In order to achieve carbon neutrality in California by 2045, it is estimated that California will need approximately eight million light-duty (LD) electric vehicles (EVs) and 1.5 million shared EV chargers by 2030. Moreover, a specific target has been set to ensure all drayage trucks operating in the state are zero emission (ZE) vehicles by 2035. Aligning EV charging with renewable energy sources and local grid requirements demands innovative charging solutions. Achieving this target will require precise modeling to determine the number, locations, and load characteristics of the EV chargers, especially for the medium-duty (MD) and heavy-duty (HD) vehicle sectors.
This dissertation introduces new novel data-driven coordination techniques through intelligent controlled charging applied to drayage trucks operations and indirect controlled charging by proposing innovative pricing strategies for LD, MD, and HD EVs. It presents a real-time, carbon-based pricing model for EV charging. This model prioritizes periods of high renewable energy availability, considers local solar photovoltaics (PV) variations, and incorporates a Time-of-Use (TOU) pricing structure. Utilizing Machine Learning (ML), it predicts day-ahead and three-hour-ahead forecasts.
Moreover, this research extensively explores trip- and tour-patterns of battery electric trucks (BETs) in drayage operations. By developing data-driven clustering techniques, this study adjusts energy efficiency values for loaded and unloaded conditions, simulates charging scenarios to determine the state-of-charge (SOC), and presents case studies for potential en-route opportunity charging at specific power levels. An intelligently controlled charging model for BETs is also developed, aiming to optimize tour completion and reduce charging costs using TOU energy rates. Based on a one-year analysis of real-world activity data from a fleet operating at the Ports of Los Angeles and Long Beach, the effectiveness of the algorithm was validated through a sensitivity analysis comparing reserved SOC scenarios of 5%, 50%, and 80%.
The research findings conclude that intelligent charging models paired with innovative pricing strategies can go a long way to help optimize EV charging coordination techniques for EV fleets, reduce energy costs, and address delay costs associated with drayage operations. The study highlights the importance of considering the variability of renewable energy sources and the need for effective coordination between LD/MD/HD charging, port operations, and TOU charging rates.
