多元测井序列的预分段逐次最优二分割法

A pre-divided optimal dimidiate partitioning method of multi-log sequence

基于有序样品聚类分析理论的最优分割法存在遍历算法的常见不足,这使它在应用中计算量巨大,严重影响其应用。文中介绍了一种最优分割法的优化方法,是通过VB语言编程实现。即通过对测井资料作预分段后的逐次最优二分割,来使计算量大大降低,从而解决最优分割的应用问题。

The optimal partitioning method based on the cluster analysis of the ordered samples possesses a familiar flaw as all the ransacked programs do. It makes the optimal partitioning method have a huge computational complexity which badly baffles its spreading of applications. This paper presents a modified method based on the optimal partitioning method, which deals with the log data with a pre-divided optimal dimidiate partitioning method. It will enormously reduce the computational complexity, so it can resolve the problem in its application. This method is actualized in VB language. It works rapidly and gives favorable results.