Designing Dispatching Rules to Minimize Total Tardiness

Abstract

Summary. We approximate optimal solutions to the Flexible Job-Shop Problem by using dispatching rules discovered through Genetic Programming. While Simple Priority Rules have been widely applied in practice, their efficacy remains poor due to lack of a global view. Composite Dispatching Rules have been shown to be more effective as they are constructed through human experience. In this work, we employ suitable parameter and operator spaces for evolving Composite Dispatching Rules using Genetic Programming, with an aim towards greater scalability and flexibility. Experimental results show that Composite Dispatching Rules generated by our Genetic Programming framework outperforms the Single and Composite Dispatching Rules selected from literature over large validation sets with respect to total tardiness. Further results on sensitivity to changes (in coefficient values and terminals) among the evolved rules indicate that their designs are optimal.

Learning to schedule job-shop problems: representation and policy learning using graph neural network and reinforcement learning

“…To cope with such difficulties, PDRs have been extensively applied to real-world manufacturing scheduling problems [31,17,45]. A PDR is a heuristic rule that assigns jobs to machines based on their priorities while considering the current status of the system.…”

Genetic programming for production scheduling: a survey with a unified framework

“…Component(s) to be evolved by GPComponent ReferencesDispatching rule or priority ruleMiyashita (2000);Dimopoulos and Zalzala (2001);Yin et al (2003);Ho and Tay (2005);Geiger et al (2006);Jakobovic and Budin (2006);Jakobovic et al (2007);Tay and Ho (2007);Beham et al (2008);Tay and Ho (2008);Yang et al (2008); Baykasoglu et al(2010); Kofler et al (2009); Hildebrandt et al (2010); Kuczapski et al (2010); Nie et al (2010); Pickardt et al (2010); Abednego and Hendratmo (2011); Nie et al (2011b); Jakobovi and Marasovi (2012); Nguyen et al (2012a); Nie et al (2012); Nguyen et al (2013b,c,d); Nie et al (2013b,a); Park et al (2013a,b); Pickardt et al (2013); Qin et al (2013); Hildebrandt and Branke (2014); Hildebrandt et al (2014); Hunt et al (2014b,a); Nguyen et al (2014e); Park et al (2014); Branke et al (2015); Chen et al (2015); Han et al (2012); Hunt et al (2015b,a); Nguyen et al (2015a,b); Park et al (2015b,a); Shi et al (2015); Sim and Hart (2015); Wang et al (2015); Branke et al (2016a); Karunakaran et al (2016b); Durasevic et al (2016); Freitag and Hildebrandt (2016); Hart and Sim (2016); Karunakaran et al (2016a); Masood et al (2016a); Mei et al (2016); Nguyen (2016); Nguyen et al (2016); Park et al (2016b,a); Riley et al (2016); Mei and Zhang ((2011); Vazquez-Rodriguez and Ochoa (2011); Mascia et al (2013); Nguyen et al (2013a, 2014b,a) Others Alsina et al (2015); Baykasoglu and Ozbakr (2015); Belisrio and Pierreval (2015); Furuholmen et al (2009); Li et al (2008) (2008) applied GP to evolve dispatching rules for flexible job shop scheduling and use the least waiting time assignment (Ho and Tay, 2004) to find a suitable machine to process an operation. Similarly, Pickardt et al (2010) evolved dispatching rules for semiconductor manufacturing and used two existing heuristics, i.e.…”

Genetic programming for production scheduling: a survey with a unified framework

“…Yin et al [142] Ho and Tay [50] Geiger et al [38] Jakobovic and Budin [59] Jakobovic et al [60] Tay and Ho [134] Beham et al [14] Geiger and Uzsoy [37] Baykasoglu [9] Li et al [73] Tay and Ho [135] Yang et al [141] Mucientes et al [83] Baykasolu and Gken [12] Kofler et al [66] Furuholmen et al [36] Hildebrandt et al [46] Kuczapski et al…”