Numerous microseismic signals are produced by rock mass fracture during earthquakes, geological disasters, or underground excavations. Moreover, a large amount of noise signals are captured during microseismic signal monitoring. Specifically, some noise signals closely resemble microseismic signals, which severely impedes the rapid and accurate detection of the latter and the assessment of geological hazards. Therefore, we propose a precise model for identifying and classifying microseismic signals based on deep learning technology and short-time Fourier transform (STFT) technology. First, the STFT time–frequency analysis reveals the unique characteristics of noise, microseismic, and blasting signals, thereby allowing noise signals that are very similar to microseismic signals in the time domain to be finely distinguished. Second, the introduced attention mechanism focuses the classification on essential signal features. Finally, because tens of thousands of actual monitoring data points are considered, the deep neural network for microseismic classification is trained and tested under complex geological engineering conditions. The results demonstrate that the neural network model has good time–frequency feature extraction ability, and the well-trained model can satisfactorily complete daily classifications. Moreover, the model performs well when classifying similar noise and low-SNR microseismic signals. We believe that this type of signal-processing method, which considers multiple perspectives, can be extended to data processing in many other data-driven fields.

The microseismic signals released by rock mass fracture can be captured via microseismic monitoring to evaluate the development of geological disasters. This is crucial for underground engineering construction, underground mining, and earthquake and geological disaster evaluation. However, extracting information effectively is difficult due to the low signal-to-noise ratio of microseismic signals caused by complex environmental factors. Therefore, denoising and detection (onset time picking) are essential to processing microseismic signals and extracting source information. To improve the efficiency and accuracy of microseismic signal processing, we propose a parallel dual-tasking network, which is an advanced deep learning model that can simultaneously perform microseismic denoising and detection tasks. The network, comprising one encoder and two parallel decoders, is customised to extract input data features, and two outputs can be simultaneously generated to denoise and detect microseismic signals. The model exhibits excellent denoising and detection performance for microseismic signals containing various types of noise. Compared with traditional methods, the signal-to-noise ratio of the denoised signal is greatly improved, and the waveform distortion of the denoised signal is small. Even when the signal-to-noise ratio is low, the proposed model can maintain good onset time pickup performance. This method obviates the need for different denoising methods for different types of noise and precludes setting thresholds artificially to improve the denoising effect and detection accuracy. Moreover, the dual processing characteristics of the model facilitate simultaneous denoising and detection, which improves the processing efficiency of microseismic data and meets the demand for automatically processing massive microseismic data. Therefore, this method has excellent data processing potential in exploration seismology, and earthquake and disaster assessment.