Elastic parameter change rate multi-stage pre-stack inversion based on immune genetic algorithm
Zhang Hai-yan1, Liu Huai-shan2, Tong Si-you2, Zhang Jin2, Wang Xin3
1. School of Information Sciences and Engineering, Ocean University of China, Qingdao City, Shandong Province, 266071, China;
2. School of Earth Sciences, Ocean University of China, Qingdao City, Shandong Province, 266071, China;
3. Geophysical Technique Research Centre, BGP Inc. of CNPC, Zhuozhou City, Hebei Province, 072750, China
Abstract:Base on the studies of self-adaptive immune genetic algorithm developed by Zhang Hai-yan (2008) and using for reference of inversion method of bi-directional reflection physical model in multi-angle remote sensing,in this paper studies on elastic parameter change rate multi-stage pre-stack inversion based on immune genetic algorithm was conducted as the following: at first part of the observed data were used to invert the most sensitive parameters of the model,so the effective application of the limited information in the observed data was estimated,secondly after the uncertainty of the parameters was reduced,other part of the observed data were used to invert the other part of parameters and etc. Therefore in multi-stage inversion not all observed data were used at one time to simultaneously invert all the unknown parameters,it is an inversion process which is divided into multi-stage,during the process the uncertainty of the sensitive parameters was reduced gradually. For the three layer oilgas bearing sandstone model,comparison experiment and field data trial calculation of one-stage inversion and multi-stage inversion were conducted,it is noticed in the experiment and calculation that the near offset trace PP wave reflectivity is sensitive to Δρ/ρ and ΔVP /VP,while the far offset trace PP wave reflectivity is more sensitive to ΔVS /VS.