Ship Hull Steel Plate Deformation Modeling Based on Gaussian Process Regression

The linear heating and formation of steel plates is one of the most critical technologies in shipbuilding. Excellent technology not only provides good hydrodynamics for the hull but also affects the whole hull construction cycle and cost. In the heating and formation of a steel plate, the material, size, and thickness of the steel plate; heating temperature; heating position; and many other factors affect the formation of a steel plate. It is a very difficult process to know the influence relationship between various factors. In this study, a steel plate model is established by the Gaussian regression method, which can predict the steel plate deformation according to the selected steel plate material, size and thickness, heating temperature, and heating position. The accuracy of the model was evaluated, and the Gaussian process regression model has a better accuracy compared to other machine learning algorithm models. Finally the model visualization; designing the UI; selecting the steel plate material, size, and thickness; and inputting the heating temperature, the deformation magnitude, and stress magnitude of the steel plate can be obtained. The model can provide guidance to field workers for the heating and formation of hull steel plates and achieve efficient and fast formation of target steel plates.