A multi objective sound quality optimization of electric motor noise in hybrid vehicles

The use of hybrid and electric powertrains in passenger vehicles bring unusual narrowband auditory stimuli into the passengers compartment. When attempting to shape and control those effects, in either a passive or active approach, it is possible that one aspect of the soundscape is addressed at the expense of evidencing other unwanted noise components. Therefore, a proper handling of the different disturbance components is needed, which would take into account the intricate interrelation between those components and its cross effects on the various sound quality metrics relevant to this application. This paper discusses the performance and outcomes of two multi-objective evolutionary algorithms in dealing with the vector optimization of four sound quality metrics, namely Loudness, Roughness, Sharpness and Tonality, when applied to an electric motor driven powertrain, as perceived in the passengers compartment of a hybrid vehicle. Some criteria concerning the identification and optimization of the responsible narrowband components for the aforementioned psychoacoustic perceptions are given, together with a discussion about the parameters of the two evolutionary algorithms which guarantee convergence and diversity of solutions over the Pareto set of trade-offs, in a single solver run. An active sound quality control is implemented through computer simulations for tackling a synthesized stationary powertrain-induced noise, hence demonstrating the viability of accomplishing the desired sound quality targets devised during the vector optimization stage.