Traditional job scheduling is an independent optimization phase without considering material flow among machines. However, in actual production, the machine layout will affect material flow time, and therefore affecting job completion time. In this paper, a novel multi-task-oriented production layout problem is proposed, with the goal to optimize material flow by adjusting machine placement. To solve this problem, a production layout evolution (PLEV) framework is established, which models it as a two-stage optimization problem. In the first stage, a concept of ‘relevance between machines (RBM)’ is proposed The RBM measures the relationship between two machines according to the processing data. And a machine placement scheme is designed based on RBM. In the second stage, a job scheduling rule based on the earliest finish time (EFT) is adopted to evaluate production layouts. Then, through an evolutionary programming mechanism, the production layout can evolve to optimize makespan. Experiments are carried out to verify the feasibility and performance of the proposed PLEV framework. The results demonstrate that promising product layouts for multi-task processing can be obtained by using this framework.

传统的作业计划是一个独立的优化阶段，无需考虑机器之间的物料流。但是，在实际生产中，机器布局会影响物料流动时间，从而影响作业完成时间。本文提出了一种新颖的面向多任务的生产布局问题，旨在通过调整机器位置来优化物料流。为了解决此问题，建立了生产布局演进（PLEV）框架，并将其建模为两阶段优化问题。在第一阶段，提出了“机器之间的相关性（RBM）”的概念。RBM根据处理数据来测量两台机器之间的关系。并基于RBM设计了一种机器布局方案。在第二阶段 采用基于最早完成时间（EFT）的作业计划规则来评估生产布局。然后，通过进化编程机制，生产布局可以进化以优化制造期。进行实验以验证所提出的PLEV框架的可行性和性能。结果表明，使用此框架可以实现用于多任务处理的有前途的产品布局。