.png){kind=icon}
返回
A self-adaptive multi-objective evolutionary algorithm assisted by non-dominated entropy combined with multi-dominated screening
.png){kind=icon}
PRE
AI
DOI：10.1016/j.eswa.2026.131284.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Multi-objective problems exist in all aspects of life and there are already some evolutionary algorithms (EAs) to solve these problems. However, for complex problems, traditional EAs, due to their inherent randomness, find it difficult to conduct detailed local explorations each time. Moreover, during the exploration process, they tend to be poor convergence and weak diversity of the population of the algorithm. To address these challenges, this paper proposes a self-adaptive multi-objective evolutionary algorithm assisted by non-dominated entropy and multi-dominated screening (MDEA-NDE). The MDEA-NDE introduces two dynamic mechanisms: adaptive crossover based on crowded entropy and screening based on dominance. During the evolution of the MDEA-NDE, the mutation factor is adjusted through crowded-entropy to help break away from enhance the convergence performance of the algorithm and the diversity of the population. Meanwhile, certain dynamic screening based on dominance is carried out on the crossed parent generations to strengthen the local search ability of the algorithm. To verify the performance of MDEA-NDE, this paper selects three benchmark functions, UF, CF and DTLZ, which cover different characteristics, to test the MDEA-NDE, which is compared with other advanced algorithms. The results show that MDEA-NDE has stronger competitiveness compared with other algorithms. Furthermore, the Friedman test statistics show that MDEA-NDE has significant differences compared with other algorithms. In addition, MDEA-NDE is further evaluated on a real-world engineering case, demonstrating competitive performance in both convergence and diversity.
期刊
IF:
7.5
论文数:
3.0W
被引数:
10.2W
机构
暂无机构信息
引用论文
分享收藏Benchmarking Real-World Many-Objective Problems: A Problem Suite With Baseline Results.png)
IEEE ACCESS
OAAI
IEEE ACCESSIF3.6
OAAI分享收藏Kriging-assisted indicator-based evolutionary algorithm for expensive multi-objective optimization
PREAI
分享收藏分享收藏Multi-objective generalized normal distribution optimization: a novel algorithm for multi-objective problems多目标广义正态分布优化: 一种求解多目标问题的新算法OAAI
OAAI分享收藏A practical tutorial on the use of nonparametric statistical tests as a methodology for comparing evolutionary and swarm intelligence algorithms关于使用非参数统计检验作为比较进化和群体智能算法的方法的实用教程
PREAI
分享收藏
查看更多内容