摘要
针对调度目标为最小化最大完工时间的并行多机批调度问题,提出了改进的基于批序列编码的混合粒子群算法。在基本粒子群算法的基础上,引入了学习因子二阶振荡、随机权重、最大速度线性递减及自然选择等方法,改善了算法本身易陷入局部最优及早熟收敛等问题,并解决了因引入新的方法造成算法收敛速度慢及寻优能力差等问题。由仿真结果可知:改进的算法均优于常规的粒子群算法,且根据批序列编码的改进算法更优于常规基于工件序列编码的改进算法。
Aiming at minimizing the maximum completion time on the parallel machines,an improved hybrid particle swarm algorithm based on batch sequence encoding was proposed; and basing on the basic particle swarm optimization algorithm,the methods like the second-order oscillative learning factor,the random weight,the maximum speed linear degressive and the natural selection were introduced to solve problems such as falling into local optimum and premature convergence,slow convergence and poor ability of optimization caused by newly-introduced methods.Simulation results demonstrate that the improved algorithm outperforms the normal particle swarm algorithm,and the improved algorithm based on batch sequence works better than the improved algorithm based on workpiece sequence.
出处
《化工自动化及仪表》
CAS
2014年第4期397-401,454,共6页
Control and Instruments in Chemical Industry
基金
中国矿业大学青年科研基金资助项目(OC090196)
关键词
粒子群算法
并行机
批调度
最大完工时间
particle swarm optimization algorithm
parallel machines
batch scheduling
maximum completion time
