【摘要】 提出了离散三群粒子群优化算法(DTHSPSO),该算法将整个粒子群分为三群,第1群粒子朝全局历史最优方向飞行,第2群粒子朝着相反方向飞行,第3群粒子在全局历史最优位置周围随机飞行。粒子的速度保持连续性,对于粒子位置的处理采用两种方法:一是通过传递函数,根据速度的大小进行离散化;二是直接通过强硬限制函数(Hardlim函数)将位置离散化。通过对两种离散函数进行测试与比较,表明两种DTHSPSO都比基本离散粒子群优化算法(DPSO)具有更好的优化性能,而且直接采用Hardlim函数的DTHSPSO算法效果更加突出。更多还原

【Abstract】 Discrete three sub-swarms particle swarm optimization algorithm(DTHSPSO) is proposed.In the new algorithm,it is assumed that the particles are divided into three sub-swarms.One sub-swarm flies toward the global best particles.The second sub-swarm flies toward the opposite direction.The last sub-swarm randomly flies around the global best particle.In DTHSPSO,two ways are used to handle the position of particles.In the first way,the corresponding velocity as a probability measure is used via the transfer function,and the second way directly utilizes the hard limit function(Hardlim function).Two well-known test functions′ optimization problems are utilized to test both DTHSPSO and general (discrete) particle swarm optimization algorithm(DPSO).The simulation results show that both of DTHSPSO have greater efficiency that DPSO.Especially,the second way of DTHSPSO has more wonderful performance.更多还原