石油物探2017，Vol.56Issue(1)：31-37,74,8.DOI:10.3969/j.issn.1000-1441.2017.01.004
基于照明预处理的分步多参数时间域声波全波形反演方法研究
The stepped multi-parameter FWI of acoustic media in time-domain by L-BFGS method with illumination analysis
摘要
Abstract
Density can predict fluid saturation of reservoir and plays an important role in reservoir interpretation and hydrocarbon prediction.Due to the cross-talk effects of velocity and density,density is difficult to reconstruct in multi-parameter full waveform inversion.To solve this problem,the strategy of stepped multi-parameter FWI is chosen which takes more accurate velocity inversion result as the initial velocity model and combines with the initial density model for next step simultaneous multi-parameter inversion.The more accurate the initial velocity model is,the more accuracy the density result is.In the meantime,to further balance the energy of gradient and reduce the influence of geometric diffusion in seismic wave propagation,the preprocessing L-BFGS method is carried out based on the illumination analysis.The numerical examples testify the feasibility of the stepped multi-parameter FWI and the pretreatment L-BFGS method that they can improve accuracy of result.关键词
多参数全波形反演/分步反演/L-BFGS法/照明分析/声波介质Key words
multi-parameter FWI/stepped inversion/L-BFGS/illumination analysis/acoustic media分类
天文与地球科学引用本文复制引用
张广智,姜岚杰,孙昌路,黄义双..基于照明预处理的分步多参数时间域声波全波形反演方法研究[J].石油物探,2017,56(1):31-37,74,8.基金项目
国家自然科学基金(41674130)、国家科技重大专项(2016ZX05027004-001,2016ZX05002-005-009)、国家重点基础研究发展计划(973计划)项目(2014CB239201-7HZ)、中央高校基本科研业务费专项资金(15CX08002A)联合资助.This research is financially supported by the National Natural Science Foundation of China (Grant No.41674130),the National Science and Technology Major Project of China (Grant Nos.2016ZX05027004-001,2016ZX05002005-009) (973 Program),the National Key Basic Research and Development Program of China (Grant No.2014CB239201-7HZ) and the Fundamental Research Funds for the Central Universities (Grant No.15CX08002A). （41674130）
