The use of neural network algorithms for recognizing vibration signal recognition requires a dataset for training. However, traditional manual dataset creation is time-consuming and labor-intensive. This paper proposes a computer-based intelligent method for constructing vibration signal datasets using the YOLOv5 target detection algorithm. The surveillance videos captured by the camera are processed using frame difference calculation, Gaussian filter, binarization, inflation, and contour recognition. The YOLOv5 algorithm is then used to detect targets and map the corresponding vibration event categories to the vibration signals captured by the geophone, resulting in the dataset. The experimental results indicate that this method produces a vibration signal dataset with distinctive features and enhances efficiency.

Seismic monitoring has been instrumental in various domains such as natural earthquake early warning, mineral mining safety assessment, and hydraulic fracturing impact evaluation. However, the monitoring data often exhibit low signal-to-noise ratio (SNR) and large volume. Therefore, developing an efficient, high-precision, and universally applicable seismic waveform automatic classification network model becomes significant and practical. We propose a physical interpretable time-frequency deep convolutional recurrent neural network (TF-DCRNN) model which consists of an integration of a time-frequency convolutional (TFconv) layer and a convolutional recurrent neural network (CRNN). Subsequently, we evaluate the classification performance by comparing five network models, including convolutional neural network (CNN) and long short-term memory (LSTM), using Ricker wavelet datasets with varying SNR levels (-15 ~ 0 dB). Our findings verify the superiority of the TF-DCRNN model in the classification of strong interference environment from both numerical and physical simulation. Moreover, integrating multiple network models or incorporating a TFconv layer can moderately enhance the classification performance, which provides the direction for network model optimization.

Scientific assessment of the transport channels and distribution of oil, gas, water, and other fluids is the key to enhancing oil and gas production capacity and ensuring the safe and efficient development of mineral resources. Traditional microseismic monitoring has been suffering from limitations such as ambiguous attribute characteristics of the seismic source and low signal-to-noise ratios of the data, leading to poor location accuracy and large uncertainties. We develop a phase change seismic source monitoring method that allows for a detailed characterization of the spatial distribution of subsurface fluid transport by accurately locating the equivalent seismic source within supportive fractures. A phase change seismic source is developed and constructed based on the phase change material, enabling control over phase change temperature, source energy, and source scale through the principle of phase change energy storage. The source introduces an active approach to seismic signal generation, diverging from the conventional passive method reliant on rock rupture. Then, 3D homogeneous models are constructed for ground monitoring and in-well monitoring using impulse source propagation theory. The numerical simulation of the wavefield uses the finite-difference method, and the equivalence of discrete sources is confirmed using the Pearson correlation function. The reliability of the monitoring method is verified through a seismic monitoring experiment and comprehensive feasibility analysis.

The Controlled Source Electromagnetic Method (CSEM) assesses electromagnetic field components across specific orientations within a field coordinate system to calculate apparent resistivity for electromagnetic detection. However, in complex terrains, capturing the x, y, or z components accurately is often impracticable, leading to discrepancies in apparent resistivity calculations and impacting detection precision. To mitigate these challenges, this paper proposes a method for observing electric and magnetic field components in any orientation, thereby enhancing wide-azimuth electromagnetic detection. Firstly, the paper analyzes the angular deviations and the limitations of zero-value bands in certain single-component observation methods. Secondly, it examines the geoelectric response of electric and magnetic fields in various orientations and the non-monotonicity in apparent resistivity calculation. These analyses validate the feasibility of the wide-azimuth electric field observation method, substantially augmenting field observation flexibility. Moreover, the study identifies the challenges in accurately calculating apparent resistivity in wide-azimuth magnetic detection. This study provides innovative insights for advancing the CSEM detection method and improving detection accuracy.

Boron is a critical strategic mineral resource. Boron deposits in the Liaodong region currently supply more than 35% of China’s domestic demand. To advance exploration efforts in this area, detailed physical property measurements were carried out on various rock formations within the Yingkou–Anshan district. Utilizing integrated geophysical approaches, including gravity, magnetic, and electrical surveys, this study introduced—for the first time—the wide field electromagnetic method (WFEM) for deep exploration testing. Measured parameters included density, magnetic susceptibility, and resistivity. The electromagnetic methods proved effective in boron mineral exploration due to their pronounced response over the Lieryu Formation, which is enriched in boron minerals. We refined inversion parameters to improve the consistency between geophysical models and actual geological characteristics by correlating physical property parameters with drilled core lithology. Comprehensive analysis indicates that boron mineralization in the Houxianyu mining area is not restricted to the Lieryu Formation. Significant boron enrichment also occurs in deeper structures, including thick granite bodies, large-scale folds, and various unconformity contacts. These deep mineralized bodies share similar physical properties with known ore deposits but exhibit distinct geochemical signatures. Through integrated interpretation of gravity, magnetic, and electrical profiles—combined with geological, mineralogical, and structural data from the Houxianyu area—it is evident that ore bodies in the study area occupy structurally complex positions, influenced by regional tectonic evolution and magmatic activity. Geophysical results further reveal a notable deep extension of mineralization, indicating promising potential for deep prospecting.