An Adaptive Generalized S-Transform Algorithm for Seismic Signal Analysis

As a powerful method in signal processing, time-frequency analysis shows the characteristics of signals in the form of a joint domain distribution of time and frequency. The S-transform is one of the most effective and frequently used algorithms for time-frequency analysis of seismic signals. To address the problems of fixed time window, such as single time-frequency resolution and poor energy aggregation, an adaptive generalized S-transform algorithm is proposed. In this method, we designed a new generalized Gaussian window function with four variable parameters to obtain the optimal time-frequency distribution. The proposed window function uses time-frequency aggregation as the objective function for parameter optimization. Combined with the calculation results of the time-frequency aggregation and time-frequency distribution, we proved that the proposed method has a satisfactory analysis effect and anti-noise performance. In addition, a simulation experiment of the seismic signal is carried out using this algorithm, and the results show that the proposed adaptive generalized S-transform enhances the time-frequency resolution and energy aggregation in seismic signal analysis.