摘要
为了改善粒子群优化算法的收敛速度,在布朗运动和伊藤过程的启示下,提出了一种混合布朗运动和粒子群优化算法这两种思想的改进算法。通过对布朗运动和伊藤过程进行抽象,设计了漂移算子和波动算子。漂移算子保留了粒子的位置属性,但没有了速度属性,并引入了吸引子的概念,借鉴差分变异算子设计了波动算子。通过解决典型的复杂函数优化问题,实验结果表明,改进算法具有收敛速度快的特点,并具有良好的健壮性和稳定性。
In order to improve the convergence rate of particle swarm optimization （ PSO）, with the inspiration of the Brownian motion and ITO process, this paper proposed a kind of hybrid algorithm,i, e. BMPSO, which mixed Brownian motion and PSO algorithm. It designed the drift operator and fluctuation operator by abstracting Brownian motion and ITO process. The drift op- erator retained the position property of particle, but there was no speed attribute for particle of BMPSO, and also introduced the attractor concept. And then, the differential mutation operator was used to design the fluctuation operator. Experimental results on typical complex function optimization problems show that BMPSO retains the fast convergence characteristics, at the same time, possessing better robustness and stability.
出处
《计算机应用研究》
CSCD
北大核心
2011年第7期2439-2442,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(61070009)
景德镇陶瓷学院博士科研启动项目
关键词
粒子群优化算法
布朗运动
伊藤过程
热力学
particle swarm optimization（PSO）
Brownian motion
/TO process
thermodynamics
