Abstract
The ability to predict the performance of novel subsystems and ship concepts is more important than ever now that the industry faces application of new systems of which little or no operational experience exist. Long-term simulation, with time horizons of months or even years, is a tool used in research and industry to predict the potentials and benchmark performance of novel concepts. Sea passage operation in the simulations, e.g., choice of speed and route, affect predictions of perfor-mance directly and can ultimately influence decisions for new-build and retrofit solutions. This paper suggests and tests a method for dynamically altering the speed of the simulated ship depending on the ship's performance, present weather, and weather forecast. Knowledge of how the ship will behave in various conditions are represented by a hydrodynamic ship simulation model, and an optimization algorithm is used as the inference engine.