Implementation of an Electrochemical Impedance Spectroscopy for the condition diagnosis of fuel cells in the DC-DC converter of vehicle

Kurzfassung

Proton exchange membrane fuel cells (PEMFCs) are regarded as a promising technology for fuel cell electric vehicles (FCEVs) because of their high efficiency and low emissions. However, durability, degradation and reliability remain significant challenges for their widespread commercialization. Electrochemical impedance spectroscopy (EIS) is a well-established diagnostic technique capable of identifying electrochemical processes and degradation mechanisms within fuel cells. Conventional EIS systems, however, require dedicated laboratory instrumentation, making their integration into automotive applications impractical. This thesis therefore investigates the technical feasibility of implementing EIS using the existing DC-DC converter of a FCEV, eliminating the need for additional impedance measurement hardware. A structured implementation methodology was developed for the Globe XLP-80 PEMFCsystem, in which the DC-DC converter was utilized to inject sinusoidal current perturbations while the corresponding voltage and current responses were acquired through the Fuel Cell Control Unit (FCCU) via the Controller Area Network (CAN). An implementation framework defining stable voltage for stability conditions, continuous conduction mode (CCM) of converter, battery state of charge (SOC) and maintaining constant operating point of the vehicle, was also established to enable converter-based impedance measurements under practical operating conditions.The measured signals were analysed using Fast Fourier Transform (FFT)-based signal processing using converter, to determine amplitude, phase angle and complex impedance, from which Nyquist and Bode plots were constructed. Experimental investigations demonstrated that the existing DC-DC converter was capable of generating sinusoidal excitation signals over the investigated frequency range of 1-600 Hz and successfully acquiring impedance spectra. Oscilloscope measurements confirmed the generated excitation signals, while FFT results obtained from the converter showed good agreement by comparing the results with independently developed Python based signal processing. Signal validation using time-domain analysis, amplitude error, signal-to-noise ratio (SNR), total harmonic distortion (THD) and Kramers-Kronig (KKT) analysis was used to assess measurement quality. Low-frequency measurements were primarily limited by operating-point drift, converter gain limitations and measurement noise, whereas high-frequency measurements were mainly affected by reduced voltage response amplitude, sensor accuracy, quantization effects and electrical noise. The measured Nyquist spectra are qualitatively consistent with PEMFC impedance characteristics reported in the literature. However, the results are subject to implementation limitations, including sensor accuracy, converter dynamics, excitation-amplitude variation and measurement noise. Also no ECM (Equivalent Circuit Model) fitting, calibrated reference EIS measurement or repeatability assessment was performed. Overall, this work demonstrates the technical feasibility of performing EIS using the existing DC-DC converter in a fuel cell system. The proposed methodology establishes a practical foundation for embedded condition monitoring using hardware already available within FCEV architectures. Future work should focus on improving measurement accuracy, validating repeatability, optimizing converter control, enhancing sensor performance, accelerated EIS (aEIS) and integrating degradation identification algorithms to enable real-time onboard fuel cell diagnostics.