Abstract
Abstract
To solve the problem of complex and time-consuming simulation calculations in ship type optimization, the application of approximate model technology based on spatial segmentation sequence sampling is becoming increasingly widespread. Based on the Kriging approximation model and the cross-leaving one error method, a new sampling strategy is proposed during the iterative sampling process. The importance of adding sample subspaces and surrounding subspaces is considered in the design space, and parallel thinking is added to develop an efficient parallel sequence sampling method to improve sample aggregation. The proposed method was applied to optimize ship types, and its effectiveness was verified.
Reference8 articles.
1. Cross validation based single response adaptive design of experiments for Kriging metamodeling of deterministic computer simulations [J];Aute;Structural and Multidisciplinary Optimization,2013
2. An accumulative error based adaptive design of experiments for offline metamodeling [J];Li;Structural and Multidisciplinary Optimization,2010
3. SURROGATES Toolbox User’s Guide Version 3.0 [EB/OL];Viana,2021
4. Virtual library of simulation experiments: test functions and datasets [EB/OL];Surjanovic,2021
5. Efficient space-filling and non-collapsing sequential design strategies for simulation-based modeling;Crombecq;European Journal of Operational Research,2011