This repository contains the open-source dataset used in the paper:
Liangkai Liu, Wei Li, Dawei Wang, Yi Wu, Ruigang Yang, and Weisong Shi, Fuel Rate Prediction for Heavy-Duty Trucks, accepted to IEEE Transactions on Intelligent Transportation Systems, March 2023.
The dataset includes sensor data from heavy-duty trucks, which can be used for fuel rate prediction and other research purposes.
The FEAD dataset is collected through Engine Management System (EMS) and Instant Fuel Meter (IFM) devices. We choose an off-the-shelf CAN bus parser to read fuel consumption from truck's EMS with 10mL measurement resolution and 6.1% - 6.5% error. Regarding the IFM, which provides more accurate fuel measurement, we use Onosokki's FP-2140H fuel meter with 0.1mL measurement resolution and 0.2% error.
The dataset consists of the following data:
- Engine’s status information (engine speed, torque, throttle, fuel rate, etc.) for every 100 milliseconds
The data is collected from multiple heavy-duty trucks during different driving conditions and environments. Please refer to the paper for more details.
| Name | Time | Trucks | Rows | Features | Download |
|---|---|---|---|---|---|
| EMS dataset-1 | 12/2019 | 9 | 10,273,969 | EMS engine data, latitude, longitude, triggertime, city, road level, etc. | link |
| EMS dataset-2 | 04/2020 | 29 | 26,145,539 | EMS engine data, latitude, longitude, triggertime, city, road level, etc. | link |
| IFM dataset | 06/2020 | 1 | 872,844 | IFM engine data | link |
If you use this dataset in your research, please cite the following paper:
@article{liu2023fuel,
title={Fuel Rate Prediction for Heavy-Duty Trucks},
author={Liu, Liangkai and Li, Wei and Wang, Dawei and Wu, Yi and Yang, Ruigang and Shi, Weisong},
journal={IEEE Transactions on Intelligent Transportation Systems},
year={2023},
month={March}
}