Transit Bus Simulation

Overview

The goal of this repository is to create a side-by-side comparison of transit buses with combustion engines with those that are electric to advise policy in Mexico City.

Obtaining all the data was completed without any on-vehicle sensors. The process started by obtaining the shapefile route of each transit bus in Mexico City. These were then projected into a different coordinate refrence system to make them easier to work with [CRS Projection]. Then, elevation for points across the route were queried to calculate the gradient [Gradient Querying]. Next, traffic and speed data were utilized to create a speed profile of the area during different seasons and times of the day [Speed Querying]. This profile and the gradient data were then spatially joined and used to create a complete drive cycle with a one second resolution [Cycle Generation]. Finally, using these drive cycles along with bus specifications, a simulation software named FASTSim was employed to predict the difference in energy consumption profiles of electric and combustion engine along these routes [Simulation].

Data Sources

The data used in this project comes from the following sources:

• Bus Route Shapefiles [add source]
• Waze Speed Data [https://wazeopedia.waze.com/wiki/USA/Traffic_data]

The software used in the project can be found here:

• Google Maps Elevation API [https://developers.google.com/maps/documentation/elevation]
• FASTSim [https://www.nrel.gov/transportation/fastsim.html]

Development Setup

macOS

1. Clone this repository and cd inside.
2. Ensure python 3.7.9 is available on the systeem via pyenv.

   brew install pyenv
   pyenv local 3.7.9

3. Create a python virtual environment and jupyter kernel.

   # Create the virtual environment, in which the package versions specified in the Pipfile will be installed
   pipenv install --python 3.7.9
   
   # Activate the virtual environment
   pipenv shell
   
   # Create a jupyter kernel tied to the active python virtual environment
   python -m ipykernel install --user --name=tbs-kernel

4. Start the jupyter lab: jupyter lab
   • The first time you do this, you'll need to specify which kernel to use via the jupyter lab browser UI. TODO include screenshot.

Windows

TODO