摘要
针对NP-hard性质的作业车间调度问题,设计了一种改进的离散粒子群优化算法。引入遗传算法交叉算子和变异算子来实现粒子的更新,并将变异思想和模拟退火算法思想融入该算法中对全局最优粒子的邻域进行局部搜索,很好地防止了算法出现早熟收敛。通过将该算法和标准粒子群优化算法用于求解典型JSP,计算结果对比表明,改进的算法具有很强的全局寻优能力;就综合解的质量和计算效率而言,改进算法优于标准粒子群优化算法。同时,将该算法结果与文献中其他相关算法结果进行比较,验证了该改进算法的有效性。该算法能够有效地、高质量地解决作业车间调度问题。
According to the Job-Shop scheduling problem which contained NP-hard feature, this paper designed a kind of im- proved discrete particle swarm optimization algorithm. It introduced the crossover operator and the mutation operator of genetic algorithm to realize the particle updating. It embodied the thought of the variation and simulated annealing algorithm into this algorithm to achieve local search for the global optimal particle neighborhood, which prevented premature convergence of the algorithm well. To solve typical JSP through using the improved algorithm and standard particle swarm optimization algorithm, the calculation results showed that the improved algorithm had a strong capability of global optimization. As for the quality and computation efficiency of integrated solution, the improved algorithm was better than the standard particle swarm optimization algorithm. At the same time, comparing the result of the improved algorithm with other related algorithms in literature, it veri- fied the effectiveness of the improved algorithm. This algorithm can solve the JSP in high quality and effectively.
出处
《计算机应用研究》
CSCD
北大核心
2013年第8期2405-2409,共5页
Application Research of Computers
关键词
改进离散粒子群
作业车间调度
遗传算法
模拟退火
局部搜索
improved discrete particle swarm
Job-Shop scheduling
genetic algorithm（GA）
simulated annealing
local search
作者简介
吴正佳（1964．），男，湖南黄梅人，教授，博士，主要研究方向为先进制造企业信息系统分析与集成、智能算法理论、设备综合管理与监测；
罗月胜（1986．），男，湖北黄石人，硕士，主要研究方向为智能优化算法、调度优化（1uoyuesheng_98@sina．com）；
周玉琼（1987-），女，湖北孝感人，硕士研究生，主要研究方向为智能优化算法、结构优化；
黄绍雄（1987-），男，湖北黄梅人，硕士研究生，主要研究方向为智能优化算法、调度优化．
