Abstract: The underground mining link of the hanging gang mine is accompanied by blasting activities,which can easily lead to final slope instability. In order to investigate the characteristics of the time-frequency energy distribution of vibration signals and assess the vibration hazards,several vibration measurement tests were carried out at the preparation iron ore mine in Hejing County. For the vibration data obtained from monitoring,the Hilbert-Huang transform ( HHT) method,combined with wavelet soft threshold denoising was applied to analyze the vibration signal of this hanging gang mine from the perspective of time,frequency and energy,and the EMD (Empirical Mode Decomposition) postponement identification method and instantaneous energy postponement identification method was applied to the electronic detonator The actual postponement time was identified. The analysis results show that:① The blasting energy is mainly concentrated in the frequency band of 0 ~ 200 Hz, and the main vibration band is 0 ~ 50 Hz;② The blasting energy develops in the direction of low frequency,and contains 51. 22% of energy in the frequency band of 0~ 10 Hz,and the low frequency band contains most of the energy,which is easy to trigger the self-vibration of the building ( structure) and has a great threat to the stability of the slope of the hanging gang mine;③ The blasting energy is in the range of 0. 495 s and 1. 119 s,i. e. ,the blasting of slotted holes and auxiliary holes generate more energy,so it is recommended to set empty holes for slotted holes and appropriately reduce the amount of auxiliary hole charge in section 5;④ Under the conventional extension,the recognition rate of detonator segments of EMD method and instantaneous energy method are 100%,and the recognition accuracy of actual extension time is 95. 2% for EMD method and only 82. 6% for instantaneous energy method. This is a preliminary analysis because the EMD method itself is not completely orthogonal decomposition,the analysis easily causes confusion between the modes,resulting in recognition accuracy not reaching 100%.