The design by multi-objectives optimization implies the optimization of several contradictory objectives simultaneously. In fact there is no optimal solution for one particular objective if the other objectives are considered, but the aim is to simultaneously minimize all the objectives in order to reach an optimal compromise. Optimum is reached if any improvement of one objective induces the degradation of one other. Such an optimum is located on a front called Pareto front. The Pareto front, a set of optimal solutions that are not equivalent, allows us to choose an optimal solution with criteria external to optimization process (economic or functional). In this study, a multi-objective particle swarm optimization (a metaheuristic) algorithm has been used to optimize a wood plastic composite for decking application. This metaheuristic, based on evolutionary techniques, applies to a great diversity of functions objectives: continuous or discrete equations, qualitative knowledge rules and algorithms. The design variables are mainly variables of raw materials production, and the incorporation of a biopolymer, the control of timber particle sizes and chemical or thermal timber changes. The objective functions are equations and an algorithm integrating discrete data in the modelling of creep behavior, water resistance and fossil resources depletion.< Leer menos