支持向量机方法在油气储层参数预测中的应用

在油气储层综合研究中，储层参数的准确求取是一项关键性技术，而储层参数和地震信息之间并不存在直接的解析关系，不能用确定的函数表达式进行描述，通常采用数学统计的方法进行储层参数的预测。针对非线性函数拟合方法存在的困难，从Fourier多项式逼近的角度对非线性函数拟合的支持向量机的计算公式进行了的分析，其结论对理解和构造核函数提供了理论依据。支持向量机方法能够解决小样本情况下非线性函数拟合的通用性和推广性的问题，是求复杂的非线性拟合函数的一种非常有效的技术。模型及实例表明，该方法对油气储层参数的预测是有效的。

APPLICATION OF SVM METHOD TO PREDICTING RESERVOIR PARAMETER

It is a key technique to accurately determine the reservoir parameters when conducting reservoir comprehensive researches. But the reservoir parameters are usually predicted by mathematical statistical methods because without the direct analytical relations between reservoir parameters and seismic information not to be described by definite function expression. In order to overcome the difficulties in the methods for nonlinear function approximation, the support vector machine method (SVM) is introduced and the formulae are educed based on Fourier polynomial, which provide a theoretical basis for understanding and constructing the kernel function. The problems of universality and extensibility in nonlinear function approximation using small samples can be solved by the method, it a very efficient technique for nonlinear function approximation. Living Examples indicate that this method can be applied to effectively predicting reservoir parameters.