A cost-effective solution to estimate fuel consumption and greenhouse gas emissions for motorcycles: a case study

ABSTRACT

The lack of country-specific data related to the real-world fuel consumption (FC) of motor vehicles is a significant impediment to the inventory of greenhouse gas (GHG) emissions in the transport sector in Vietnam. This study aims to develop a cost-effective solution for estimating the on-road FC of motorcycles (MC) to overcome this lack. An Artificial Neural Network (ANN)-based model (R² >0.85 and MAPE < 20%) was developed to predict the MC’s instant FC rate (FR_inst) based on the experimental data that were continuously collected using the data logger installed directly on the test MC. The developed ANN-based model was utilized to estimate FR_inst according to the real-world driving characteristics that were collected using the Global positioning system (GPS) device. The discrepancy between the estimated and measured FC (in terms of liters per kilometer) was only 6.8%. Consequently, the proposed approach could help improve the consumed fuel-based GHG emission inventory in the transport sector.

KEYWORDS:

• ANN
• fuel consumption
• GPS
• GHG
• motorcycle
• Hanoi

Nomenclature

AF	=	Activation function
ANN	=	Artificial neural networks
CSEF	=	Country-specific emission factor
EF	=	Emission factor
FC	=	Fuel consumption
$\overline{\mathrm{F}\mathrm{C}}$	=	Average fuel consumption for each trip
FRinst	=	Fuel consumption rate
G	=	Grade
GHG	=	Greenhouse gas
GPS	=	Global positioning system
GWP	=	Global warming potential
IPCC	=	Intergovernmental Panel on Climate Change
MAPE	=	Mean absolute percentage error
MC	=	Motorcycles
Qfuel	=	Lower heating value of gasoline
R	=	Correlation coefficient
R2	=	Determination coefficient
RMSPE	=	Root mean squared percentage error
ρfuel	=	Gasoline density
SAFDdiff	=	Speed – acceleration distribution deviation
TA	=	Training algorithm
TFC	=	Total consumed fuel
Vinst	=	Instant speed
VKT	=	Traveled kilometers
VSP	=	Vehicle-specific power

Acknowledgements

Khanh Nguyen Duc was funded by the Master, PhD Scholarship Programme of Vingroup Innovation Foundation (VINIF), code VINIF.2022.TS058.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research is funded by University of Transport and Communications (UTC) under grant number T2023-MT-003.

Notes on contributors

Yen-Lien T. Nguyen

Yen-Lien T. Nguyen is a Doctor at the Faculty of Transport Safety and Environment, University of Transport and Communications. She is an expert in environmental and transportation issues.

Khanh Nguyen Duc

Khanh Nguyen Duc is a lecturer at the School of Mechanical Engineering, HUST. He is an expert in electronic control, engine upgrades, alternative fuels, and transportation.

Anh-Tuan Le

Anh-Tuan Le is a Professor at the School of Mechanical Engineering, HUST. He is an expert in alternative fuels, conventional and electric vehicles.

Hai-Yen T. Than

Hai-Yen T. Than works at the Faculty of Transport Safety and Environment, University of Transport and Communications. Her search tendency is environmental pollutants from vehicles.

Quy Cao Minh

Quy Cao Minh is a Doctor at the Faculty of Transport Safety and Environment, University of Transport and Communications. He is an expert in transportation environment management.