Acoustic Optimisation at Workplaces Based on Sound Quality: The Case of Stationary Noise Signals

Numerical optimisation procedures offer effective methods to analytically forecast the modifications of the system being examined. This paper describes the use of a multiobjective optimisation code in order to identify solutions which lead to an improvement of the sound quality at the operator station of earth moving machinery in stationary conditions. For this purpose, it seems appropriate to focus on loudness and sharpness parameters as previous studies showed that they are the ones which correlate most closely to subjective annoyance judgments. The 1/3 octave band sound pressure levels recorded at the operator position were used to represent the system, numerical modules for the correct calculation of the psychoacoustic parameters were specifically developed and a multi-objective genetic algorithm was used to find the modifications of the input system which lead to the minimisation of the above psychoacoustic parameters. In order to verify whether the optimisation process leads to noise signals subjectively considered less annoying than the original one, listening tests were carried out using both the original and some optimised signals. The results confirmed the suitability of the optimisation procedure and also the subjective relevance of the simultaneous reduction of loudness and sharpness.