Ship Multidisciplinary Robust Design Optimization Under Multi-dimensional Stochastic Uncertainties

At present,most of the ship optimization design models are based on the deterministic optimization theory.However,in practical engineering problems,the optimization scheme will be affected by many uncertain factors.For example,the design parameters can not reach or exceed the design value.During to the influence of wind,wave and many other factors,the actual speed cannot reach the designed speed.The existence of these uncertain variables often makes the optimal solution very sensitive to the input variables and produces larger deviations or even causes some constraints to fail.So taking the existence of uncertainties in practical problems into account is very important.Multidisciplinary Robust Design Optimization was applied for ship optimization in this paper.As the ship is a long design cycle,covering a wide range of complex engineering design issues,which involves the construction of the economy,rapidity,maneuverability and other aspects.However,the disciplines do not exist independently.If the traditional serial optimization design is adopted,the synergies between different disciplines will be ignored and the optimization will fall into the local optimal solution.In order to ensure the feasibility of the ship forms,the interval analysis method is applied to the uncertain robust optimization design.We take the conceptual design of bulk carriers as an example,it is divided into several sub-disciplinary systems to construct a multidisciplinary framework of uncertainties.Through the analysis of the influence of uncertain variables on the ship’s optimization plan,a multi-disciplinary and robust design optimization study is completed.It proves the importance of uncertain multidisciplinary design optimization methods into the overall ship design,which can effectively reduce the failure rate of the ship design optimization.The paper is mainly studied as follows:1.The uncertainties that may exist in the ship was taken into account,multidisciplinary Robust Design Optimization method was introduced in the ship optimization.The uncertainty theory is studied.Uncertainty classification and approximate model methods,uncertainty analysis methods and uncertainty quantification methods are introduced.In order to provide reasonable and effective analysis and calculation of the uncertain factors,the calculation provides a solid theoretical basis and technical support.2.In order to solve the uncertainty optimization problem of single variable and multiple variables coexistence,the robust design optimization of ship uncertainty based on interval analysis is studied.In fact,a ship may have many uncertain variables in the actual optimization design,it is very difficult to obtain the exact probability distribution or fuzzy membership function of these uncertainties,in order to ensure the feasibility of the ship forms,the interval analysis method is applied to the uncertain robust optimization design.We take the conceptual design of bulk carriers as an example,the study conduct the uncertain variables single existence and multiple co-existence respectively.The final optimization results prove that using interval analysis method opens up an effective new way to solve the problem of ship robust design optimization.3.The uncertainty collaborative optimization framework with interval analysis is constructed,and the problem of consistency condition is improved.As most of the current frameworks are only applicable to deterministic optimization and lack of analysis and screening processes for uncertain variables,an uncertainty interval optimization framework has been constructed for this research gap.With the help of dynamic relaxation factor and dynamic penalty function,an improved uncertain collaborative optimization framework is proposed to solve the problem of consistency condition.It is applied to the optimization calculation example of gear box reducer,which proves the effectiveness of the above optimization framework.4.Experimental design and approximate model methods are studied in the paper.By using approximate model technique,the number of samples can be reduced while ensuring the accuracy of approximation,thus reducing the computational complexity of the optimization process,shortening the calculation period and reducing the computational cost.The Latin hyperelevation method is applied to the experimental design.Three approximate models are constructed and the precision of the three approximate models are checked.5.Construction of 45000 t Bulk Carriers Based on Uncertain Interval Analysis Collaborative Optimization Framework.Divide the bulk carrier into multiple subsystems,fully consider the performance of the ship and the uncertain factors that may occur during the actual navigation,and apply the interval analysis method to the uncertain multidisciplinary design optimization of a 45000 t bulk carrier.The final results show that the robust performance of the ship is reduced,but the performance of the optimization objectives of various disciplines has been greatly improved,and a more effective ship type scheme has been obtained.