Читать книгу Optimization and Machine Learning - Patrick Siarry - Страница 30

The set of instances used for the 3L-CVRP is available at http://www.or.deis.unibo.it/research.html and was introduced by Gendreau et al. (2006). There are in total 27 3L-CVRP instances available on the web which provide an interesting test bed for the comparison of different heuristic and metaheuristic solutions. The vehicle characteristics are W = 25, H = 30 and L = 60, respectively. The demand of each customer is between 1 and 3. The capacity of vehicles is 0.75.

Table 1.2. Comparative study of the 3L-CVRP

Author	Problem	Routing problem	Loading problem
		Solution methods	Solution methods
Gendreau et al. (2006) Apile et al. (2007) Tarantilis et al. (2009) Fuellerer et al. (2010) Ren et al. (2011) Massen et al. (2012) Bortfeldt (2012) Wisniewski et al. (2011) Zhu et al. (2012) Miao et al. (2012) Ruan et al. (2013) Bortfeldt and Homberger-2013 Tao and Wang (2015) Junqueira etal. (2013) Hokima et al. (2016) Vega et al. (2020) Moura (2008) Moura and Oliveira (2009) Zhang et al. (2017) Moura et al. (2019) Vega et al. (2019) Ceschia etal. (2013) Pace etal. (2015) Pollaris et al. (2017) Bortfeldt et al. (2015) Koch et al. (2018) Reil et al. (2018) Koch et al. (2020) Bartok and Imreh (2011) Mannel and Bortfeldt (2016) Yi and Bortfeldt (2016) Li et al. (2018) Bortfeldt and Yi (2020)	3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRP 3L-CVRPTW 3L-CVRPTW 3L-CVRPTW 3L-CVRPTW 3L-CVRPTW 3L-HCVRP 3L-HFCVRPTW CVRP with pallet loading and axle weight constraints 3L-VRPB 3L-VRPBTW 3L-VRPBTW 3L-VRPMB 3L-VRPPD 3L-VRPD 3L-SDVRP 3L-SDVRP 3L-SDVRP	TS SA LS-GLS ACO Branch-and- bound black box algorithm TS TS TS GA HBMO P1R2 TS integer linear programming model Branch-and-Cut GRASP-CWS GA LS, GRASP TS-ABC MILP NLMIP LNS ILS and SA ILS LNS/VNS LNS TS TS LS LNS TS Data-driven 3-layer GA-LS	TS SA LS-GLS ACO Branch-and-bound black box algorithm TRSA bottom-left Deepest-Bottom-Left-Fill heuristic and the Maximum Touching Area TS six heuristics P1R2 least waste algorithm Branch-and-Cut GRASP-CWS GA LS, GRASP TS-ABC MILP NLMIP LNS ILS and SA ILS TSH LNS TS TS LS TSH TS Data-driven 3-layer GA-LS