An improved collaborative optimization algorithm of ship structures' static and dynamic subject based on the mixed and dynamic penalty function

Based on the standard collaborative optimization and aimed at the limitation of existing improved collaborative optimization like the relaxation factor method and the penalty function method, the collaborative optimization algorithm was improved by introducing the relaxation factor to construct the mixed dynamic penalty function. A hybrid algorithm based on the Non-dominated Sorting Genetic Algorithm and Adaptive Simulated Annealing was proposed for system-level optimization through the Isight software. The improved collaborative optimization algorithm was applied to the multi-objective optimization design of ship structures to optimize the static and dynamic characteristics of the marine engine room. The optimal solution was obtained and compared with the results of existing collaborative optimization algorithm based on the dynamic penalty function. It was shown that the improved algorithm has fewer iterations, better target value, and smaller interdisciplinary inconsistency information, which has certain value for multi-objective and multi-disciplinary structural optimization in actual ship engineering application. © 2019, Editorial Office of Journal of Vibration and Shock. All right reserved.