Data-driven, personalized route planning for connected electric vehicles: Optimizing time, energy, and charging stops

Electric vehicles (EVs) present a sustainable alternative to traditional vehicles, yet challenges like range anxiety caused by limited driving range and inaccurate energy predictions hinder widespread adoption. This study introduces a solution using connected electric vehicles (CEVs) equipped with sensors and onboard computers for real-time data collection. The data support a personalized energy model that adapts to individual driving behaviors and environmental conditions, ensuring accurate energy consumption predictions and alleviating range anxiety. A comprehensive framework for time- and energy-optimal route planning is proposed, featuring intelligent charging station (CS) recommendations. This framework integrates factors such as ambient temperature, traffic density, road gradients, and driver-specific speed profiles to predict energy requirements for both traction and auxiliary systems. Charging station recommendations are dynamically optimized using a bidirectional search within an adapted Bellman-Ford algorithm. This search balances energy and time efficiency by evaluating energy constraints across the route, from the source to intermediate charging stations and the final destination. The method provides both energy-optimal and sub-optimal paths that consider real-time CS availability. Simulation tests on routes from Paris to surrounding cities demonstrate up to 25 % energy efficiency improvement over traditional time- or distance-optimized routes. The energy model achieves a high accuracy, with a 2.66 % error margin for State of Charge (SoC) and 0.6 kWh for energy consumption. By combining advanced predictive capabilities, real-time optimization, and dynamic CS recommendations, this research addresses critical EV adoption barriers and promotes environmental sustainability.