Strong Amplitude Suppression in Seismic Data Using PSO-VMD

Strong-amplitude reflections generated by coal seams and hydrocarbon source rocks often obscure weak reservoir signals and hinder accurate seismic interpretation. These geological features generate high-energy reflections, while the overlapping seismic signals exhibit complex, nonstationary characteristics that challenge traditional signal separation techniques. To address this problem, a method termed PSO-VMD is proposed in this paper. PSO-VMD integrates Particle Swarm Optimization (PSO) with Variational Mode Decomposition (VMD) to adaptively determine key parameters such as the number of modes and penalty factor, improving decomposition accuracy. By formulating an optimization objective based on envelope entropy, PSO-VMD enhances the separation between strong and weak components in the time-frequency domain. Furthermore, the method ensures better lateral continuity and suppresses mode mixing by refining VMD’s variational constraints through global search. Comparative experiments on synthetic and field data verify that PSO-VMD achieves superior performance over wavelet transform singular value decomposition (WT-SVD) and empirical mode decomposition (EMD) in strong reflection suppression, weak signal preservation, and stratigraphic boundary continuity. The proposed method demonstrates effective suppression of strong reflections and enhances seismic signal quality in complex stratigraphic settings.