Modeling and Mitigating the Cycle Aging Cost of Vehicle Batteries in Energy Transportation Nexus
Shuangqi Li, Pengfei Zhao, Chenghong Gu, Siqi Bu, Jianwei Li, Shuang Cheng
Abstract
There is a critical need to protect batteries from degradation to achieve the benefit of electric vehicles (EVs) in the energy and transportation sectors. However, existing battery degradation models have either failed to reflect the actual battery cycle aging mechanism, or are too complex for practical vehicle energy management. This paper presents a new approach for the efficient integration of battery cycles aging mechanisms into vehicle energy management by a prioritized experience-driven imitative learning (PExp-IL) framework. A prioritized degradation experience pool is constructed by comprehensively analyzing battery cycle aging mechanisms in vehicle standard driving schedules. Battery cycle aging features are mathematically presented by learning the expert experience that characterizes its degradation under different working conditions imitatively. The battery protective target is achieved by integrating the established gradient aging model into a vehicle energy management optimization problem. With the anti-aging policy gradient, battery aging mitigation can be easily incorporated into vehicle energy management scenarios that highly depend on algorithm computation efficiency. Two detailed demonstrative cases are further constructed to realize the anti-aging battery management in hybrid vehicles and grid integration of EVs. This approach provides a new practical solution for improving vehicle total economy by mitigating battery aging costs, which can contribute to net zero in the energy-transportation nexus.