工业工程 ›› 2015, Vol. 18 ›› Issue (3): 98-104.

• 实践与应用 • 上一篇    下一篇

基于三阶段解码机制的作业车间JIT调度算法研究

  

  1. (西安理工大学 1.机械与精密仪器工程学院; 2.高等技术学院,陕西 西安 710082)
  • 出版日期:2015-06-30 发布日期:2015-10-15
  • 作者简介:王荪馨(1974-),陕西省人,副教授,博士研究生,主要研究方向为先进制造技术、智能优化调度方法.
  • 基金资助:

    国家自然科学基金资助项目(51175420)

A Three-stage Decoding Method for Just-In-Time Job-Shop Scheduling

  1. (1.School of Mechanical and Precision Instrument Engineering, Xi'an University of Technology, Xi'an 710082, China; 2. School of Technology, Xi'an University of Technology, Xi'an 710082, China)
  • Online:2015-06-30 Published:2015-10-15

摘要: 零部件加工过程的精细化管控是保证大型复杂装备制造企业高效、均衡和低成本运营的重要手段。针对现有JIT调度方法普遍采用两阶段求解策略导致调度算法计算效率较低的问题,本文提出一种单阶段的改进型遗传算法解决思路:针对每一迭代优化过程中的染色体,设计了一种包含半主动解码、拖期工件贪婪插入、提前工件贪婪插入的三阶段渐进式解码机制,以引导待调度工序集尽可能在各自的交货时间点准时完工。这种三阶段解码方法具有与非正规调度指标吻合、解码机制简单和解码速度快等独特优势,最后采用72个JIT调度标准算例验证了该方法在优化性能、计算效率等方面的有效性。

关键词: 作业车间, JIT调度, 提前/拖期, 遗传算法, 三阶段解码机制

Abstract: Some parts of large and complex products are very complicated in manufacturing since they have many operations to go through and are time-consuming. The just-in-time (JIT) scheduling is suitable to lean control these manufacturing process. In order to cope with the time consuming problem, the two-tage strategies were employed to solve JIT scheduling problem in most researches. A novel genetic algorithm (GA) with three-stage decoding mechanism is introduced in this research. In each iteration process of GA, the chromosomes, which are represented by an operationbased scheme, are decoded successively by three stages. Firstly, the semi-active decoding method is induced to explore the more search space of solutions. Secondly, the greedy insertion for tardy operations is executed to swap the tardy operations left to the machine idle time to reduce the tardiness costs of solution. Thirdly, the greedy insertion for early operations is proposed to swap the early operations rightly to the suitable machine idle time to cut the more earliness costs. The experimental results finally demonstrate the effectiveness and efficiency of our algorithm over a wide range of benchmarks.

Key words: job-shop, jit scheduling, earliness/tardiness, genetic algorithm, three-stage decoding