A statistical analysis to compare results of different on-road vehicle performance testing

Abstract

Assessing the energy performances of a vehicle on the road requires collecting large data samples to ensure repeatability of the measurements despite the many variables which influence them. This paper proposes a method to dimension the campaign and compares various campaigns despite differences in the influencing variables through statistical analysis. This methodology has been applied to the Italian testing campaign of Toyota Yaris Hybrid 2020. The results are compared to those collected in a similar campaign for the previous model. The pandemic of COVID-19 forced the acquisitions to be made in months in which the weather and the traffic were different, and the reduced mobility lowered the traffic intensity. The proposed method was able to make the campaigns comparable besides the differences. It can also be applied to design future campaigns to reduce the quantity of collected data while maintaining the same statistical significance. The newest model of Toyota Yaris Hybrid performs better than the previous one, and it was analysed how.