Dynamically Modeling Bus Routes for Informed Transit and Electric Vehicle Development

Abstract

Metropolitan Transit Agencies incorporating Battery-Electric Buses face significant hurdles when developing these programs due to the uncertainty and variability inherent to mass transit and the substantial energy requirements involved. To expedite the process and support transit electrification, I have developed an open-source python-based modeling tool called reRoute_Dynamics that couples Longitudinal Dynamics Models and GIS data to generate realistic energy storage load profiles. I validated this model through comparing collected GPS data to aggregated Monte Carlo Simulations for several transit routes in King County, Washington. I then demonstrated that this tool is capable of providing information transit agencies pursue in their pilot programs - how vehicle energy mileage depends on different aspects of a trip, route, and geography, as well as provide insights into the cell-level behavior of the energy storage system. Geography, vehicle mass, and driving behavior are all significant driving factors that reduce vehicle mileage from a manufacturer’s quote, conclusions that are validated by existing literature. This work provides a tool and framework for electric vehicle transit research, and will enable more transit agencies to pursue electric vehicle pilot programs at lower cost.