Ship Structure Optimization Based On AMPSO-BP-GA

Due to the complexity of the hull structure,the structural optimization task of the ship presents the characteristics of a large number of design variables,a high-dimensional objective function,and multiple extreme values.With the improvement of computer software and hardware performance,intelligent algorithms with different characteristics have gradually been applied to the field of ship structure optimization.For example,particle swarm optimization(PSO)has strong global search capabilities and fast convergence.It is often combined with BP neural networks to form PSO-BP The neural network acts as a proxy model.Genetic algorithm(GA)has strong convergence ability and is often used in the optimization of the extreme value of the objective function,and can achieve better structural optimization results.However,the above single traditional structural optimization algorithm is prone to problems such as falling into local optimum or slow solution speed.Therefore,based on adaptive mutation particle swarm optimization(AMPSO),BP neural network,genetic algorithm(GA),combined with orthogonal experiments designed by Isight/Nastran,this thesis proposes an AMPSO-BP-GA structural optimization method.And this method was applied to the optimization design of ship structure and achieved good results.The main research contents are as follows:(1)By optimizing the extremum of multiple multi-dimensional nonlinear functions,it is verified that the AMPSO algorithm can balance the global search ability and the local convergence ability,and has the characteristics of fast search speed and not easy to fall into the local extremum.Then through the analysis and comparison of the prediction error of the calculation results of the finite element software,it is proved that AMPSO-BP is more suitable as a proxy model of the finite element software in ship structure optimization than the BP neural network,PSO-BP neural network,and GA-BP neural network..(2)Effectively integrate the AMPSO-BP neural network and GA algorithm,and propose a structural optimization method for AMPSO-BP-GA.Subsequently,the ten-bar truss and the structure optimization of the vehicle ferry boat springboard are taken as examples to verify the effectiveness of the structure optimization method.The results show that under the lightest weight as the objective function and the same constraints,the proposed AMPSO-BP-GA method has a better optimization effect than other structural optimization schemes referred to in this article,indicating that the AMPSO-BP-GA structural optimization method has Effectiveness.(3)The AMPSO-BP-GA structural optimization method was applied to the structural optimization of the tanker compartment.The results showed that the optimized compartment was reduced by 8.1%compared to the initial design weight.This method considered the oil tanker’s loading conditions during the structural optimization process.In order to meet the strength requirements as the premise,make full use of the strength margin of the cabin structure,and at the same time consider the corrosion increase and the discrete optimization of the bone materials to meet the actual engineering needs.It is proved that the method is feasible,effective and popular,which can provide reference for other ship structure design.