Scholarlink

<p>Industry 5.0 emphasizes human–machine collaboration and focuses on resilience and sustainability, as defined by the European Commission. Recent trends in technology and the abundance of available data have driven the digital transformation of manufacturing processes. However, achieving the goals of Industry 5.0 requires resilient manufacturing systems capable of adapting to disruptions and uncertainties. Often, when a bottleneck machine exists in a manufacturing system, the problem simplifies to single-machine scheduling, specifically, scheduling on the bottleneck machine. This paper addresses this challenge by first presenting a quadratic linear programming (QLP)-based mathematical model to obtain a lower bound (LB) on the completion time variance (CTV), a key indicator of system resilience to variations in processing times. This is followed by two quadratic integer programming (QIP) based mathematical models for optimally solving the single-machine CTV problem (1||CTV). Second, we develop a tight lower bound on the completion time variance of jobs for a partial sequence (considering both scheduled and unscheduled jobs) and then present a branch and bound (B&amp;B) algorithm for optimally solving the single-machine CTV minimization problem. Minimizing CTV enhances the system's ability to handle unexpected variations in demand or machine failures, thus improving resilience. We also demonstrate the dominance of this proposed lower bound over existing ones. Both the proposed B&amp;B algorithm and mathematical models are compared with existing exact methods for optimally solving the 1||CTV problem. We also propose two lemmas to determine job sequencing order and subsequently develop a heuristic schedule based on these lemmas. This heuristic, a construction heuristic, is computationally very fast. Finally, a detailed computational study demonstrates the superiority of the proposed methods, with separate analyses for each algorithm, ultimately contributing to more resilient manufacturing operations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Exact algorithms and resilient heuristic approaches to minimize the completion time variance of jobs on a single machine

  • 期刊论文

  • 发表日期：

    2025年3月5日

查看全文
  • Raju Rajkanth,
  • Sakthivel Madankumar,
  • Chandrasekaran Rajendran,
  • Hans Ziegler

摘要

Industry 5.0 emphasizes human–machine collaboration and focuses on resilience and sustainability, as defined by the European Commission. Recent trends in technology and the abundance of available data have driven the digital transformation of manufacturing processes. However, achieving the goals of Industry 5.0 requires resilient manufacturing systems capable of adapting to disruptions and uncertainties. Often, when a bottleneck machine exists in a manufacturing system, the problem simplifies to single-machine scheduling, specifically, scheduling on the bottleneck machine. This paper addresses this challenge by first presenting a quadratic linear programming (QLP)-based mathematical model to obtain a lower bound (LB) on the completion time variance (CTV), a key indicator of system resilience to variations in processing times. This is followed by two quadratic integer programming (QIP) based mathematical models for optimally solving the single-machine CTV problem (1||CTV). Second, we develop a tight lower bound on the completion time variance of jobs for a partial sequence (considering both scheduled and unscheduled jobs) and then present a branch and bound (B&B) algorithm for optimally solving the single-machine CTV minimization problem. Minimizing CTV enhances the system's ability to handle unexpected variations in demand or machine failures, thus improving resilience. We also demonstrate the dominance of this proposed lower bound over existing ones. Both the proposed B&B algorithm and mathematical models are compared with existing exact methods for optimally solving the 1||CTV problem. We also propose two lemmas to determine job sequencing order and subsequently develop a heuristic schedule based on these lemmas. This heuristic, a construction heuristic, is computationally very fast. Finally, a detailed computational study demonstrates the superiority of the proposed methods, with separate analyses for each algorithm, ultimately contributing to more resilient manufacturing operations.