Analysis of Sensor Placement and Signal Processing Enhancement for Vehicle Speed Measurement Through Rumble Strip-Induced Ground Vibrations

Abstract

This research studies four sensor positions for measuring vehicle speed through ground vibrations induced by transverse rumble strips. The study also introduces an automatic signal collection and processing method that utilizes amplitude thresholds in the time domain to differentiate between random background noise and the actual rumbling caused by vehicles crossing the rumble strips by looking for periodicity. The collected signal undergoes a 4-band band-pass filter and zero-padding before applying a Fast Fourier Transform (FFT). The frequency with the highest magnitude among the four spectra is used to calculate the vehicle's speed. Additionally, the research proposes using the calculated speed and the time-domain length of the signal to measure the vehicle’s wheelbase. Experimental results show a correlation with analytical findings.