4. Conclusion
In this study, a methodology is developed to integrate Bayesian approaches with the hydrodynamics analysis of marine floating structures to improve their safety. For this purpose, the frequency domain approach is applied for hydrodynamic analysis given that this method provides an efficient solution to compute, either numerically or experimentally, the stochastic wave loads on structures. BNayesian network is adopted for estimating the probability of failure to identify the best design point. A floating tensioned cylinder is considered as a case study to demonstrate the application of the methodology. The structure is subjected to 12 sea states and the reliability of the mooring system is examined with respect to the allowable horizontal elongation. It is found that the structure can tolerate the extreme wave height with optimum critical surge response of XC = 3.5 m, corresponding to reliability index of almost β = 3.50. This methodology can be applied to effectively perform reliability analysis of a floating structure with tensioned mooring system. In order to use the proposed methodology for another type of failure, firstly it is necessary to develop a suitable limit state function for a particular failure scenario. The same approach should then be followed for developing the BN and estimation probability of the failure. For this purpose, a suitable limit state function, G, for a particular failure scenario (such as capsizing a vessel due to extreme roll angle) should firstly be developed and then follow the same approach proposed in Section 2.3 for developing related BN and estimation probability of the failure. Results of this research confirm that the methodology is successful in identifying the critical design point of the system with respect to hydrodynamic response of the structure in different sea states which can assist in maintaining an acceptable level of failure risk during the operational time.