Optimizing Sensor Arrangement for Strain Mode-based Hull Monitoring System Application for Navy Ships

Ships and naval vessels must have their structural strength assessed by structural analysis by classification society rules.However, these assessments are only calculated at the design stage, and it is necessary to monitor the structural health of the vessel in real-time during actual operation.This paper proposes a method for estimating the structural responses of naval vessels to assess their structural health.The proposed method employs the conversion matrix technique, which is constructed from the strain modes of structure.This method identifies suitable strain modes through hydrodynamic analysis, and a conversion matrix is created by conducting a direct global structural analysis.The prediction model was confirmed with numerical analysis results, a selection method for strain modes was established, and the model's performance was analyzed according to the number of modes.Additionally, a data-driven quantitative analysis was conducted to investigate the influence of sensor quantity and placement on the prediction model's accuracy.A genetic algorithm was utilized to optimize sensor placement to achieve the best performance with a limited number of sensors.This optimization significantly improved prediction accuracy, demonstrating an enhancement of approximately 20.9% compared to the initial design.The findings from this analysis offer guidance for installing strain sensors on ships.They are expected to assist in the early detection and repair of structural defects, ensuring continuous operational deployment and cost-efficient maintenance for naval vessels.