摘要
为了克服人工鱼群算法容易陷入局部最优以及求解精度较低的弱点,提出了一种基于logistic模型的自适应人工鱼群算法。算法运行过程中,步长和视野两个参数自适应调整,在算法的早期保持种群多样性,并具有较高的收敛速度;在算法后期加强局部搜索能力,算法具有较高的求解精度。通过在5个经典Benchmark函数上的实验及相关的聚类实验,表明该算法具有较好的收敛速度和求解精度。
To overcome the weakness of the artificial fish swarm algorithm （AFSA）, AFSA is more likely to plunge into local op timum and has lower precision, an improved AFSA based Logistic mode[ is proposed. The algorithm can automatically adiust the visual and step during the running time. Therefore, it can keep the individuals diversity and has faster convergence speed at the initial generations. However, the algorithm has stronger searching ability of local optimum and has higher precision of result at a later time. Some experiments on five classical Benchmark functions and some pertinent clusters show that this improved AFSA al- gorithm can be convergent to global optimization with faster convergence speed and higher precision.
出处
《中国科技论文》
CAS
北大核心
2013年第7期668-671,共4页
China Sciencepaper
基金
高等学校博士学科点专项科研基金资助项目(20110023110002)
