The mooring line failure can cause significant drift, continuous mooring line failures, riser collision and production shutdown, but mooring line repairs and replacements are time consuming, and there is a lack of associated short-term emergency measures. To solve the above problems, a short-term recovery strategy is proposed in this paper to suppress low-frequency motions. The strategy contains a control process consisting of five controllers and a thrust distribution optimization algorithm and a prediction process consisting of the LSTM and Dense neural network. The simulation results show that all five controllers can suppress the low-frequency motion after mooring line failure, and the control efficiency of the controller can be improved and the control force and power consumption can be reduced with the help of the predictive correction algorithm. The combination of the predictor-corrector algorithm and the model-free adaptive controller provides the best comprehensive performance. The direction of the control force and the average propulsion angle of the thrusters are independent of the wave direction and are related to the direction of the failed mooring line.