摘要
由一名新闻记者-机器人与机器学习的工作人员新闻编辑每日新闻-机器学习-计算智能的最新研究结果已经发表。根据NewsRx驻北京记者的新闻报道，研究表明：“快速发展的工业物联网(IIoT)正在推动传统制造向智能制造的转变。智能车间调度是智能制造的重要组成部分之一，它要求在不同的机器上分配工作以达到特定的生产目标。”本研究经费来源于国家自然科学基金(NSFC)。我们的新闻编辑引用了北京邮电大学的一项研究:“在实际生产中广泛存在的具有批处理机(FSSP-BPM)的流水车间调度问题,需要两个不同但相互依赖的决策:批处理和作业调度。现有的方法依赖于固定的搜索范式,利用专家知识来寻找满意的解决方案。”针对这一问题，本文提出了一种基于深度强化学习(DRL)的FSP-BPM决策过程，首先将FSP-BPM决策过程构造为一个Markov决策过程(MDP)，并设计相应的状态、行为和报酬；本文提出了一个基本的schedu-ling框架，该框架基于一个具有注意机制的编解码模型，并设计了一个批形成模块和一个基于未标记多维数据的调度模块。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Current study results on Machine Learn ing - Computational Intelligence have been published. According to news reportin g originating from Beijing, People's Republic of China, by NewsRx correspondents , research stated, "The rapidly evolving Industrial Internet of Things (IIoT) is driving the transition from conventional manufacturing to intelligent manufactu ring. Intelligent shop scheduling, as one of the essential components of intelli gent manufacturing in IIoT, is desired to allocate jobs on different machines to achieve specific production targets." Financial support for this research came from National Natural Science Foundatio n of China (NSFC). Our news editors obtained a quote from the research from the Beijing University of Posts and Telecommunications, "The flow-shop scheduling problem with batch pr ocessing machines (FSSP-BPM), which widely exists in real-world manufacturing, r equires two distinct but interdependent decisions: batch formation and job sched uling. Existing approaches rely on fixed search paradigms that utilize expert kn owledge to find satisfactory solutions. However, these methods struggle to ensur e solution quality under real-time constraints due to the varying data distribut ion and the complexity of large-scale practical problems. To address this challe nge, we propose a deep reinforcement learning (DRL) based method. First, we form ulate the FSSP-BPM decision process as a Markov Decision Process (MDP) and desig n the corresponding state, action, and reward. Second, we propose a basic schedu ling framework based on an encoder-decoder model with the attention mechanism. F inally, we design a batch formation module and a scheduling module trained on un labeled multi-dimensional data."
NETL
NSTL