基于改进匹配追踪算法的时频属性在薄储层沉积微相研究中的应用

Time-frequency attribute based on modified matching pursuit algorithm and its application to sedimentary microfacies of thin reservoir area

匹配追踪算法具有较高的时频分辨率和局部自适应性,能同时在时间域和频率域获得较准确的定位,但仍存在低频阴影和运算效率低等问题,影响了该技术在实际工作中的运用。为提高匹配追踪算法的计算精度和运算效率,在基于Ricker子波匹配追踪方法的基础上,通过优化迭代方案,改进原子主频计算方法和压缩原子旁瓣的方式,提高了频率定位的准确性和时频谱的时间分辨率;采用自适应获取原子参数的方法,大大提高了算法的运算效率。模型实验结果和实际地震资料证明了该算法的有效性,在某薄储层地区应用该改进算法提取时频属性,切片结果很好地表征了砂岩的展布特征,有助于沉积微相的刻画。

The matching pursuit algorithm has higher time-frequency resolution and local adaptive capability and obtains precise position both in time and frequency domain.It is difficult to apply it in practical work due to the low-frequency shadow and low computation efficiency.This paper proposes a method to improve the calculation precision and operation efficiency of Ricker wavelet matching pursuit algorithm.The accuracy of frequency localization and time resolution of time-frequency spectrum is improved by optimizing iteration scheme which improves the calculating method of atom main frequency and compressed atom＇s sidelobe. The efficiency of algorithm is greatly improved by adaptive acquisition of atomic parameters.By using model experiment and actual seismic data,the effect of improved algorithm is proved.And in some thin reservoir area,the distribution of sandstone is characterized well by the time-frequency slice calculated by improved algorithm.It would help the sedimentary microfacies characterization.