Stochastic modeling of ship dynamics in irregular waves by explainable artificial intelligence and free-running model tests
Abstract
To achieve high level of operational safety, ship maneuverability in irregular waves should be addressed. Seakeeping and maneuveTo achieve high level of operational safety, ship maneuverability in irregular waves should be addressed.Seakeeping and maneuverability of a ship have been separated and analyzed by dynamics modeling in low frequency (maneuvering) and high frequency (wave-inducedlity of a ship have been separated and analyzed by dynamics modeling in lowfrequency (maneuvering) and high frequency (wave-induced motion) regions. The present research project aims to suggest a new approach of ship dynamics analysis, based on the data-driven mo motion) regions. The present research project aimsto suggest a new approach of ship dynamics analysis, based on the data-driven modeling of integrating maneuvering and seakeeping and stochastic analysis.The final product of the proposed research project is a dadeling of integratingmaneuvering and seakeeping and stochastic analysis.The final product of the proposed research project is a data-driven modeling of ship dynamics in irregular waves. The new approach can save resources in analysis of ship dynamics in irregulata-driven modeling of ship dynamics in irregularwaves. The new approach can save resources in analysis of ship dynamics in irregular waves. The stochastic simulation based on probability propagation is a proper method in making decision of ship maneuvering real sr waves. The stochasticsimulation based on probability propagation is a proper method in making decision of ship maneuvering realseas, where prompt decision-making and response are required in both manual and autonomous operation.The technical approach concernseas, where prompt decision-making and response are required in both manual and autonomous operation.The technical approach concerns collection of input data for data-driven modeling, establishing data-driven modeling, and verification and validation of the modeli collection of input data for data-driven modeling, establishing data-drivenmodeling, and verification and validation of the modeling. From collected data of free-running model tests andseakeeping tests, deterministic modeling of ship dynamics is derived. The mong. From collected data of free-running model tests andseakeeping tests, deterministic modeling of ship dynamics is derived. The modeling outputs hydrodynamic force and moments acting onto the hull, from ship motion and control parameters.The stochastic modelingdeling outputs hydrodynamicforce and moments acting onto the hull, from ship motion and control parameters.The stochastic modeling aims to derive probability density function of hydrodynamic force and moment onto the hull in certain sea state condition. The mode aims to derive probability density function of hydrodynamic force and momentonto the hull in certain sea state condition. The modeling is founded from the hydrodynamic force and moment modeling in regular wave and the probability density function of the wave perling is founded from the hydrodynamic force and momentmodeling in regular wave and the probability density function of the wave period and height in irregular waves.The verification of data-driven modeling concerns to find the optimal algorithm, in terms of robuiod and height in irregular waves.The verification of data-driven modeling concerns to find the optimal algorithm, in terms of robustness and fast computation. Various artificial neural network algorithms and setting parameters are tested for the same datasets.stness andfast computation. Various artificial neural network algorithms and setting parameters are tested for the samedatasets.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 31, 2020
- Source ID
- N629092012069
Entities
People
- Jeonghwa Seo
Organizations
- Chungnam National University
- Office of Naval Research
- United States Navy