Wavelet-Guided Hybrid Attention Neural Network for Identification of Microseismic Event

Microseismic monitoring technology enables accurate and real-time identification of the spatial location and type of geological hazards, serving as a crucial component of early warning systems for geological disaster prevention and mitigation. Due to the complex and unpredictable physical conditions of highway slopes, effective identification of microseismic signals remains a challenging task. To address this issue, a wavelet-guided hybrid attention intelligent framework for microseismic event recognition is proposed. Initially, a wavelet domain attention-guided filtering module is designed, which maps time-domain signals to the wavelet domain and employs an attention mechanism to filter and extract features from the signals, thereby enhancing noise suppression and robustness. Subsequently, a physical-knowledge-guided loss function composed of Hilbert transform and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">l</i>2 / <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">l</i>4 norm is designed to optimize the neural network model. Finally, experimental validation is conducted on a dataset of microseismic events from highway slopes. The experimental results show that the average recognition accuracy of the proposed method reached 96.31%, outperforming comparative methods and demonstrating the potential application of the proposed method in practical engineering.