Abstract: With the deep development of the “dual carbon” strategy and green logistics, the demand for decarbonization in the freight transportation area has become increasingly urgent. Although the widespread adoption of electric vehicles (EVs) plays a crucial role in achieving green transportation, it is still hindered by challenges such as “charging anxiety” and “the limited availability of charging piles”. In such context, the Electric Road Systems (ERS) technology, which enables vehicles to obtain electric power from the road while driving, offers a feasible solution to overcome the above bottlenecks. Compared to pure electric vehicles, hybrid electric vehicles (HEVs) not only exhibit lower emissions and higher efficiency, but also offer longer driving ranges and stronger environmental adaptability, making them an ideal choice for logistics transportation in ERS scenarios. Given that ERS will alter the energy replenishment mode of HEVs, this study constructs a route optimization model for HEVs based on ERS. The model takes into account the impact of ERS on the energy distribution of HEVs, and sets the overall objective as minimizing the sum of total electricity and fuel costs. We develop an Improved Marine Predators Algorithm (IMPA) to solve the model. The algorithm integrates improving strategies from Large Neighborhood Search (LNS) and Genetic Algorithm (GA), in which LNS could help to escape the local optima through its destroy-and-repair mechanism effectively, and GA enhances population diversity through crossover and mutation operations, thereby preventing premature convergence. IMPA leverages the fast-solving nature of MPA and incorporates these improvements to enhance the overall performance of its soulution. Thereafter, through small-scale and large-scale numerical experiments, our IMPA is compared with the original Marine Predators Algorithm (MPA), the GA, and the CPLEX 22.10 solver. The results demonstrate that IMPA outperforms MPA, GA, and CPLEX in terms of solution quality, especially in large-scale scenarios, showing its efficiency and practicality for complex route optimization problems. Additionally, we conduct sensitivity analysis experiments to reveal in-depth managerial insights: 1) ERS can effectively reduce fuel consumption and total energy costs, but its marginal benefits gradually diminish as the number of deployed ERS segments increases; 2) There is a negative correlation between battery capacity and total cost, meaning that a larger battery capacity leading to a smaller total cost; 3) in terms of the distribution strategy of ERS segments, in the early stages of deployment, an intensive distribution strategy is more effective in reducing energy costs than a uniform distribution strategy; while in the later stages of deployment, a uniform distribution strategy can achieve lower cost. Keywords: electric road system; hybrid electric vehicles; improve the marine predator algorithm; routing optimization

Key words: electric road system, hybrid electric vehicles, improve the marine predator algorithm, routing optimization