<p>Lithium polymer batteries serve as the initial startup power source for stable railway vehicle operation. However, their operational characteristics, in which complete charging and discharging do not occur in each cycle, make it difficult for existing methods to estimate their State of Health (SoH). In this paper, we propose In-Boundary Capacity (IBC) based on Incremental Capacity Analysis (ICA) that can quantitatively evaluate the degradation state of batteries without the need for full charging and discharging. ICA was performed based on actual railway vehicle operation data, and the IBC was derived by accumulating the Incremental Capacity (IC) values for specific voltage ranges. The results showed that IBC had a high correlation with SoH, confirming its potential as an alternative indicator of SoH. Additionally, the main characteristic values of the battery were used as input variables to construct the LSTM, TCN, AutoFormer, N-Linear, and BasisFormer models, and their predictive performances were compared. The results showed that BasisFormer achieved the lowest average error among the models evaluated in this study. The proposed method can effectively evaluate degradation without battery separation or full charge/discharge experiments, demonstrating its suitability for efficient evaluation and life prediction, even in complex operating environments of railway vehicles.</p>