A hybrid genetic algorithm for multi-depot homogenous locomotive assignment with time windows

Ghoseiri, Keivan; Ghannadpour, Seyed Farid

doi:10.1016/j.asoc.2009.06.004

Cited by 46 publications

(24 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Unlike PMX and HMX, which were used in our recent research (Ghoseiri and Ghannadpour 2010), the REX that is suggested in this paper produces better solution and works properly than them. Eventually, produced results show that suggested approach is quiet effective in respect of quality and time.…”

Section: Resultsmentioning

confidence: 99%

“…In this paper, the cluster-first, route-second approach is used to convert the multi-depot locomotive assignment to P single depot problems (for more details, see Ghoseiri and Ghannadpour (2010)). By this approach, at first stage, trains are assigned to each depot to obtain P clusters, and after that each cluster is solved by the suggested hybrid genetic algorithm.…”

Section: Solution Proceduresmentioning

confidence: 99%

“…This paper uses a λ-interchange mechanism to move trains between routes to generate neighborhood solution for the problem (for more details see Ghoseiri and Ghannadpour (2010)). In one version of the algorithm called global best, the whole neighborhood is explored, and the best move is selected.…”

Section: Initial Populationmentioning

confidence: 99%

“…Ghoseiri and Ghannadpour (2009a) suggested the solving of the general locomotive planning (or scheduling) that is one of the most attractive topics in operation research by vehicle routing problem. Ghoseiri and Ghannadpour (2010) extended the previous research and tried to study the multi-depot homogenous locomotive assignment with more real-life assumptions. This model was formulated by vehicle routing problem with time windows (VRPTW) and solved heuristically by an efficient hybrid genetic algorithm.…”

Section: Introductionmentioning

confidence: 99%

“…According to these levels of planning, locomotive scheduling is the final stage of the railroad scheduling and depends on incoming requests, but with this assumption, engines and crews are rostered and appropriate cars are available. It must be mentioned that this paper is a complementary version of our recent researches (Ghoseiri and Ghannadpour 2009a;Ghoseiri and Ghannadpour 2010) about the locomotive assignment problem. Ghoseiri and Ghannadpour (2009a) suggested the solving of the general locomotive planning (or scheduling) that is one of the most attractive topics in operation research by vehicle routing problem.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Locomotive assignment problem with train precedence using genetic algorithm

Noori

Ghannadpour

2012

J Ind Eng Int

Self Cite

View full text Add to dashboard Cite

This paper aims to study the locomotive assignment problem which is very important for railway companies, in view of high cost of operating locomotives. This problem is to determine the minimum cost assignment of homogeneous locomotives located in some central depots to a set of pre-scheduled trains in order to provide sufficient power to pull the trains from their origins to their destinations. These trains have different degrees of priority for servicing, and the high class of trains should be serviced earlier than others. This problem is modeled using vehicle routing and scheduling problem where trains representing the customers are supposed to be serviced in pre-specified hard/soft fuzzy time windows. A two-phase approach is used which, in the first phase, the multi-depot locomotive assignment is converted to a set of single depot problems, and after that, each single depot problem is solved heuristically by a hybrid genetic algorithm. In the genetic algorithm, various heuristics and efficient operators are used in the evolutionary search. The suggested algorithm is applied to solve the medium sized numerical example to check capabilities of the model and algorithm. Moreover, some of the results are compared with those solutions produced by branch-andbound technique to determine validity and quality of the model. Results show that suggested approach is rather effective in respect of quality and time.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Solution Proceduresmentioning

confidence: 99%

Section: Initial Populationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations