“…4) The benchmark of Paraskevopoulos et al (2008) for the HFTW is a subset of the FSMTW instances, in which the fleet size is set equal to that found in the best known solutions of Liu and Shen (1999a). In total, there are 24 benchmark instances derived from Liu and Shen (1999a) Ochi et al (1998a) DE, PS 12 Ochi et al (1998b) DE, PS 13 Gendreau et al (1999) TS 14 Liu and Shen (1999a) TW CT 15 Liu and Shen (1999b) TW CT 16 Irnich (2000) PD • LB, SP 17 Dullaert et al (2002) TW CT 18 Han and Cho (2002) CT 19 Renaud and Boctor (2002) • SP, CT 20 Wassan and Osman (2002) TS 21 • AP 40 Xu et al (2012) TW, MD • VNS 41 Belfiore and Yoshizaki (2013) TW, PD • SS 42 De la Cruz et al (2013) TW, OV • ACO 43 Franceschelli et al (2013) MV • CT 44 Kopfer and Kopfer (2013) • MIP 45 Kritikos and Ioannou (2013) TW, OV • CT 46 Leung et al (2013) TDL SA 47 Salhi et al (2013) BC Euchi and Chabchoub (2010) • TS 15 Li et al (2010) AMP 16 Tütüncü (2010) BC CH 17 Bettinelli et al (2011) TW, MD • BCP 18 Brandão et al (2011) TS 19 Potvin and Naud (2011) • TS 20 Li et al (2012) OP AMP, TS 21 Naji-Azimi and Salari (2013) • CH 22 Belmecheri et al (2013) TW, BC • PSO 23 Kwon et al (2013) GR • Koç et al (2015) TW PS A.1, A.2 and A.3 in the Appendix. In these tables, the first column provides the references, and for each reference two performance indicators are shown: the average percentage deviation (Dev) from the value of the best-known solution (BKS) for each instance retrieved from the articles surveyed, and the average computation time in seconds (Time).…”