针对连续优化问题,提出了一种改进的CSO算法。该算法思想借鉴了物学领域的“瀑布效应”原理,通过Fg(food global)投射食物,吸引具有简单智慧和行为规则的蟑螂在解空内爬行,完成搜索。实验结果表明,改进的CSO算法寻优率高、收敛速度快,尤其是搜索到了LevyNo.5测试函数的“新解”。
Aimed at the continuous optimization, an improved cockroach swarm optimization (CSO) is presented. The algorithm accords with "waterfall effect" and have Fg (food global) throwing food in solution space. The cockroaches with low intelligence and simple action will crawl to these food, search for optimal solutions. Simulation experiments show that CSO has high optimization rate and rapid speed of convergence, especially find a new value of LevyNo. 5.