Int. J. Production Economics 112 (2008) 192–201 Multicriteria approach to rank scheduling strategies $ O. Roux a,Ã , D. Duvivier b , V. Dhaevers a , N. Meskens a , A. Artiba c a Catholic University of Mons, 151 Chausse´e de Binche, 7000 Mons, Belgium b LIL, Universite´du Littoral Coˆte d’Opale, 50, rue Ferdinand Buisson, BP 719, 62228 Calais Cedex, France c Institut Supe´rieur de Me´canique de Paris (Supmeca), 3, rue Fernand Hainaut, 93407 Saint-Ouen, France Received 6 December 2005; accepted 28 August 2006 Available online 10 April 2007 Abstract This paper deals with multicriteria decision-making applied to discrete-continuous scheduling problems. The need for reusability and modularity leads us to build a ‘‘generic’’ optimization and simulation framework, while the need to quickly generate good compromises between conflicting objectives requires the implementation of multicriteria scheduling models. This paper describes the practical possibilities of two hybrid models within this framework, the first one uses a lexicographical sort and the second one a multicriteria method to rank scheduling strategies. The two hybrids are applied to a real-life highly constrained industrial problem. r 2007 Elsevier B.V. All rights reserved. Keywords: Multicriteria decision-making; Optimization; Simulation; Scheduling 1. Introduction In the particular context of discrete or discrete- continuous production processes organized as gen- eralized jobshops, we are developing a ‘‘generic’’ framework for the optimization and simulation of real-life multicriteria planning and scheduling (Chre´tienne et al., 1995). When considering these problems, it clearly appears that exact methods are limited in the size of the problems they can handle because of their time/space requirements. The selected method must be able to overcome the difficulty of accurately simulating the details of the functioning of real-life problems in a reasonable amount of time. Consequently, our framework is based on several hybrid models combining a simulator with several optimizations tools. The scheduling unit comprises a set of widely used dispatching rules such as shortest processing time (SPT) (Chre´tienne et al., 1995), as well as dedicated dispatching rules (Duvivier et al., 2005). The simulation model is used to capture all the details of the functioning of real-life problems. Considering discrete-continuous processes, two approaches have been investigated in our framework. According to its complexity, the ‘‘continuous part’’ of the problem is either directly embedded in the simula- tion model using simple mathematical functions, or modeled using hybridization of our simulator with an external solver dedicated to this continuous part. Concerning the optimization aspects of the pro- blem, our work aims at assessing which solving ARTICLE IN PRESS www.elsevier.com/locate/ijpe 0925-5273/$ - see front matter r 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.ijpe.2006.08.020 $ This work is subsidized by the Walloon Region of Belgium and the European Social Fund. The authors would like to thank the staff at Fontainunion who contributed to this work. Ã Corresponding author. Fax: +32 65323363. E-mail address: roux@fucam.ac.be (O. Roux).