118 Ana Maria A. C. Rocha 1 , Marina A. Matos 2 , M. Fernanda P. Costa 3 , A. Gaspar-Cunha 4 , and Edite M. G. P. Fernandes 5 POLYMER SINGLE SCREW EXTRUDER OPTIMIZATION USING TCHEBYCHEFF SCALARIZATION METHOD AND SIMULATED ANNEALING ALGORITHM Abstract. The single screw extrusion optimal design involves the optimization of six criteria that can be efficiently handled by a weighted Tchebycheff scalarization method. The performance of the method has been analyzed for three different methods to generate weight vectors. The experimental results show that the tested strategies provide similar and reasonable solutions and supply a valuable procedure to identify good trade-offs between conflicting objectives. Keywords: Single screw extrusion, Multi-objective optimization, Tchebycheff scalarization, Simulated annealing method 1. Introduction The single screw extrusion (SSE) design is concerned with the definition of the optimal screw operating conditions and geometry in such a way that some selected criteria achieve their best values. The screw operating conditions and geometry can be established using empirical knowledge, combined with a trial-and-error approach until the desirable criteria values are attained. However, a more efficient approach is to handle the SSE design as an optimization problem. The optimization of the SSE design is a very difficult task since it deals with the optimization of several criteria that are conflicting [1,2,3,4], which means that the improvement of one criterion leads to another criterion degradation. The SSE design has been addressed in the past and the resulting multi-objective optimization problem has been solved by a multi-objective evolutionary algorithm (MOEA) named reduced Pareto set genetic algorithm (RPSGA) [2,3,5]. Most MOEA treat the multi-objective optimization problem as a whole and find the entire set of promising and desirable solutions in a single run of the algorithm. They are, in general, stochastic methods that generate, handle and mutate a population of solutions at each 1 ALGORITMI Center, University of Minho, 4710-057 Braga, Portugal, arocha@dps.uminho.pt 2 ALGORITMI Center, University of Minho, 4710-057 Braga, Portugal, aniram@live.com.pt 3 Centre of Mathematics, University of Minho, 4710-057 Braga, Portugal, mfc@math.uminho.pt 4 Institute for Polymer and Composites, University of Minho, 4800-058 Guimarães, Portugal, agc@dep.uminho.pt 5 ALGORITMI Center, University of Minho, 4710-057 Braga, Portugal, emgpf@dps.uminho.pt