Accurate estimation of battery State of Health (SOH) is of significant importance for the normal operation of batteries and guiding their recycling. Currently, the mainstream SOH estimation methods are data-driven approaches. However, these methods face challenges in effectively utilizing multi-dimensional feature data and addressing long-sequence modeling problems. Therefore, this paper proposes a lithium-ion battery SOH estimation method based on the Mamba architecture and multi-dimension feature fusion. The method extracts multi-dimensional features from battery charging sequence data and subsequently trains the Mamba model to estimate the battery SOH. Simulation experiments were conducted on a public dataset, and the results demonstrate that the proposed method achieves high accuracy in battery SOH estimation tasks, validating the potential of the Mamba architecture in the field of SOH estimation.