Research On The Automatic Zonation Methods Of Well Logs And Its Application On The Lithology Identification

Currently the mainly methods of zonation of well logs is handwork,the method not only consumes time and labor ,the result of identification of rock bundaries may be affacted by subjective factors of interpretater.but automatic zonation methods of well logs could prevent the approximate definition of rock boundaries and make the process of zonation more effective.As the introduction of the computer, data process and mathematical algorithms,the maily idea and implementaion of the automatic zonation of well logs have been changing all the time.Because the principle of automatic zonation methods of well logs based on activity is simple and require less calculation,the method had been applied extensively in the automatic zonation.While, only one log could be used in automatic zonation process based on activity,and the quality of zonation relates to the quality of log data and its application, the single log often include noise and limited amount of information,therefore,in order to reach a good zonation result by automatic zonation method of well logs based on activity,it is essential to process well logs data,the well logs blend is advanced under such a background.This thesis deals with the meaning of the well logs blend process from the Mathematical Statistics and the signal processing ,and the basic principle of the automatic zonation method of well logs based on activity. Furthermore, the corresponding program has been prepared with C language and was used to process the practical well logs data and identify the lithology based on the result of the automatic zonation,finally derive the parameter settings empirical formulas from summarizing the log interpretation process.It is not the best choice to choose log at random in the logs blend process,but the log whose sum of correlation coefficient with other well logs is larger,and the SP log may not be better choices. The parameter settings is very important in automatic zonation process based on acitivity, Especially the window length and threshold of the activity.Activity was a normal distribution,the best parameter settings of the threshold of the activity is the value whose cumulative distribute frequency is 68.2%.