Accurately predicting the effects of blasting vibration is the core challenge in controlling its hazards and ensuring safe production in mines. Traditional empirical formulas are difficult to characterize the non-stationary characteristics of vibration waveforms, which limits further improvement in prediction accuracy. Therefore, a vibration prediction model was constructed for open-pit single hole blasting based on time-frequency decoupling. This model characterized the non-stationary spectral properties of vibration waves through a frequency attenuation function and employed the Gamma function to describe signal intensity envelopes. Particle swarm optimization was utilized to determine model parameters from on-site measured waveforms. Combined with support vector machine methods, a multivariate regression relationship between model parameters and key influencing factors (charge, blast-center distance, longitudinal wave velocity) was established. The program was implemented on the Python platform and validated through field experiments at an open-pit mine in Jiangxi. Results indicate that the developed single-hole blasting vibration prediction program exhibits significant advantages in parameter determination methods, validation effectiveness, and comprehensive superiority.