A Fatigue Life Prediction Model for Ship Core Storage Structures Under Navigation Sea Conditions by Integrating Finite Element Simulation and Data-Driven Methods

In order to study the fatigue life of the core storage structure of the drilling vessel under navigation sea conditions and ensure the long-term operation of the core storage, considering that it is difficult to conduct engineering fatigue failure experiments on its structure and there are multiple nonlinearly superimposed fatigue influencing factors, this paper proposes a fatigue life prediction model for the core storage structure by integrating finite element simulation and data-driven methods. This method takes the actual working conditions as the condition, uses finite element simulation experiments to obtain fatigue performance data, and further enriches the data set. Then, it analyzes the influence weight relationship of fatigue performance parameters and incorporates it into the training of the deep convolutional neural network (DCNN) model to finally obtain the load-life prediction model for the marine core storage structure. Finally, the prediction ability and accuracy of the model are verified using the pseudo-simulation data set and the simplified structural stress method.