Balancing multiple objectives with anarchic manufacturing

被引:3
|
作者
Ma, Andrew [1 ]
Nassehi, Aydin [1 ]
Snider, Chris [1 ]
机构
[1] Univ Bristol, Design & Manufacture Futures Lab, Bristol, Avon, England
关键词
Distributed systems; simulation; multiple objectives;
D O I
10.1016/j.promfg.2020.01.142
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Balancing multiple conflicting objectives is a problem faced by all manufacturers, with on-time performance and cash position as two important examples. This study compares the effect of changing demand and priority of conflicting objectives within conventional, centralised, hierarchical systems and anarchic distributed systems, via agent-based simulation and modelling. The anarchic manufacturing system, based on free-market principles, delegates decision making authority and autonomy to the lowest possible level. Its proposed inherent flexibility and self-healing characteristics can accommodate complex scenarios. Decision-makers within the hierarchical system selected an appropriate dispatch rule to improve performance against objectives within its span of control. The anarchic system is shown to trade-off on-time and cash performance as respective objectives change priority, while the hierarchical systems do not. This warrants further investigation for distributed and anarchic systems as a system able to flexibly deal with complexity that also reacts to balance multiple objectives is highly desirable. (C) 2019 The Authors. Published by Elsevier B.V.
引用
收藏
页码:1453 / 1460
页数:8
相关论文
共 50 条