Gilles Pesant | École Polytechnique de Montréal (original) (raw)

Papers by Gilles Pesant

Annals of Mathematics and Artificial Intelligence, 2002

A common telecommunications network design strategy, in order to protect it against failures, is ... more A common telecommunications network design strategy, in order to protect it against failures, is to achieve a decomposition into a set of rings covering the network. We propose an algorithm for the corresponding constrained cycle cover problem which draws from constraint programming and graph algorithms. Optimal solutions are obtained efciently on networks of realistic size.

Lecture Notes in Computer Science, 2004

The stretch constraint occurs in many rostering problems that arise in the industrial and public ... more The stretch constraint occurs in many rostering problems that arise in the industrial and public service sectors. In this paper we present an efficient algorithm for domain consistency propagation of the stretch constraint. Using benchmark and random instances, we show that this stronger consistency sometimes enables our propagator to solve more difficult problems than a previously proposed propagation algorithm for the stretch constraint. We also discuss variations of the stretch constraint that seem simple and useful, but turn out to be intractable to fully propagate.

Lecture Notes in Computer Science, 2003

This paper presents a constraint programming model and search strategy to formulate and solve sta... more This paper presents a constraint programming model and search strategy to formulate and solve staff scheduling problems in health care. This is a well-studied problem for which many different approaches have been developed over the years but it remains a challenge to successfully apply any given instance of a method to the various contexts encountered. We show how the main categories of rules involved may be expressed using global constraints. We describe a modular architecture for heuristic search. The resulting flexible and rather general constraint programming approach is evaluated on benchmark problems from different hospitals and for different types of personnel.

Journal of Computer Science, 2005

Lecture Notes in Computer Science, 2008

... Département de génie informatique et génie logiciel École Polytechnique de Montréal CP 6079, ... more ... Département de génie informatique et génie logiciel École Polytechnique de Montréal CP 6079, succ. Centre-ville Montreal, Canada H3C 3A7 pesant@crt.umontreal.ca, claude-guy.quimper@ polymtl.ca Abstract. ... If lim n→∞ ( ∑n i=1 r3 i ) 1 3 ( ∑n i=1 σ2 i ) 1 2 = 0, ...

Constraint Programming (CP) offers a rich modeling language of constraints embedding efficient al... more Constraint Programming (CP) offers a rich modeling language of constraints embedding efficient algorithms to handle complex and heterogeneous combinatorial problems. To solve hard combinatorial optimization problems using CP alone or hybrid CP-ILP decomposition methods, costs also have to be taken into account within the propagation process. Optimization constraints, with their costbased filtering algorithms, aim to apply inference based on optimality rather than feasibility. This paper introduces a new optimization constraint, cost-regular. Its filtering algorithm is based on the computation of shortest and longest paths in a layered directed graph. The support information is also used to guide the search for solutions. We believe this constraint to be particularly useful in modeling and solving Column Generation subproblems and evaluate its behaviour on complex Employee Timetabling Problems through a flexible CP-based column generation approach. Computational results on generated benchmark sets and on a complex real-world instance are given.

Lecture Notes in Computer Science, 1997

This paper describes the combination of a well-known tsp heuristic, genius, with a constraint pro... more This paper describes the combination of a well-known tsp heuristic, genius, with a constraint programming model for routing problems. The result, genius-cp, is an e cient heuristic single-vehicle routing algorithm which is generic in the sense that it can solve problems from many di erent contexts, each with its particular type(s) of constraints. The heuristic quickly constructs high-quality solutions while the constraint model provides great exibility as to the nature of the problem constraints involved by relieving that heuristic of all constraint satisfaction concerns. We show how those two components are integrated in a clean way with a well-de ned, minimal interface. We also describe di erent routing problems on which this algorithm can be applied and evaluate its performance.

Lecture Notes in Computer Science, 2009

Context-free grammar constraints enforce that a sequence of variables forms a word in a language ... more Context-free grammar constraints enforce that a sequence of variables forms a word in a language defined by a context-free grammar. The constraint has received a lot of attention in the last few years as it represents an effective and highly expressive modeling entity. Its application has been studied in the field of Constraint Programming, Mixed Integer Programming, and SAT to solve complex decision problems such as shift scheduling. In this theoretical study we demonstrate how the constraint can be linearized efficiently. In particular, we propose a lifted polytope which has only integer extreme points. Based on this result, for shift scheduling problems we prove the equivalence of Dantzig's original set covering model and a lately introduced grammar-based model.

19th IEEE International Conference on Tools with Artificial Intelligence(ICTAI 2007), 2007

Distributed Constraint Optimization is increasingly used for problem solving by multiple agents. ... more Distributed Constraint Optimization is increasingly used for problem solving by multiple agents. However, there are situations where the system is made up of heterogeneous agents, for which the context, the structure, and the business rules define the interactions that are possible between them. As an example, supply chains are made up of interdependent business units having some form of customer-supplier hierarchical relationships. The coordination space for these hierarchical situations can be described as a tree. Therefore, we propose a distributed algorithm (MacDS) that performs discrepancy-based search which is known to perform well for centralized problems. The proposed algorithm is complete and aims at producing good solutions in a short amount of time. It allows concurrent computation and is tolerant to message delays. It has been evaluated using real industrial supply chain problems, for which it showed good performance.

Abstract We introduce a new approach for encoding STRIPS planning problems as Constraint Satisfac... more Abstract We introduce a new approach for encoding STRIPS planning problems as Constraint Satisfaction Problems: the encoding makes use of automata for modelling the dynamics of the objects involved in the domain. We describe a total order planner based on Constraint Programming that takes advantage of this encoding, using global constraints to model the automata and to reduce the search space significantly. The planner can be easily extended, using soft global constraints, in order to deal with preferences among the goals ...

ABSTRACT We study a hybrid MIP/CP solution approach in which CP is used for detecting infeasibili... more ABSTRACT We study a hybrid MIP/CP solution approach in which CP is used for detecting infeasibilities and generating cuts within a branch-and-cut algorithm for MIP. Our framework applies to MIP problems augmented by monotone constraints that can be handled by ...

The paper investigates the relationship between counting the lattice points belonging to an hyper... more The paper investigates the relationship between counting the lattice points belonging to an hyperplane and the separation of Chvátal-Gomory cutting planes. In particular, we show that counting can be exploited in two ways: (i) to strengthen the cuts separated, e.g., by Gomory classical procedure, and (ii) to heuristically evaluate the effectiveness of those cuts and possibly select only a subset of them. Empirical results on a small set of 0-1 Integer Programming instances are presented.