关于柔性作业车间调度问题的仿真研究  

Simulation and Research of Flexible Job Shop Scheduling Problem

作　　者：周恺[1] 纪志成[1] 

机构地区：[1]江南大学物联网工程学院,江苏无锡214122

出　　处：《计算机仿真》2016年第3期282-287,375,共7页Computer Simulation

基　　金：国家高技术研究发展计划(863计划)(2013AA040405)

摘　　要：研究柔性作业车间调度优化问题。由于传统的一些方法在解决柔性作业车间调度问题时,面临着早熟、精度低等缺点,导致调度性能降低。针对上述难题,提出了一种改进的量子粒子群算法,结合反向学习策略和边界变异策略的优势,在增加种群多样性的同时避免了陷入边界最优。经过5个标准测试函数和一个柔性作业车间调度优化模型的仿真测试,结果表明改进的算法可增强全局寻优能力,提高收敛精度,避免搜索过程过早陷入局部最优,在解决调度问题中可获得较小的加工完工时间,具有优良的调度优化性能。Flexible job shop scheduling problem was studied to optimize in the paper. Solving flexible job shop scheduling problem with some traditional method is confronted with many disadvantages, such as premature conver- gence and low precision, these disadvantages reduce scheduling performance. To solve the problem, an improved quantum-behaved particle swarm optimization algorithm was proposed. The improved algorithm combines the advan- tage of opposition-based learning strategy and the advantage of bounded mutation strategy, and it can expand popula- tions and avoid algorithm into the optimum of boundary. Five benchmark functions and a scheduling optimization ex- ample have been used to test the proposed method. The results indicate that the improved algorithm can enhance glob- al optimization ability, convergence precision and avoid getting into local optimum. It also obtains a smallest makes- pan in solving f/exible job-shop scheduling problem.

关 键 词：柔性作业车间 量子粒子群算法 调度优化 边界变异 反向学习 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]