Optimum Design Of Oil Tanker Structures Based On Particle Swarm Optimization

Reducing the steel consumption of hull structure can increase ship's dead weight capacity and bring more economic interests to shipyards and shipowner.So how to accomplish it always be an important part in ship design.Without changing ship's hull structure style,structure optimum become a kind of dimensions combinatorial optimization problem.Particle swarm optimization has been proved to be an effective way of solving this problem in recent yesrs.An improved particle swarm optimization,which reference the crossover and mutation operation of genetic algorithm,is adapted to optimize an oil tanker structure in this thesis.Four benchmark functions and Traveling Salesman Problem are used to verify the validity of this improved strategy.The improved particle swarm optimization performances better than the standard.The experiential formulas selected from rules can not get the characteristics and response of hull structures accurately.While FEM can solve the problem mentioned above effectively.And using FEM to analyze hull structure has been recognized by classification society worldwide.In this thesis,we select the stress calculated by direct strength analysis of this ship to evaluate every dimensions combination.For getting structure stress and its deformation truly,this thesis use AQWA to simulate P-M wave spectrum to obtain the maximum bending moment in this ship's lifetime.In addition,all operations in ANSYS are finished by ANSYS Parametric Design Language.This measure can achieve automatic calculation between ANSYS and MATLAB.In the optimization process,we choose the weight of a cabin in midship as fitness function and select the stress calculated by direct strength analysis should be less than permissible stress as constraint condition.The results shows that last optimum design reduces by 6.3 percent of steel consumption.Therefore,particle swarm optimization is an effective algorithm to optimum design hull structures.