Shared autonomous electric vehicle system design and optimization under dynamic battery degradation considering varying load conditions

Shared autonomous electric vehicles (SAEVs) that combine electric vehicles (EVs) with autonomous driving technologies and car-sharing services are expected to become a significant component of future transportation systems. However, the issue of battery degradation, which directly affects performance and efficiency, may limit their widespread adoption. To address this issue, this study proposes a comprehensive framework for SAEV system design and optimization in accordance with the dynamic nature of battery degradation under varying load conditions. The framework incorporates a semi-empirical method, which is a predictive approach capable of estimating capacity changes based on factors such as temperature, C-rate, cycle number, and depth of discharge (DOD) for each cycle. The analysis confirmed that battery degradation significantly affected the actual simulation and system design outcomes, which further highlights the significance of dynamic input conditions. Notably, the DOD limit and charging power were identified as the key design variables, warranting separate scenario analyses because of their critical impact on battery degradation. When battery degradation was considered and both the DOD and charging power were used as design variables, the total cost was reduced by 51.3% ($776,708) compared to the case where battery degradation was not considered. Furthermore, the necessary adaptations in the SAEV design and operation under different scenarios were determined. For this, additional parametric studies were conducted by altering various operating conditions, such as the waiting time constraints, fleet cost, and battery degradation rate. The accurate prediction of battery degradation using the current model provides valuable insights and guidance for SAEV design and operation, thereby ensuring longevity and optimized performance.