.png){kind=icon}
.png){kind=icon}
Return
A time-sequence-guided multi-objective time-varying optimization algorithm
.png){kind=icon}
PRE
AI
DOI：10.1007/s10586-026-06506-x.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Existing multi-objective optimization algorithms improve performance primarily by spatial partitioning, multi-search strategies, and population structure partitioning. However, performance of the algorithms shows significant differences at various stages due to dynamic changes in the population over time. In response to this challenge, a multi-objective time-varying optimization algorithm is proposed, which integrates the temporal factor into population evolution. In the proposed algorithm, two strategies are designed, named time-varying search strategy and hierarchical pre-selection method. The time-varying search strategy explores the guiding role of the temporal factor in the search process to dynamically adjust the search range of individuals. The hierarchical pre-selection method adopts rank-specific offspring selection to dynamically balance population convergence and diversity. The proposed algorithm is compared with six advanced multi-objective evolutionary algorithms on 31 benchmark functions. Experimental results show that the proposed algorithm outperforms the competitors in terms of HV, IGD, and IGDp values. In addition, the proposed algorithm is applied to four widely used real-world engineering design problems to further verify its application ability.
Keywords:
Multi-objective optimization
Mutation strategy
Offspring selection
Swarm intelligence
Journal
C
IF:
4.1
Papers:
5.0K
Citations:
7.5K