基于逆算子估计的高阶AVO非线性反演

High-order nonlinear AVO inversion based on estimated inverse operator

Zoeppritz方程线性近似式已被广泛应用于AVO反演方法研究中,但在界面差异明显时线性近似式难以满足精度要求,而高阶非线性近似式有助于解决此类问题。鉴于大多数优化算法并不太适用于求解非线性反演问题,本文提出一种新的AVO非线性反演方法:利用逆算子估计算法求解AVO反演问题,其直接求逆的思路有别于传统优化类反演算法的搜索方式,多解性显著减弱;应用Zoeppritz方程高阶近似式建立反演目标函数,提高了界面两侧差异明显时线性近似精度,还兼顾了纵横波速度比对AVO反演的影响。模型试算得知反演结果与模型值相符;实际应用结果表明,三参数反演结果与地震、测井信息吻合较好。这套基于逆算子估计的AVO反演方法具有较强的稳定性和可靠性,运用高阶近似式及选取合适的纵横波速度比是确保反演结果具有高精度的关键。

In an AVO inversion process, Zoeppritz linear approximation is widely utilized. However, the approximation shows a significant disparity with the exact equation when the two media across the interface vary dramatically because these equations are under the assumption of minor differences between the media. One possible approach to improve the accuracy is developing high-order approximations. This paper presents a new AVO nonlinear inversion method in which inverse operator estimation algorithm is utilized. Compared with conventional optimization-class inversion algorithms, this method is a direct inverse method and it owns its unique advantages. The inversion objective function uses high-order approximation Zoeppritz equation to improve precision when the elastic parameters between both sides of the interface vary dramatically. Ratio of the P- and S-waves' velocity is taken into account as well. Model tests and real data results show that AVO inversion based on inverse operator estimation algorithm owns a high stability and reliability. A better inversion result can be obtained when high order approximation and a changing ratio of the P- and S-waves' velocity are utilized. © 2016, Editorial Department OIL GEOPHYSICAL PROSPECTING. All right reserved.