Trend Features Extraction of Apparent Resistivity in Tunnel Engineering

Troublesome incidents like sudden water inflows increase the risk of collapse accidents in tunnel excavation. In this study, a data-driven underground water prediction method is proposed based on trend features extracted from apparent resistivity. A novel framework is developed for extracting trend features from the contour lines of apparent resistivity. These trend features are subsequently integrated with numerical features from the resistivity matrix for classification. The effectiveness of the proposed method is demonstrated by apparent resistivity data from real tunnel engineering. The result indicates that the classification accuracy of the proposed method outperforms the method without feature extraction.