Adaptive microseismic data compressed sensing method based on dictionary learning

The theory of compressed sensing brings a revolutionary breakthrough to signal acquisition technology, it can sample the signal far below the Nyquist frequency, and accurately reconstruct the original signal through numerical optimization method. In order to improve the precision of microseismic signal after reconstruction and improve the sampling efficiency, an adaptive microseismic data compressed sensing method based on dictionary learning is proposed in this paper. Firstly, the adaptive redundancy dictionary is constructed according to the characteristics of the microseismic signal. Then the energy and the sparsity under redundant dictionary of signal are calculated. Finally, according to the comprehensive index of energy and sparseness, an adaptive sampling strategy of compressed sensing is developed. Simulation experiment show that, the algorithm reduces the number of samples by 10% compared with traditional compressed sensing, and improves the reconstruction accuracy.