基于粒子群和支持向量机的油藏物性参数拟合方法

摘要

Using the traditional methods of reservoir physical parameters to fit,the computa-tion complexity is high,and the fitting is overhead.A method of reservoir physical property pa-rameter fitting based on particle swarm optimization and support vector machine algorithm is thus proposed.The data structure characteristics of reservoir physical parameters are analyzed, and the clustering process and feature extraction of physical parameters of reservoir are carried out by using particle swarm optimization algorithm.The support vector machine algorithm is used to improve the characteristic parameter of restructuring and least squares fitting.It reali-zes the accurate data mining of the physical parameters of the reservoir,and improves the accu-racy of fitting.The simulation results show that using this method to analyze and fit the data characteristics of reservoir physical parameters is better.The extracted parameter can effective-ly reflect the physical characteristics of the oil reservoir,and it has good application value in guiding the exploration and oil exploitation of the oil field.%采用传统方法对油藏物性参数进行拟合时，计算过程复杂度较高，拟合开销较大，提出一种基于粒子群和支持向量机算法的油藏物性参数拟合方法。对油藏物性参数的数据结构特征进行分析，采用粒子群差分扰动进化算法，进行油藏物性参数的聚类处理和特征提取，通过支持向量机算法完成参数的特征重组和最小二乘拟合，实现对油藏物性参数特征量的准确挖掘，提高拟合精度。仿真结果表明，采用该方法进行油藏物性参数的数据特征分析和拟合，准确性较好，提取的参数特征能有效反应油藏的物理特性，在指导油田勘探和石油开采中，具有较好的应用价值。