文献类型：会议论文

中文题名：PREDICTING SANDSTONE THICKNESS OF UNDERGROUND RESERVOIR BASED ON RADIAL BASIC FUNCTION NETWORK

作者：JINGWEN TIAN [1];MEIJUAN GAO [1];KAI LI [2];LISHUAN HU [3];

机构：[1]DEPARTMENT OF AUTOMATIC CONTROL,BEIJING UNION UNIVERSITY,BEIJING 100101,P.R.CHINA;SCHOOL OF INFORMATION SCIENCE,BEIJING UNIVERSITY OF CHEMICAL TECHNOLOGY BEIJING 100029,P.R.CHINA;[2]SCHOOL OF INFORMATION SCIENCE,BEIJING UNIVERSITY OF CHEMICAL TECHNOLOGY BEIJING 100029,P.R.CHINA;[3]DEPARTMENT OF AUTOMATIC CONTROL,BEIJING UNION UNIVERSITY,BEIJING 100101,P.R.CHINA;

第一机构：北京联合大学城市轨道交通与物流学院

会议论文集：2007 INTERNATIONAL SYMPOSIUM ON COMPUTER SCIENCE AND TECHNOLOGY(2007国际计算机科学与技术学术大会)(ISCST2007')论文集

会议日期：2007年1月1日

会议地点：宁波

主办单位：全国35所示范性软件职业技术学院院长联合会

语种：英文

中文关键词：RADIAL BASIC FUNCTION NETWORK;SEISMIC CHARACTERISTIC PARAMETER;SANDSTONE THICKNESS;UNDERGROUND RESERVOIR;PREDICTING

摘要：TODAY USING MULTIPLE SEISMIC CHARACTERISTIC PARAMETERS TO PREDICT THE SANDSTONE THICKNESS OF UNDERGROUND THIN INTERBEDDED RESERVOIR IS AN IMPORTANT RESEARCH SUBJECT IN THE PETROLEUM EXPLORATION.TRADITIONAL BP NETWORK CAN NOT GET CONTENTMENT RESULT TO PREDICT THE SANDSTONE THICKNESS.IN POINT OF THE THIN INTER-BEDDED RESERVOIR THE FEATURES OF RESERVOIR THICKNESS THIN AND THE GREAT CHANGE IN HORIZONTAL AND COMPLEXITY OF THE RESERVOIR,THIS PAPER OFFERS A NEW SANDSTONE THICKNESS PREDICTION METHOD OF UNDERGROUND RESERVOIR--RADIAL BASIC FUNCTION NETWORK.RADIAL BASIC FUNCTION NETWORKS IS A LOCAL APPROACHING NETWORKS,IT HAS SOME ADVANTAGES OF LEARNING SPEED RAPIDLY AND FINE ABILITY OF FUNCTION APPROACHING AND MODEL CLASSIFY,THEREFORE USING THE NEW METHOD IN THIS PAPER CAN IMPROVE THE PREDICTION ACCURACY OF UNDERGROUND RESERVOIR SANDSTONE THICKNESS AND ENHANCE THE NETWORK CONVERGENCE SPEED.THE PREDICTION RESULT OF AN EXAMPLE PROVES THIS METHOD FEASIBLE AND PRACTICABLITY.