Abstract: The analysis of blasting vibration signals is of vital importance for enhancing the accuracy and safety of blasting operations. The core issue lies in the non-stationary nature of the signals and the limi⁃ tations of traditional stationary signal analysis methods. Through a systematic review of the development process of analysis methods in this field，it is found that the early Fourier transform and short-time Fourier transform based on the assumption of stationarity were limited by their fixed time-frequency resolution. Methods such as wavelet transform effectively capture the characteristics of non-stationary signals; the Hil⁃ bert-Huang transform，with its adaptive time-frequency analysis mechanism，demonstrates unique advan⁃ tages in handling complex transient signals，but requires the combination of improved algorithms such as EEMD and EWT to avoid end-point effects and mode aliasing. Intelligent information processing and multi-source data fusion are the key to breaking through the existing analysis bottlenecks and achieving precise assessment and real-time control of vibration effects. In the future，signal analysis based on deep learning，multi-modal data collaboration，and dynamic feedback systems will jointly drive the develop⁃ ment of blasting vibration analysis towards intelligence，precision，and real-time.

Key words: blasting vibration signals, analysis methods, non-stationary signals, intelligent informa? tion processing