Resilience-Oriented Predictive Energy Management and Route Adaptation for Long-Distance Electric Vehicles Under Charging Infrastructure and Road Network Uncertainties
The increased deployment of electric vehicles (EVs) has increased the demand for intelligent energy management strategies to ensure reliable and efficient operation during long-distance travel. Existing EV energy management methods are mainly concerned with energy optimization of the batteries and charging scheduling; however, they often assume reliable charging infrastructure and fixed travel routes. In real-world environments, charging stations can be congested, unavailable, or out of service, and traffic incidents and road closures can affect route feasibility and energy usage. These uncertainties may lead to energy depletion (vehicle stranding) and may compromise successful trip completion. Therefore, this paper proposes a resilience-oriented predictive energy management and route-adaptation framework for long-distance EVs operating under uncertainties in charging infrastructure and the transportation network. The proposed unified decision-making framework incorporates the ability to predict the battery state of charge, estimate the energy requirement of an EV, and determine the availability of charging stations, traffic conditions, disruptions in the road network, and regenerative energy recovery opportunities. A safety-constrained predictive controller is developed to dynamically manage energy consumption while maintaining a minimum energy reserve that guarantees access to feasible charging alternatives under adverse operating conditions. Further, real-time route adaptation is achieved via continuous monitoring of charging station conditions, projected wait times, and road network conditions to determine the most energy-efficient and resilient routes. To assess operational robustness, a resilience index is introduced to quantify the vehicle’s capability to complete a trip while satisfying energy safety constraints in the presence of infrastructure and traffic disturbances. Monte Carlo simulation results on a representative long-distance corridor demonstrate that the proposed framework achieves a 99.3% mission success rate versus 67.3% for a naive shortest-distance baseline, while reducing generalized operating costs by 21.5% and charging stops by 42% relative to a conservative full-charge baseline. These results are simulation-based case study outcomes for the representative corridor and disturbance model studied here and are not measured or expected real-world performance. The proposed approach provides a practical pathway toward resilient, intelligent, and reliable next-generation electric mobility systems.