Abstract: Unconstrained optimization is an ancient mathematics issue, due to the development of the intelligent computing science, there are many of techniques that can cope with such problems as well, besides classic mathematics’ method. This paper takes advantage of the cultural algorithm’s double-layers architecture and embeds the differential evolution in updating operation of the knowledge space to achieve situationable knowledge updating by differential evolution during evolution of the whole architecture, then population space makes use of all these information to assure correct evolution direction and to fulfill individual evolution efficiently in order to improve the algorithm’s performance. This paper selects 6 benchmark functions to test the classic algorithm and the improved algorithm. The results demonstrate that the later improves the performance.


Key words: unconstrained optimization, cultural algorithm, differential evolution