Shape optimization of noise barriers using genetic algorithms

This article presents a method to find optimal shapes for noise barriers by coupling a boundary element solution of the sound pressure around the barrier and an optimization process by genetic algorithms to minimize the sound pressure level in a domain behind the barrier. The objective is not to provide geometries with immediate practical applications but to estimate the improvement that could be obtained if noise barriers with improved shapes were used instead of the traditional barriers built today. The method supposes given source and receiver positions and the calculation provides an optimal shape for the barrier to reduce the sound pressure at receiver points over a specified frequency band. Different examples are presented to estimate the influence of the source and receiver positions, of the frequencies and the influence of the size of the barrier. The main conclusion is an estimate of the potential improvement of noise barriers efficiency by using better geometries.