.png){kind=icon}
.png){kind=icon}
Return
Dynamic Multiobjective Squirrel Search Algorithm Based on Decomposition With Evolutionary Direction Prediction and Bidirectional Memory Populations
.png){kind=icon}
OA
AI
DOI：10.1109/ACCESS.2019.2932883.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
In order to improve the optimization effect of dynamic multiobjective problems (DMOPs), this paper proposes dynamic multiobjective squirrel search algorithm based on decomposition with evolutionary direction prediction and bidirectional memory populations (DMOISSA/D-P&M). To enhance the adaptability of the changing environments, DMOISSA/D-P&M assigns every individual a modification vector, a positive memory population, and a reverse memory population, all of them are updated in real-time with evolution. The modification vector is used to predict the evolutionary direction and the memory populations are used to retain the evolutionary information in historical environments. The predicted evolutionary direction and the memory individuals take part in the optimizing process in the new environment, which improves the convergence speed. To enhance the optimizing ability in every transient environment, DMOISSA/D-P&M designs two searching strategies for Squirrel Search Algorithm (SSA), the improved SSA satisfies different requirements of the multiobjective evolutionary algorithm based on decomposition (MOEA/D) at different evolutionary stages, which improves the convergence and the distribution of the obtained Pareto front in each transient environment. The experimental results on test functions of DMOPs show that DMOISSA/D-P&M has much better convergence, better distribution, and greater capabilities on coping with environmental changes compared with other dynamic multiobjective optimization algorithms.
Keywords:
Bidirectional memory populations
dynamic multiobjective optimization
evolutionary direction prediction
squirrel search algorithm
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
3.6
Papers:
9.8W
Citations:
29.4W