自适应最优核时频分布在地震储层预测中的应用

摘要

提高时频分辨率是频谱成像技术研究的重点.自适应最优核时频分布采用随信号特征自适应变化的核函数,在模糊域对远离原点的互分量进行抑制,并尽可能的保留集中在原点附近的自分量.通过理论模型验证,该方法较好地抑制了交叉项干扰,同时较连续小波变换和平滑伪Wigner-Ville分布等方法具有更高时频分辨率.最后在营尔凹陷长沙岭地区的实例应用中,利用自适应最优核时频分布对该区目标储层进行了频谱成像处理,并结合沉积相特征对长沙岭地区进行有利区带预测.结果表明:该方法适用于实际地震信号的时频分析,对储层刻画优于传统方法,且对研究区储层预测具有有效性.%To improve resolution is the key point of spectral imaging. Adaptive optimal kernel time-frequency representation used a kernel function which can change adaptively with signal characteristics to weight the ambiguity function. The cross-components located away from the origin of the ambiguity plane are suppressed, and the auto-components centered at the origin are passed. Adaptive optimal kernel time-frequency representation decreases the impact of cross terms obviously and obtains a better time-frequency resolution tban the methods as CWT and SPWVD, which is proved by the model data. Based on adaptive optimal kernel time-frequency representation, the seismic data of the target layer in Changshaling area of Ying'er Sag are processed by spectral imaging. On the basis of spectral imaging results and sedimentary environment, a prediction of favorable zones in Changshaling area was made. The result shows that this method is not only suitable for the time-frequency analysis of seismic signal and excels the traditional methods in reservoir description, but also takes an effectiveness in reservoir prediction.