Microscopic energy consumption modelling of electric buses: model development, calibration, and validation. (September 2021)

• Record Type: Journal Article
• Authors: Fiori, Chiara; Montanino, Marcello; Nielsen, Sune; Seredynski, Marcin; Viti, Francesco
• Is Part Of: Transportation research. Volume 98(2021)
• Publication Date: 2021-09
• Subjects: Electric bus -- Energy consumption -- SOC -- Calibration -- Validation -- Eco-driving; Transportation -- Research -- Periodicals; Transportation -- Environmental aspects -- Periodicals; 354.76
• DOI: 10.1016/j.trd.2021.102978
• Languages: English
• ISSNs: 1361-9209

Abstract

Highlights: A microscopic power-based energy consumption model of electric buses is proposed. The model leverages on the analysis of e-Bus data from two months of operations. The simulation of auxiliary systems energy consumption is integrated in the model. The model was calibrated against field data, yielding an average error of 2.5%. Model validation was conducted at both the microscopic and macroscopic levels. Abstract: In this paper a microscopic energy consumption model of electric buses (e-Buses) is developed, leveraging on a vast data collection campaign of e-Bus daily operations. The model is composed of two modules, which estimate the instantaneous power required for traction and the average power required by the auxiliary systems. The model simulates the instantaneous energy consumption as a function of line and vehicle characteristics, and operating conditions including ambient temperature, bus load and encountered traffic. The model was calibrated against trajectory data from 435 trips, showing an error in reproducing the instantaneous battery state-of-charge of 5.5% in 95% of experiments (2.5% on average). A two-level model validation was performed against data from 110 trips. At a microscopic level, results proved that the calibrated model is robust against variability of line operating conditions. At a macroscopic level, the model was validated against the macroscopic speed-consumption function, showing an error of 6%.