This paper proposes a low-complexity, cost-effective in-situ SOH estimation system. Since SOH is not directly measurable, it must be inferred from observable parameters using model-based estimation. The proposed system applies small perturbations to the battery, calculates its electrochemical impedance spectroscopy (EIS), and performs complete in-situ SOH estimation directly on a low-cost embedded microcontroller unit (MCU), without reliance on external computing resources. Measuring a complete impedance spectrum is often time intensive. This work explores two independent approaches to shorten the SOH estimation time. First, the correlation between SOH and the full frequency spectrum is analyzed to identify key frequency ranges that allow faster computation. Second, conventional stepped sine and multi-sine perturbation methods for EIS acquisition are compared to confirm that the latter achieves comparable accuracy with shorter measurement time.
.jpg "Jingyuan Liang photo")