Destination labels for battery electric vehicles in eVMT dataset
Data files
Mar 13, 2024 version files 8.27 MB
-
bev_trips_labels.csv
-
LSTM-1087_final_-lstm_final.py
-
LSTM-1091_final_-lstm_final.py
-
LSTM-1092_final_-lstm_final.py
-
LSTM-1093_final_-lstm_final.py
-
LSTM-1094_final_-lstm_final.py
-
LSTM-1098_final_-lstm_final.py
-
LSTM-1100_final_-lstm_final.py
-
LSTM-1109_final_-lstm_final.py
-
LSTM-1111_final_-lstm_final.py
-
LSTM-1112_final_-lstm_final.py
-
LSTM-1122_final_-lstm_final.py
-
LSTM-1123_final_-lstm_final.py
-
LSTM-1125_final_-lstm_final.py
-
LSTM-1125a_final_-lstm_final.py
-
LSTM-1127_final_-lstm_final.py
-
LSTM-1131_final_-lstm_final.py
-
LSTM-1132_final_-lstm_final.py
-
LSTM-1135_final_-lstm_final.py
-
LSTM-1137_final_-lstm_final.py
-
LSTM-1140_final_-lstm_final.py
-
LSTM-1141_final_-lstm_final.py
-
LSTM-1143_final_-lstm_final.py
-
LSTM-1144_final_-lstm_final.py
-
LSTM-1146_final_-lstm_final.py
-
LSTM-1217_final_-lstm_final.py
-
LSTM-1253_final_-lstm_final.py
-
LSTM-1257_final_-lstm_final.py
-
LSTM-1260_final_-lstm_final.py
-
LSTM-1267_final_-lstm_final.py
-
LSTM-1271_final_-lstm_final.py
-
LSTM-1272_final_-lstm_final.py
-
LSTM-1279_final_-lstm_final.py
-
LSTM-1280_final_-lstm_final.py
-
LSTM-1281_final_-lstm_final.py
-
LSTM-1285_final_-lstm_final.py
-
LSTM-1288_final_-lstm_final.py
-
LSTM-1294_final_-lstm_final.py
-
LSTM-1295_final_-lstm_final.py
-
LSTM-1296_final_-lstm_final.py
-
LSTM-1304_final_-lstm_final.py
-
LSTM-1307_final_-lstm_final.py
-
LSTM-1352_final_-lstm_final.py
-
LSTM-1353_final_-lstm_final.py
-
LSTM-1357_final_-lstm_final.py
-
LSTM-1367_final_-lstm_final.py
-
LSTM-1368_final_-lstm_final.py
-
LSTM-1370_final_-lstm_final.py
-
LSTM-1371_final_-lstm_final.py
-
LSTM-1375_final_-lstm_final.py
-
LSTM-1376_final_-lstm_final.py
-
LSTM-1384_final_-lstm_final.py
-
LSTM-1388_final_-lstm_final.py
-
LSTM-1393_final_-lstm_final.py
-
LSTM-1403_final_-lstm_final.py
-
LSTM-1409_final_-lstm_final.py
-
LSTM-1412_final_-lstm_final.py
-
LSTM-1414_final_-lstm_final.py
-
LSTM-1419_final_-lstm_final.py
-
LSTM-1421_final_-lstm_final.py
-
LSTM-1422_final_-lstm_final.py
-
LSTM-1423_final_-lstm_final.py
-
LSTM-1424_final_-lstm_final.py
-
LSTM-1427_final_-lstm_final.py
-
LSTM-1429_final_-lstm_final.py
-
LSTM-1435_final_-lstm_final.py
-
README.md
Abstract
This dataset provides detailed information on the destination labels for the trip trajectory and charging of 65 Battery Electric Vehicles in California in the eVMT dataset. This dataset includes the Tesla Model S and Chevrolet Bolt only. Additionally, the repository contains a Python script that trains a deep-learning model to predict the driving behavior of the drivers in this dataset. The aim of this model is to forecast the charging needs of the drivers, so that they can align their charging needs with renewable energy resources availability. This way, the impact of fossil fuel resources in charging the vehicle can be decreased, and carbon emission per mile driven can be reduced.
README: eVMT dataset
https://doi.org/10.5061/dryad.6wwpzgn60
Description of the data and file structure
The research conducted in this study involved configuring models using a subset of data from the eVMT project's dataset. This dataset was created as part of a California-wide study that spanned five years (2015-2020). The primary objective of the study was to gain insights into the driving and charging behaviors of battery-electric vehicles. Data was collected from approximately 400 households and 800 vehicles, out of which 132 were BEVs. The BEV dataset includes around 182,000 trips and 39,000 charges from 132 EVs, and it provides second-by-second on-road data. The data logger recorded important driving and charging characteristics like speed and GPS coordinates at a second-by-second interval. For this study, a subset of data from the 24 Chevrolet Bolts and 42 Tesla Model S vehicles was selected for training and testing the LSTM model. However, due to the data management agreement and privacy of the vehicle's owner, only the destination label is shareable with the public.
eVMT Dataset Highlights:
- Study Period: 2015-2020.
- Focus: Insights into BEV driving and charging behaviors.
- Data Source: About 400 households with 800 vehicles, including 132 BEVs.
- Data Points: Approximately 182,000 trips and 39,000 charging sessions from 132 BEVs.
- Details: Second-by-second on-road data, including speed and GPS coordinates.
Available Data:
- Due to privacy agreements, only the destination label of trips and charging sessions is presented in this repository.
- Subset Used: Data from 24 Chevrolet Bolts and 42 Tesla Model S vehicles.
This Repository Contents:
- Destination Labels: CSV files containing trip and charging session destination labels for 65 BEVs.
- Python Scripts: Training and testing models for each vehicle in the dataset.
About the CSV Dataset:
Contents: Destination labels for vehicle trips including trip IDs, vehicle models, clustering results, and actual labels.
Clustering: Uses DB Scan Analysis to group GPS coordinates into 'Home', 'Work', or 'Other Location'.
Python Scripts Details:
- Files Naming: Each script is named after the vehicle number, e.g.,
LSTM-1087_final-lstm_final.py
. - Matching: Each vehicle number in the CSV file corresponds to a specific Python script.
Methods
This dataset was collected by the Electric Vehicle Research Center at the University of California, Davis (2015-2020). A data logger was connected to each vehicle to gather information about the trip trajectory and charging details of the vehicles. The destination labels were generated using the DB-Scan clustering technique to cluster the destinations based on their number at each location within a certain radius into Home, Work, and Other locations.