Permeability prediction using PSO-XGBoost based on logging data
GU Yufeng1, ZHANG Daoyong1, BAO Zhidong2
1. Strategic Research Center of Oil and Gas Resources, Ministry of Natural Resources, Beijing 100034, China; 2. China University of Petroleum (Beijing), Beijing 102249, China
Abstract:Models for permeability prediction generally can be classified into two major types, physical and fitting models. Universally, physical models are wel-comed by geophysicists since the predicted values are calculated on the basis of logging theory, but they show bad generalization on application due to strict requirements on logging data. Fitting models repre-sented by stepwise regression are capable to make quick prediction, but they are difficult to accurately and analytically explain the relationship between permeability and logging curves because of their cal-culation mechanisms, thus also presenting bad gen-eralization. In order to create a new and more pow-erful fitting model, XGBoost, a widely used fitting model at present, is selected and modified by PSO to optimize hyper-parameter tuning. Then the hybrid model PSO-XGBoost is proposed. In this paper, taking the tight sandstone reservoirs of the Chang 4+5 members as a case, the prediction capability of the PSO-XGBoost mo-del are validated by three well-designed experiments. The experiment results show that:①Compared with physical models, fitting models utilize a fewer parameters to complete prediction, and present better applicability on permeability prediction when modeling data are insufficient, but they have limits on generalization since the prediction is sensitive to the quality of mode-ling data and thereby usually unstable; ②SVR, GBDT, and XGBoost can be improved by PSO, and the formed PSO-SVR, PSO-GBDT and PSO-XGBoost can figure out permeability rapidly. In comparison, PSO-SVR and PSO-GBDT show relatively unstable prediction due to their sensitivities on the quality of learning samples, while PSO-XGBoost displays better performances in predicting efficiency, reliability of predicted results, and prediction stability. Therefore, PSO-SVR is deemed to be unsuitable on permeability prediction, and PSO-XGBoost suitable; ③The prediction capabilities of stepwise regression, PSO-SVM, PSO-GBDT, and PSO-XGBoost can be enhanced when more learning samples are trained.
窦占斌.致密含气砂岩储层测井评价方法研究[D].北京:中国地质大学(北京),2014.DOU Zhanbin.Logging Evaluation Study of Tight Gas Sandstone Reservoir[D].China University of Geosciences (Beijing),Beijing,2014.
[2]
周新波,段迎利,袁伟,等.M油田渗透率计算方法研究[J]. 科技创新与应用,2014,(32):76.ZHOU Xinbo,DUAN Yingli,YUAN Wei,et al. Permeability computation method research of M oilfield[J].Technology Creation and Application,2014,(32):76.
[3]
张冲,张占松,张超谟.基于等效岩石组分理论的渗透率解释模型[J].测井技术,2014,38(6):690-694.ZHANG Chong,ZHANG Zhansong,ZHANG Chaomo.A permeability interpretation model based on equivalent rock element theory[J].Well Logging Technology,2014,38(6):690-694.
[4]
陈俊,谢润成,刘成川,等.中江气田侏罗系致密砂岩气藏测井流体识别及定量评价[J].天然气工业,2019,39(增刊1):136-141.CHEN Jun,XIE Runcheng,LIU Chengchuan,et al.Flow Characterization and quantitative evaluation of the tight gas-bearing sandstone reservoirs in the Jurassic member of Zhongjiang gas field[J].Natural Gas Industry,2019,39(S1):136-141.
[5]
柴愈坤,冯沙沙,王华.致密砂岩储层物性参数建模方法探讨[J].中外能源,2017,22(5):39-43.CHAI Yukun,FENG Shasha,WANG Hua.Discussion on the physical parameter modeling method for tight sandstone reservoir[J].Sino-Global Energy,2017,22(5):39-43.
[6]
张鹏,张小莉.低孔低渗储层渗透率测井解释模型研究[J]. 地下水,2014,36(2):74-76.ZHANG Peng,ZHANG Xiaoli.Study on the well-logging interpretation model of reservoirs of low porosity and permeability[J].Ground Water,2014,36(2):74-76.
[7]
苏海波,王晓宏,张世明,等.低渗透油藏油水相对渗透率模型的分形表征方法[J].东北石油大学学报,2019,43(5):88-94.SU Haibo,WANG Xiaohong,ZHANG Shiming,et al.Fractal characterization method of oil-water relative permeability model in low permeability reservoirs[J]. Journal of Northeast Petroleum University,2019,43(5):88-94.
[8]
陈俊,沙里锞,王新海,等.用压覆岩心渗透率优化测井渗透率计算模型[J].断块油气田,2016,23(2):189-192.CHEN Jun,SHA Like,WANG Xinhai,et al.Optimization of logging permeability calculation model using overburden pressure core permeability[J].Fault-Block Oil and Gas Field,2016,23(2):189-192.
[9]
廖东良,吴海燕.基于流动单元改进的渗透率解释模型[J]. 测井技术,2015,39(6):802-806.LIAO Dongliang,WU Haiyan.Modified permeability model based on flow units[J].Well Logging Techno-logy,2015,39(6):802-806.
[10]
刘建建,赵军龙,屈晓荣.鄂尔多斯盆地S区长6储层测井解释模型的建立及应用[J].中外能源,2016,21(8):32-38.LIU Jianjian,ZHAO Junlong,QU Xiaorong. Establishment and application of logging interpretation model of Chang 6 reservoir in area S of Ordos Basin[J]. Sino-Global Energy,2016,21(8):32-38.
[11]
刘敏.长庆T气田致密砂岩气层测井评价方法[D].山东青岛:中国石油大学(华东),2016.LIU Min.Method Research for Well Logging Evaluation of Tight Gas-Bearing Sandstone in T Gas Field of Changqing[D].China University of Petroleum (East China),Qingdao,Shandong,2016.
[12]
由嘉雨.榆树林油田葡萄花油层储层参数精细评价[D].黑龙江大庆:东北石油大学,2016.YOU Jiayu.Refined Evaluation of Reservoir Parameters of Putaohua Oil Layer in Yushulin Oil Field[D]. Northeast Petroleum University,Daqing,Heilongjiang,2016.
[13]
邓浩阳.高孔低渗碳酸盐岩储层孔隙结构及物性表征方法研究[D].四川成都:西南石油大学,2018.DENG Haoyang.The Evaluation Method of Pore Structure and Physical Property in Carbonate Rock Reservoir with High Porosity and Low Permeability[D].Southwest Petroleum University,Chengdu,Sichuan,2018.
[14]
李佳.基于机器学习的多孔介质渗透率预测研究[D].浙江杭州:浙江大学,2019.LI Jia.A Machine Learning-Based Approach for Permeability Prediction of Porous Media[D].Zhejiang University,Hangzhou,Zhejiang,2019.
[15]
Majid B,Hadi R.Reservoir rock permeability prediction using SVR based on radial basis function kernel[J].Carbonates and Evaporites,2019,34(3):699-707.
[16]
Zhang G Y,Wang Z Z,Li H J,et al.Permeability prediction of isolated channel sands using machine lear-ning[J].Journal of Applied Geophysics,2018,159(9):605-615.
[17]
Subasi A,El-Amin M F,Darwich T,et al.Permeability prediction of petroleum reservoirs using stochastic gradient boosting regression[J].Journal of Ambient Intelligent and Human Computing,2020,53(2):147-153.
[18]
韩启迪,张小桐,申维.基于梯度提升决策树(GBDT)算法的岩性识别技术[J].矿物岩石地球化学通报,2018,37(6):1173-1180.HAN Qidi,ZHANG Xiaotong,SHEN Wei.Lithology identification technology based on gradient boosting decision tree(GBDT) algorithm[J].Bulletin of Minera-logy,Petrology and Geochemistry,2018,37(6):1173-1180.
[19]
谢云欣.四川盆地雷口坡组油气聚集带特征及分布评价[D].四川成都:成都理工大学,2018.XIE Yunxin.Oil and Gas Accumulation Zones Characteristics and Distribution Evaluation of Leikoupo Formation,Sichuan Basin[D].Chengdu University of Technology,Chengdu,Sichuan,2018.
[20]
Zhang C S,Zhang Y,Shi X J,et al.On incremental lear-ning for gradient boosting decision trees[J].Neural Processing Letters,2019,50(1):957-987.
[21]
Chen T,Guestrin C.XGboost:A scalable tree boosting system[C].ACM SIGKDD International Confe-rence on Konwledge Discovery and Data Mining,2016,785-794.
[22]
Torlay L,Perrone-Bertolotti M,Thomas E.Machine lear-ning-XGBoost analysis of language networks to classify patients with epilepsy[J].Brain Informatics,2017,4(3):159-169.
[23]
闫星宇,顾汉明,肖逸飞,等.XGBoost算法在致密砂岩气层测井解释中的应用[J].石油地球物理勘探,2019,54(2):447-455.YAN Xingyu,GU Hanming,XIAO Yifei,et al.XGBoost algorithm applied in the interpretation of tight-sand gas reservoir on well logging data[J].Oil Geophysical Prospecting,2019,54(2):447-455.
[24]
杨维,李歧强.粒子群优化算法综述[J].中国工程科学,2004,6(5):87-94.YANG Wei,LI Qiqiang.Survey on particle swarm optimization algorithm[J].Engineering Science,2004,6(5):87-94.
[25]
刘建华.粒子群算法的基本理论及其改进研究[D].湖南长沙:中南大学,2009.LIU Jianhua.The Research of Basic Theory and Improvement on Particle Swarm Optimization[D].Center South University,Changsha,Hunan,2009.
[26]
温阳东,李龙剑.基于LDIW-PSO算法的BP神经网络在压力传感器中的应用[J].化工自动化及仪表,2014,41(9):1031-1034.WEN Yangdong,LI Longjian. Application of LDIW-PSO algorithm-based BP neural network in pressure sensor[J]. Control and Instruments in Chemical Industry,2014,41(9):1031-1034.
[27]
赵冰瑶.姬塬油田王盘山地区长4+5储层成岩相微观孔隙结构及渗流特征研究[D].陕西西安:西北大学, 2018.ZHAO Bingyao.The Study on Microscopic Pore Structure and Percolation Characteristics of Diagene-tic Facies in Chang 4+5 Reservoir of Wangpanshan Area, Jiyuan Oilfield[D].Northwest University, Xi'an, Shaanxi,2018.
[28]
王文枫,岳大力,赵继勇,等.利用地震正演模拟方法研究地层结构——以鄂尔多斯盆地合水地区延长组三段为例[J].石油地球物理勘探,2020,55(2):411-418 WANG Wenfeng,YUE Dali,ZHAO Jiyong,et al.Research on stratigraphic structure based on seismic forward modeling:A case study of the third member of the Yanchang Formation in Heshui area, Ordos Basin[J].Oil Geophysical Prospecting,2020,55(2):411-418.
[29]
李慧琼,张盟勃,蒲仁海,等.黄257井区叠前纵波方位各向异性裂缝分布预测[J].石油地球物理勘探,2017,52(2):350-359.LI Huiqiong,ZHANG Mengbo,PU Renhai,et al.Late Triassic fracture detection with seismic azimuth anisotropics in Huang 257 survey,Ordos Basin[J].Oil Geophysical Prospecting,2017,52(2):350-359.
[30]
周荔青,刘忠群,蒲仁海,等.镇泾地区长8段三维地震强振幅异常带成因探讨[J].石油地球物理勘探,2017,52(2):371-380.ZHOU Liqing,LIU Zhongqun,PU Renhai,et al.Strong amplitude anomaly on 3D seismic survey in the Southwestern Ordos Basin[J].Oil Geophysical Pro-specting,2017,52(2):371-380.