船舶管件智能化派工系统  

Intelligent Dispatch System for Ship Pipe Fittings

作　　者：胡敏[1] Hu Min(China Shipbuilding NDRI Engineering Co.,Ltd.,Shanghai 200090,China)

机构地区：[1]中船第九设计研究院工程有限公司,上海200090

出　　处：《科学与信息化》2022年第10期90-93,共4页Technology and Information

摘　　要：在船舶管件的生产加工过程中,需要将不同批次不同类型的各类任务派发给各个生产部门,这一过程既要保证每一项生产加工任务都能够顺利完成,还要保证管件在规定的上架日期内交货。文章首先结合各部门的生产状态信息和待生产加工的任务信息包括管件信息、加工工艺等,建立船舶管件生产数据分析模型;其次基于马尔科夫的强化学习获得各部门的动态派工信息;最后通过数据训练得出派工模型,模型能够给出合理派工推荐,提升部门的生产效率。In the production and processing of ship pipe fittings,different batches and types of tasks need to be dispatched to each production department.This process should not only ensure that each production and processing task can be completed smoothly,but also ensure that the pipe fittings are delivered within the specified shelf date.Firstly,this article combines with the production status information of each department and the task information to be produced and processed,including pipe fitting information and processing technology,to establish a production data analysis model of ship pipe fittings.Secondly,this article uses Markov based reinforcement learning to obtain the dynamic dispatching information of each department.Finally,after data training,the dispatching model can give reasonable dispatch recommendations and improve the production efficiency of the departments.

关 键 词：强化学习 智能化派工系统 船舶制造管理 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]