期刊
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
卷 110, 期 -, 页码 -出版社
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.engappai.2022.104727
关键词
Global optimization; Interplanetary transfer trajectory; Differential evolution; Global optimization; Interplanetary transfer trajectory; Differential evolution
类别
资金
- Fundamental Research Funds for the Central Universities, China [JAI210003]
- National Natural Science Foundation of China [62133015]
This paper proposes a powerful differential evolution algorithm called G_DE to tackle the global optimization problem in interplanetary transfer trajectory design. With a two-stage evolutionary process and four guidance strategies, G_DE is able to find the current known best solutions and demonstrates comparable performance to other differential evolution algorithms in experimental tests.
The extremely sensitive and highly non-linear search space of interplanetary transfer trajectory design brings about big challenges on global optimization. As a representative, the current known best solution (fuel consumption) of the global trajectory optimization problem designed by the European space agency is very hard to be found. To deal with this difficulty, a powerful differential evolution with the guided movement for population, named G_DE is proposed in this paper. G_DE employs a two-stage evolutionary process, which concentrates on learning global structure in the earlier process, and tends to self-adaptively probe the structures of numerous local spaces at a later stage. During these two stages, four guidance strategies related to the learning from the population distribution are proposed. The experimental test results show that G_DE can find the current known best solutions of Cassini1 and Sagas directly. For the newly proposed GTOP-X problem, G_DE has found state-of-the-art solutions for four encounter sequences, and using five gravity assists, G_DE found encounter sequences with fuel consumption as low as 1.5676 km/s. The experimental test on 10D CEC2017 problems proves G_DE has a comparable performance with recently proposed differential evolution variant.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据