创造性驱动优化算法

Creativity Driven Optimization Algorithm

模拟人类的创造性思维来求解问题一直是人工智能研究的热点和难点之一。基于当前的创造性思维研究理论,同时借鉴现有自然启发算法的建模过程,提出了一种新的智能优化算法——创造性驱动优化算法(Creativity Driven Optimization Algorithm,CDOA)。首先,构建出创造性驱动优化模型,并且为其5个子模型设计出具体的操作算子。而后,根据各子模型之间的联系,给出创造性驱动优化算法的执行步骤。为验证创造性驱动优化算法的有效性,使用8个CEC-2013实参数优化基准函数对CDOA进行了测试,并与当前最先进的3个同类算法进行对比。实验结果显示,CDOA在复杂函数上具有较好的寻优能力。最后,对CDOA进行的计算复杂度实验及分析表明,在相同实验条件下,与其他3个对比算法相比,CDOA具有更快的执行速度。

Solving problem by mimicking human＇s creative thinking process has been one of the hotspots in artificial in- telligence. According to the current researches about creative thinking, and also learning from the present nature-in- spired algorithms,a novel intelligent algorithm named creativity driven optimization algorithm（CDOA） was proposed. First, the creativity driven optimization model was constructed and special operators were designed for its five sub-mo- dels. Then, the execution steps of CDOA were given out. In order to test CDOA＇s effectiveness, eight CEC-2013 real- parameter benchmark functions were used. The optimization results of CDOA were compared with three state-of-the-art algorithms. The result shows that CDOA has an appealing optimization performance, especially on the complex composi- tion functions. At last, an experiment was carried out to analyze the computation complexity of CDOA. The results indi- cate that CDOA has lower computation complexity than the other three comparison algorithms.