Sensing and Monitoring of Underground Rock Permeation Damage Using Musicalized MAE GMM Analysis

Abstract

Moisture could promote the crack propagation in underground rock and increase the accident risk. It is of great significant to sense the moisture permeation state in real time. In this research, a GMM (Gaussian Mixed Model) analysis of musicalized MAE (Modal Acoustic Emission) method for moisture permeation monitoring of layered rock was proposed. First, the effective AW (A Wave) component was extracted by means of frequency decomposition. Then, the envelope was sparsely sampled through musicalized model. The MAE was encoded into 4 musicalized indicators to compress the data with high fidelity. Further, GMM was used to fit MAE parameters. The moisture permeation was evaluated by the signal correlation and the reliability theory. Numerical simulation and experiment demonstrated that moisture could affect the MAE ware. The proposed method could process the monitoring data and evaluate the moisture permeation levels. The continuous MAE signal could be converted into discrete data, which only requires little storage resources. Thus, it is meaningful in the narrow and harsh underground environment.