The Multiple-factor Prediction Of Favorable Area For Oil And Gas Resources Based On GIS

The prediction of oil and gas favorable area is the core work of oil and gas exploration and development,which is related to the cost and efficiency of oil and gas resource development.With the development of digital reservoir,the technology of spatial analysis and prediction based on GIS has been used in oil and gas field for years.The prediction of oil and gas favorable area involves the comprehensive analysis of multidisciplinary results,which will be an important issue that needs to be further studied in the exploration and development of oil and gas.Using GIS to predict the favorable area of oil and gas resources has the characteristics of automation,fast,timeliness,update.It also can reduce the dependence of professional researchers on geological study.In this paper,the research on the automatic selection of multiple-factor prediction and the method of the automatic prediction model of oil and gas favorable area have been carried out.The multiple-factor prediction system of oil and gas resource favorable areas has been developed by R language,GIS technology and.NET.The following are the main contents of this paper:(1)The spatial database of the gas and oil factors information and many geological map based on GIS has been created.Layers about multiple-factor favorable areas is also created.Complex sample data include Wells data can be stored well.(2)The research about multidisciplinary work datum standardization method has been completed,which unify vast amounts of data about oil and gas resources in different formats such as map spatial reference,cell,mapsheet and so on.(3)The multiple-factor prediction model based on GIS has been established.With the help of a lot of oil and gas resources data and many kinds of factor selection methods and the optimal prediction algorithm researching,the Multiple-factor automated Prediction of Favorable Area For oil and gas Resources has been realized by randomforest factor screening method and randomforest decision classification scheme.Taking the data of a block in Sulige gas field as an example,this system is used to analyze and deal with the sample data of hundreds of wells in the research block and the multiple-factor results about oil and gas fields,and establishing algorithm model.The system can achieve the desired effect by comparing with the prediction of professional staff.It is hoped that it will help professional researchers to draw conclusions by providing some reference.