Tonal Annoyance Metric Development for Automotive Electric Vehicles

Historical metrics intended to drive the development of vehicle powertrains have focused on sounds that are characteristic of IC engines. The interior noise contribution of the propulsion system in electric vehicles has significantly more tonal noise (and much less impulsive and broadband noise) than their IC engine counterparts. This tonal noise is not adequately represented by current propulsion systems metrics. While metrics exist today that were developed to represent the presence of tones in sounds most have focused on the level aspect of the tones relative to the surrounding noise or masking level, some examples include tonality, tone-to-noise ratio, and prominence ratio. A secondary, but also important aspect of tones is the annoyance as a function of frequency. This paper will highlight the development of a tonal annoyance weighting curve that can be used to account for the frequency aspect of tonal annoyance relative to electric vehicles. This weighting curve can then be used in conjunction with the current tonal metrics to represent perceived tonal annoyance of electric vehicles. The benefit of the tonal annoyance weighting curve is demonstrated by comparing the correlation of a common commercially available tonality metric before and after the weighting curve was applied to a subjective study of 8 electric vehicle sounds with significantly different frequency content.