Random Noise Attenuation based on Scale Transformation Stochastic Resonance System and FXVMD

Abstract Random noise in seismic data seriously affect subsequent inversion and interpretation. Therefore, effectively reducing random noise is a crucial link in seismic data processing. Traditional random noise attenuation methods struggle to select appropriate parameters and have poor adaptability, leading to ineffective noise reduction. To address the limitations of traditional methods, this paper combines the theory of scale transformation stochastic resonance system (STSRS) in nonlinear physics with f-x variational mode decomposition (FXVMD), proposing a new method that integrates dual-domain synergy advantages with strong multi-band adaptability. The proposed method utilizes STSRS capable of detecting effective signals across the entire frequency domain. Firstly, the STSRS processes each trace of seismic data to obtain resonance frequencies. Then, the signal is transformed into the frequency domain for variational mode decomposition. Within the frequency band centered at the resonance frequency and extending half-bandwidth to both sides, intrinsic mode functions with frequencies within this range are filtered. The appropriate bandwidth is determined through multiple rounds of validation on the experimental data. Finally, the denoised signal is obtained through reconstruction. The effectiveness of the proposed method compared to traditional methods has been demonstrated through numerical and field examples.