A methodology to determine drivetrain efficiency based on external environment

Abstract

In this paper a statistical method for establishing the efficiency of the drivetrain under different real-world usage conditions has been proposed. The method is based on real-world driving data from an electric vehicle (EV) trial conducted in the UK. It was found that the external environment (road-type and traffic) causes distinct operating regions in the drivetrain. This paper makes use of a neural network to predict the road-type and introduces two new variables (start-stop index and congestion index) to establish the external environment. Based on this external environment a new metric called frequency weighted distribution is introduced to evaluate the performance of the drivetrain. This methodology of design based on the driving environment is of importance to newer advanced powertrains such as hybrids and EVs. The end result would be a design which caters to a specific usage profile.