EV Freight Charging Problem: a Two Body Problem of Vehicle and Station

The rapid transformation from fossil fuel to battery power in land transportation amid the global warming threat raises various challenges. One such challenge is in freight transportation, which is responsible for around 6% of the total GHG emissions in the USA. However, the size and the power requirement of these MD/HD Vehicles are significantly higher than passenger vehicles. Their need for high-capacity batteries and powerful engines to transport goods for thousands of miles require a unique charging strategy and infrastructure. This multi-faceted problem of vehicles, charging stations, and power grids is the basis of this research. We approach this problem from three angles: EV, station, and their joint operations. Our first contribution is to minimize the cost of a long-distance trip with multiple charging actions and stochastic waiting times. We develop a stochastic dynamic programming based path-planning model under uncertainty and reduce waiting times by 24% while increasing the system performance by 27%. We show that battery utilization at 80% is very efficient given the trade-off between vehicle range and charging times. We further enhance the performance via V2I communication and reservation. Our second contribution is from charging station perspective. Due to excessive energy needs of MD/HD vehicles, charging stations should provide power in the order of megawatts to provide charging in a timely manner. However, supplying such power in a matter of minutes creates a huge strain on power grids. We design a green charging station and analyze its operations using a stylized model and show optimal cost regions of using battery and power grid. We then numerically analyze the long-term investment models of a charging station and perform sensitivity analysis to determine when to invest for solar panels and batteries.