Vibration signal analysis of coal-gangue based on Hilbert marginal spectrum in clean coal technology

With the increase of social demand on environmental quality, the clean coal technology (CCT) is one of the most effective measures for environmental control. It is an important means in CCT to decrease the gangue content. Targeting at distinguishing the top-coal and gangue, this paper discussed coalgangue interface detection on fully mechanised mining face. The Hilbert marginal spectrum based on empirical mode decomposition (EMD) is applied to the vibration feature extraction for the top-coal and gangue. The vibration signals of coal and gangue contain non-stationary characteristics. The EMD is used to decompose the original vibration signals into a number of intrinsic mode components (IMFs), among which the first seven IMFs have strong correlation with original signals. The Hilbert transformation is carried on these seven IMFs and the Hilbert marginal spectrum is obtained. From the Hilbert marginal spectrum of each IMF, we find the IMF C 1 contains the main difference of top-coal caving and coal-gangue caving. Thus we present the marginal energy function to distinguish the caving state. Experimental results show that the vibration feature extraction based on Hilbert marginal spectrum can be served as coal-gangue interface detection effectively in CCT.