Electrochemical Impedance Spectroscopy (EIS), combined with equivalent circuit modeling, enables detailed quantification of series resistance R_s, parallel resistance R_p, and parallel capacitance C_p, which serve as indicators of internal stress. In contrast, current voltage characterization (I-V) accesses the power loss. A modeling framework is necessary to translate EIS-derived parameters under mechanical stress into performance outcomes. To address this, three 10 W PV panels were subjected to mechanical drop tests. The panels started with different initial ranges of R_s, R_p and C_p, and each was exposed to a different number of steel ball drops to impose varying stress levels. EIS measurements were carried out using a portable STEMlab 125-14 platform, and a parameter extraction algorithm was applied to obtain R_s, R_p and C_p, In parallel, the Keithley 2400 was used to record I-V data. A Bayesian Gaussian Process Regression model was then introduced to map the nonlinear relation between EIS parameters and maximum power point metrics. Preliminary analysis yielded low prediction errors of 4.7% for V_mp 7.2% for I_mp, and 7.9% for ć Pć_mp.
