The reservoir of Gudong oilfield has experienced four periods of development,Logs can only reflect reservoir characteristics in the certain period that the oilfield is experiencing.So how to infer log response in other development period with change rule of logs and reservoir,for enriching logging information during oilfield description and simulation in different stages,is the research aim of this study.Characteristics of log is consisted of log's value and pattern.Representative value is a focus symbol of a curve.Curve shape,including overall shape and sawtooth level.This paper discusses all kind of mathematical ways to calculate representative value,filtering,automatic dividing layering,quantitative characterization shape and sawtooth level.Combined coring and logging data of nineteen core wells,summarizing the logging response characteristics of different lithofacies,summing up the physical properties,oil-bearing properties and permeability of different lithofacies before and after flooding,dividing the reservoirs into three types of petrophysical phases,and conducting the logging response characteristics in different periods.Using data mining algorithms to identify the logging curves in different periods,complete the characteristics of logging curves in different periods,and convert the curves in different periods. |