Research On Big Data Disk Storage And Data Mining For Low Permeability Oil Well

Data storage technology and data mining technology have been promoted vigorously in the field of petroleum industry in recent years.On the one hand, oilfield enterprises are undergoing reform of information technology, and in practical production, lots of production data are accumulated, which should be managed scientifically.On the other hand, With the development of data mining theory, data mining technology has been applied to the practical application widely,such as commercial, engineering and so on. For oil field enterprises, the establishment of integrated storage system will solve the problem of massive data management,and useful information and knowledge could be extracted from the massive data with the help of data mining technology. Data storage technology and data mining technology will provide more help for the development of the oil field.This paper aims to complete the design of oil field data storage system and data mining applications, the main research work is as follows:(1)Put forward the solution of massive optical disk data storage system in oil field which is combined with the optical disk library and PLC,after analyzed the data storage requirement of the oil field.The storage system is divided into four levels, which are customer application layer, service management layer, data storage layer and data collection layer.(2)Put forward an improved disc file system which is based on existing UDF file system.The improved disc file system improved the disc scanning speed and enhance the disc recording fault tolerance.(3)Discussed the general design scheme of oil field data warehouse according to the actual demand of oil field production,then the structure design of the fact table and dimension table of oil field data warehouse is given.(4)Studied the cluster analysis algorithm which is based on density—DBSCAN,according to the characteristics of oil field production data. The DBSCAN algorithm is optimized by using spatial index technology,in this way,the time performance of the improved DBSCAN algorithm is enhanced when processing large scale data,which lay the foundation for the application of data mining in oil field production.(5)Used the improved clustering algorithm to achieve the two types of oilfield data mining applications: oil well classification and reservoir classification.Then,the results obtained by clustering the proposed policy recommendations for the development of oilfield production.This paper built the model of oilfield data management through the application of data storage and data mining,and realized the whole process of production data acquisition and storage.At the same time, used data mining techniques to extract rules and knowledge from the production data,which provides a reliable basis for decision makers to optimize the production system and improve the production efficiency.