A concurrent estimation framework for multiple aging parameters of lithium-ion batteries for eVTOL applications

Electric vertical take-off and landing (eVTOL) aircraft has emerged as one of the promising platforms for next-generation low-altitude aircraft due to the low noise and the free emission. To guarantee safety and reliability during flight, it is crucial to accurately estimate the key parameters of eVTOL batteries. However, the continuously high-rate current during the flight mission poses significant challenges, hindering the direct application of existing battery state estimation algorithms from terrestrial electric vehicles to eVTOL applications. In addition, most state of health estimation methods for eVTOL applications lack the in-depth understanding of battery aging, such as the battery degradation modes (DMs). To overcome the above issues, this paper presents a concurrent estimation framework for multiple aging parameters of eVTOL batteries using the specific flight data. First, the evolution of measurements during flight throughout the aging is investigated, and the coupling relationships among battery parameters are analyzed. Secondly, the main DMs are calculated based on differential voltage curves. Then, the impacts of different flight scenarios on battery DMs are discussed. Thirdly, three measured sequences, including the current, the voltage, and the integral current, are selected to construct the input matrix through the correlation analysis and the consistency evaluation. Subsequently, a multi-parameter co-estimation model is trained by a bottleneck-architecture residual network. Lastly, a publicly available eVTOL battery dataset is employed to verify the effectiveness and the generalizability of the proposed method. The results show that percentage versions of the mean absolute error and the root mean squared error are within 2.5 % and 3.0 %, respectively.