The research concentrates on the dynamic response issue of the deep-sea ship-type aquaculture platform in the complex marine environment and proposes a rapid prediction model based on machine learning, which is capable of precisely predicting the structural safety status of the ship-type cage under severe wave conditions. Specifically, through establishing numerical models of ship - type cages with diverse structures, numerical simulation is employed to comprehensively analyze the hydrodynamic characteristics of the aquaculture platform, validating the accuracy of the numerical simulation through experiments, and then using the hydrodynamic numerical results as training data, an artificial neural network (ANN) model for early-warning of disaster-induced damage to the ship-type cage is successfully constructed. By utilizing the established ANN model, the hydrodynamic results of the cage under various wave conditions are predicted, including key indicators like the maximum tension of the cable and the maximum stress of the floating frame. Furthermore, the research also employs the grey correlation analysis method to effectively identify the dominant disaster-causing factors that lead to the occurrence of damage. Validation indicates that the prediction results are highly consistent with the experimental results, which is of crucial guiding significance for farmers to take preventive measures prior to disasters.