Research on Operational Data Driven Ship Fouling Assessment Method

Objectives Hull fouling can significantly affect the sailing efficiency of ships. In actual operations, the dry-docking time for ships is usually determined based on experience, which often makes it difficult to clean the fouling at the optimal time, leading to increased fuel consumption costs. Therefore, a real-time hull fouling assessment method is proposed, which can evaluate the performance loss caused by hull fouling in real time, providing a basis for cleaning decisions. Methods Based on the data collected from the ship during the period without fouling and meteorological forecast data, a multi-layer neural network model is established to achieve high-precision prediction of fuel consumption at sea. The condition of hull fouling is assessed by comparing the deviation between the model's predicted results and actual measurements. The distance threshold filtering method is used to filter the data in the assessment segment to avoid prediction errors caused by model drift, ensuring the accuracy of the assessment results. Results Three sets of non-training data were selected for validation, corresponding to the periods before fouling, during fouling, and after cleaning. For the data before fouling and after cleaning, the model's prediction deviation percentage was approximately 7%. For the data during the fouling period, the model's deviation percentage exceeded 13%. Conclusions The validation results indicate that this method can effectively assess the fouling condition of ships. The percentage deviation between the model's predicted values and the measured values can be considered as the incremental fuel consumption caused by fouling. This also facilitates further calculation of the benefits of cleaning.