Optimizing the ship constructions by automatic line heating forming process based in numerical simulation and artificial intelligence

Abstract

This paper presents the development of a novel automatic lineheating forming

machine based on intensive application of the numerical simulation and artificial intelligence. The forming of certain parts of the shell of the ships can be done by heating forming or mechanical forming. Line heating forming is usually a more flexible process, and therefore, it is more useful. The principal problem with line heating forming is that it is very time consuming, manual and needs very qualified workers. In this research, line heating forming is studied looking for the best way to automate the process.

Numerical models were developed based on finite element methods to simulate the process of heating, cooling and finally forming of the plates. The models were tested and validated against experimental test over small plates. However, as numerical models have extremely long computational times demanding high computational capacities; they are not directly applicable in a real time scenario, as it is needed.

Finally, after the mathematization of this knowledge and making use of it, it was developed a software based on artificial intelligence. Using an informed heuristic search strategy, this software predicts the optimal set of heating lines, their sequence and velocity to be applied over the plate to obtain the final shape. The developed system was tested to show its feasibility. As a consequence, there is a significative reduction in computational time allowing the system to be applied in a soft real-time environment. This model will help shipyard manufacturers determine the positions and trajectory of torch for the flame heating lines and their heating parameters to form a desired shape of plate.