多产品克隆选择算法在间歇过程调度中的应用  

Multi-product clonal selection algorithm and its application to batch plants scheduling

作　　者：林可鸿[1,2] 贺益君[1] 陈德钊[1] 

机构地区：[1]浙江大学化学工程与生物工程系,浙江杭州310027 [2]台州学院医药化工学院,浙江临海317000

出　　处：《浙江大学学报（工学版）》2010年第2期338-343,共6页Journal of Zhejiang University：Engineering Science

基　　金：国家自然科学基金资助项目(20276063)

摘　　要：针对传统处理多产品多步骤多设备零等待的间歇过程调度方法(MMMSZ)只适用于求解小型问题的这一特点,构建了多产品克隆选择算法(MCSA).该算法从生产计划出发,生成较优的生产方案和各种全产品批次,并以全产品批次作为抗体,调用改进的克隆选择算法进行优化,其适用面宽,搜索空间大,能有效地处理约束,适用于规模不很大的问题.对于大规模问题,可基于周期调度策略,开发大规模多产品克隆选择算法(LMCSA),以克服维数灾难.将它们用于2个间歇调度示例,实验结果表明,2种算法能在合理的时间内给出较优解,综合性能良好.The traditional multi-product multi-stage and multi-machine batch plants scheduling under zero-wait policy （MMMSZ） was only applied to small-scale practical issues.A novel multi-product clonal selection algorithm （MCSA） was proposed according to the feature.The better production program and a lot of entire product batches were created from the production plan.The entire product batches were considered as the antibodies,and the improved clonal selection algorithm was called.MCSA can solve many kinds of problem,search more solution space,efficiently deal with the constraints,and is applicable to the small-scale problem.Large-scale multi-product clonal selection algorithm （LMCSA） was created by employing the periodic scheduling strategy in order to conquer the dimension disaster.MCSA and LMCSA were used to solve two examples of batch plants scheduling.Experimental results show that both algorithms can get the sub-optimal solution with appropriate time and their comprehensive performance is good.

关 键 词：多产品多步骤多设备 间歇过程 零等待 克隆选择算法 周期调度策略 

分 类 号：TQ021.8[化学工程]