The launch vehicle is highly sensitive to the launch window and the self-alignment accuracy.The sea launch platform is affected by rough sea conditions such as wind,wave and surge,which causes low frequency,wide amplitude and time-varying sway.The traditional static base alignment method cannot adapt to this situation,and the limited supporting facilities are inconvenient for carrying out the light-sight auxiliary alignment and transfer alignment.Therefore,a coarse alignment method based on swaying identification and a fine alignment method based on online estimation of measurement noise array are proposed.The influence factors of coarse alignment are quantitatively analyzed,the swaying period and amplitude are identified in real time,and the coarse alignment time is adjusted online according to the sensor characteristics and precision requirements,so as to create good initial conditions for fine alignment.In the process of fine alignment,the measurement noise is estimated,and the Kalman filter is dynamically modified to adapt to the large time-varying swaying of the launching platform.The proposed method was verified by flight experiments on CERES-1 sea-launch vehicle in September 2023.Under rough sea conditions where the shaking angle is greater than 2.1° and the angular velocity is greater than 2.35(°)/s,the coarse alignment algorithm can achieve the accuracy of less than 2.0° within 300.0 s,and the fine alignment algorithm is better than 0.3°,which guarantees the success of the launch vehicle's first flight.