摘要
提出一种简便高效的二元离散粒子群算法.其每个粒子的各元素在新位置取二元值0或1的概率正比于其当前位置、历史最优位置和邻域内历史最优位置的取值,而负比于其前一个位置的取值.它不涉及在离散粒子群算法中难以解释的"速度"的概念.在算法中引入一个领袖粒子,有效地加快了算法的收敛速度,且没有增加函数的评估计算量.
A fast and easy binary discrete particle swarm optimization （PSO） algorithm is proposed. In this algorithm,the probability of a certain particle element assuming a value of 0 or 1 is in positive proportion to value 0 or 1 of this element in the current position of the particle,the historic best position it experienced,and the best point found by any member of its topological neighborhood,but in negative proportion to value of the former position of it. This algorithm doesn＇t involve the meaning of velocity which is usually hard to be defined in the discrete PSO. A queen informant is also introduced,which doesn＇t increase the number of function evaluations,but speeds up the convergence.
出处
《控制与决策》
EI
CSCD
北大核心
2010年第2期255-258,共4页
Control and Decision
基金
国家自然科学基金项目(60873058)
关键词
二元离散粒子群算法
基础构件
比例概率
领袖粒子
Binary discrete particle swarm optimization Essential components Proportion probability Queen informant
