淤泥质海底航道边坡失稳滑塌声纹特征分析

边坡稳定是淤泥质海底航道安全运营的关键问题之一。海底航道边坡失稳滑塌过程的不同阶段浅地层剖面图像声纹特征不同。采用灰度共生矩阵和小波多尺度分解方法分析了航道边坡失稳滑塌过程浅地层剖面图像纹理特征。结果表明,对比度、能量、相关度和逆差距在边坡失稳滑塌过程中具有明显的差异,分别从不同方面刻画了浅地层剖面图像声纹的清晰度、纹理粗细、主要方向和规则性,以定量的形式显示其内在异质性。能量、相关度和逆差距的水平和垂直方向变化大于对角方向,水平方向波动最明显。浅地层剖面声强时间序列和垂向空间变化序列小波多尺度分解突出了声强变化的局部细节及其在各尺度上变化的强弱分布和突变点位置,克服了常规人工识读的困难,这些声纹特征是边坡失稳滑塌预测预警的重要依据,有助于对海底航道边坡稳定性的探测。

Analysis on characteristics of sound print of slope instability and collapse in muddy submarine channel

Slope stability is crucial to the safe operation of the muddy submarine channel. The acoustic characteristics of sub-bottom profile images are different at different stages of submarine channel slope instability and collapse. Gray level cooccurrence matrix and wavelet multi-scale methods were used to analyze the texture characteristics of sub-bottom profile images of channel slope instability and collapse. The results show that there are obvious differences in contrast, energy, correlation and homogeneity during the process of slope failure and collapse, which describe the clarity, thickness, main direction and regularity of sound prints of sub-bottom profile images from different aspects, and show the inherent heterogeneity quantitatively. The energy, correlation and homogeneity show obvious changes in horizontal and vertical directions than that in diagonal direction, especially in horizontal direction. The wavelet multi-scale decomposition of time series and vertical spatial change series of sub-bottom profile images highlights the local details of the sound intensity change, the intensity distribution and the mutation point position in each scale, which is difficult for conventional manual identification. These characteristics of sound print variation are important evidence for prediction and early warning, which is helpful for the detection of submarine channel slope instability and collapse.