Abstract:

As a new meta-heuristic swarm intelligence optimization algorithm, the flower pollination algorithm has the characteristics of simple structure, strong robustness, few control parameters, good global optimization performance and high execution efficiency. Its characteristics and advantages have attracted domestic and foreign many scholars have carried out research on it. It briefly explaines the design ideas of the flower pollination algorithm, and analyzes the existing improvements to the algorithm’s initial solution quality, population diversity and search capabilities, as well as the algorithm’s performance in various fields such as classification, positioning, path planning, and economic regulation application. It analyzes the direction and content of the flower pollination algorithm that can be further studied in the future.

Key words: flower pollination algorithm, parameter optimization, search method, hybrid algorithm

摘要：

作为一种新型元启发式群智能优化算法，花授粉算法具有结构简单、鲁棒性强、控制参数少，全局寻优性能好和执行效率高等特点，其具有的特性和优势，吸引了国内外众多学者对其展开研究。简单阐述了花授粉算法的设计思路，并分析了现有的对算法初始解质量、种群多样性以及搜索能力的改进以及算法在各个领域如分类、定位、路径规划、经济调控等方面的应用。分析了花授粉算法未来的可以进一步研究的方向与内容。

关键词: 花授粉算法, 参数优化, 搜索方式, 混合算法