Real-World Autonomous Driving: Exploring Energy Dynamics in Gradually Evolving Traffic Flows

Autonomous vehicles (AVs) in real-world traffic environments increase the complexity of traffic flow and energy consumption, due to the combined effects of powertrain configurations, driving behaviors, and traffic propagation. This study developed a mixed platoon energy consumption estimation model (MP-ECEM) for autonomous and human-driven vehicles to examine energy consumption across four scenarios, integrating both ideal and real-world driving behaviors, and accounting for abrupt and gradual platoon formations. The results indicated that gradual platoon intensity causes broader power fluctuations due to complex vehicle interactions, and real-world power distribution is smoother compared to theoretical models. The energy consumption exhibits a more stable variation pattern, reflecting real-world traffic dynamics. For different powertrains, the potential energy savings are 11.4%, 18.1%, and 41.8% for internal combustion engine vehicle, recovery electric vehicle, and non-recovery electric vehicle, respectively, under gradual platoon intensity with a 0.5 AV penetration rate. In contrast, the actual savings based on testing AVs data in China are 2.9%, 4.7%, and 11.5%. Additionally, AVs have a greater impact on energy consumption at low speeds than at high speeds, where air resistance is the dominant factor. Regional energy performance is comparable between China and the United States, both slightly outperforming Europe. This study offers key insights into the effects of AV technology on energy consumption across powertrain systems and regions in real-world conditions.