An effective multi-objective discrete grey wolf optimizer for a real-world scheduling problem in welding production. (September 2016)

• Record Type: Journal Article
• Authors: Lu, Chao; Xiao, Shengqiang; Li, Xinyu; Gao, Liang
• Is Part Of: Advances in engineering software. Volume 99(2016)
• Page Start: 161
• Page End: 176
• Publication Date: 2016-09
• Subjects: Welding scheduling -- Multi-objective evolutionary algorithm -- Controllable processing times -- Sequence dependent setup times -- Transportation times -- Grey wolf optimizer; Computer-aided engineering -- Periodicals; Engineering -- Computer programs -- Periodicals; Engineering -- Software -- Periodicals; Periodicals; 620.0028553
• DOI: 10.1016/j.advengsoft.2016.06.004
• Languages: English
• ISSNs: 0965-9978

Abstract

Highlights: A multi-objective mathematical model for welding scheduling problem (WSP) is constructed. A new multi-objective discrete grey optimizer is proposed to solve this WSP. A new encoding scheme is designed to accommodate the characteristic of the WSP. A strategy that reduces machine load is proposed to improve the quality of solution. Abstract: This paper aims to provide a solution method for a real-world scheduling case from a welding process, which is one of the important processes in modern industry. The unique characteristic of the welding scheduling problem (WSP) is that multiple machines can process one operation at a time. Thus, WSP is a new scheduling problem. We first formulate a new multi-objective mixed integer programming model for this WSP based on a comprehensive investigation. This model involves some realistic constraints, controllable processing times (CPT), sequence dependent setup times (SDST) and job dependent transportation times (JDTT). Then we propose a multi-objective discrete grey wolf optimizer (MODGWO) considering not only production efficiency but also machine load on this real-world scheduling case. The solution is encoded as a two-part representation including a permutation vector and a machine assignment matrix. A reduction machine load strategy is used to adjust the number of machines aiming to minimize the machine load. To evaluate the effectiveness of the proposed MODGWO, we compare it with other well-known multi-objective evolutionary algorithms including NSGA-II and SPEA2 on a set of instances. Experimental results demonstrate that the proposed MODGWO is superior to the compared algorithms in terms of convergence, spread and coverage on most instances. Finally, MODGWO is successfully applied to this real-world WSP. This implies that the proposed model is feasible and the proposed algorithm can solve this real-world scheduling problem very well.