Optimal design of internal pressure resistant square cabin based on strength analysis

Objectives  In order to make the marine internal pressure resistant square cabin meet the requirements of strength and lightweight design, the neural network proxy model is combined with heuristic intelligent optimization algorithms and applied to the shape and size optimization of the internal pressure resistant square cabin components. Methods  The the corner chamfer radius, plate thickness, and the model number of beam are selected as design variables to carry out three-dimensional parametric modeling, and sample points are selected according to the optimal Latin hypercube experiment design method, then the response values of these sample points are calculated to build a radial basis functions (RBF) neural network proxy model. The surrogate model are combined respectively with three heuristic optimization algorithms: adaptive simulated annealing algorithm (ASA), multi-island genetic algorithm (MIGA), and particle swarm optimization algorithm (PSO) to perform global optimization. Results  The results show that three hybrid optimization methods can all reduce the structural weight on the basis of meeting the allowable strength requirements, and the optimal solution sought by the RBF-ASA method in the overall situation has a relatively good weight reduction effect. Conclusions  This study can provide a reference for the optimization design of the internal pressure resistant square cabin structure, and is of great significance for overcoming the key technical problems faced by the ship using nuclear power plants.