物探化探计算技术2016,Vol.38Issue(2):212-218,7.DOI:10.3969/j.issn.1001-1749.2016.02.11
基于粒子群优化算法的多元线性拟合方法研究及其应用
The particle swarm optimization research and application based on multivariate linear fitting method
摘要
Abstract
In the actual logging ,density curve is the most vulnerable to be influenced by hole enlargement .In order to elim‐inate the influence ,the particle swarm optimization is introduced to the multivariate linear fitting method .Particle swarm opti‐mization algorithm is used to optimizing the objective function ,which is the intelligent evolutionary algorithm with adaptive control .In this paper ,building the multivariate linear fitting model based the particle swarm optimization with some logging curves (such as gammar ray ,resistivity and acoustic) ,which is affected relatively small by the borehole environment in the ref ‐erence layer having a level borehole and the same lithology with the position in the bad bore environment .Then we use this model to reconstruct the density curve in the layer where the borehole is enlarging .Finally ,the reconstructed density curve compares with original density curve ,the density curve calculated by gardner formula .The research results show that the cor‐relation coefficient between the seismic traces near the well and the synthetic seismic record using the density reconstructed by the multivariate linear fitting methods reached 0 .84 .It indicated that the proposed multivariate linear fitting method can effec‐tively improve the quality of density logging curves .关键词
粒子群算法/多元线性拟合/井间扩径/校正密度Key words
PSO/multivariate linear fitting/bore expanding/correction density分类
天文与地球科学引用本文复制引用
韩家兴,吴施楷,田仁飞,李杰,杨宽..基于粒子群优化算法的多元线性拟合方法研究及其应用[J].物探化探计算技术,2016,38(2):212-218,7.基金项目
国家自然科学基金 ()