Search

The Cumulative Capacitated Vehicle Routing Problem is a variant of the classic routing problem in which the objective function is to minimize the sum of arrival times to customers. This article proposes a model for the problem that uses position indexes in order to calculate the contribution of the travel time of an edge to the arrival times of the remaining customers on a route. The model is implemented and solved by the branch-cut-and-price (BCP) algorithm in the VRPSolver package. Computational experiments indicate that the proposed BCP model is superior to the literature, being able to solve many open instances. Good results were also obtained for the Multi-Depot variant of the problem.

We consider the shortest path problem with resource constraints arising as a subproblem in state-of-the-art branch-cut-and-price algorithms for vehicle routing problems. We propose a variant of the bidirectional label-correcting algorithm in which the labels are stored and extended according to the so-called bucket graph. This organization of labels helps to significantly decrease the number of dominance checks and the running time of the algorithm. We also show how the forward/backward route symmetry can be exploited and how to eliminate arcs from the bucket graph using reduced costs. The proposed algorithm can be especially beneficial for vehicle routing instances with large vehicle capacity and/or with time window constraints. Computational experiments were performed on instances from the distance-constrained vehicle routing problem, including multidepot and site-dependent variants, on the vehicle routing problem with time windows, and on the “nightmare” instances of the heterogeneous fleet vehicle routing problem. Significant improvements over the best algorithms in the literature were achieved, and many instances could be solved for the first time.

International audience; In this paper, we are interested in the exact solution of the vehicle routing problem with back-hauls (VRPB), a classical vehicle routing variant with two types of customers: linehaul (delivery) and backhaul (pickup) ones. We propose two branch-cut-and-price (BCP) algorithms for the VRPB. The first of them follows the traditional approach with one pricing subproblem, whereas the second one exploits the linehaul/backhaul customer partitioning and defines two pricing sub-problems. The methods incorporate elements of state-of-the-art BCP algorithms, such as rounded capacity cuts, limited-memory rank-1 cuts, strong branching, route enumeration, arc elimination using reduced costs and dual stabilization. Computational experiments show that the proposed algorithms are capable of obtaining optimal solutions for all existing benchmark instances with up to 200 customers, many of them for the first time. It is observed that the approach involving two pricing subproblems is more efficient computationally than the traditional one. Moreover, new instances are also proposed for which we provide tight bounds. Also, we provide results for benchmark instances of the heterogeneous fixed fleet VRPB and the VRPB with time windows.

Major advances were recently obtained in the exact solution of vehicle routing problems (VRPs). Sophisticated branch-cut-and-price (BCP) algorithms for some of the most classical VRP variants now solve many instances with up to a few hundreds of customers. However, adapting and reimplementing those successful algorithms for other variants can be a very demanding task. This work proposes a BCP solver for a generic model that encompasses a wide class of VRPs. It incorporates the key elements found in the best existing VRP algorithms: ng-path relaxation, rank-1 cuts with limited memory, path enumeration, and rounded capacity cuts; all generalized through the new concepts of “packing set” and “elementarity set”. The concepts are also used to derive a branching rule based on accumulated resource consumption and to generalize the Ryan and Foster branching rule. Extensive experiments on several variants show that the generic solver has an excellent overall performance, in many problems being better than the best specific algorithms. Even some non-VRPs, like bin packing, vector packing and generalized assignment, can be modeled and effectively solved.

This work proposes a partial optimization metaheuristic under special intensification conditions (POPMUSIC) for the classical capacitated vehicle routing problem. The proposed approach uses a branch-cut-and-price algorithm as a powerful heuristic to solve subproblems whose dimensions are typically between 25 and 200 customers. The whole algorithm can be seen as the application of local search over very large neighborhoods, starting from a single initial solution. The main computational experiments were carried out on instances having between 302 and 1000 customers. Using initial solutions generated by some of the best available metaheuristics for the problem, POPMUSIC was able to obtain consistently better solutions for long runs of up to 32 h. Starting from the best known solutions available in CVRP library (CVRPLIB), POPMUSIC was able to find new best solutions for several instances, including some very large ones. In a final experiment, we successfully applied the same POPMUSIC framework to tackle the heterogeneous fleet vehicle routing problem and the vehicle routing problem with backhauls.

International audience; We propose branch-cut-and-price algorithms for the classic bin packing problem and also for the following related problems: vector packing, variable sized bin packing and variable sized bin packing with optional items. The algorithms are defined as models for VRPSolver, a generic solver for vehicle routing problems. In that way, a simple parameterization enables the use of several branch-cut-and-price advanced elements: automatic stabilization by smoothing, limited-memory rank-1 cuts, enumeration, hierarchical strong branching and limited discrepancy search diving heuristics. As an original theoretical contribution, we prove that the branching over accumulated resource consumption (Gélinas et al. 1995), that does not increase the difficulty of the pricing subproblem, is sufficient for those bin packing models. Extensive computational results on instances from the literature show that the VRPSolver models have a performance that is very robust over all those problems, being often superior to the existing exact algorithms on the hardest instances. Several instances could be solved to optimality for the first time.

This paper studies two polytopes: the complete set packing and set partitioning polytopes, which are both associated with a binary n -row matrix having all possible columns. Cuts of rank 1 for the latter polytope play a central role in recent exact algorithms for many combinatorial problems, such as vehicle routing. We show the precise relation between the two polytopes studied, characterize the multipliers that induce rank 1 clique facets and give several families of multipliers that yield other facets.

The vehicle routing problem with time windows VRPTW consists of finding least-cost vehicle routes to satisfy the demands of customers that can be visited within specific time windows. We introduce two enhancements for the exact solution of the VRPTW by branch-price-and-cut BPC. First, we develop a sharper form of the limited-memory subset-row inequalities by representing the memory as an arc subset rather than a node subset. Second, from the elementary inequalities introduced by Balas in 1977, we derive a family of inequalities that dominate them. These enhancements are embedded into an exact BPC algorithm that includes state-of-the-art features such as bidirectional labeling, decremental state-space relaxation, completion bounds, variable fixing, and route enumeration. Computational results show that these enhancements are particularly effective for the most difficult instances and that our BPC algorithm can solve all 56 Solomon instances with 100 customers and 51 of 60 Gehring and Homberger instances with 200 customers. The online appendix is available at https://doi.org/10.1287/ijoc.2016.0744 .

Pecin etal. (2016) introduced a limited memory technique that allows an efficient use of Rank-1 cuts in the Set Partitioning Formulation of Vehicle Routing Problems, motivating a deeper investigation of those cuts. This work presents a computational polyhedral study that determines the best possible sets of multipliers for cuts with up to 5 rows. Experiments with CVRP instances show that the new multipliers lead to significantly improved dual bounds and contributes decisively for solving an open instance with 420 customers.

The recent research on the CVRP is being slowed down by the lack of a good set of benchmark instances. The existing sets suffer from at least one of the following drawbacks: (i) became too easy for current algorithms; (ii) are too artificial; (iii) are too homogeneous, not covering the wide range of characteristics found in real applications. We propose a new set of 100 instances ranging from 100 to 1000 customers, designed in order to provide a more comprehensive and balanced experimental setting. Moreover, the same generating scheme was also used to provide an extended benchmark of 600 instances. In addition to having a greater discriminating ability to identify “which algorithm is better”, these new benchmarks should also allow for a deeper statistical analysis of the performance of an algorithm. In particular, they will enable one to investigate how the characteristics of an instance affect its performance. We report such an analysis on state-of-the-art exact and heuristic methods.

The best performing exact algorithms for the capacitated vehicle routing problem developed in the last 10 years are based in the combination of cut and column generation. Some authors only used cuts expressed over the variables of the original formulation, in order to keep the pricing subproblem relatively easy. Other authors could reduce the duality gaps by also using a restricted number of cuts over the master LP variables, stopping when the pricing becomes prohibitively hard. A particularly effective family of such cuts are the subset row cuts. This work introduces a technique for greatly reducing the impact on the pricing of these cuts, thus allowing much more cuts to be added. The newly proposed branch-cut-and-price algorithm also incorporates and combines for the first time (often in an improved way) several elements found in previous works, like route enumeration and strong branching. All the instances used for benchmarking exact algorithms, with up to 199 customers, were solved to optimality. Moreover, some larger instances with up to 360 customers, only considered before by heuristic methods, were solved too.

Major advances were recently obtained in the exact solution of Vehicle Routing Problems (VRPs). Sophisticated Branch-Cut-and-Price (BCP) algorithms for some of the most classical VRP variants now solve many instances with up to a few hundreds of customers. However, adapting and reimplementing those successful algorithms for other variants can be a very demanding task. This work proposes a BCP solver for a generic model that encompasses a wide class of VRPs. It incorporates the key elements found in the best recent VRP algorithms: ng-path relaxation, rank-1 cuts with limited memory, and route enumeration; all generalized through the new concept of “packing set”. This concept is also used to derive a new branch rule based on accumulated resource consumption and to generalize the Ryan and Foster branch rule. Extensive experiments on several variants show that the generic solver has an excellent overall performance, in many problems being better than the best existing specific algorithms. Even some non-VRPs, like bin packing, vector packing and generalized assignment, can be modeled and effectively solved.

International audience; This paper considers a family of Vehicle Routing Problem (VRP) variants that generalize the classical Capacitated VRP by taking into account the possibility that vehicles di↵er by capacity, costs, depot allocation, or even by the subset of customers that they can visit. This work proposes a branch- cut-and-price algorithm that adapts advanced features found in the best performing exact algorithms for homogeneous fleet VRPs. The original con- tributions include: (i) the use of Extended Capacity Cuts, defined over a pseudo-polynomially large extended formulation, together with Rank-1 Cuts, defined over the Set Partitioning Formulation; (ii) the concept of vehicle-type dependent memory for Rank-1 Cuts; and (iii) a new family of lifted Extended Capacity Cuts that takes advantage of the vehicle-type dependent route enumeration. The algorithm was extensively tested in in- stances of the literature and was shown to be significantly better than pre- vious exact algorithms, finding optimal solutions for many instances with up to 200 customers and also for some larger instances. A new set of set of benchmark instances is also proposed.

International audience; This paper deals with the Minimum Latency Problem (MLP), a variant of the well-known Travel- ing Salesman Problem in which the objective is to minimize the sum of waiting times of customers. This problem arises in many applications where customer satisfaction is more important than the total time spent by the server. This paper presents a novel branch-and-price algorithm for MLP that strongly relies on new features for the ng-path relaxation, namely: (1) a new labeling algo- rithm with an enhanced dominance rule named multiple partial label dominance; (2) a generalized definition of ng-sets in terms of arcs, instead of nodes; and (3) a strategy for decreasing ng-set size when those sets are being dynamically chosen. Also, other elements of efficient exact algorithms for vehicle routing problems are incorporated into our method, such as reduced cost fixing, dual stabilization, route enumeration and strong branching. Computational experiments over TSPLIB instances are reported, showing that several instances not solved by the current state-of-the-art method can now be solved.

This work proposes a Branch-cut-and-price (BCP) approach for the Vehicle Routing Problem with Simultaneous Pickup and Delivery (VRPSPD). We also deal with a particular case of the VRPSPD, known as the Vehicle Routing Problem with Mixed Pickup and Delivery. The BCP algorithm was tested in well-known benchmark instances involving up to 200 customers. Four instances were solved for the first time and some LBs were improved.

This paper deals with the Heterogeneous Fleet Vehicle Routing Problem (HFVRP). The HFVRP generalizes the classical Capacitated Vehicle Routing Problem by considering the existence of different vehicle types, with distinct capacities and costs. The objective is to determine the best fleet composition as well as the set of routes that minimize the total costs. The proposed hybrid algorithm is composed by an Iterated Local Search (ILS) based heuristic and a Set Partitioning (SP) formulation. The SP model is solved by means of a Mixed Integer Programming solver that interactively calls the ILS heuristic during its execution. The developed algorithm was tested in benchmark instances with up to 360 customers. The results obtained are quite competitive with those found in the literature and new improved solutions are reported.

We propose a branch-and-cut algorithm for the VRPSPD where the constraints that ensure that the capacities are not exceeded in the middle of a route are applied in a lazy fashion. The algorithm was tested in 87 instances with 50–200 customers, finding improved lower bounds and several new optimal solutions.

This paper presents an algorithm to obtain lower bounds for the Split Delivery Vehicle Routing Problem. An extended formulation over a large set of variables is provided and valid inequalities are identified. The algorithm combined column and cut generation and improved the best known lower bounds for all instances from the literature. Some reasonably sized instances are solved to optimality for the first time.

A well-known transformation by Pearn, Assad and Golden reduces a capacitated arc routing problem (CARP) into an equivalent capacitated vehicle routing problem (CVRP). However, that transformation is regarded as unpractical, since an original instance with r required edges is turned into a CVRP over a complete graph with 3r + 1 vertices. We propose a similar transformation that reduces this graph to 2r + 1 vertices, with the additional restriction that a previously known set of r pairwise disconnected edges must belong to every solution. Using a recent branch-and-cut-and-price algorithm for the CVRP, we observed that it yields an effective way of attacking the CARP, being significantly better than the exact methods created specifically for that problem. Computational experiments obtained improved lower bounds for almost all open instances from the literature. Several such instances could be solved to optimality.Scope and purpose The scope of this paper is transforming arc routing problems into node routing problems. The paper shows that this approach can be effective and, in particular, that the original instances may generate node routing instances that behave as if the size is not increased. This result is obtained by slightly modifying the well-known transformation by Pearn, Assad and Golden from capacitated arc routing problem (CARP) to the capacitated vehicle routing problem (CVRP), that is regarded as unpractical. The paper provides a computational experience using a recent branch-and-cut-and-price algorithm for the CVRP. The results are significantly better than the exact methods created specifically for that problem, improving lower bounds for almost all open instances from the literature. Several such instances could be solved to optimality.

The best performing exact algorithms for the Capacitated Vehicle Routing Problem are based on the combination of cut and column generation. Some authors could obtain reduced duality gaps by also using a restricted number of cuts over the Master LP variables, stopping separation before the pricing becomes prohibitively hard. This work introduces a technique for greatly decreasing the impact on the pricing of the Subset Row Cuts, thus allowing much more such cuts to be added. The newly proposed Branch-Cut-and-Price algorithm also incorporates and combines for the first time (often in an improved way) several elements found in previous works, like route enumeration and strong branching. All the instances used for benchmarking exact algorithms, with up to 199 customers, were solved to optimality. Moreover, some larger instances with up to 360 customers, only considered before by heuristic methods, were solved too.

This work deals with the Vehicle Routing Problem with Simultaneous Pickup and Delivery. We propose undirected and directed two-commodity flow formulations, which are based on the one developed by Baldacci, Hadjiconstantinou and Mingozzi for the Capacitated Vehicle Routing Problem. These new formulations are theoretically compared with the one-commodity flow formulation proposed by Dell’Amico, Righini and Salani. The three formulations were tested within a branch-and-cut scheme and their practical performance was measured in well-known benchmark problems. The undirected two-commodity flow formulation obtained consistently better results. Several optimal solutions to open problems with up to 100 customers and new improved lower bounds for instances with up to 200 customers were found.

This chapter presents techniques for constructing robust Branch-Cut-and-Price algorithms on a number of Vehicle Routing Problem variants. The word ‘‘robust’’ stresses the effort of controlling the worst-case complexity of the pricing subproblem, keeping it pseudo-polynomial. Besides summarizing older research on the topic, some promising new lines of investigation are also presented, specially the development of new families of cuts over large extended formulations. Computational experiments over benchmark instances from ACVRP, COVRP, CVRP and HFVRP variants are provided.

This paper presents a robust branch-cut-and-price algorithm for the Heterogeneous Fleet Vehicle Routing Problem (HFVRP), vehicles may have various capacities and fixed costs. The columns in the formulation are associated to q-routes, a relaxation of capacitated elementary routes that makes the pricing problem solvable in pseudopolynomial time. Powerful new families of cuts are also proposed, which are expressed over a very large set of variables. Those cuts do not increase the complexity of the pricing subproblem. Experiments are reported where instances up to 75 vertices were solved to optimality, a major improvement with respect to previous algorithms.

Udgivelsesdato: JUL The best exact algorithms for the Capacitated Vehicle Routing Problem (CVRP) have been based on either branch-and-cut or Lagrangean relaxation/column generation. This paper presents an algorithm that combines both approaches: it works over the intersection of two polytopes, one associated with a traditional Lagrangean relaxation over q-routes, the other defined by bound, degree and capacity constraints. This is equivalent to a linear program with exponentially many variables and constraints that can lead to lower bounds that are superior to those given by previous methods. The resulting branch-and-cut-and-price algorithm can solve to optimality all instances from the literature with up to 135 vertices. This more than doubles the size of the instances that can be consistently solved.

The best exact algorithms for the Capacitated Vehicle Routing Problem (CVRP) have been based on either branch-and-cut or Lagrangean relaxation/column generation. This paper presents an algorithm that combines both approaches: it works over the intersection of two polytopes, one associated with a traditional Lagrangean relaxation over q-routes, the other defined by bound, degree and capacity constraints. This is equivalent to a linear program with exponentially many variables and constraints that can lead to lower bounds that are superior to those given by previous methods. The resulting branch-and-cut-and-price algorithm can solve to optimality all instances from the literature with up to 135 vertices. This doubles the size of the instances that can be consistently solved.

Orientador: Cid Carvalho de Souza Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Computação Resumo: Nesta tese,dois problemas de otimização combinatória em grafos são modelados por programação linear inteira e resolvidos através de técnicas de geração de colunas. Os dois casos correspondem a generalizações de problemas clássicos em grafos e que ocorrem em muitas situações práticas. O primeiro, chamado problema dos anéis-estrelas capacitados, é uma generalização do problema de roteamento de veículos e modela situações reais encontradas nas áreas de logística de distribuição e de transporte. O segundo, conhecido por problema da coloração particionada, generaliza o problema da coloração de vértices em grafos e ocorre em aplicações no projeto de redes ópticas. As formulações de programação linear inteira desenvolvidas neste trabalho para modelar ambos os problemas estão relacionadas 'a técnica da decomposição de Dantzig-Wolfe e usam uma quantidade exponencialmente grande de variáveis de decisão . Nestas formulações, cada uma das variáveis representa uma estrutura específica do problema sendo estudado. No problema dos anéis-estrelas capacitados, cada variável está associada a um anel-estrela e, no problema da coloração particionada, a um conjunto independente. As relaxações lineares destes tipos de modelos, em geral, apresentam limitantes duais mais apertados que outros modelos compactos, isto é, definidos para um número polinomial de variáveis. Nesta tese, nós avaliamos estas novas formulações, comparando-as com outros modelos conhecidos para os problemas estudados. Além disso, nos dois casos, projetamos e implementamos algoritmos exatos do tipo branch-and-price e/ou branch-and-cut-and-price capazes de computar os modelos propostos. Experimentos computacionais foram realizados com estes algoritmos que confirmaram a adequação das técnicas aqui empregadas. Tanto para o problema dos anéis-estrelas capacitados quanto para o problema da coloração particionada, os resultados alcançados por nós foram comparados com aqueles reportados na literatura e mostraram que os algoritmos baseados em geração de colunas tiveram desempenho melhor que os algoritmos propostos anteriormente Abstract: In this thesis, two combinatorial optimization problems are modeled by integer linear programming and solved using the column generation technique. Both cases correspond to generalizations of classical problems in graphs that occur in many practical situations. The first, called capacitated ring-star problem is a generalization of the vehicle routing problem and models real situations found in logistics and transportation. The second, known as the partition coloring problem, generalizes the vertex coloring problem in graphs and arises in design of fiber optics networks. The integer linear programming formulations developed in this work to model both problems are related to the Dantzing-Wolfe decomposition and use exponential number of decision variables. In these formulations, each decision variable represents a specific structure of the problem under study. For the capacitated ring-star problem, each variable is assigned to a ring-star and, for the partition coloring problem, to an independent set. The linear relaxation of this kind of model in general leads to tighter dual bounds than the ones obtained from compact models, i.e., defined over a polynomial number of variables. In this thesis, we evaluated both new formulations, comparing them to other known models for the respective problems. Moreover, in both cases, we designed and implemented exact branch-and-price and/or branch-and-cut-and-price algorithms that are able to solve the proposed models. Computational experiments were performed with these algorithms and showed that the used techniques were adequate. Both for the capacitated ring-star problem and for the partition coloring problem, we compared our results with those reported in the literature and showed that the algorithms based on column generation outperformed the previous ones Doutorado Otimização Combinatória Doutor em Ciência da Computação

Orientador: Vinícius Amaral Armentano Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Elétrica e de Computação Resumo: Este trabalho aborda três variantes multiatributo do problema de roteamento de veículos. A primeira apresenta frota heterogênea, janelas de tempo invioláveis e tempos de viagem determinísticos. Para resolvê-la, são propostos algoritmos ótimos baseados na decomposição de Benders. Estes algoritmos exploram a estrutura do problema em uma formulação de programação inteira mista, e três diferentes técnicas são desenvolvidas para acelerá-los. A segunda variante contempla os atributos de frota heterogênea, janelas de tempo flexíveis e tempos de viagem determinísticos. As janelas de tempo flexíveis permitem o início do serviço nos clientes com antecipação ou atraso limitados em relação às janelas de tempo invioláveis, com custos de penalidade. Este problema é resolvido por extensões dos algoritmos de Benders, que incluem novos algoritmos de programação dinâmica para a resolução de subproblemas com a estrutura do problema do caixeiro viajante com janelas de tempo flexíveis. A terceira variante apresenta frota heterogênea, janelas de tempo flexíveis e tempos de viagem estocásticos, sendo representada por uma formulação de programação estocástica inteira mista de dois estágios com recurso. Os tempos de viagem estocásticos são aproximados por um conjunto finito de cenários, gerados por um algoritmo que os descreve por meio da distribuição de probabilidade Burr tipo XII, e uma matheurística de busca local granular é sugerida para a resolução do problema. Extensivos testes computacionais são realizados em instâncias da literatura, e as vantagens das janelas de tempo flexíveis e dos tempos de viagem estocásticos são enfatizadas Abstract: This work addresses three multi-attribute variants of the vehicle routing problem. The first one presents a heterogeneous fleet, hard time windows and deterministic travel times. To solve this problem, optimal algorithms based on the Benders decomposition are proposed. Such algorithms exploit the structure of the problem in a mixed-integer programming formulation, and three algorithmic enhancements are developed to accelerate them. The second variant comprises a heterogeneous fleet, flexible time windows and deterministic travel times. The flexible time windows allow limited early and late servicing at customers with respect to their hard time windows, at the expense of penalty costs. This problem is solved by extensions of the Benders algorithms, which include novel dynamic programming algorithms for the subproblems with the special structure of the traveling salesman problem with flexible time windows. The third variant presents a heterogeneous fleet, flexible time windows and stochastic travel times, and is represented by a two-stage stochastic mixed-integer programming formulation with recourse. The stochastic travel times are approximated by a finite set of scenarios generated by an algorithm which describes them using the Burr type XII distribution, and a granular local search matheuristic is suggested to solve the problem. Extensive computational tests are performed on instances from the literature, and the advantages of flexible windows and stochastic travel times are stressed. Doutorado Automação Doutor em Engenharia Elétrica CNPQ 141064/2015-3

This study considers a vehicle routing problem with time windows, accessibility restrictions on customers and a fleet that is heterogeneous with regard to capacity, average speed and cost. A vehicle can perform multiple routes per day, all starting and ending at a single depot, and it is assigned to a single driver, whose total work hours are limited. The available fleet is divided into an owned fleet, for which a variable cost is incurred, and a chartered fleet, for which only a fixed cost is incurred for each vehicle used. A column generation algorithm embedded in a branch-and-bound framework is proposed. The column generation pricing subproblem required a specific elementary shortest path problem with resource constraints algorithm to address the possibility for each vehicle performing multiple routes per day and to address the need to determine the workdays start time within the planning horizon. To make the algorithm efficient, a constructive heuristic and a learning metaheuristic algorithm based on tabu search were also developed. Both were used on branch-and-bound tree nodes to generate a good initial solution to the linear restricted master problem; particularly, to find a good initial primal bound to the branch-and-bound tree. Este estudo aborda um problema de roteirização de veículos com janelas de tempo, restrições de acessibilidade nos clientes e uma frota que é heterogênea em relação à capacidade de carga, velocidade média de deslocamento e custo. Um veículo pode percorrer múltiplas rotas por dia, todas começando e terminando em um mesmo depósito, e está designado a um único motorista, cujo total de horas trabalhadas no dia está limitado a um valor máximo. A frota disponível é dividida em uma frota própria, para a qual um custo variável é incorrido, e uma frota de freteiros, para a qual apenas um custo fixo é incorrido para cada veículo utilizado. Um algoritmo baseado em geração de colunas, integrado a um procedimento de branch-and-bound, é proposto neste estudo. O subproblema de precificação da geração de colunas requereu um algoritmo específico para o problema do caminho mínimo elementar com restrições sobre recursos capaz de lidar com a possibilidade de cada veículo percorrer múltiplas rotas por dia e capaz de lidar com a necessidade de determinar o instante de início do dia de trabalho do motorista dentro do horizonte de planejamento. Para tornar o algoritmo eficiente, uma heurística construtiva e uma heurística de melhoria baseada em busca tabu também foram desenvolvidos. Ambos são utilizados nos nós da árvore de branch-and-bound para gerar boas soluções iniciais para o problema mestre restrito da geração de colunas; particularmente, para encontrar um bom limitante primal inicial para a árvore de branch-and-bound.