Research on Coal-Rock Recognition Based on the EMD-EM-BP Model

This article presents a coal-rock recognition method based on a BP neural network. It involves testing acoustic emission (AE) signals during the shearing and cutting of coal-rock, analyzing coal-rock characteristics, and applying Empirical Mode Decomposition (EMD) to the AE signals to obtain several Intrinsic Mode Functions (IMFs). Energy moments of the IMFs, with significant correlation coefficients, are then extracted. These energy moments serve as inputs for the BP neural network to identify the acoustic emission signals of coal-rock. Experimental results indicate that after using the EMD algorithm to decompose the AE signals and calculate the energy moments of the IMFs in relation to the original AE signals, distinctive features become apparent. The BP neural network achieves a high accuracy rate of 95% in recognizing coal-rock characteristics. This model offers exceptional recognition precision and lays the theoretical and technological groundwork for realizing automated and intelligent mining in fully mechanized mining operations.