基于启发式遗传基因算法的半导体制造系统库存控制研究

Inventory Control for Semiconductor Manufacturing System Based on Heuristic Genetic Algorithm

针对外部订单变动对制造系统库存数量的影响,提出了一种面向半导体制造系统的双闭环库存优化模型.该模型将外部订单的实时信息及时反馈至半导体生产线,将外部订单需求与生产线在制品库存进行优化:首先,对半导体制造系统进行了分析,提出了基于恒定在制品和实时库存的订单反馈机制,建立了基于原材料闭环和半成品闭环的双闭环库存模型;其次,针对模型所涉及的多目标参数优化问题,提出启发式遗传基因算法并进行多目标寻优.为了验证该方法的有效性,将所提出的双闭环模型和启发式遗传基因算法应用在半导体封装测试工厂进行实例验证.结果表明,所提出的双闭环库存优化机制具有较好的系统性能参数.

Aiming at the influence of external order variability on the inventory of manufacturing system, this paper introduces a double close loop model to optimize inventory for semiconductor manufacturing sys- tem. This model timely sends the real-time information of external orders to semiconductor production line. Meanwhile, the model realizes the collaborative optimization of the external order demands and the in-pro- cess inventory on production line. First all, the semiconductor manufacturing system is analyzed. Order feed- back mechanism based on constant work-in-process products and real-time invertory is proposed. And the double close loop mechanism based on raw material loop and semi-finished product loop is established. Subsequently, directing at the target parameter optimization which is involved in the model, a heuristic ge- netic algorithm is proposed for muki-objective optimization. In order to verify the effectiveness of this meth- od, the proposed double close loop model and the heuristic genetic algorithm are used in the semiconductor packaging factory for example verification. The results show that the proposed double close loop inventory optimization mechanism has comparatively good system performance parameters.