The disturbance factors in the landing stage of carrier-based aircraft are very complex and the control accuracy requirements are very high, so the landing accidents occur frequently. In the Automatic Carrier Landing System (ACLS), the adjustment of many parameters makes it a repetitive and tedious work. In this paper, a direct lift control parameter optimization method based on offline high-precision optimal fitting and online aerodynamic identification correction is proposed, and it is applied to the controller of the carrier-based aircraft to optimize the control parameters of the controller. The neural network is used to fit the control parameter optimization model offline, and the aerodynamic derivative is identified online to calculate the difference, which is brought into the fitting model to solve the optimization increment. The simulation results show that compared with the ordinary off-line control gain method, the off-line and on-line hybrid control parameter optimization method proposed in this paper has fast response speed, small dynamic error, and improves control performance and accuracy.