Artificial neural network based detectability prediction of synthesized exterior electric vehicle sounds

Research output: Contribution to conferencesPaperContributedpeer-review



The safety for traffic participants like pedestrians and cyclists might be effected by the current electrification of vehicles. Detectability of electric vehicle sounds is an important attribute for safety reasons. The aim of this work is to determine the detectability of different electric vehicle sounds for a constant speed, single car pass-by situation. For this purpose, the differences in detection time are investigated with perception studies. The correlation between physical-psychoacoustical parameters and detection time estimations obtained from jury testing is also investigated in this study. Moreover, an artificial neural network (ANN) is also used as a prediction tool of detectability estimations for further evaluations of different possible stimuli. Lastly, advantages and shortcomings of using ANNs for detectability estimations are also discussed.


Original languageEnglish
Publication statusPublished - 2017


Title46th International Congress and Exposition on Noise Control Engineering
SubtitleTaming Noise and Moving Quiet
Abbreviated titleINTER-NOISE 2017
Conference number46
Duration27 - 30 August 2017
CityHong Kong

External IDs

ORCID /0000-0002-0803-8818/work/142257113


ASJC Scopus subject areas


  • Detectability, Electric vehicle, Neural networks, Psychoacoustics, Safety, Sound quality