Identification and extraction of lateral target signals in tunnel geological prediction with the Karhunen-Loéve beamforming method

Abstract

Before the construction of a tunnel, effective and accurate advanced prospecting techniques are required to detect unexpected geological heterogeneity around the tunnel. However, because the data collected during tunnel advance prediction include waves reflected from different directions and a large amount of environmental noise, the seismic records collected are extremely complex, and it is difficult to extract the target signals from the lateral area of the tunnel effectively. The beamforming method can create a focused wave front through the stack of the wavefield, improving the data quality of the target signal. This paper incorporates the Karhunen-Loéve method in the beamforming procedure to extract the lateral signal of the tunnel. First, the test principle of Karhunen-Loéve beamforming (KLBF) based on the tunnel seismic prediction (TSP) observation system is discussed, and the corresponding delay parameters are estimated according to different application conditions. In addition, ray tracing and three-dimensional wave equation methods are used to carry out numerical simulation research on the models of faults, karst caves and underground rivers. Finally, the corresponding KLBF method can be used to improve the quality of effective signals by different delay time calculation methods, which can better identify and extract the lateral reflected signals.