混合离散粒子群的任务调度算法及应用  

Task Scheduling Algorithm and Application Based on Hybrid Discrete Particle Swarm Optimization

作　　者：简平 邹鹏 熊伟 

机构地区：[1]装备学院,北京101416

出　　处：《火力与指挥控制》2014年第5期146-149,160,共5页Fire Control & Command Control

基　　金：湖南省研究生科研创新基金资助项目(CX2010B025)

摘　　要：针对粒子群算法存在易陷入局部极值、精度差的缺点,引入了模拟退火的思想,研究了一种混合粒子群优化算法,防止早熟现象的产生,使种群在更新迭代中保持了多样性,并将混合算法应用到典型的任务规划调度问题求解中,通过仿真验证了混合算法的有效性,与单一的离散粒子群算法相比,提高了调度问题的求解效果。In view of the defects of PSO such as easy to get into local extremum and low computational precision,a improved hybrid algorithm is designed by introducing the selection strategy based on SA algorithm to DPSO. The hybrid algorithm keeps the population diversity during the iterative process and overcomes the defect of premature convergence,which is applied in the typical task scheduling problem. Simulation results show the validity of the hybrid algorithm,and the algorithm improves the efficiency for settling the problem comparing the single DPSO.

关 键 词：离散粒子群优化算法 模拟退火算法 混合算法 任务调度 低轨预警系统任务规划 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]