Prestack Gather Analysis And Optimization Method Research

Prestack gathers optimization technique is the preparatory work of prestack avo inversion, and is important technology of the lithologic reservoir exploration. The way of the optimized set can help explain the AVO characteristics of underground medium. In the view of the real project data obtained for various reasons lead to the image quality is not high, for pre-stack seismic gathers rich in random noise, curved noise and having the problem of unflatten gathers, missing some traces so we need to optimize the quality of prestack data set.First, this article discussed the development of optimization technology and the research status at home and abroad. Then this article introduces the whole process of the optimization technology and principle, the process of comparing with various techniques and the improvement of my optimization. Finally this paper proves the technique has better effects from theory and actual application. In this paper, the main innovation points are as follows:1. For prestack data is rich of noise, to suppress the irregular random noise, this paper adopted L2 norm and L4 norm regularization term hybrid method, combining with L2 norm of good suppress gaussian noise ability, four norm of good suppress sub-gaussian noise ability, can effectively remove the random noise. For the interference of multiple wave, which appears as the curve. this article adopts the method of the Radon transform, in transform domain we can easily remove the curve noise and inverse transform back to the original domain.2. This thesis put forward a algorithm Segmental DTW which is based on DTW(Dynamic Time Warp) and introduce the specific process of this algorithm. The algorithm incorporates DTW which is widely used in speech recognition into matching similar section in a sequence of seismic signal sampling points, get the best matching path, and then based on the best matching path to find the best matching key section, shift the best matching section, eliminate the phenomenon of excessive tension or compression. By the test data, this method has a good effect to achieve the event flatten and retain the original seismic feature.3. This paper proposes an adaptive tight wavelet framework to recover the missing traces data. Based on the structural characteristics of seismic data, we formulate a proper tight framework,through effective sparse transformation, the original image can be represented by part information in the transform domain data, using split bregman iterative algorithm to construct the objective function, this can both reduce the computational complexity and recover missing information of seismic signals. Besides holding the original useful information, this can have a good effect of recovering the random missing traces and non-random missing traces.In this paper,the optimization algorithm is used to different work areas. From the point of practical effect, for noise removal, multiple wave suppression, flattening and the lack of information has a good recovery, gathers quality significantly improved.