基于小波分析的蚂蚁追踪技术

Ant tracking technology based on wavelet analysis

黄健良 俱炜农 蔺美丽 李斌 李博文4

中海油研究总院有限责任公司，北京 100028 中国石油长庆油田分公司第七采油厂，陕西 西安 710201 中国地质调查局广州海洋地质调查局，广东 广州 510075 西京学院，陕西 西安 710123

CNOOC Research Institute Co., Ltd., Beijing 100028, China No.7 Oil Production Plant, Changqing Oilfield Company, PetroChina, Xi′an 710201, China Guangzhou Marine Geological Survey, China Geological Survey, Guangzhou 510075, China Xijing Unversity, Xi′an 710123, China

三参数小波是近年来新发展的小波分析方法，文中介绍了该小波分析的基本原理。模型一的研究展示了每个参数的影响，并总结出处理地震资料常用的参数组合；模型二的研究展示了利用小波域分频处理提高地震资料分辨率的原理；模型三的时频分析，对比了三参数小波变化、Morlet连续小波变化、S变换的时?鄄频聚焦性。研究表明，三参数小波灵活性高，在时?鄄频域有比较理想的局部化性质。相对于传统相干检测技术，蚂蚁算法对断裂等非连续性边界的刻画效果更佳。文中将蚂蚁追踪技术与三参数小波分析结合，形成了基于小波分析的蚂蚁追踪技术流程及方法，可获取多种尺度级别的地震边缘信息，更深入有效地帮助解释多序级别断裂及多尺度岩性体边界。实际应用表明，该技术对断裂系统的刻画与解释方面有良好的效果。

Three-parameter wavelet is a new method of wavelet analysis. The basic principle of three-parameter wavelet analysis is introduced. Through the analysis and the test of model 1, the influence of each parameter on wavelet is studied, and the commonly used parameter combinations for processing seismic data are summarized. The forward test of model 2 shows the principle of improving the resolution of seismic data by using wavelet domain frequency division processing. By using the time-frequency analysis test of model 3, the time-frequency focusing ability of the three-parameter wavelet transform, Morlet continuous wavelet transform and S transform is compared. It is found that three-parameter wavelet has high flexibility and ideal localization in time-frequency domain. Ant algorithm can much better get the feature of fracture system with the algorithm applied in seismic data interpretation. This paper combines ants tracking technology and the three-parameter wavelet, and a new ant tracking technique based on wavelet analysis is developed. The technology can obtain multiscale seismic edge information, and is useful to detect fracture system. Practical application shows that the technology has good effects in interpretation of fracture system.

10.6056/dkyqt202001010