Research And System Development Of Lithologic Identification Algorithm Based On Log Data

Traditional logging data identification methods are slow,low-precision,human-oriented and unfavorable for practical engineering practice.Therefore,the construction of a lithologic automatic identification method with high speed,high recognition rate and good promotion ability has become an urgent need in the field of well log data interpretation.At the same time,with the rapid development of well logging technology and computer technology,how to do more immediate and direct analysis of logging data for lithologic results is a matter of current consideration.Based on the purpose of lithologic identification,this paper mainly proceeds from the following aspects: logging data processing and storage,logging data segmentation,lithologic identification method research,system development and implementation,and evaluation of the final application results.analysis.(1)Firstly,the structure of the logging data is discussed.Then,the dirty data in the logging data is processed and eliminated.Finally,it is imported into the CIFLog platform for storage and management.(2)Study the logging data segmentation method,select and improve the logging data segmentation algorithm,and finally select the moving average algorithm to segment the logging data and implement the segmentation method.The segmentation of logging data is mainly to segment the existing well logging data in accordance with the lithology categories,which is conducive to adopting suitable lithology identification methods for different logging data segments.(3)This paper mainly discusses three kinds of lithologic identification methods:crossplot recognition method,neural network method,support vector machine method,and class structure design of its algorithm,and finally achieve the above lithology recognition algorithm.(4)In the part of system development and implementation,the requirements of the system are first analyzed,the overall architecture of the system is designed,and each functional module is designed in detail.Finally,the above research was implemented and a relatively complete lithology identification subsystem was developed on the CIFLog platform.(5)The well logs that need to be identified are properly processed and imported into the CIFLog platform.Then they are segmented and lithologically identified.At the same time,MATLAB software is used to identify the same data.Finally,the two types of identification results are mapped to a lithology profile and analyzed based on the existing borehole core.This paper studies the structure,storage,and management of logging data,andanalyzes lithologic identification methods,and designs and implements them.It provides scientific method support and theoretical basis for lithologic identification,and implements application to engineering at the same time.Among them,the exploration of the entire oil field has played an important practical significance.