ABSTRACT
When a ship is approaching the end of its designed service life, shipowners often seek to extend the ship’s service life to maximize economic benefits. Decision-making for Service Life Extension (SLE) is complex due to degradation effects such as corrosion. It not only requires consideration of economic benefits but also the impact of corrosion to ensure the ship’s safety during the extended service period. This paper proposes a corrosion-sensitive SLE decision-making method for ships based on Deep Reinforcement Learning (DRL). The method predicts future hull degradation considering corrosion states and optimizes the economic benefits to determine the optimal SLE duration and maintenance plan. DRL is an advanced and effective decision-making tool that maximizes rewards during the interaction between the agent and its environment. A Very Large Crude Carrier (VLCC) case study demonstrates its effectiveness in determining the optimal SLE duration for ships.
