Multidisciplinary Synthesis Optimization Process in Multihull Ship Design

Abstract

This paper describes the development and application of a synthesis‐level multidisciplinary design and optimization (MDO) method for multihull ships. This method is unique in utilizing advanced multi‐objective optimization methods, neural network (NN)‐based response surface methods (RSM), and in its broad scope, integrating powering, stability, seakeeping, hullforms definition, structural weights, and cost and payload capacity into a single design tool. This MDO method is developed in the context of a multilevel hierarchy system approach where the results of the synthesis level optimization can be used for subsystem optimization and overall coordination of multilevel design system. NN‐based RSM for seakeeping and powering is developed and used in the optimization process. This RSM approach moves the computational cost of such performance evaluations out of the optimization cycle, substantially reduces the optimization cost, and allows for using results of physics‐based methods, such as advanced computational fluid dynamics, at the synthesis‐design level of design hierarchy. Details of these methods are delineated and multi‐objective optimization results are presented in the form of Pareto optimum solutions for multihull ship concepts such as trimaran sealift support ships and catamaran high‐speed connector ships.

An inverse method to determine the mechanical properties of the iris in vivo

“…In MIGA each population of individuals is divided into several islands on which all traditional genetic operations are performed separately including selection, crossover and mutation, individuals among the islands can migrate. This approach leads to the Pareto optimum solution in the most efficient manner [ 42 ]. The specifications for the MIGA algorithm were shown in Table 2 .…”

Hull form optimization research based on multi‐precision back‐propagation neural network approximation model

“…Introducing the approximate model into ship form optimization can reduce the number of high-intensity simulation calculations, greatly upgrade the efficiency of optimization. With the continuous development of machine learning methods, model forms have gradually increased, and scholars have also studied the application of some approximate models in ship design, such as Kriging, 5,11,18,32 response surface methodology (RSM), 13,23,24 radial basis function (RBF). 10,29 In recent years, the application of neural networks in approximation technology has become a hot topic.…”

Application of Multi-Objective Optimization Algorithm Based on Physical Programming in Ship Conceptual Design

“…The optimization quality is able to be improved. Finally, the improved MOOA-PP is applied to ship conceptual design optimization with three objectives and compared with the multi-objective evolutionary algorithm (MOEA) [5][6][7][8] to verify the effectiveness of the improved algorithm. Messac A.…”