Optimisation of opencast mining machinery noise using differential evolutionary algorithm

Abstract

Noise levels produced by various mining equipment's are high and exposure to such levels is considered as a severe problem. The aim of this paper is to develop and analyse the ability of a differential evolution (DE) algorithm to locate global optima of the far field noise levels produced by mining machineries in the mine. The objective function formulated is maximisation of sound pressure level (SPL) so as to determine optimal distance, optimal directivity index, optimal sound power level (SWL) and other optimal attenuation parameters. The most essential challenge in optimisation problems is CPU time. Comparison with the best known variants of DE over the objective function reflects the superiority of the parameter tuning scheme in terms of accuracy, convergence speed and robustness. Results show that DE/RAND/2 is able to converge, find optimum values faster compared to other mutation variants and seems to be a promising approach for machinery noise optimisation problems.