Detection of vehicle mass using dynamic response to sinusoidal steering excitation

Abstract

This study proposes a method for estimating vehicle load using sinusoidal steering excitation and analysis of dynamic responses: lateral velocity (LV), roll velocity (RV) and yaw velocity (YV). The approach enables reliable discrimination between unloaded (W1) and loaded (W2) vehicle states under real driving conditions. Lateral velocity increases with vehicle speed and exhibits higher magnitudes for the loaded state, particularly within the 0.5–1.5 Hz range. Phase responses for W2 show stronger negative shifts and earlier transitions, indicating increased dynamic delay, which makes LV a useful supplementary indicator, especially at medium speeds. Roll velocity (RV) is identified as the most sensitive and repeatable response, and serves as the primary diagnostic variable. The largest differences between loading states occur within the 0.9–1.3 Hz range, where W2 responses exhibit higher amplitudes and a shift toward lower frequencies, directly reflecting increased vehicle mass. Above 2 Hz, system damping increases and load-related differences diminish. Yaw velocity (YV) shows amplitudes 1–3 dB higher for W2 within the 0.6–1.6 Hz range, together with phase shifts, but its sensitivity is lower than that of RV, classifying it as a supplementary indicator. Unlike methods based on complex identification algorithms or high-frequency measurements, this study adopts a simplified linear single-degree-of-freedom (1-DoF) oscillator model. Defined road-test conditions, including speeds of 45–56 km/h and excitation near 1.0 Hz, ensure dominance of a single dynamic mode. The method is computationally efficient and cost-effective, although its linear formulation may limit accuracy under strongly nonlinear dynamics or varying road conditions.