A methodology for the multi-objective shape optimization of thin noise barriers

Abstract The approach of this paper is based on the evolutionary multi-objective optimization (EMO) of very thin noise barrier models with improved performance idealized as single-wire designs. To assume such a simplification of reality, the dual boundary element (DBE) formulation for assessing the acoustic efficiency arises as the most appropriate strategy involving BE to avoid drawbacks associated with the exclusive implementation of the standard formulation (SBE). The 2D analysis performed in this work focuses on the simultaneous optimization of two objectives in conflict using the Non-dominated Sorting Genetic Algorithm (NSGA-II): the maximization of noise attenuation and the minimization of the amount of material used in manufacturing the barrier, represented by the overall length of its elements (this function is closely related to the final cost of the device). Under this framework, two optimization strategies are compared for each model with equal number of fitness evaluations: (1) when considering a random initial population and (2) when including the best single-objective optimal design in the initial population. The results obtained show wide and uniformly spread-out non-dominated fronts, reflected in the geometric diversity featured by optimal designs; statistical analysis confirm the advantages of the latter initial population strategy.