Entity

Time filter

Source Type

Montréal, Canada

Hajji A.,Laval University | Gharbi A.,CIRRELT | Artiba A.,University of Valenciennes and HainautCambresis
2011 4th International Conference on Logistics, LOGISTIQUA'2011 | Year: 2011

This paper considers a stochastic optimal control problem of unreliable three stages manufacturing systems. The supplier and the transformation stage are both subject to random events. Moreover, due to the periods of unavailability of the supplier, a random delay could postpone the reception of the order. Our objective is to find a control policy for the supply and production activities that minimizes the incurred cost and to propose a practical approach aiming to evaluate and quantify the control policy. Stochastic dynamic programming and numerical methods combined to a simulation based approach are thus proposed to achieve a close approximation of the production and supply policy. To illustrate the usefulness of the combined approach extensions to cover more complex systems, were optimal control theory may not be easily used, are developed and analyzed. To illustrate the practical usefulness of the approach, an application aiming to develop a quantitative tool to help establishing and negotiating order costs is presented. © 2011 IEEE. Source


Perboli G.,Polytechnic University of Turin | Crainic T.G.,CIRRELT | Tadei R.,Polytechnic University of Turin
IEEE International Conference on Automation Science and Engineering | Year: 2011

In this paper, we introduce GASP - Greedy Adaptive Search Procedure, a metaheuristic able to efficiently address two and three-dimensional multiple container packing problems. GASP combines the simplicity of greedy algorithms with learning mechanisms aimed to guide the overall method towards good solutions. Extensive experiments indicate that GASP attains near-optimal solutions in very short computational times, and improves state-of-the-art results in comparable computational times. © 2011 IEEE. Source


Vidal T.,Pontifical Catholic University of Rio de Janeiro | Crainic T.G.,CIRRELT | Gendreau M.,Ecole Polytechnique de Montreal | Prins C.,University of Technology of Troyes
Networks | Year: 2015

Timing problems involve the choice of task execution dates within a predetermined processing sequence, and under various additional constraints or objectives such as time windows, time-dependent costs, or flexible processing times, among others. Their efficient resolution is critical in branch and bound and neighborhood search methods for vehicle routing, project and machine scheduling, as well as in various applications in network optimization, resource allocation, and statistical inference. Timing-related problems have been studied for years, yet research on this subject suffers from a lack of consensus, and most knowledge is scattered among operations research and applied mathematics domains. This article introduces a classification of timing problems and features, as well as a unifying multidisciplinary analysis of timing algorithms. In relation to frequent application cases within branching schemes or neighborhood searches, the efficient resolution of series of similar timing subproblems is also analyzed. A dedicated formalism of reoptimization "by concatenation" is introduced to that extent. The knowledge developed through this analysis is valuable for modeling and algorithmic design, for a wide range of combinatorial optimization problems with time characteristics, including rich vehicle routing settings and emerging nonregular scheduling applications, among others. © 2015 Wiley Periodicals, Inc. Source


Rancourt M.-E.,CIRRELT | Bellavance F.,HEC Montreal | Goentzel J.,Massachusetts Institute of Technology
Socio-Economic Planning Sciences | Year: 2014

Empirical research characterizing transportation markets in developing countries is scarce. By analyzing contracts between the World Food Programme and private carriers, we identify the determinants of transportation tariffs in Ethiopia and quantify their relative importance. The econometric models devised from our unique dataset provide insights for shippers to improve procurement processes, for carriers to develop competitive business models and for government authorities to define effective investments and policies. Results indicate that the number of carriers on a given lane significantly reduces transportation tariffs and that policies stimulating competition may be as important as road infrastructure investments in Ethiopia's development strategy. © 2014 Elsevier Ltd. Source


Crainic T.G.,CIRRELT | Mancini S.,Polytechnic University of Turin | Perboli G.,Polytechnic University of Turin | Tadei R.,Polytechnic University of Turin
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2011

In this paper we address the Two-Echelon Vehicle Routing Problem (2E-VRP), an extension of the classical Capacitated VRP, where the delivery from a single depot to the customers is managed by routing and consolidating the freight through intermediate depots called satellites. We present a family of Multi-Start heuristics based on separating the depot-to-satellite transfer and the satellite-to-customer delivery by iteratively solving the two resulting routing subproblems, while adjusting the satellite workloads that link them. The common scheme on which all the heuristics are based consists in, after having found an initial solution, applying a local search phase, followed by a diversification; if the new obtained solutions are feasible, then local search is applied again, otherwise a feasibility search procedure is applied, and if it successful, the local search is applied on the newfound solution. Different diversification strategies and feasibility search rules are proposed. We present computational results on a wide set of instances up to 50 customers and 5 satellites and compare them with results from the literature, showing how the new methods outperform previous existent methods, both in efficiency and accuracy. © 2011 Springer-Verlag. Source

Discover hidden collaborations