Research on crow swarm intelligent search optimization algorithm based on surrogate model
- PDF / 433,803 Bytes
- 7 Pages / 595.22 x 842 pts (A4) Page_size
- 88 Downloads / 235 Views
DOI 10.1007/s12206-020-2215-8
Journal of Mechanical Science and Technology 34 (10) 2020 Original Article DOI 10.1007/s12206-020-2215-8 Keywords: · Crow search algorithm · Kriging · Optimization algorithm · Surrogate model
Research on crow swarm intelligent search optimization algorithm based on surrogate model Huanwei Xu, Liangwen Liu and Miao Zhang University of Electronic Science and Technology of China, Chengdu 611731, China
Correspondence to: Huanwei Xu [email protected]
Abstract
Accepted June 26th, 2020
A large amount of calculation exists in a complex engineering optimization problem. The swarm intelligence algorithm can improve calculation efficiency and accuracy of complex engineering optimization. In the existing research, the surrogate model and the swarm intelligence algorithm are only two independent tools to solve the optimization problem. In this paper, we propose the surrogate-assisted crow swarm intelligent search optimization algorithm (SACSA) by combining the characteristics of swarm intelligence algorithm and surrogate model. The proposed algorithm utilizes the initial samples to construct the surrogate model, and then the improved crow search algorithm (CSA) is applied to obtain optimal solution. Finally, the proposed algorithm is compared with EGO, MSSR, ARSM-ISES, AMGO and SEUMRE, MPS, HAM algorithms. The comparison results show that the proposed algorithm can find a global optimal solution with fewer samples and is beneficial to improving the efficiency and accuracy of calculation.
† This paper was presented at ICMR2019, Maison Glad Jeju, Jeju, Korea, November 27-29, 2019. Recommended by Guest Editor Insu Jeon
1. Introduction
© The Korean Society of Mechanical Engineers and Springer-Verlag GmbH Germany, part of Springer Nature 2020
Traditional optimization methods tend to fall into a local optimum when solving complex engineering optimization problems, while engineering designers want to obtain a global optimal solution of the optimization problem. Swarm intelligence algorithms use random search mechanisms to evaluate multiple individuals in each iteration process, which is beneficial to increasing the probability of obtaining a global optimal solution. In addition, swarm intelligence algorithms do not need to use gradient information of the optimization problem, so it can solve black box problems. However, complex engineering optimization is actually a computationally expensive optimization problem. It is inevitable to call a large number of fitness functions when performing population evaluation, which hinders the popularization and application of swarm intelligence algorithms in the field of practical engineering [1]. To reduce the computational burden, some scholars have introduced a surrogate model into the evaluation of swarm intelligence algorithms and developed several surrogate-assisted intelligent algorithms [2-5]. Commonly used surrogate-assisted intelligent algorithms include genetic algorithm and particle swarm algorithm and so on. Fonseca et al. [6] took locally weight
Data Loading...
