A Gaussian wavelet-based method for extracting rockfall motion information in consecutive impact tests on flexible barrier systems

Abstract

Flexible barriers are vulnerable to consecutive rockfall impacts, particularly in mountainous areas. The barriers deformation, energy absorption, and anti-impact performance can be evaluated from the rockfall motion information. However, a slight interference in rockfall displacement signals, generated during the actual measurement, could result in a distortion of the rockfall velocity and acceleration extracted through conventional derivative calculations. Consecutive impact tests on steel-wire ring nets were conducted, and the motion of the impact rock was captured from a high-speed video. The random noise in the impact signal was quantified using the signal-to-noise ratio and was distinguished in the frequency domain. Furthermore, to increase the derivative calculation accuracy, a Gaussian wavelet-based method is proposed, and an amplitude parameter is introduced into the wavelet transform procedure. The self-consistency of the proposed method was achieved using a combined differentiation–integration iteration process. Analytical signals of a single-degree-of-freedom system under a sine-wave impulse were used to validate the proposed method. We observed that the wavelet scale parameters were strongly related to the impact signal frequency band, which accounted for 90% – 99% of the total energy in the frequency domain. Finally, the Gaussian wavelet-based method was applied to the real impact signals of a rockfall. The results indicated that the proposed method is an effective and reliable tool for extracting the velocity and acceleration information of falling rocks during the consecutive impact process.