Research And Practice Of Data Mining In Oil And Gas Resources Information

With the development of the oil industry and the accrual of time,a large amount of oil data has accumulated in the oil fields.How to make full use of these data and find out the knowledge and laws is an important research topic,Because they can help improve the efficiency of oil companies and bring them economic benefits.But due to the amount of data is too big and data formats are different,which makes it very difficult to analyze these data through artificial methods,and it is difficult to find the hidden information and laws in these data.Data mining can help us solve this problem at this time.Data mining has become a familiar vocabulary in modern society.It has advantages over other technologies in dealing with some knowledge discovery and analysis of massive data.It has great significance to apply data mining technology to the oil industry.Data mining technology and the booming big data technology will make great contributions to the analysis of oilfield data.This paper uses the analysis method of fishbone map,collects oil and gas resource information data,and comprehensively uses data mining technology,GIS technology,big data technology and other high technologies,Through these technologies,some research and practice of data mining for oil and gas resources have been done.The main research results are divided into the fllowing sections:1.Based on reading a large amount of literature,this article summarizes the application algorithms on oil and gas resources.The application of these algorithms in the field of oil and gas resources is summarized into four aspects: predictive simulation,classification and discrimination,comprehensive evaluation,and inductive clustering.Each aspect includes different algorithms,Finally,based on the summary results,this paper designs and implements a B/S-based oil and gas resource information analysis algorithm library.Compared with the popular statistical analysis software in the market,this algorithm library is dedicated to data mining for petroleum resource data is more targeted.2.Tne traditional data mining lacks the visualization part.This article uses GIS as a visualization tool for data mining.The vector map data of 468 basins and 12,610 oilfields from the United States IHS company were collected as the original analysis data.Interpolation of the discovery time data of global oil and gas fields through GIS and superposition analysis with global basin data reveals that the number of newly discovered oil and gas fields has been decreasing year by year in terms of the discovery time of oil fields.In 2017,it even reached nearly 60 years.The lowest value.This wasattributed to the decrease in the number of undiscovered large-scale oil and gas fields and,on the other hand,to the oil and gas companies’ intentional reduction of oil and gas exploration operations due to the collapse of oil and gas prices in 2014.3.The analysis of China’s oil safety assessment based on principal component analysis is taken as an example.This paper collected raw data of the six indicators that affecting China’s oil safety from 1993-2015 through the China Statistical Yearbook,BP World Energy Statistics and other databases(including the proportion of oil in energy consumption,the intensity of oil consumption over the years,the degree of foreign dependence on oil consumption,the average price of crude oil in the three major markets,the ratio of storage and production,the replacement rate of oil reserves)Through the dimension reduction,the original indicators are compressed into environmental factors,oil supply capacity and supply risk level these three new comprehensive indicators.According to the index weights,the annual comprehensive score of petroleum safety is calculated.Finally,it is found that the petroleum safety level in recent years is showing a decreasing trend.From the analysis results,the country should focus on adjusting the level of foreign dependence of oil to improve oil security degree.4.Based on the status quo of the "data explosion and lack of knowledge" in the oilfield,a data mining framework for oil and gas resources based on the perspective of big data was designed,and the oil and gas data was analyzed from the perspective of big data.The oil and gas resources big data mining platform is divided into three layers,including the basic support layer,data mining layer,and presentation layer.Through this big data oil and gas resource data mining platform can help users provide fuel-efficient solutions,oil big data visualization,global oil consumption analysis,scheduling forecasting,Oilfield exploration forecasting,oil price forecasting,and more.Applying data mining technology and big data technology to the oil and gas industry will inevitably contribute to the future realization of smart oil fields.