河北地质大学学报2023，Vol.46Issue(6)：28-34,7.DOI:10.13937/j.cnki.hbdzdxxb.2023.06.004
基于常规测井资料的储层流体识别
Reservoir Logging Fluid Identification Based on Conventional Logging Data
摘要
Abstract
Due to factors such as physical properties and fluid properties of the long 3 reservoir in the M oilfield resulting in complex logging response characteristics information,it is difficult to accurately identify reservoir fluid properties using single logging information.In order to improve the accuracy of identifying the fluid properties of the long 3 reservoir in the study area,the general methods of fluid identification analysis were sorted out based on a large amount of literature,and a comparison of the application of multiple methods of fluid identification was carried out in combination with the existing actual logging information,commonly used discriminatory methods to achieve the purpose of rapid and accurate identification of fluid properties that are difficult to achieve.Therefore,this paper applies the random forest principle,which can combine multiple information analysis and less human interference,to the fluid property identification of reservoirs in the study area,and the method has been tested to achieve the best identification results with an accuracy rate of 90.47%,which provides an idea for intelligent algorithms to solve geological problems.关键词
测井解释/流体识别/随机森林Key words
well logging interpretation/fluid identification/random forest分类
能源科技引用本文复制引用
许鉴源,赵军龙,白倩,黄千玲..基于常规测井资料的储层流体识别[J].河北地质大学学报,2023,46(6):28-34,7.基金项目
陕西省自然科学基础研究计划(2019JM-359). （2019JM-359）
