Mine data fusion technology based on multi-feature joint decision making

BackgroundIn the field of geological exploration of coal mines, methods such as in-seam seismic exploration, the transient electromagnetic method, and audio-frequency electrical penetration offer unique technical advantages. However, each of these methods can only reflect a single physical property of a medium, suffering from limitations including the one-sidedness of information and the multiple solutions of interpretations, thus failing to adapt to the new situations of current coal mine exploration. Multi-source data fusion promotes a technological development by leaps from one-sided detection to comprehensive analysis through mechanisms of information complementation, feature enhancement, and field-source integration. Therefore, this technology can enhance the accuracy of both the positioning and morphological identification of anomalous geobodies. MethodsThe data fusion technology based on wavelet decomposition, which employs principal component analysis (PCA) for low-frequency components and multi-feature joint decision-making for high-frequency components, was used to integrate geophysical data obtained using the three methods (i.e., in-seam seismic exploration, transient electromagnetic method, and audio-frequency electrical penetration). The purpose is to enhance the comprehensive analytical capacity for geological information and to achieve accurate detection of geological structures and anomalies. First, raw data obtained using the three methods were preprocessed to eliminate noise interference and normalize data formats. Subsequently, using the wavelet transform, the resulting data from various sources were then decomposed into the coefficients of subbands corresponding to different frequencies. Based on their importance in geological characterization, the coefficients of low- and high-frequency subbands were integrated using targeted fusion rules. Finally, the fused geophysical data were determined through inverse wavelet transform. Results and ConclusionsCompared to data from a single source, the fused data incorporated the characteristics of structural boundaries delineated using the in-seam seismic exploration, supplemented by information on underground electrical contrast obtained using the electrical method. Therefore, these data allow for the presentation of more abundant details of geological structure characteristics. The proposed data fusion technology holds broad application prospects in fields such as geologic hazard prediction and resource exploration, providing a novel, effective approach for the development of geophysical exploration technology.