混沌差分文化算法及其仿真应用研究

Differential Evolution-based Cultural Algorithm Combined with Chaotic Search and Its Application

针对差分进化算法(DE)全局寻优能力差,无法有效的求解工程中复杂的高维非线性优化问题等缺点,提出一种混沌差分文化算法(CDECA)。该算法模型将DE嵌入文化算法的框架作为主群体空间的进化过程,同时,引入具有较强局部搜索性能的混沌搜索来进行信念空间的进化,并通过设计一组联系操作实现文化算法模型中两个空间的互相影响互相促进,提高算法的寻优效率。几个典型测试函数的测试结果表明CDECA的搜索能力优于DE,将其应用于某大型水库的优化调度,也取得满意的效果。

In view of the critical shortcomings such as local convergence and lack of precision when applied to solve actual nonlinear engineering problem with complicated constraints of differential evolution algorithm （DE）, a differential evolution-based cultural algorithm combined with chaotic Search was proposed. In order to improve the search ability of DE, the algorithm proposed inserts DE to the framework of cultural algorithm model as the evolution method of the population space and chaotic search which behaved well in local searching was adopted to lead the evolvement of the belief space. A set of communication operation was designed to implement the interaction between the two spaces of the cultural algorithm. Several Benchmark functions were tested and the results show that the proposed algorithm has better global convergence ability than that of DE. The algorithm is applied to the optimal operation of a reservoir and satisfying results are obtained.