A new approach to evaluate personalized energy consumption is presented in this paper. The method consists of identifying driver-vehicle dynamics using the probability weighted autoregressive model, which is one of the multi-mode ARX models, and then of reproducing the driver-vehicle behavior in a vehicle-following task. The energy consumption of the vehicle is estimated from the velocity profile calculated by using the driver-vehicle model. In this paper, driving simulator and real-world driving data were recorded to identify the driver-vehicle model in various situations. As a result, real-world energy consumption could be reproduced in a variety of situations with an average error of 1.9% and a standard deviation within 1.5%. Several promising applications of the energy consumption evaluation are introduced in this paper, such as an online energy consumption prediction, a powertrain choice-assistance system for car buyers, and a solution to estimate the macroscopic energy consumption of aggregated vehicles in a traffic flow.
Energy Consumption Evaluation Based on a Personalized Driver-Vehicle Model
IEEE transactions on intelligent transportation systems ; 18 , 6 ; 1468-1477
2017
Article (Journal)
English
Personalized Prediction of Vehicle Energy Consumption Based on Participatory Sensing
Online Contents | 2017
|