• 论文 •    

基于模糊控制与遗传算法的串行生产控制系统最优设计

莫巨华,黄敏,王兴伟   

  1. 1.东北大学 信息科学与工程学院,辽宁沈阳110004;2.东北大学 流程工业综合自动化教育部重点实验室,辽宁沈阳110004
  • 出版日期:2009-11-15 发布日期:2009-11-25

Optimal design of serial production control system based on fuzzy control and genetic algorithm

MO Ju-hua, HUANG Min, WANG Xing-wei   

  1. 1.School of Information Science & Engineering, Northeastern University, Shenyang 110004, China;2.Ministry of Education Key Lab of Integrated Automation of Process Industry, Northeastern University, Shenyang 110004, China
  • Online:2009-11-15 Published:2009-11-25

摘要: 为实现系统的最优设计,在建立基于模糊控制的生产存储控制仿真模型和详细描述控制器设计细节的基础上,建立以极小化在制品量和投放波动水平为目标、以满足顾客满意率要求为约束的多目标规划,并提出一种改进遗传算法与过程仿真相结合的求解方法。在该求解方法中,多目标规划通过加权平均转化为单目标规划。为考察模糊控制系统的性能,在实例分析中通过仿真与定量在制品法、看板及全面拉式几种系统进行了比较。结果表明,模糊控制系统不仅能够保持较低的在制品水平,而且能够保持较低的投放波动水平。

关键词: 生产控制, 模糊控制, 多目标规划, 遗传算法, 仿真

Abstract: To realize the optimal design of the control system, a multi-objective programming model was set up based on the construction of fuzzy Production Inventory Control (PIC) simulation model and the detailed description of the fuzzy controller design. This model was designed to minimize the Work-in-Process (WIP) and order placement fluctuation level while satisfying customers'requirements. An improved Genetic Algoritwohm (GA) integrating with process simulation was proposed to optimize the design of the fuzzy control system. Hereinto, the multi-objective program was translated into a single-objective one by weighted average conversion. To examine the performance of the fuzzy control system, it was compared to CONstant Work-In-Process (CONWIP), Kanban and generic pull systems via simulation. Results showed that fuzzy system could maintain its WIP, especially its order placement fluctuation at lower level.

Key words: production control, fuzzy control, multi-objective programming, genetic algorithm, simulation

中图分类号: