Application of multi-objective optimization techniques to the design of sound diffusers
* Presenting author
Multi-objective optimization (MO) techniques have proven to be a very interesting methodology for solving design problems in different areas of engineering. Diffuser design problems are another example. When designing a diffuser, different features are to be optimized simultaneously, which are usually in opposition. It is precisely the latter that encourages the use of MO techniques. The designer chooses the particular features to be optimized independently, carries out the optimization, studies the resulting Pareto fronts, and finally chooses the optimal diffuser using his preferences a posteriori. This results in more informed and therefore more conscientious decision making. This work shows how the MO techniques can be applied in the design of diffusers through a series of examples where different characteristics are taken into account: the maximization of the diffusion coefficient at a particular frequency range or the minimization of the depth of the diffuser. In the last case, we will explore the possibility of changing not only the depth of the wells but as well their widths. The study of the Pareto fronts presented shows how the optimal diffuser to choose can be susceptible to be different depending on the characteristics most wanted by the designer.