Abstract: Cultural algorithms are derived from the cultural evolution.Based on the population space in which individual evolves，the belief space is constructed，in which effective implicit information during the evolution process are abstracted and utilized.The dual evolution structure provides a universal model for the use of knowledge in evolutionary algorithms，which can improve the performance of algorithms.The algorithms have been used to solve many complex optimization problems successfully.In the paper，essential theory of cultural algorithms is reviewed.Developments of the algorithms about the types of knowledge are discussed.Different applications of the algorithms are summarized，and the further possible research directions are suggested.

Key words: cultural algorithms, dual evolution structure, knowledge, evolutionary algorithms

摘要： 文化算法模拟文化进化过程，在实现个体进化的种群空间基础上，构建信度空间，用于对进化过程中有效隐含信息的挖掘和利用。该双层进化机制为进化计算中的知识引导提供了通用框架，已证明能有效提高算法性能，并被成功用于解决诸多实际复杂优化问题。文章介绍了文化算法基本原理，从提取知识类型角度详细阐述了算法研究进展，总结了其在不同领域的应用，并展望了算法未来可能的研究方向。

关键词: 文化算法, 双层进化结构, 知识, 进化算法