Inverse Modeling of Electrified Road Surface Roughness Based on Vehicle Dynamic Response

Abstract

To support the analysis of vibration excitation characteristics in electrified road systems, this study proposes an inverse modeling approach for reconstructing road surface roughness in heavy-duty transport scenarios based on vehicle dynamic responses. The method segments the measured suspension displacement signal using RMS values in the time domain and initializes road profile parameters using a parameterized harmonic superposition model. Road spectrum strength and spectral characteristics are iteratively optimized in each segment by minimizing the RMS error between simulated and measured responses, ultimately generating a 3D roughness model of the road surface. To assess the engineering applicability of the reconstructed model, a frequency-domain analysis is conducted on the translational and rotational excitations at the pantograph base. Results reveal that pitch and vertical responses dominate in the low-frequency range and exhibit strong coupling, whereas roll and lateral responses show multi-peak characteristics in the mid- to high-frequency range, indicating possible local resonance phenomena. This research provides a novel methodology and theoretical basis for modeling complex road surfaces and their application in the dynamic analysis of electrified highway systems.