The 14th Typhoon Yagi in 2018 influenced Shandong with severe precipitation and gale wind. In this paper, numerical studies of the influences of different ensemble covariance weight and different AMDAR data assimilation time windows on the typhoon forecast are conducted based on the Hybrid-3DVAR system of WRF (Weather Research and Forecasting) model. The results show that increasing ensemble covariance weight can improve the forecast of path of Typhoon Yagi. The path and precipitation forecasts are best when the ensemble covariance equals 100%. AMDAR data assimilation can improve the Yagi typhoon path and precipitation forecast, while 100% ensemble covariance Hybrid-3DVAR data assimilation can give further improvement. Different AMDAR data assimilation time windows can affect the assimilation data amount, thus influencing the fine feature forecast of precipitation. The 45 minutes assimilation time windows can give smallest errors of the analysis and forecast fields, and then can give the most satisfied precipitation forecast. Different AMDAR data assimilation time windows mainly affect the location of the forecasted rainfall and the improvement of the typhoon path forecast is limited.