Single-objective and bi-objetive parallel heuristics for the travel planning problem
Ano de defesa: | 2016 |
---|---|
Autor(a) principal: | |
Orientador(a): | |
Banca de defesa: | |
Tipo de documento: | Dissertação |
Tipo de acesso: | Acesso aberto |
Idioma: | eng |
Instituição de defesa: |
Universidade Federal de Viçosa
|
Programa de Pós-Graduação: |
Não Informado pela instituição
|
Departamento: |
Não Informado pela instituição
|
País: |
Não Informado pela instituição
|
Área do conhecimento CNPq: | |
Link de acesso: | http://www.locus.ufv.br/handle/123456789/9489 |
Resumo: | In this study we apply single-objective and bi-objective parallel heuristics to solve broad and realistic formulations of the travel planning problem. Given a travel time window and a set of destinations with their corresponding dwelling times, the goal of our single-objective approach is to find a route that produces a budget travel’s itinerary, involving flights, hotels and departure/arrival times. In turn, our bi-objective approach adds a complexity level in the problem’s formulation once we are seeking for a Pareto set of detailed travel itineraries, which are both cost and time efficient. When the sequence of cities is fixed, the single-objective version of the problem is commonly modeled in literature as a time-dependent network and the best itinerary is computed using shortest path algorithms. However, in this study, finding the order of cities that minimizes the total cost, and besides that, a set of good trade-off solutions, are also goals. Therefore, our single-objective formulation stands for a TDSPP (Time Dependent Shortest Path Problem) embedded in the TSP (Travel Salesman Problem) whereas our bi-objective formulation stands for a TDSPP embedded in a bi-objective TSP. On the first formulation we apply an ILS (Iterated Local Search) heuristic and on the second formulation we apply the NSGA-II (Nondominated Sorting Genetic Algorithm II) framework. For performance assessing, the results of both heuristics were compared to the results of corresponding exact methods with no time constraints. All test cases simulate realistic travel itineraries and run upon real-world travel data collected in advance, besides having to comply with an execution threshold of approximately 1 minute. For 285 single-objective test cases, our ILS heuristic was able to reach solutions in average less than 4.1% divergent from an exact implementation, besides reaching the optimal solution in about 30% of the test cases. In turn, for 180 bi-objective test cases, our NSGA-II implementation was able to reach an approximated solution in average up to 8% divergent from an exact implementation. |
id |
UFV_8d28a902d281dacf59f0b15ba88f520c |
---|---|
oai_identifier_str |
oai:locus.ufv.br:123456789/9489 |
network_acronym_str |
UFV |
network_name_str |
LOCUS Repositório Institucional da UFV |
repository_id_str |
|
spelling |
Beirigo, Breno Alveshttp://lattes.cnpq.br/6762636102657092Santos, André Gustavo dos2017-02-10T15:29:30Z2017-02-10T15:29:30Z2016-09-21BEIRIGO, Breno Alves. Single-objective and bi-objetive parallel heuristics for the travel planning problem. 2016. 92f. Dissertação (Mestrado em Ciência da Computação) - Universidade Federal de Viçosa, Viçosa. 2016.http://www.locus.ufv.br/handle/123456789/9489In this study we apply single-objective and bi-objective parallel heuristics to solve broad and realistic formulations of the travel planning problem. Given a travel time window and a set of destinations with their corresponding dwelling times, the goal of our single-objective approach is to find a route that produces a budget travel’s itinerary, involving flights, hotels and departure/arrival times. In turn, our bi-objective approach adds a complexity level in the problem’s formulation once we are seeking for a Pareto set of detailed travel itineraries, which are both cost and time efficient. When the sequence of cities is fixed, the single-objective version of the problem is commonly modeled in literature as a time-dependent network and the best itinerary is computed using shortest path algorithms. However, in this study, finding the order of cities that minimizes the total cost, and besides that, a set of good trade-off solutions, are also goals. Therefore, our single-objective formulation stands for a TDSPP (Time Dependent Shortest Path Problem) embedded in the TSP (Travel Salesman Problem) whereas our bi-objective formulation stands for a TDSPP embedded in a bi-objective TSP. On the first formulation we apply an ILS (Iterated Local Search) heuristic and on the second formulation we apply the NSGA-II (Nondominated Sorting Genetic Algorithm II) framework. For performance assessing, the results of both heuristics were compared to the results of corresponding exact methods with no time constraints. All test cases simulate realistic travel itineraries and run upon real-world travel data collected in advance, besides having to comply with an execution threshold of approximately 1 minute. For 285 single-objective test cases, our ILS heuristic was able to reach solutions in average less than 4.1% divergent from an exact implementation, besides reaching the optimal solution in about 30% of the test cases. In turn, for 180 bi-objective test cases, our NSGA-II implementation was able to reach an approximated solution in average up to 8% divergent from an exact implementation.Nesse trabalho são aplicadas heurísticas paralelas mono-objetivas e bi-objetivas para solucionar formulações abrangentes e realistas do problema de planejamento de viagens. Dado o intervalo de tempo que uma viagem pode ocorrer e um conjunto de destinos com seus respectivos tempos de permanência, a abordagem mono-objetiva procura determinar um itinerário de baixo custo que compreenda voos, hotéis e horários de partida/chegada. Por sua vez, a abordagem bi-objetiva adiciona complexidade a formulação do problema, uma vez que pretende determinar o conjunto Pareto de itinerários de viagem capazes de equilibrar custo e tempo. Quando a sequência de cidades é fixa, a versão mono-objetiva do problema é comumente modelada na literatura como uma rede dependente do tempo e o melhor itinerário é calculado usando algoritmos de caminho míınimo. Contudo, nesse trabalho, determinar a ordem de visitação das cidades também é um objetivo. Portanto, a formulação mono-objetiva proposta representa um TDSPP (Time Dependent Shortest Path Problem) incorporado ao TSP (Travel Salesman Problem) e a formulação bi-objetiva representa um TDSPP incorporado em um TSP bi-objetivo. Na primeira formulação foi aplicada a heurística ILS (Iterated Local Search) e na segunda formulação o framework NSGA-II (Nondominated Sorting Genetic Algorithm II). Os resultados de ambas as heurísticas foram comparados com os resultados produzidos por métodos exatos executados sem restrições temporais. Todos os casos de teste simulam itinerários de viagem realistas e foram executados em um banco de dados de viagens e hospedagens coletadas com antecedência. Além disso, independentemente da abordagem utilizada, estabeleceu-se que o tempo de execução de cada caso deve ser de aproximadamente 1 minuto. A heurística ILS proposta para a versão mono-objetiva do problema foi executada em 285 instâncias e alcançou, em média, soluçõoes no máximo 4.1% divergentes de uma implementa ̧ao exata, além de atingir a melhor solução em cerca de 30% dos casos de teste. Por sua vez, o framework NSGA-II foi capaz de produzir soluções no máximo 8% divergentes da implementação exata para 180 instâncias.Coordenação de Aperfeiçoamento de Pessoal de Nível SuperiorengUniversidade Federal de ViçosaPesquisa operacionalOtimização combinatóriaSolução de problemasHeurísticaCiência da ComputaçãoSingle-objective and bi-objetive parallel heuristics for the travel planning problemHeurísticas paralelas para o problema de planejamento de viagens mono- objetivo e bi-objetivoinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisUniversidade Federal de ViçosaDepartamento de InformáticaMestre em Ciência da ComputaçãoViçosa - MG2016-09-21Mestradoinfo:eu-repo/semantics/openAccessreponame:LOCUS Repositório Institucional da UFVinstname:Universidade Federal de Viçosa (UFV)instacron:UFVORIGINALtexto completo.pdftexto completo.pdftexto completoapplication/pdf2736980https://locus.ufv.br//bitstream/123456789/9489/1/texto%20completo.pdff41dd8d979519bbcc7b9d7350c591432MD51LICENSElicense.txtlicense.txttext/plain; charset=utf-81748https://locus.ufv.br//bitstream/123456789/9489/2/license.txt8a4605be74aa9ea9d79846c1fba20a33MD52THUMBNAILtexto completo.pdf.jpgtexto completo.pdf.jpgIM Thumbnailimage/jpeg3537https://locus.ufv.br//bitstream/123456789/9489/3/texto%20completo.pdf.jpgb830253629eec7d1e73169622ca8aa56MD53123456789/94892017-02-10 22:00:28.319oai:locus.ufv.br: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Repositório InstitucionalPUBhttps://www.locus.ufv.br/oai/requestfabiojreis@ufv.bropendoar:21452017-02-11T01:00:28LOCUS Repositório Institucional da UFV - Universidade Federal de Viçosa (UFV)false |
dc.title.pt-BR.fl_str_mv |
Single-objective and bi-objetive parallel heuristics for the travel planning problem |
dc.title.en.fl_str_mv |
Heurísticas paralelas para o problema de planejamento de viagens mono- objetivo e bi-objetivo |
title |
Single-objective and bi-objetive parallel heuristics for the travel planning problem |
spellingShingle |
Single-objective and bi-objetive parallel heuristics for the travel planning problem Beirigo, Breno Alves Pesquisa operacional Otimização combinatória Solução de problemas Heurística Ciência da Computação |
title_short |
Single-objective and bi-objetive parallel heuristics for the travel planning problem |
title_full |
Single-objective and bi-objetive parallel heuristics for the travel planning problem |
title_fullStr |
Single-objective and bi-objetive parallel heuristics for the travel planning problem |
title_full_unstemmed |
Single-objective and bi-objetive parallel heuristics for the travel planning problem |
title_sort |
Single-objective and bi-objetive parallel heuristics for the travel planning problem |
author |
Beirigo, Breno Alves |
author_facet |
Beirigo, Breno Alves |
author_role |
author |
dc.contributor.authorLattes.pt-BR.fl_str_mv |
http://lattes.cnpq.br/6762636102657092 |
dc.contributor.author.fl_str_mv |
Beirigo, Breno Alves |
dc.contributor.advisor1.fl_str_mv |
Santos, André Gustavo dos |
contributor_str_mv |
Santos, André Gustavo dos |
dc.subject.pt-BR.fl_str_mv |
Pesquisa operacional Otimização combinatória Solução de problemas Heurística |
topic |
Pesquisa operacional Otimização combinatória Solução de problemas Heurística Ciência da Computação |
dc.subject.cnpq.fl_str_mv |
Ciência da Computação |
description |
In this study we apply single-objective and bi-objective parallel heuristics to solve broad and realistic formulations of the travel planning problem. Given a travel time window and a set of destinations with their corresponding dwelling times, the goal of our single-objective approach is to find a route that produces a budget travel’s itinerary, involving flights, hotels and departure/arrival times. In turn, our bi-objective approach adds a complexity level in the problem’s formulation once we are seeking for a Pareto set of detailed travel itineraries, which are both cost and time efficient. When the sequence of cities is fixed, the single-objective version of the problem is commonly modeled in literature as a time-dependent network and the best itinerary is computed using shortest path algorithms. However, in this study, finding the order of cities that minimizes the total cost, and besides that, a set of good trade-off solutions, are also goals. Therefore, our single-objective formulation stands for a TDSPP (Time Dependent Shortest Path Problem) embedded in the TSP (Travel Salesman Problem) whereas our bi-objective formulation stands for a TDSPP embedded in a bi-objective TSP. On the first formulation we apply an ILS (Iterated Local Search) heuristic and on the second formulation we apply the NSGA-II (Nondominated Sorting Genetic Algorithm II) framework. For performance assessing, the results of both heuristics were compared to the results of corresponding exact methods with no time constraints. All test cases simulate realistic travel itineraries and run upon real-world travel data collected in advance, besides having to comply with an execution threshold of approximately 1 minute. For 285 single-objective test cases, our ILS heuristic was able to reach solutions in average less than 4.1% divergent from an exact implementation, besides reaching the optimal solution in about 30% of the test cases. In turn, for 180 bi-objective test cases, our NSGA-II implementation was able to reach an approximated solution in average up to 8% divergent from an exact implementation. |
publishDate |
2016 |
dc.date.issued.fl_str_mv |
2016-09-21 |
dc.date.accessioned.fl_str_mv |
2017-02-10T15:29:30Z |
dc.date.available.fl_str_mv |
2017-02-10T15:29:30Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/masterThesis |
format |
masterThesis |
status_str |
publishedVersion |
dc.identifier.citation.fl_str_mv |
BEIRIGO, Breno Alves. Single-objective and bi-objetive parallel heuristics for the travel planning problem. 2016. 92f. Dissertação (Mestrado em Ciência da Computação) - Universidade Federal de Viçosa, Viçosa. 2016. |
dc.identifier.uri.fl_str_mv |
http://www.locus.ufv.br/handle/123456789/9489 |
identifier_str_mv |
BEIRIGO, Breno Alves. Single-objective and bi-objetive parallel heuristics for the travel planning problem. 2016. 92f. Dissertação (Mestrado em Ciência da Computação) - Universidade Federal de Viçosa, Viçosa. 2016. |
url |
http://www.locus.ufv.br/handle/123456789/9489 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.publisher.none.fl_str_mv |
Universidade Federal de Viçosa |
publisher.none.fl_str_mv |
Universidade Federal de Viçosa |
dc.source.none.fl_str_mv |
reponame:LOCUS Repositório Institucional da UFV instname:Universidade Federal de Viçosa (UFV) instacron:UFV |
instname_str |
Universidade Federal de Viçosa (UFV) |
instacron_str |
UFV |
institution |
UFV |
reponame_str |
LOCUS Repositório Institucional da UFV |
collection |
LOCUS Repositório Institucional da UFV |
bitstream.url.fl_str_mv |
https://locus.ufv.br//bitstream/123456789/9489/1/texto%20completo.pdf https://locus.ufv.br//bitstream/123456789/9489/2/license.txt https://locus.ufv.br//bitstream/123456789/9489/3/texto%20completo.pdf.jpg |
bitstream.checksum.fl_str_mv |
f41dd8d979519bbcc7b9d7350c591432 8a4605be74aa9ea9d79846c1fba20a33 b830253629eec7d1e73169622ca8aa56 |
bitstream.checksumAlgorithm.fl_str_mv |
MD5 MD5 MD5 |
repository.name.fl_str_mv |
LOCUS Repositório Institucional da UFV - Universidade Federal de Viçosa (UFV) |
repository.mail.fl_str_mv |
fabiojreis@ufv.br |
_version_ |
1794528709852528640 |