Joint optimization of ship speed and trim based on machine learning method under consideration of load

Abstract

The maritime sector has diligently endeavored to mitigate fuel consumption to curtail emissions and expenditures within the sustainable development framework. To comprehensively analyze ships’ fuel consumption, considering the combined influence of multiple factors, we develop an integrated optimization approach encompassing speed, trim, and speed–trim adjustments under different loading conditions. Firstly, we employ the fuel consumption prediction models established before, conduct a detailed analysis of the ship’s route in distinct segments, and formulate optimization models by scrutinizing the ship’s actual voyage. Secondly, we conducted single-parameter optimization for speed and trim to achieve the minimum fuel consumption for the entire route. We also implemented a joint optimization approach for simultaneously optimizing speed and trim to enhance fuel efficiency further. Lastly, we applied a smoothing method to the model’s prediction results to solve the step problem; and compared the optimization results before and after smoothing to assess the approach’s effectiveness. The results show that the joint optimization in the ballast condition yielded fuel consumption savings of 12.30% and 11.70% before and after smoothing, respectively; The fuel savings achieved under full load conditions were 10.18% and 9.47%. Demonstrate that joint optimization can overcome the limitations of single parameter optimization and enhance the fuel-saving rate.