The Application of a Partheno-Genetic Algorithm's Pattern Clustering in Complicated Lithology Identificati

摘要

Considering the classification indetermination and diversity in petroleum reservoir geology's complicated lithology partition, a method based on Partheno-Genetic Algorithm's pattern clustering is proposed in this article. This method realized evolving operation by genetic operatiors such as genetic transposition, genetic shift and so on which hide Traditional Genetic Algorithm's crossover operators' function, predigest heredity process and avoids "prematurity convergence' problem at the same time. Using PGA to solve pattern clustering problem make clustering's result independence of initial clustering. By coping with Daquing oil field's real data, the method is proved effective.