XGBoost算法在致密砂岩气储层测井解释中的应用
|
摘要
传统单一模型的机器学习方法用于致密砂岩气储层测井解释时存在多解性,为此,将XGBoost算法应用于致密砂岩气储层测井解释。基于A工区测井解释资料,以不同种类的测井资料作为输入变量,通过XGBoost算法建立回归预测模型,预测该区孔隙度与渗透率参数,并探讨了XGBoost算法中各类参数的优化。以准确率指标为评价标准,通过XGBoost算法建立的分类预测模型对该区储层类型进行预测,同时与随机森林方法和支持向量机算法进行比较,XGBoost算法的预测效果较好。结果表明XGBoost算法能准确地预测孔隙度、渗透率并对该工区致密砂岩气层进行有效识别。
Conventional single-model machine learning methods used in tight-sand gas reservoir interpretation on well logging data have the multi-solution problem.To overcome this problem,we use the XGBoost algorithm.Based on logging data in the Area A,different types of well logging data are used as input variables,and a regression prediction model is established by XGBoost algorithm.The porosity and permeability in this area are predicted.The optimization of various parameters in XGBoost algorithm is also discussed.The classification prediction model established by XGBoost algorithm predicts reservoir types in the area.Based on our prediction results,the XGBoost algorithm achieves a better porosity & permeability prediction and tight-sand gas reservoir identification in the area compared with the random forest method and vector-supported machine algorithms.
Conventional single-model machine learning methods used in tight-sand gas reservoir interpretation on well logging data have the multi-solution problem.To overcome this problem,we use the XGBoost algorithm.Based on logging data in the Area A,different types of well logging data are used as input variables,and a regression prediction model is established by XGBoost algorithm.The porosity and permeability in this area are predicted.The optimization of various parameters in XGBoost algorithm is also discussed.The classification prediction model established by XGBoost algorithm predicts reservoir types in the area.Based on our prediction results,the XGBoost algorithm achieves a better porosity & permeability prediction and tight-sand gas reservoir identification in the area compared with the random forest method and vector-supported machine algorithms.
引文
[1] 罗瑞兰,雷群,范继武,等.低渗透致密气藏压裂气井动态储量预测新方法——以苏里格气田为例[J].天然气工业,2010,30(7):28-31.LUO Ruilan,LEI Qun,FAN Jiwu,et al.A new approach to the dynamic reserve estimation of fractured gas wells in low-permeability tight gas reservoirs:The Sulige Gas Field as a case study[J].Natural Gas Industry,2010,30(7):28-31.
[2] 赵彦超,吴春萍,吴东平.致密砂岩气层的测井评价——以鄂尔多斯盆地大牛地山西组一段气田为例.地质科技情报[J],2003,22(4):65-70.ZHAO Yanchao,WU Chunping,WU Dongping.Logging evaluation to tight gas sandstone:a case study from the first member of Shanxi formation in Daniudi gas pool,Ordos basin,China.Geological Science and Technology Information,2003,22(4):65-70.
[3] 杨双定.鄂尔多斯盆地致密砂岩气层测井评价新技术[J].天然气工业,2005,25(9):45-47.YANG Shuangding.New methods of Log Evaluation of the Tight Sandstone Gas Reservoirs in E’erduosi Basin[J].Natural Gas Industry,2005,25(9):45-47.
[4] 魏钦廉,肖玲,李康悌.常规测井识别致密地层含气性[J].特种油气藏,2007,14(4):22-25.WEI Qinlian,XIAO Ling,LI Kangti.Identify gas potential in tight reservoir by conventional logging[J].Special Oil and Gas Reservoirs,2007,14(4):22-25.
[5] 章雄,潘和平,骆淼,等.致密砂岩气层测井解释方法综述[J].工程地球物理学报,2005,2(6):431-436.ZHANG Xiong,PAN Heping,LUO Miao,et al.Summary of logging interpretation for gas-bearing tight sandstone[J].Chinese Journal of Engineering Geophysics,2005,2(6):431-436.
[6] Rogers S J,Fang J H,Karr C L,et al.Determination of lithology from well logs using a neural network[J].AAPG Bulletin,1992,76(5):731-739.
[7] Baldwin J L,Bateman R M,Wheatley C L.Application of a neural network to the problem of mineral identification from well logs[J].Log Analyst,1990,3(5):279-293.
[8] 张向君,张晔.基于支持向量机的交互检验储层预测[J].石油物探,2018,57(4):597-600.ZHANG Xiangjun,ZHANG Ye.Reservoir prediction through cross-validation based on support vector machine[J].Geophysical Prospecting for Petroleum,2018,57(4):597-600.
[9] Anazi A,Gates I D.On the capability of support vector machines to classify lithology from well logs[J].Natural Resources Research,2010,19(2):125-139.
[10] Sebtosheikh M A,Motafakkerfard R,Riahi M A,et al.Support vector machine method,a new technique for lithology prediction in an Iranian heterogeneous carbonate reservoir using petrophysical well logs[J].Carbonates Evaporites,2015,30(1):59-68.
[11] Shi N,Li H,Luo W.Data mining and well logging interpretation:application to a conglomerate reservoir[J].Applied Geophysics,2015,12(2):263-272.
[12] Li H,Tan F,Xu C,et al.Lithology identification of conglomerate reservoir based on decision tree method[J].Well Logging Technology,2010,34(1):16-21.
[13] 柴明锐,程丹,张昌民,等.机器学习方法对砂砾岩岩屑成分的预测——以西北缘X723井百口泉组为例[J].西安石油大学学报(自然科学版),2017,32(5):22-28.CHAI Mingrui,CHENG Dan,ZHANG Changmin,et al.Prediction of debris composition in glutenite by machine learning method:a case study in Baikouquan formation of well X723 in the NW Margin of Junggar Basin[J].Journal of Xi’an Shiyou University(Natural Science Edition),2017,32(5):22-28.
[14] 李诒靖,郭海湘,李亚楠,等.一种基于Boosting的集成学习算法在不均衡数据中的分类[J].系统工程理论与实践,2016,36(1):189-199.LI Yijing,GUO Haixiang,LI Ya’nan,et al.A boosting based ensemble learning algorithm in imbalanced data classification[J].Systems Engingeering Theory & Practice,2016,36(1):189-199.
[15] 宋建国,杨璐,高强山,等.强容噪性随机森林算法在地震储层预测中的应用[J].石油地球物理勘探,2018,53(5):954-960.SONG Jianguo,YANG Lu,GAO Qiangshan,et al.Strong tolerance random forest algorithm in seismic reservoir prediction [J].Oil Geophysical Prospecting,2018,53(5):954-960.
[16] Cracknell M J,Reading A M.The upside of uncertainty:Identification of lithology contact zones from airborne geophysics and satellite data using random forest and support vector machines[J].Geophysics,2013,78(3):WB113-WB126.
[17] 周雪晴,张占松,张超谟,等.基于粗糙集——随机森林算法的复杂岩性识别[J].大庆石油地质与开发,2017,36(6):127-133.ZHOU Xueqing,ZHANG Zhansong,ZHANG Chaomo,et al.Complex lithologic identification based on rough set-random forest algorithm[J].Petroleum Geology and Oilfield Development in Daqing,2017,36(6):127-133.
[18] Chen T,Guestrin C.XGBoost:A scalable tree boos-ting system[C].ACM SIGKDD International Conference on Knowledge Discovery and Data Mining,2016,785-794.
[19] Torlay L,Perrone-Bertolotti M,Thomas E,et al.Machine learning-XGBoost analysis of language networks to classify patients with epilepsy[J].Brain Informa-tics,2017,4(3):159-169.
[20] 贾文慧,孙林子,景英川.基于XGBoost模型的股骨颈骨折手术预后质量评分预测[J].太原理工大学学报,2018,49(1):174-178.JIA Wenhui,SUN Linzi,JING Yingchuan.Surgical prognosis quality score of femoral neck fracture base on XGBoost model [J].Journal of Taiyuan University of Technology,2018,49(1):174-178.
[21] Friedman J H.Greedy function approximation:A gradient boosting machine[J].The Annals of Statistics,2001,29(5):1189-1232.
[22] 范铭涛,沈全意,吴辉,等.复杂岩性裂缝—孔隙型储层孔隙度计算方法研究[J].天然气工业,2005,25(5):29-30.FAN Mingtao,SHEN Quanyi,WU Hui,et al.Porosity calculation method of complex lithological fractured-porous reservoir[J].Natural Gas Industry,2005,25(5):29-30.
[23] 牛聪,张益明,王迪,等.LA地区盒8段优质储层的特征与分布预测[J].石油地球物理勘探,2017,52(3):591-598.NIU Cong,ZHANG Yiming,WANG Di,et al.Prediction of high-quality reservoir characteristics and distribution in the area LA[J].Oil Geophysical Prospecting,2017,52(3):591-598.
[24] 方兴,孙夕平,张明,等.基于AVO流体反演的储层孔隙度预测技术[J].石油地球物理勘探,2012,47(3):469-472.FANG Xing,SUN Xiping,ZHANG Ming,et al.Re-servoir porosity prediction based on AVO fluid inver-sion[J].Oil Geophysical Prospecting,2012,47(3):469-472.
[25] 潘豪杰,张研,李红兵,等.基于贝叶斯理论的天然气水合物储层弹性—电性数据联合反演[J].石油地球物理勘探,2018,53(3):568-577,586.PAN Haojie,ZHANG Yan,LI Hongbing,et al.Joint inversion of elastic-electrical data for gas hydrate re-servoirs based on Bayesian theory[J].Oil Geophysical Prospecting,2018,53(3):568-577,586.
[2] 赵彦超,吴春萍,吴东平.致密砂岩气层的测井评价——以鄂尔多斯盆地大牛地山西组一段气田为例.地质科技情报[J],2003,22(4):65-70.ZHAO Yanchao,WU Chunping,WU Dongping.Logging evaluation to tight gas sandstone:a case study from the first member of Shanxi formation in Daniudi gas pool,Ordos basin,China.Geological Science and Technology Information,2003,22(4):65-70.
[3] 杨双定.鄂尔多斯盆地致密砂岩气层测井评价新技术[J].天然气工业,2005,25(9):45-47.YANG Shuangding.New methods of Log Evaluation of the Tight Sandstone Gas Reservoirs in E’erduosi Basin[J].Natural Gas Industry,2005,25(9):45-47.
[4] 魏钦廉,肖玲,李康悌.常规测井识别致密地层含气性[J].特种油气藏,2007,14(4):22-25.WEI Qinlian,XIAO Ling,LI Kangti.Identify gas potential in tight reservoir by conventional logging[J].Special Oil and Gas Reservoirs,2007,14(4):22-25.
[5] 章雄,潘和平,骆淼,等.致密砂岩气层测井解释方法综述[J].工程地球物理学报,2005,2(6):431-436.ZHANG Xiong,PAN Heping,LUO Miao,et al.Summary of logging interpretation for gas-bearing tight sandstone[J].Chinese Journal of Engineering Geophysics,2005,2(6):431-436.
[6] Rogers S J,Fang J H,Karr C L,et al.Determination of lithology from well logs using a neural network[J].AAPG Bulletin,1992,76(5):731-739.
[7] Baldwin J L,Bateman R M,Wheatley C L.Application of a neural network to the problem of mineral identification from well logs[J].Log Analyst,1990,3(5):279-293.
[8] 张向君,张晔.基于支持向量机的交互检验储层预测[J].石油物探,2018,57(4):597-600.ZHANG Xiangjun,ZHANG Ye.Reservoir prediction through cross-validation based on support vector machine[J].Geophysical Prospecting for Petroleum,2018,57(4):597-600.
[9] Anazi A,Gates I D.On the capability of support vector machines to classify lithology from well logs[J].Natural Resources Research,2010,19(2):125-139.
[10] Sebtosheikh M A,Motafakkerfard R,Riahi M A,et al.Support vector machine method,a new technique for lithology prediction in an Iranian heterogeneous carbonate reservoir using petrophysical well logs[J].Carbonates Evaporites,2015,30(1):59-68.
[11] Shi N,Li H,Luo W.Data mining and well logging interpretation:application to a conglomerate reservoir[J].Applied Geophysics,2015,12(2):263-272.
[12] Li H,Tan F,Xu C,et al.Lithology identification of conglomerate reservoir based on decision tree method[J].Well Logging Technology,2010,34(1):16-21.
[13] 柴明锐,程丹,张昌民,等.机器学习方法对砂砾岩岩屑成分的预测——以西北缘X723井百口泉组为例[J].西安石油大学学报(自然科学版),2017,32(5):22-28.CHAI Mingrui,CHENG Dan,ZHANG Changmin,et al.Prediction of debris composition in glutenite by machine learning method:a case study in Baikouquan formation of well X723 in the NW Margin of Junggar Basin[J].Journal of Xi’an Shiyou University(Natural Science Edition),2017,32(5):22-28.
[14] 李诒靖,郭海湘,李亚楠,等.一种基于Boosting的集成学习算法在不均衡数据中的分类[J].系统工程理论与实践,2016,36(1):189-199.LI Yijing,GUO Haixiang,LI Ya’nan,et al.A boosting based ensemble learning algorithm in imbalanced data classification[J].Systems Engingeering Theory & Practice,2016,36(1):189-199.
[15] 宋建国,杨璐,高强山,等.强容噪性随机森林算法在地震储层预测中的应用[J].石油地球物理勘探,2018,53(5):954-960.SONG Jianguo,YANG Lu,GAO Qiangshan,et al.Strong tolerance random forest algorithm in seismic reservoir prediction [J].Oil Geophysical Prospecting,2018,53(5):954-960.
[16] Cracknell M J,Reading A M.The upside of uncertainty:Identification of lithology contact zones from airborne geophysics and satellite data using random forest and support vector machines[J].Geophysics,2013,78(3):WB113-WB126.
[17] 周雪晴,张占松,张超谟,等.基于粗糙集——随机森林算法的复杂岩性识别[J].大庆石油地质与开发,2017,36(6):127-133.ZHOU Xueqing,ZHANG Zhansong,ZHANG Chaomo,et al.Complex lithologic identification based on rough set-random forest algorithm[J].Petroleum Geology and Oilfield Development in Daqing,2017,36(6):127-133.
[18] Chen T,Guestrin C.XGBoost:A scalable tree boos-ting system[C].ACM SIGKDD International Conference on Knowledge Discovery and Data Mining,2016,785-794.
[19] Torlay L,Perrone-Bertolotti M,Thomas E,et al.Machine learning-XGBoost analysis of language networks to classify patients with epilepsy[J].Brain Informa-tics,2017,4(3):159-169.
[20] 贾文慧,孙林子,景英川.基于XGBoost模型的股骨颈骨折手术预后质量评分预测[J].太原理工大学学报,2018,49(1):174-178.JIA Wenhui,SUN Linzi,JING Yingchuan.Surgical prognosis quality score of femoral neck fracture base on XGBoost model [J].Journal of Taiyuan University of Technology,2018,49(1):174-178.
[21] Friedman J H.Greedy function approximation:A gradient boosting machine[J].The Annals of Statistics,2001,29(5):1189-1232.
[22] 范铭涛,沈全意,吴辉,等.复杂岩性裂缝—孔隙型储层孔隙度计算方法研究[J].天然气工业,2005,25(5):29-30.FAN Mingtao,SHEN Quanyi,WU Hui,et al.Porosity calculation method of complex lithological fractured-porous reservoir[J].Natural Gas Industry,2005,25(5):29-30.
[23] 牛聪,张益明,王迪,等.LA地区盒8段优质储层的特征与分布预测[J].石油地球物理勘探,2017,52(3):591-598.NIU Cong,ZHANG Yiming,WANG Di,et al.Prediction of high-quality reservoir characteristics and distribution in the area LA[J].Oil Geophysical Prospecting,2017,52(3):591-598.
[24] 方兴,孙夕平,张明,等.基于AVO流体反演的储层孔隙度预测技术[J].石油地球物理勘探,2012,47(3):469-472.FANG Xing,SUN Xiping,ZHANG Ming,et al.Re-servoir porosity prediction based on AVO fluid inver-sion[J].Oil Geophysical Prospecting,2012,47(3):469-472.
[25] 潘豪杰,张研,李红兵,等.基于贝叶斯理论的天然气水合物储层弹性—电性数据联合反演[J].石油地球物理勘探,2018,53(3):568-577,586.PAN Haojie,ZHANG Yan,LI Hongbing,et al.Joint inversion of elastic-electrical data for gas hydrate re-servoirs based on Bayesian theory[J].Oil Geophysical Prospecting,2018,53(3):568-577,586.