Electric Vehicles Can Now Charge While Driving

Electric vehicles face a major hurdle: long charging times and limited charging stations can force drivers to detour and wait, especially on long trips. This is critical as transportation accounts for 28% of U.S. greenhouse gas emissions, with over 94% from fossil fuels, and policies like the European Parliament's ban on new internal combustion engines by 2035 push for faster EV adoption. A new study introduces a dynamic charging solution that could transform how EVs power up, using modular vehicles that connect and share electricity while driving, potentially cutting energy use and travel delays significantly.

The researchers developed a system called platoon-based vehicle-to-vehicle charging (PV2VC), where electricity request vehicles (ERs) with low batteries can be charged by electricity supplier vehicles (ESs) while moving together in attached platoons. This approach avoids the need for ERs to detour to stationary charging stations, which typically take 20-30 minutes for DC fast chargers or 4-6 hours for Level 2 chargers, and often involve nonlinear charging rates that slow above 80% battery. In tests, the PV2VC technology saved up to 11.07% in energy consumption, 11.65% in travel time, and 11.26% in total cost compared to a benchmark scenario where vehicles only charge at fixed stations. For example, in an illustrative case with two ERs and one ES, ERs reduced electricity use by 34.55% and 23.85% and travel time by 27.27% and 15.38%, respectively, by charging on the move instead of stopping at a station.

To evaluate this technology, the team formulated a mixed integer linear programming (MILP) model that optimizes routes and charging schedules for ERs and ESs, minimizing total energy use and travel time. They compared three operational scenarios: the electric vehicle routing problem (EVRP), where vehicles charge only at stationary locations; the electric vehicle platooning problem (EVPP), where vehicles can platoon to save energy but still charge at stations; and the PV2VC scenario. The model includes constraints for vehicle flow, arrival times, platoon synchronization, and energy conservation, with assumptions like a constant 10% energy saving from platooning and a 90% efficiency for power transfer between vehicles. For large-scale problems, they proposed a customized genetic algorithm (GA) with solution representations in multiple layers, such as visiting sequences and dwell times, and genetic operators to modify charging locations and platoon assignments.

Numerical experiments on a modified Sioux Falls network with five scenarios involving 2 to 4 ERs and 1 to 2 ESs showed the GA's effectiveness, achieving an average optimality gap of 0.34% compared to commercial solver Gurobi, with much faster computation times. For instance, in scenario S5 with 4 ERs and 2 ESs, the GA found solutions with 11.26% total cost savings from the EVRP benchmark, while the solver timed out after two hours. , detailed in Table 8 and Figure 9, indicate that PV2VC outperforms EVPP, which only saved up to 3.65% in total cost, as platooning alone often adds wait times for synchronization. The PV2VC benefits were most pronounced for long-distance routes with low initial battery levels, sparse charging facilities, and when travel time is valued higher than energy costs, such as in emergency situations.

Of this research are substantial for real-world EV adoption, offering a way to reduce range anxiety and charging infrastructure gaps. By enabling charging on the move, PV2VC could make EVs more practical for long trips and in areas with limited stations, potentially lowering operational costs and emissions. However, the study notes limitations: the MILP model and GA assume constant parameters like platoon savings and transfer efficiency, which may vary in dynamic environments, and real-world implementation would require advanced modular vehicle technology and coordination. Future work could explore larger-scale tests, dynamic online applications, integration with vehicle relocation, and combining ES services with other tasks like deliveries to enhance feasibility.