainty disturbance on control accuracy, NESO is designed, which can effectively estimate the impact of external disturbance on the movement of carrier-based UAV, and exploratory gray wolf optimization is proposed for parameter tuning of the observer. To achieve decoupling between control inputs, MPC is used to establish a carrier-based UAV landing control system. Simulation comparative results demonstrate that the carrier-based UAV control system proposed in the paper performs better than the sliding mode controller scheme that employs a sliding mode observer for landing accuracy.