Abstract：With the development of high-efficiency acquisition and wide frequency, wide azimath, high density seismic exploration technology, seismic acquisition data shows exponential growth, which brings challenges to the field quality evaluation of such data. The present theoretical field evaluation method of massive seismic data and its adaptability in quality control of massive seismic acquisition data are studied in this paper. It is difficult for single attribute models to fully characterize the quality of seismic data. A multi-attribute discriminant analysis model for single shot records and its production process are designed. Given the severe subjectivity of the multi-attribute discriminant analysis model that completely relies on the standard records and the threshold value, an intelligent quality classification model for massive seismic data is put forward. With feature analysis of massive seismic data, an intelligent evaluation process of single shot records based on the random forest is proposed. Three sample enhancement techniques are used to solve the problem of small and unbalanced samples of single shot records. The random forest classification algorithm of single shot records and its key technologies are studied, including the branch node construction based on continuous seismic attri-butes, the selection of modeling parameters, and the evaluation of the classification results. The application of experimental data shows that the results of the new method are correct and ready to be highly parallelized. Finally, according to the ana-lysis of the correlations of these models as well as their adaptability and timeliness, the combined application of multiple models can meet the requirements of field quality control of massive seismic acquisition data.
冷广升.地震数据采集质量控制方法研究与应用[J]. 中国煤炭地质, 2010, 22(增刊1):67-72, 76.LENG Guangsheng. Study on seismic data acquisition quality controlling method and its application[J]. Coal Geology of China, 2010(S1):67-72, 76.
[3]
张翊孟, 刘秋林, 张永科.地震资料品质定量分析和采集参数优选[J]. 石油地球物理勘探, 2008, 43(增刊2):1-5.HANG Yimeng, LIU Qiulin, ZHANG Yongke. Quantitative analysis of seismic data quality and optimization of acquisition parameters[J]. Oil Geophysical Prospecting, 2008, 43(S2):1-5.
[4]
段云卿.地震资料自动评价系统[J]. 勘探地球物理进展, 2006, 29(3):221-224.DUAN Yunqing. Automatic seismic data evaluation system[J]. Progress in Exploration Geophysics, 2006, 29(3):221-224.
[5]
潘树林, 周熙襄, 钟本善.地震资料采集监控及评价系统的开发[J]. 物探化探计算技术, 2007, 29(1):12-14, 91.PAN Shulin, ZHOU Xixiang, ZHONG Benshan. De-velopment of seismic data acquisition monitoring and evaluation system[J]. Computing Techniques for Geophysical and Geomechnical Exploration, 2007, 29(1):12-14, 91.
[6]
蓝宣.SACS地震采集质量控制评价系统介绍[J]. 物探装备, 2003, 13(3):209-210.LAN Xuan. Introduction of SACS seismic acquisition data quality control appreciation system[J]. Equipment for Geophysical Prospecting, 2003, 13(3):209-210.
[7]
岩巍, 夏颖, 李铮铮, 等.G3i仪器野外数据采集质量监控软件简介[J]. 物探装备, 2015, 25(4):274-279.YAN Wei, XIA Ying, LI Zhengzheng, et al. Brief introduction of QC software in G3i instrument[J]. Equipment for Geophysical Prospecting, 2015, 25(4):274-279.
Mjolsness E and DeCoste D. Machine learning for science:State of the art and future prospects[J]. Scie-nce, 2001, 293(5537):2051-2055.
[10]
Howard W R. Pattern recognition and machine lear-ning[J]. Kybernetes, 2007, 36(2):275.
[11]
赵贤正, 邓志文, 白旭明.基于视觉特征的地震资料品质分析方法[J]. 石油地球物理勘探, 2016, 51(增刊1):42-46.ZHAO Xianzheng, DENG Zhiwen, BAI Xuming. Seismic data quality analysis based on visual features[J]. Oil Geophysical Prospecting, 2016, 51(S1):42-46.
周志尧, 石慧敏, 吴勇.基于地质导向的三维地震资料品质评价方法[C].吉林省科学技术协会学术部会议论文集, 吉林省科学技术协会学会学术部, 2014, 179-180.ZHOU Zhirao, SHI Huimin, WU Yong. The quality evaluation method of 3D seismic data based on geological orientation[C]. Proceedings of academic Department Meeting of Jilin Association for Science and Technology, 2014, 179-180.
张全胜, 罗春波, 杨宝珍, 等.Reland.SeisQC系统在地震采集质量监控中的应用[J]. 物探装备, 2013, 23(1):67-69.ZHANG Quansheng, LUO Chunbo, YANG Baozheng, et al. Application of Reland. SeisQC system for seismic acquisition quality control[J]. Equipment for Geo-physical Prospecting, 2013, 23(1):67-69.
[19]
杨振邦, 屈邵忠, 周卉丽.克浪软件在煤田地震勘探试验资料评价中的应用[J]. 煤炭技术, 2014, 33(3):61-63.YANG Zhenbang, QU Shaozhong, ZHOU Huili. Application of Kelang software in coalfield seismic exploration and evaluation of test data[J]. Coal Technology, 2014, 33(3):61-63.
[20]
黄有晖, 朱运红, 蔡明, 等.实时质量监控技术在复杂山地三维地震采集中的应用[J]. 天然气勘探与开发, 2015, 38(2):31-34.HUANG Youhui, ZHU Yunhong, CAI Ming, et al. Application of real-time quality control technology to 3D seismic acquisition in complex mountains[J]. Na-tural Gas Exploration and Development, 2015, 38(2):31-34.
[21]
Stuarl J R, Peter N.人工智能:一种现代的方法(第二版)[M]. 北京:清华大学出版社, 2006.
[22]
徐维秀.地震属性优化分析和预测及有效性方法研究[D].上海:同济大学, 2007:38-39.XU Weixiu. Study on Seismic Attribute Optimization Analysis and Prediction and Validity Method[D]. Tongji University, Shanghai, 2007:38-39.
[23]
周志华.Machine Learning机器学习[M].北京:清华大学出版社, 2019.
[24]
张玉玺, 刘洋, 张浩然, 等.基于深度学习的多属性盐丘自动识别方法[J]. 石油地球物理勘探, 2020, 55(3):475-483.ZHANG Yuxi, LIU Yang, ZHANG Haoran, et al. Multi-attribute automatic interpretation of salt domes based on deep learning[J]. Oil Geophysical Prospecting, 2020, 55(3):475-483.
[25]
Sebastian R, Vahid M. Python Machine Learning(2nd Edition影印版)[M]. 江苏南京:东南大学, 2018.
[26]
Scikit learn. Scikit-learn machine learning in python[OL]. https://scikit-learn.org/stable,2021.
[27]
Wu Q, Burges C, Svore K, et al. Adapting boosting for information retrieval measures[J]. Information Retrieval Journal, 2010, 13(3):254-270.
[28]
魏新建, 李书平, 陈德武, 等.复杂区域地震采集质量评价技术及其应用[J]. 石油物探, 2019, 58(1):27-33.WEI Xinjian, LI Shuping, CHEN Dewu, et al.Evaluation of seismic acquisition quality in complex area[J]. Geophysical Prospecting for Petroleum, 2019, 58(1):27-33.
[29]
孙哲, 杜清波, 翟金浩, 等.超高效混叠地震采集实时质控技术[J]. 石油物探, 2020, 59(2):177-185.SUN Zhe, DU Qingbo, ZHAI Jinhao, et al. Real-time quality control for ultra efficient blended seismic acquisition[J]. Geophysical Prospecting for Petroleum, 2020, 59(2):177-185.
