Author(s): LUO Bo-wen, YANG Hua-ke, ZHOU Zhi-jin, WANG Zhao-hui, Marine Mineral Resources Exploration and Mining Equipment and Technology of Hunan Province Engineering Laboratory, Hunan University of Science and Technology, Guizhou Institute of Technology College of Mechanical Engineering
Pages: 51-56
Year: 2016 Issue:  5
Journal: China Offshore Platform
Keyword:  deep sea; high frequency transformer; multi-objective optimization; NSGA-Ⅱ algorithm;
Abstract: Aimed at the deep-sea high-frequency transformer,the multi-objective optimization mathematical model of the volume and the depletion can be established.It deals with the optimization design of contradiction between loss and volume target of the deep-sea highfrequency transformer,taking the advantage of genetic algorithm into solving nonlinear multiobjective optimization problems.By changing and improving the non-dominated sorting and choosing truncation strategies in the NSGA- Ⅱ algorithm,individuals with low density in one non-dominated level can possess larger genetic probability,which improves the global search ability of the algorithm and increases its speed.The improved NSGA- Ⅱ algorithm was used to make optimization designs to 6 kW high-frequency transformer,and the high-frequency transformer entitycould be manufactured according to the results of optimization.The test proves that the improved NSGA- Ⅱ algorithm can effectively reduce the total loss and the volume of deep-sea high-frequency transformer.