The current research of blasting vibration signal is mainly focused on the time-frequency analysis，but not from the perspective of energy. In this paper，the blasting vibration signal was analyzed based on the time-energy density method of wavelet transform. Firstly，the Mallat algorithm was used to decompose the local information of the signal with wavelet，and then the energy of each moment was obtained according to the time-energy density curve. The time-varying of energy density of signal in each frequency band was counted. Finally，five groups of blasting vibration signals of the No.2 double-hole tunnel in the Liyanlongkeng were analyzed using the method. The results show that the closer the distance to the blasting center，the stronger the blasting energy is released. The energy coefficient ratio of wavelet decomposition frequency band d2 and d3 accounts for 93.33%. The maximum energy ratio of wavelet decomposition frequency band is d2，and the corresponding vibration velocity and frequency are utilized to judge with the blasting safety regulation. This method provides a new way for blasting vibration safety assessment from the perspective of energy，and that also has a theoretical guiding significance for the study of blasting vibration mechanism.