Abstract

Swarm algorithms belong to the class of population metaheuristic optimization methods. Despite the use of various metaphors, most swarm algorithms have similar structures, where one can distinguish common components such as the decision population initialization, decision diversification, and decision intensification. Based on the concept of generality, an analysis of key approaches to, methods for, and ways of increasing the efficiency of swarm optimization algorithms was carried out. In the survey, swarm optimization algorithms are viewed as a set of operators without a detailed discussion of each algorithm. The main focus is on the analysis of the key components of the algorithms. The main idea behind efficiency improvement is to maintain a balance between diversification and intensification. In this context, we consider mechanisms for supporting population diversity, methods for tuning and adjusting the swarm algorithm parameters, and approaches to hybridization of algorithms. We also indicate several open problems related to the topic of the survey.

中文翻译：

---

提高群优化算法效率的方法。一项调查

摘要

群算法属于群体元启发式优化方法类。尽管使用了各种比喻，但大多数群算法具有相似的结构，可以区分常见的组件，例如决策种群初始化、决策多样化和决策强化。基于通用性的概念，对提高群优化算法效率的关键途径、方法和途径进行了分析。在调查中，群优化算法被视为一组算子，而没有对每个算法进行详细讨论。主要重点是分析算法的关键组成部分。效率提升背后的主要思想是在多元化和集约化之间保持平衡。在这种情况下，我们考虑支持种群多样性的机制，调整和调整群算法参数的方法，以及算法混合的方法。我们还指出了与调查主题相关的几个未解决的问题。