A Bi-Level Genetic Algorithm for Budget-Constrained Dynamic Wireless Charging Lanes Deployment

To reduce greenhouse gas emissions from road transport, many countries are promoting electric vehicles (EVs). However, as of 2023, global EV adoption remains low (3%) due to limited range, insufficient charging infrastructure, and high costs. Dynamic wireless charging (DWC) offers a potential solution by charging EVs in motion via electromagnetic coupling, thereby extending range and reducing battery size. Given DWC’s high construction cost, optimal lane placement is a key challenge. This study proposes a bi-level model: the upper level selects DWC lanes to minimize total system cost (construction and travel time) while ensuring trip completion, while the lower-level models user equilibrium flow considering vehicle range and stochastic initial states of charge. The results showed that installing DWC along major overlapping corridors used by multiple origin–destination pairs improved resource efficiency and traffic balance. In contrast, dispersed placements led to detours and higher travel costs. As budget constraints tighten, achieving equilibrium flow becomes more difficult, underscoring the interactions between model levels. In low-budget cases, increasing the iteration count is advised to improve solution quality. These results provide concrete guidance for governments and infrastructure providers on prioritizing high-traffic corridors and planning DWC deployment within budget constraints.