Ship Structure Optimization Based On PSO-BP Neural Network

With the accelerated process of economic globalization, increasing demand for crude oil, oil tankers are widespread concernas the main carrier of transportation of crude oil.According to the structure optimization of oil tanker,reducingthe use of hull steel, reducingconstruction costs,improving the competitiveness of product tanker has very important engineering significance. In the field of structural optimization of ships, using fast running model based on BP neural network as an alternative model to the structures involved in the optimization process., under the premise to meet the accuracy of the finite element analysis can effectively reduce the number of times.Aiming at the defects of BP neural network model, the particle swarm optimization algorithm makes the optimized improvement to the BP neural network model. And the using of BP neural network model based on particle swarm optimization for ship structure optimization verifies the feasibility of the model.The main research contents of this paper are as follows:(1) This paper introduces the optimization theory and method of ship structure, and analyzes various optimization methods, and concludes the features and limitations of various optimization methods.(2) According to the specification, the finite element model of the ship structure is established, and the training samples and samples are acquired by the method of sensitivity and orthogonal test.(3) The basic principle of BP neural network and PSO is studied, and the programming of the initial weight and threshold value of BP neural network is accomplished.The neural network trained by the BP neural network and the neural network based on the particle swarm optimization is trained by the training sample, and the generalization ability of the two models is detected by the detection sample.(4) The PSO-BP neural network model is used to optimize the oil cabin section, and the results are verified by the optimizationThe results show that PSO-BP neural network is a kind of effective ship structure optimization proxy model, which can effectively solve the problem of local extreme value of BP neural network model.