An Adaptive Sampling Algorithm and Approximate-Model-Based Optimization Method

• Published in: Journal of physics. Conference series 2018-07, Vol.1060 (1), p.12080
• Main Authors: Tian, J L, Tong, X Y
• Format: Article
• Language: English
• Subjects: Adaptive algorithms; Adaptive sampling; Algorithms; Approximation; Design optimization; Engine design; Optimization; Physics; Rocket engine design; Rocket engines; Search process; Thrust
• ISSN: 1742-6588; 1742-6596
• DOI: 10.1088/1742-6596/1060/1/012080

The approximate model based optimization algorithm can improve the speed of optimization convergence and save the time of calculation, which is very consistent with the development requirements of modern industrial rapid design. In this work, an adaptive and approximate model based optimization method is introduced, which is improved on the basis on HAM. This method uses three different types of approximate models at the same time to the full advantage of the three models' characteristics. Besides, an adaptive sampling algorithm is applied to achieve the self-setting mechanism for new sampling points in the search process. By testing on several standard test functions, the method performed better than GA, PSO and HAM in computational efficiency and accuracy. The method is also applied in the optimization design of rocket engine design to get the maximum thrust. Only after fifty-seven times of simulation calculation, optimal is obtained and the thrust is improved by 2.56%. the method is particularly suitable for design problems involving computation intensive analyses and simulations.