基于离散小波分析的水淹特征提取方法

EXTRACTING METHODS OF THE WATERFLOODED CHARACTERISTICS BASED ON DISCRETE WAVELET ANALYSIS

针对现有水淹层解释存在多解性的问题,提出了一种基于离散小波变换的水淹特征提取方法.该方法利用分解后小波系数高频信息与测井曲线光滑度之间的对应关系,进而反映不同沉积韵律类型中各水淹级别的水淹强度,建立了基于小波分解系数的不同沉积韵律水淹级别精细判别方法.以沉积韵律中部为标准层,结合待判小层在韵律中的空间位置、时频特征及其相对电性特征,自动判别各小层的水淹级别.密闭取心井测井曲线解释结果表明：小波系数高频分量可对水淹层电性特征进行有益补充,提高水淹层精细判别符合程度.

In the light of the existing multiple-solution problem of the waterflooded layer interpretation, the meth- ods of the waterflooded features are proposed on the basis of the discrete wavelet transformation. By means of the corresponding relationship between the high-frequency information after the wavelet coefficient decomposition and smoothness of the logging curves, the waterflooded strengths of each waterflooded level in the different sedimentary rhythms are reflect for the methods. Thus the fine discriminating and identifying methods of the waterflooded levels in the different sedimentary rhythms are established based on the wavelet decomposition coefficients. Taking the middle of the rhythms as the standard layer and integrating with the spatial positions, time-frequency properties and relative electrical characteristics for the sublayers that need to identify, the waterflooded levels of each sublayer are automatically discriminated. The interpreted results of the well logs for the sealed coring wells show that the high- frequency component of the wavelet coefficient can be a useful supplement to the electrical characteristics of the wa- terflooded reservoirs and furthermore enhance the fine discrimination coincidental degree of the waterflooded layer.