The efficiency and precision of traditional particle swarm optimization were always influenced deeply by initial positions of particles and the initialization of random numbers, especially in the optimization of the best principal dimensions of ship. The numerical distributions of different chaotic sequences were analyzed in this paper, and the chaotic sequences were introduced into particle swarm optimization to improve the ergodicity of particles in feasible region. After analysing all the constraints, the method of controlling the principal components of variables by controlling the constraints of the problem was proposed, and it was used to improve the efficiency of chaotic particle swarm optimization. Besides, the neglect of some key conditions was eliminated with the ergodicity of chaotic map. The numerical experiments showed that, the chaotic particle swarm optimization which based on tent map and improved logistic map had shorter running time but higher efficiency than basic particle swarm optimization. After all the analysis, the improved chaotic particle swarm optimization was used to search the best principal dimensions of a real ship in this paper, and the calculating example showed that the new method was feasible for optimizing the principal dimensions of ship.
展开▼