Numerous algorithms have been introduced in solving multi-/many-objective optimization problems. However, some rigorous elite definitions may result in the side-effect that new solutions are hard to be admitted by the elite archive, which will restrain the searching ability. In this paper, a universal strengthened searching (SS) module is proposed as an accessorial searching procedure to improve the performance of existing algorithms. By concentrating part of the computational resources, this strategy can enhance the convergence and diversity searching strengths in separate areas. These areas are determined by classifying decision variables according to a dynamic spread-based procedure. Moreover, incorporating SS with existing algorithms can improve the overall performance of original algorithms while persisting their inherent advantages. In this paper, the structure of an SS-embedded algorithm is illustrated and the comparison experiments have been established among several couples of algorithms. The results demonstrate that the algorithms with the SS module have better overall performance compared to the original algorithms towards different problem properties. Meanwhile, the proposed strategy can accelerate the searching procedure for the time-consuming algorithms.

在解决多/多目标优化问题中引入了许多算法。但是，某些严格的精英定义可能会带来副作用，即精英档案馆难以接受新的解决方案，从而限制了搜索能力。为了提高现有算法的性能，本文提出了一种通用的增强搜索（SS）模块作为辅助搜索程序。通过集中部分计算资源，该策略可以增强单独区域中的收敛性和多样性搜索强度。通过根据基于动态扩展的过程对决策变量进行分类来确定这些区域。此外，将SS与现有算法合并可以提高原始算法的整体性能，同时保留其固有优势。本文阐述了一种SS嵌入算法的结构，并在几对算法之间建立了比较实验。结果表明，与原始算法相比，带有SS模块的算法在不同问题特性方面具有更好的整体性能。同时，提出的策略可以加快算法的搜索过程。