| Title |
Quantitative Relationship Analysis of Fuel-Cell Evaluation Equipment Variables : A Pearson-Spearman-Kendall Methodology |
| Authors |
최원칠(Won-Chil Choi) ; 강일호(Il-Ho Kang) ; 배원규(Won-Gyu Bae) |
| DOI |
https://doi.org/10.5370/KIEE.2025.74.9.1628 |
| Keywords |
FuelCell; Testbench; Pearson; Spearman; Kendall; Correlation; Preprocessing |
| Abstract |
This study presents a quantitative framework for analyzing test-bench, facility, and environmental variables from fuel-cell test-bench data. Of 200+ channels, 29 variables were grouped into controlled, pressure?temperature, and lab/external streams. After detecting load steps, anomalies, and skewed distributions via line plots and histograms, IQR-based outlier removal and standardization were applied. Dependency analysis using Pearson’s r, Spearman’s ρ, and Kendall’s τ revealed that stack voltage is a complex function of load, supply pressure, and coolant-outlet temperature, that control-system variables exhibit near-perfect rank agreement, and that environmental variations significantly affect pressure and flow stability. Overall, this methodology provides essential data for real-time anomaly detection and operational-logic optimization in fuel-cell evaluation equipment. |