摘要
From the perspective of the equivalent overall system generated from the seismic convolution and deconvolution,a maximum kurtosis criterion of deconvolution output with the constraint of variance is derived as objective function of seismic blind deconvolution by making the impulse response in the form of unit impulse sequence.Aiming at optimizing the deconvolution operator of objective function,Particle Swarm Optimization(PSO)was applied to proceed a global search for deconvolution operator in order to realize deconvolution of seismic data.The results of numerical simulation and actual seismic data processing prove that,the method proposed in this paper can be suitable for minimum phase wavelet deconvolution and for mixed phase wavelet deconvolution.In contrast to the deconvolution results of gradient method with optimizing kurtosis criterion,the proposed method can obtain a better estimation of reflectivity from the seismic data,and broaden the spectrum of seismic data more effectively resulting in higher resolution of seismic data.