Interval many-objective evolutionary algorithm guided by dynamic dual-sequence mechanism

doi:10.1016/j.swevo.2025.101870

Interval many-objective evolutionary algorithm guided by dynamic dual-sequence mechanism

2025-04-01

0

PRE

AI

DOI：10.1016/j.swevo.2025.101870

Original

Original request for help

Save

Abstract

En 中文

Interval many-objective evolutionary algorithms (IMaOEAs) have received significant achievements in recent years. However, it is difficult for the algorithm to quickly select good interval individuals since the uncertainty affects the definition of interval dominance relations. To further reduce the computational burden of uncertainty on the interval optimization process, this paper proposes a dual-sequence mechanism-guided interval many-objective evolutionary algorithm. First, the interval binary R2 evaluation indicator IR2 was designed, which can effectively evaluate the convergence and diversity of interval individuals. Second, an uncertainty dominance relation for interval individuals is proposed and uncertainty is quantified using the weighted LP norm (WLP). Finally, the dynamic dual-sequence (DDS) mechanism was ultimately employed to retain the most exceptional individuals within the population, while simultaneously eliminating those with subpar performance in terms of convergence, diversity, and uncertainty. To extensively evaluate the performance of the proposed approach, 16 benchmark problems were used as the test suite. The experimental results demonstrate that the approach outperforms five advanced interval many-objective evolutionary algorithms, showcasing its superior performance and competitiveness.

Keywords:

Interval many-objective optimization

R2 indicator

Interval uncertainty dominance relationship

Dynamic dual-sequence

Interval many-objective evolutionary algorithm guided by dynamic dual-sequence mechanism

Abstract

Journal

Organization