Review of research progress on methods to improve the longitudinal resolution of thin reservoir logging curves
The longitudinal resolution of the logging curve will be reduced under the influence of adjacent surrounding rocks,which will make the logging response values of the logging curve differ from the true logging values of the formation,and these gaps will affect the identification and tapping of thin reservoirs in the process of logging interpretation.In order to understand the current status of research on methods to improve the longitudinal resolution of logging curves,this paper summarizes these methods into three major categories through literature research:(1)Methods based on logging principles,which are now the mainstream methods to improve the longitudinal resolution of logging curves,and these methods have geophysical significance and are theoretically mature and easy to implement in operation,but these methods are generally suitable for dealing with logging with linear response characteristics;(2)Mathematical function and formula methods,which are supported by mature mathematical theories,easy to program and fast to process data,and can be combined with other methods to process logging curves,but there is basically no effect of increasing logging curve resolution when applying conventional interpolation methods to logging curves.Moreover,the phenomenon of peak shift may occur,so this type of method needs more theoretical exploration and research;(3)Time-frequency analysis and machine learning methods,such methods are diverse and can be combined with a variety of time-frequency analysis and machine learning theory,which is the trend of future development,but not much is applied to improve the resolution of the logging curve,and several experiments are needed to determine the optimal resolution of the logging curve,where the time-frequency analysis method may show peak shift in the application process.The paper concludes with a comparison of the advantages and disadvantages of the three methods and gives recommendations for the selection of the corresponding methods.
万方数据
维普
