A hybrid genetic algorithm for the single row layout problem

Fast local search for single row facility layout

“…The size of each instance is 110. The second column displays the best known values reported in the paper by Ozcelik (2012). The obtained results indicate that VNS-LS3 again significantly outperforms VNS-LS-INSERT.…”

“…Their size ranges from 64 to 100 facilities. Both datasets have been used in several recent studies, including Yen (2009), Hungerländer andRendl (2013), Kothari and Ghosh (2013a, 2013b, 2014a, 2014b and Ozcelik (2012).…”

“…From the literature, it can be seen that many algorithms for the SR-FLP incorporate a local search procedure (Amaral & Letchford, 2013;Heragu & Alfa, 1992;Heragu & Kusiak, 1991;Kothari & Ghosh, 2013a, 2014a, 2014bKumar et al, 2008;Ozcelik, 2012;Samarghandi & Eshghi, 2010;Solimanpur et al, 2005;Teo & Ponnambalam, 2008). Two main types of local searches have been used for this problem (Kothari & Ghosh, 2013b).…”

“…There are also several layout methods available which follow the genetic paradigm. These include genetic algorithms of Ficko, Brezocnik, and Balic (2004) and Datta et al (2011) as well as hybrid genetic algorithms of Ozcelik (2012) and Kothari and Ghosh (2014a). A similar evolutionary technique called the imperialist competitive algorithm was presented by Lian, Zhang, Gao, and Shao (2011).…”

“…Two main types of local searches have been used for this problem (Kothari & Ghosh, 2013b). The first of them is based on pairwise interchanges of facilities (Amaral & Letchford, 2013;Heragu & Alfa, 1992;Heragu & Kusiak, 1991;Kothari & Ghosh, 2013a;Samarghandi & Eshghi, 2010;Solimanpur et al, 2005;Teo & Ponnambalam, 2008), whereas the second one proceeds by executing insertion moves, where a facility is moved from one position to another in the permutation (Kothari & Ghosh, 2013a, 2014a, 2014bKumar et al, 2008;Ozcelik, 2012). The performance of local search (LS) algorithms greatly depends on the neighborhood exploration (NE) procedures (Kothari & Ghosh, 2013a).…”