An adaptive local range sampling method for reliability-based design optimization using support vector machine and Kriging model

2016 ◽  
Vol 55 (6) ◽  
pp. 2285-2304 ◽  
Author(s):  
Xin Liu ◽  
Yizhong Wu ◽  
Boxing Wang ◽  
Jianwan Ding ◽  
Haoxiang Jie
Keyword(s):  
Support Vector Machine ◽  
Design Optimization ◽  
Sampling Method ◽  
Kriging Model ◽  
Support Vector ◽  
Reliability Based Design Optimization ◽  
Reliability Based Design ◽  
Local Range

An important boundary sampling method for reliability-based design optimization using kriging model

2014 ◽  
Vol 52 (1) ◽  
pp. 55-70 ◽  
Author(s):  
Zhenzhong Chen ◽  
Siping Peng ◽  
Xiaoke Li ◽  
Haobo Qiu ◽  
Huadi Xiong ◽  
...  
Keyword(s):  
Design Optimization ◽  
Sampling Method ◽  
Kriging Model ◽  
Reliability Based Design Optimization ◽  
Reliability Based Design

An Adaptive Directional Boundary Sampling Method for Efficient Reliability-Based Design Optimization

2018 ◽  
Vol 140 (12) ◽  
Author(s):  
Zeng Meng ◽  
Dequan Zhang ◽  
Zhaotao Liu ◽  
Gang Li
Keyword(s):  
Design Optimization ◽  
Computational Efficiency ◽  
Sampling Method ◽  
Engineering Application ◽  
Structural Response ◽  
Kriging Model ◽  
Reliability Based Design Optimization ◽  
Computational Accuracy ◽  
Reliability Based Design ◽  
Sample Points

Due to the nested optimization loop structure and time-demanding computation of structural response, the computational accuracy and cost of reliability-based design optimization (RBDO) have become a challenging issue in engineering application. Kriging-model-based approach is an effective tool to improve the computational efficiency in the practical RBDO problems; however, a larger number of sample points are required for meeting high computational accuracy requirements in traditional methods. In this paper, an adaptive directional boundary sampling (ADBS) method is developed in order to greatly reduce the computational sample points with a reasonable accuracy, in which the sample points are added along the ideal descending direction of objective function. Furthermore, only sample points located near the constraint boundary are mainly selected in the vicinity of the optimum point according to the strategy of multi-objective optimization; thus, substantial number of sample points located in the failure region is neglected, resulting in the improved performance of computational efficiency. Four numerical examples and one engineering application are provided for demonstrating the efficiency and accuracy of the proposed sampling method.

Sign in / Sign up

Export Citation Format

Share Document