Excitation Signal Design for Fast Electrochemical Impedance Spectroscopy in Battery Testing

Electrochemical impedance spectroscopy (EIS) is a widely used technique for analyzing battery dynamics over a broad frequency spectrum. Conventional state-of-the-art EIS methods involve applying a sequence of sinusoidal excitation signals, ranging from very low to very high frequencies, to capture the impedance response of the battery. However, this process is time-consuming, often requiring several hours to complete. Alternatively, approaches using pulse-based excitation have shown promise in reducing test time but often suffer from challenges in handling measurement noise and poor frequency resolution, especially at low frequencies. This work presents an improved rectangular pulse-based impedance characterization technique that enhances low-frequency resolution, increases robustness to noise, and reduces experimental time. This is accomplished through the following three key contributions of this paper: First, it establishes statistical noise properties in the Fourier-transformed signals, enabling effective noise reduction through averaging. Second, it proposes a log-frequency clustering approach to average impedance data, enhancing the accuracy of the impedance spectrum. Third, it presents a systematic pulse design method using the knowledge of the approximate time constants of the system to select the sampling interval, pulse width, and rest duration for reduced test time, improved low-frequency resolution, and enhanced signal-to-noise ratio (SNR). Together, the proposed approach enables faster and more accurate impedance characterization. Simulation analysis and experimental results confirm that the proposed approach enhances spectral resolution at low frequencies, mitigates the impact of noise at high frequencies, and significantly improves the reliability of impedance estimates at a faster measurement time frame.